Landownership in Eastern Germany Before the Great War: A Quantitative Analysis

L A N D OW N E R S H I P I N E A S T E R N G E R M A N Y B E F O R E T H E G R E AT WA R This page intentionally left...

Author: Scott M. Eddie

244 downloads 994 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

L A N D OW N E R S H I P I N E A S T E R N G E R M A N Y B E F O R E T H E G R E AT WA R

This page intentionally left blank

Landownership in Eastern Germany before the Great War A Quantitative Analysis SCOT T M. EDDIE

1

1

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With ofﬁces in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Scott Eddie 2008

The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by Laserwords Private Limited, Chennai, India Printed in Great Britain on acid-free paper by Biddles Ltd, King’s Lynn, Norfolk ISBN 978–0–19–820166–3 1 3 5 7 9 10 8 6 4 2

for David, Kathleen, and Paul

Preface S T RU C T U R E O F T H E B O O K This book is a quantitative study of landownership in the seven ‘core provinces’ (Kernprovinzen) of eastern Prussia (Pomerania, East Prussia, West Prussia, Brandenburg, Posen, Province of Saxony, and Silesia) in the period before 1914. The principal sources for the data used in this book are (1) a series of directories of large landed properties produced by a private publisher with active cooperation from the Ministry of Agriculture of Prussia, and (2) the published volumes of results of the land tax cadastre (Grundsteuerveranlagung) of the early 1860s. I have taken data from many other sources as well, including the German government’s censuses of agriculture and industry carried out in 1882, 1895, and 1907, but I show early on why the data from the censuses of agriculture cannot be used to infer anything about landownership—hence the reliance on the directories mentioned above. Because generalizations or conclusions require a solid quantitative base to be believable, I spend an entire chapter, Chapter 2, on a discussion of the data themselves and the efforts I have made to make these data as accurate, complete, and internally consistent as possible. (If you really trust me, you could skip Chapter 2, but I think a critical understanding of the data is vitally necessary to a critical understanding of what I have done with those data. In fact, I think this form of ‘source criticism’ should be a general requirement for all quantitative studies.) Motivation for this work was initially the contradiction between the pervasive folklore about the dominance of the great landlords, the Junker, in the basically agrarian society of these seven provinces, and the continuing (and indeed parallel) story of agricultural crisis and the loss of landed estates to the bourgeoisie during the nineteenth century. To my mind, this contradiction could only be resolved through the gathering and analysis of relevant information, in other words, by a quantitative study of such questions as ‘who were the great landlords?’ and ‘how dominant were they?’ and ‘did their degree of dominance wax or wane over time?’ Toward such resolution, one should also examine questions about the extent to which the bourgeoisie became large landowners and the degree to which they were able to penetrate the very highest ranks of the landed elite. As a start toward providing an answer, I have classiﬁed the owners of large properties into socially deﬁned categories: royalty and high nobility, lesser nobility, non-noble persons, religious bodies, the State, communities, banks and business ﬁrms, and a catch-all—and very minor—category of ‘other owners’.

Preface

vii

Although the focus is on the performance of the groups, there are many references and even whole sections devoted to individual landowners, especially in Chapters 4 and 5, which deal with owners by rank and station and with the top of the pyramid, the greatest of the great landholders, respectively. Here the reader will ﬁnd the extent of the holdings of various groups of royalty, nobility, and others, along with lists of the ﬁfty biggest landowners by name and classiﬁed two ways—according to the total acreage of their holdings and to the total land tax assessments of those holdings, a rough proxy for value of the land. Because some land, especially in the province of Saxony, was much more valuable than other land, the overlap in the two top-ﬁfty lists—area and value—for any given year is far from complete. Since the data both permit and suggest further questions, in subsequent chapters I also deal with a certain privileged class of properties, the knight’s estates (Ritterg¨uter, Chapter 6), and with industrial enterprises on rural estates in the seven provinces (Chapter 7). Individual owners pop up again in Chapter 7. In an earlier chapter (Ch. 3), on the other hand, I set out the distribution of large properties as such, from several different points of view, including an examination of regional patterns in landholding and in quality differences in land—this latter both in general for districts and in the case of the large properties in particular. Just as I grouped owners by social position, so have I grouped properties by their mix of economic activities—the share of their land devoted to (a) arable agriculture, (b) animal husbandry, (c) forestry, and (d) other activities. The large properties fell into six quite easily deﬁnable and generally mutually exclusive clusters, allowing me to examine the size distribution, regional location, and ownership status of these clusters in Chapter 8. The ﬁnal chapter of the book takes the form of a Schlusswort. ‘Epilogue’ hardly captures the ﬂavour of the German term, since Schluss can—and often does in German books—imply both a closing and a conclusion or result. My closing chapter combines a brief summary of the main things that I have done in the book with a consideration of the question of whether the procedures of the land tax cadastre, on whose data I rely so heavily, might have resulted in a systematic under-assessment of the land tax for certain classes of large properties or landowners relative to other large properties or classes of owners. The reader who is already familiar with the literature on landownership in Germany will note that I do not cover another special class of property, the entailed estates (Fideikommisse), nor do I devote more than an occasional mention to the single biggest buyer of land during the later nineteenth century, the Royal Settlement Commission (K¨onigliche Ansiedlungskommission). I do not have anything to add to the ﬁne account of Fideikommisse by Klaus Hess (see Ch. 1), and I have treated the land purchases of the Settlement Commission in some detail in other works. I also do not touch the issue of debt: so long as it is not clear whether a high level, or increase in the level, of debt relative to land value or income is a sign of economic illness or economic health, I have let my

viii

Preface

general disinclination to treat the question of debt rule. This question is best left to the efforts of other scholars. As an organizational tool and an aid to the reader, I begin each chapter except the ﬁrst—which is itself an introduction, with a review of relevant literature and data considerations—with a preview of what is in the chapter. I hope that the readers of this book will ﬁnd these previews useful, that they will make reading the body of the chapters both quicker and easier. If the previews do their job properly, they should also provide a guide through some perhaps unfamiliar thickets of statistical methodology, so that the importance of the outcome would not be obscured by the density of the thicket. THANKS A study whose gestation period is measured in decades rather than weeks or months is bound to accumulate a long list of debts to persons and institutions that have helped it on its way. Such is clearly the case with this book, and I can only fervently hope that, over the years, I have not forgotten anyone who should be thanked for such help. My greatest material debts are to those institutions that have provided ﬁnancial support for this project. The Social Sciences and Humanities Research Council of Canada provided essential funding at both the beginning and the ending stages of the research, and my own university and several of its parts have been most generous in their help—particularly the Connaught Fund, the former Centre for Russian and East European Studies, Erindale College, and the Joint Initiative in German and European Studies. Other granting bodies, most notably the Fulbright Program, the National Academy of Sciences, and the American Council of Learned Societies in the USA have also made important ﬁnancial contributions toward the completion of this project. There are also institutions that have been generous hosts to me personally, where I have resided for a time in the course of this research. In particular, I am grateful to All Souls College in Oxford for an academic year as visiting fellow; and for stays of a semester or less to the University of Rostock, the Historical Commission in Berlin, and the Technical University of Berlin in Germany; and to the Europa Institut in Budapest. At these institutions I owe my sincerest thanks to those who made room for me and smoothed and guided my way while there: Professors Peter Mathias, Georg Moll, Wolfram Fischer, Heinz Reif, and Ferenc Glatz, respectively. At my own university, I am particularly grateful to Professor Robert Bothwell and the International Relations Programme of Trinity College for providing ofﬁce space and support after my retirement from teaching; this made completion of the project incomparably easier. I also owe thanks to my departmental colleague Prof. Angelo Melino for much-needed economic advice.

Preface

ix

Research for this book has been carried out in a large number of libraries and archives, to whose helpful and knowledgeable personnel I owe a particular debt of gratitude: besides libraries at the institutions mentioned above, I have also worked in the library of the University of Greifswald; the Staatsbibliothek and the libraries of Humboldt University, the Free University, the former Historical Institute of the Academy of Sciences, and the Evangelical Church of the Union in Berlin; the Deutsche B¨ucherei in Leipzig, and in the library of the Central Statistical Bureau in Budapest. I learned to love archival research working in the Prussian State archives, ﬁrst in their former location in Merseburg, and later in the now central and combined location in Berlin-Dahlem. The Landesarchiv of the state of Brandenburg, in Potsdam, also provided many hours of fruitful research and many instances of helpful advice, and the archive of the University of Halle gave me insight into the scholarly career of Johannes Conrad, in whose footsteps I hope I am following in this book. The list of people who have been instrumental in whatever success my research might have is long, indeed, and it is difﬁcult to know where to start. Perhaps a roughly chronological order might be best: At the beginning of the study I beneﬁted from discussions with Drs Ilona Buchsteiner, Hartmut Harnisch, and Hans-Heinrich M¨uller in particular. At a somewhat later stage I received help and inspiration from Dr Ren´e Schiller while he was still writing his dissertation, and Prof. Heinz Reif has encouraged my research into Prussian landownership—even as I am willing to bet he was frustrated by its slow progress—for more years than either of us might want to count. At every stage I have relied on, and beneﬁted from, the advice of my friend and colleague, Dr Iv´an Sz´ekely, who also organized and supervised much of the original data entry and its checking and correction. He also taught me a lot about what one can do with spreadsheets. No one except myself has put more time and effort into the research for this study than another friend and colleague, Dr Christa Kouschil, who would have earned my undying gratitude for even a tenth of what she in fact has done. Dr Robert Nelson, who also has made a considerable contribution, especially to my understanding of the context in which I have placed this study, deserves many thanks, as well. A number of graduate students have served as research assistants at various stages of this project: Ralph Czychun, Aneta Szyjka Turkiewicz, Silvija Jestrovic, Heike Frenzel, James Stewart, and Jennifer Umlauf have all earned my respect for the quality of their work and their contribution to the completion of this research. The University of Toronto’s Work-Study Program has provided me with a succession of really remarkable undergraduate research assistants who have made material contributions to the outcome of this project. I am particularly grateful to Svetlana Panikian, Douglas Hughey, Joanna Sobala, Barbara Rado´n, Saˇsa Jurak, Paul Brykczynski, Huma Usmani, and Yuliya Lesnyak, all of whom worked

x

Preface

part-time during two or more academic years on this and other projects, to my great beneﬁt. Others who made signiﬁcant contributions during a summer and/or a single academic year include Kim Pagel, Alizeh Hussain, Connie Tram, Julia Muravska, David Leaman, Bridget Daley, Viktoria Csozik, Salman Saaed, Susan Gaal, Trevor Johnston, Jennifer Snively, Kristine Kjeldsen, Ahmed Gulzar, Gabriella Szolnoki, and Orsolya Soos. When someone uses masses of data, he incurs a massive debt to those people who enter the data into computer ﬁles. Chief among those to whom I owe this debt are Ute Aul, Sylke Oehming, Bence Berei, and Krisztina K´ad´ar. For permission to use the map of Central Europe in 1910 to show the location of the provinces here covered I am indebted to my friend and colleague, Professor Paul Robert Magocsi, Chair of Ukrainian Studies at the University of Toronto. I am grateful for the patience, cordiality, helpfulness, and impressive professionalism of all the members of Oxford University Press who have been involved in the publication of this book, starting with a succession of understanding editors, deputy editors, and commissioning editors: Robert Faber, Anthony Morris, Ivon Asquith, Ruth Parr, Timothy Saunders, Rupert Cousens, and Seth Cayley. The production team has been outstanding, and always right on schedule, under the direction of Kate Hind and Kate Walker. Veronica Ions as copy editor gently forced consistency on me and my manuscript while ferreting out my errors; Andrew Hawkey as proofreader found and corrected all the bugs in the typeset version. Last in the book, but certainly not least, comes the index, very professionally prepared by Noeline Bridge. These are the people who are making me look good. Acknowledgements often end with ritual thanks to the author’s spouse, much as scholarly works in communist countries often used to begin with ritual obeisance to Marx and Lenin. In my case, thanks are due more than just ritually to my wife Sharon for encouragement, patience, prodding, and even proofreading. If what I have written is clear and understandable, give her the credit. If not, blame my stubbornness. T R A N S L AT I O N S A N D T E R M I N O LO G Y All Prussian provinces were divided into either two or three Regierungsbezirke, the direct translation of which would be ‘administrative districts’. That is, for example, the only translation given in the 1901 Muret–Sanders German to English dictionary as well as in the more modern Langenscheidt dictionary of 1981. I have chosen to deviate slightly from this translation because of the translation I use for the name of the subunit of a Regierungsbezirk, namely the Kreis, as explained below: A Regierungsbezirk typically consisted of from a half-dozen to twenty Kreise, the direct translation of which would be ‘circles’. The usual translation of Kreis is ‘county,’ which I believe to be wrong on three principal counts:

Preface

xi

1. The English term ‘county’ has an exact historical and linguistic equivalent in the German word Grafschaft. 2. A typical Prussian Kreis was smaller than an English or American county: a. The average area of the ﬁfty administrative counties in England in 1891 was 260,578 hectares, almost exactly triple the average area of an East Elbian Landkreis in 1896/7 (87,307 ha; I have left cities which were independent Kreise out of the reckoning here). Only four of the English counties (Soke of Peterborough, London, Isle of Wight, and Rutland) were smaller than the average eastern Prussian Landkreis, whereas not a single Landkreis was as big as the average English county. The largest district, Stolp in the K¨oslin region of Pomerania, covered 226,691 hectares in 1896/7.¹ b. Only the very largest of the nineteenth-century Kreise in these provinces were comparable in size to an American county. Had Stolp, the largest district in area, been square, it would have measured about 47.6 kilometres (or 29.4 miles) on a side, equivalent to a county of just under twenty-ﬁve townships (a Midwestern American township is 6 miles on a side, or 36 sections, just over 93 square km). For comparison, Eddy and Foster counties, neighbouring small counties in central North Dakota, each contain eighteen townships.² Only ten of the 259 Kreise were as large or larger than either of these two counties, whereas a nearmajority (127) were less than half their size. The very smallest of the 259 Landkreise, Beuthen in the Oppeln region of Silesia, covered only 10,408 hectares in 1896/7, so was barely more than the size of a single township. 3. The Kreisstadt (typically translated as ‘county seat’) did not have to be in the Kreis; it could lie outside. Moreover, one city could be the Kreisstadt of more than one Kreis, one of which might be the city itself.³ For these reasons I have chosen to translate Kreis as ‘district’ and therefore translate Regierungsbezirk as ‘administrative region’.⁴ I believe this pair of translations also to be in accord with most people’s intuitive idea that a region would be larger than a district. ¹ Calculated from data for England taken from http://www.answers.com/topic/administrativecounties-of-england and data for Prussia taken from Statistisches Handbuch f¨ur den Preussischen Staat, 1898, 15–21. ² The total area of Eddy County is 630 sq. miles (1632 sq. km), and that of Foster County, 635 sq. mi. (1645 sq. km). Data from http://www.fedstats.gov/qf/states/38/. ³ For example, the city of Posen (Pozna´n) was the Kreisstadt of three Kreise: Posen-Ost, PosenWest, and Stadtkreis Posen, just as the city of Danzig (Gda´nsk) was for Danziger H¨ohe, Danziger Niederung, and Stadtkreis Danzig. ⁴ Richard Blanke employs the translation ‘regency’ for Regierungsbezirk. This is unusual, and conveys less information to a reader unfamiliar with Prussia than does the translation I have chosen. See Richard Blanke, Prussian Poland in the German Empire, East European Monographs, 86 (Boulder: University of Colorado Press, 1981).

xii

Preface

In the text, notes, and tables of this book I have tried to limit myself to only three abbreviations of German terms: 1. GSRE for Grundsteuerreinertrag (‘land tax net yield’), the basis of assessment for the land tax; 2. RG for Rittergut (knight’s estate), a property originally intended to provide sustenance for a knight in return for his obligation to serve his monarch in time of war. Later, ownership of a knight’s estate became the basis for membership in the ﬁrst estate in those years when the ﬁrst estate had an allotted number of seats in the provincial legislatures (please see Ch. 6); and 3. GStA for Geheimes Staatsarchiv Preussischer Kulturbesitz, the Prussian state archive. S.M.E.

Contents List of Tables List of Figures Map: Central Europe in 1910, with seven East Elbian provinces outlined in white 1. Introduction

xvii xxiii xxiv 1

1.1. The Setting 1.2. Who Owned the Land? How do we Know? 1.3. Address Books of Landed Properties as Sources for Scholars

1 4 6

1.3.1. The Pioneer: Johannes Conrad 6 1.3.2. Conrad’s Students and Followers 9 1.3.3. Recent Work in the Conradian Tradition: Hess, Nabert, Buchsteiner, Schiller 10

1.4. Farm Statistics are not Property Statistics 1.4.1. 1.4.2. 1.4.3. 1.4.4.

20

Incomparabilities in the Data of the Agricultural Censuses 23 Owning, Renting, and Ownership Statistics 27 Can Census Data Serve as a Proxy for Property Data? 30 Adjustment of the 1882 Census Data 30

2. Data Considerations 2.1. Preview 2.2. Principal Sources and Uses of Ownership Data 2.3. Correcting and Adjusting the Data

37 37 37 41

2.3.1. Internal Consistency of the Data 41 2.3.2. Problems with Individual Properties 43 General 43 The Principality of Pless 44 2.3.3. Incomplete Coverage of the Data 46 2.3.4. Adjusting for the Anomalies in Reporting State-Owned Properties 48 Introduction 48 Adjustments to State and Crown Properties for Pomerania, Brandenburg, and East Prussia 48

3. The Geographical and Size Distribution of Large Landed Properties 3.1. Preview 3.2. Introduction

54 54 55

xiv

Contents 3.3. The Land Endowment of the Eastern Provinces 3.4. The Regional Distribution of Large Properties 3.5. The Size of Large Properties

55 59 67

3.5.1. Traditional View 1: Distribution by Total Area of Properties 67 3.5.2. Traditional View 2: Distribution by Total Area Owned 69 3.5.3. A Non-Traditional View 1: Distribution by Tax Value 73 3.5.4. A Non-Traditional View 2: Logarithmic Distributions 76 3.5.5. Summary 78 3.5.6. Distribution in the Provinces 83

4. Ownership of Land by Rank or Class of Owner

86

4.1. Preview 4.2. Introduction 4.3. Ownership by Rank and Type of Owner

86 87 91

4.3.1. Royalty 93 4.3.2. The Nobility 98 Princes and Dukes 98 Counts and Countesses 102 Barons and Baronesses 103 Untitled Nobility 106 Summary 108 4.3.3. Non-Noble Persons (‘Bourgeoisie’) 109 4.3.4. Another Look at Rank: Military Ofﬁcers 111 4.3.5. The Prussian State 116 4.3.6. Non-Physical Juridical Persons (Other than the Prussian State) 117

4.4. Summary 5. The Biggest Landowners 5.1. 5.2. 5.3. 5.4.

Preview Introduction: Conrad’s Twenty Largest Landowners The Top Fifty Landowners by Total Area and Total Tax Value The Super Elite: The Top Ten Landowners

6. Knight’s Estates: A Special Class of Property 6.1. Preview 6.2. Introduction 6.3. Numbers and Area of Knight’s Estates in East Elbia

119 123 123 124 129 146 152 152 153 155

6.3.1. General Considerations 155 6.3.2. Coverage and Inclusion 157

6.4. East and West Prussia, 1834–1884/5

158

Contents 6.5. Ownership Changes in All Seven Provinces after c.1882 6.6. Who Owned How Much?

xv 163 165

6.6.1. Was there Specialization in Ownership? 167 6.6.2. Grouping Properties by their Land-Use Proﬁle 168 6.6.3. Did Different Classes of Owners Specialize in What Types of Properties They Owned and Where? 170 6.6.4. Why did Knight’s Estates Appear to Command a Premium over Other Large Properties in the Market? 173

6.7. Summary of Examination of Knight’s Estates 7. Industry on the Land 7.1. Preview 7.2. Introduction

173 175 175 176

7.2.1. The Data 177 7.2.2. Summary of Distribution of Industrial Establishments 178 7.2.3. Numbers of Certain Industrial Establishments from Tax Data 180 7.2.4. Industrial Establishments according to the Industrial Censuses 182

7.3. Industrial Establishments on the Larger Estates

185

7.3.1. Share of Census-Listed Industrial Establishments on Large Estates 187 7.3.2. Distribution by Province and Size Category over Time 188 7.3.3. Who Owned the Properties with Industrial Establishments on Them? 190 7.3.4. Owners with Twelve or more Industrial Establishments on their Land 193

7.4. Factors Inﬂuencing the Location of Rural-Based Industrial Enterprises 7.5. Summary and Conclusions 8. Classifying Land by its Use 8.1. Preview 8.2. Introduction: The Distribution of Land Quality and Value 8.3. Classifying Properties by Land-Use Proﬁles 8.3.1. Calculating Land-Use Proﬁles 224 8.3.2. Cluster Analysis and the Number of Properties 225 8.3.3. Cluster Analysis and the Area and Tax Value of Properties 229 8.3.4. Regional Concentration of Properties by Land-Use Proﬁles 230 8.3.5. Was there Specialization in Ownership of Large Properties? 232

200 207 210 210 211 222

xvi

Contents

9. Concluding Remarks 9.1. 9.2. 9.3. 9.4.

236

The Book Started with a Sermon What has this Study Shown? Was the Land Tax Biased to Favour Big Holdings? Auf Wiedersehen

236 236 242 249

Appendix. Coding the Handbook Data A.1. Location Coding A.2. Owner Codes A.2.1. Physical Persons 251 A.2.2. Non-Physical Legal Persons

250 250 250 252

A.3. Other Codes A.3.1. Additional Characteristics of Owner or Property A.3.2. Reference Variables and Fields 256

254 254

List of Works Cited 1. Archival Sources and Prussian Laws 2. Ofﬁcial Government Publications 3. Directories of Landowners and Landed Properties 4. Articles, Book Chapters, Lectures, Monographs, Reference Works

258 258 258 259

Index

265

260

List of Tables 1.1. Amending Buchsteiner’s comparison of farm census and Handbook (HB) data, c.1907 1.2. Summary differences between Buchsteiner’s and Eddie’s calculations from the same sets of data for Pomerania 1.3. Woodland area of large farm enterprises in 1882 and 1895 censuses 1.4. Agricultural area and total area per farm, all seven eastern provinces, 1882 1.5. Farm enterprises with owned and leased land, all seven eastern provinces, 1895 1.6. Census data concerning ownership relations on farms in Pomerania, 1907 1.7. Adjusted areas for large agricultural enterprises in seven eastern provinces, 1882 census 1.8. Agricultural census data and Handbook data on number of farms and properties, seven eastern provinces 1.9. Ratio of number of properties to number of farms by size, seven eastern provinces 1.10. Agricultural census data and Handbook data on agricultural area of farms and properties, seven eastern provinces 1.11. Ratio of agricultural area of properties to agricultural area of farms by size, seven eastern provinces 1.12. Agricultural census data and Handbook data on total area of farms and properties, seven eastern provinces 1.13. Ratio of total area of properties to total area of farms by size, seven eastern provinces 1.14. Agricultural census data and Handbook data on average total area of farms and properties, seven eastern provinces 1.15. Ratio of average total area of properties to average total area of farms by size, seven eastern provinces 2.1. Dates of data ‘snapshots’ for the seven East Elbian provinces 2.2. Total area of province and of various categories of properties, c.1882 (unadjusted data) 2.3. Share of area and tax value of properties of 100 hectares or more in area and tax value of all land outside of cities, by province (unadjusted ﬁgures) 2.4. Brandenburg: original data for State and Crown properties (before adjustment) 2.5. Forest land in all principal properties in the Handbooks, for all owners, Brandenburg and Pomerania (unadjusted data)

15 17 25 26 28 29 31 32 32 33 34 34 34 35 35 39 40 46 50 50

xviii 2.6. 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 3.7. 3.8. 3.9. 3.10. 3.11. 3.12. 3.13. 3.14. 3.15. 3.16. 3.17. 3.18. 4.1. 4.2. 4.3A. 4.3B. 4.3C. 4.4. 4.5. 4.6. 4.7. 4.8.

List of Tables 1907 properties added to 1884 and 1895, East Prussia Cluster means of land-use proﬁles from land tax cadastre Number of districts in each cluster by province Distribution of large properties by province, adjusted data Total tax value and tax value per hectare of large properties, by province, adjusted data Number of properties of 100 or more hectares, their area and tax assessments, by administrative region, c.1882 Share of area and tax value of properties of 100 hectares or more in area and tax value of all land outside of cities, by province, adjusted data Size distribution of properties 100 hectares or larger, all seven provinces Size distribution of properties 100 hectares or larger, Baltic provinces Size distribution of properties 100 hectares or larger, central and south-west provinces Size distribution of properties 100 hectares or larger, east and south-east provinces Size distribution of total ownership area in properties 100 hectares or larger, all seven provinces Effect of including or excluding Prussian State as landowner Size distribution comparison for large landed properties, c.1882 Distribution of large properties by size and owner (numbers) Distribution of large properties by size and owner (total area) Distribution of large properties by size and owner (total tax value) Size of property and the share of the three largest owner categories in total ownership c.1882 Concentration of ownership: total area owned by owner category, c.1882 Ownership of large landed properties by physical and juridical persons Large landed properties owned by non-state owners Distribution of ownership by category and rank of owner, all seven provinces, c.1882 Distribution of ownership by category and rank of owner, all seven provinces, c.1895 Distribution of ownership by category and rank of owner, all seven provinces, c.1907 Landownership of the Hohenzollern family in East Elbia German sovereigns and their ownership of land in Prussia Landownership of foreign and other German royal families Landholdings of princes and dukes, by province Landholdings of count(esse)s, by province

53 56 56 60 61 63 66 67 68 69 70 72 74 74 79 80 80 82 82 87 88 91 92 93 94 95 96 98 103

List of Tables 4.9. Landholdings of baron(esse)s, by province 4.10. Landholdings of untitled nobility, by province 4.11. Properties owned by untitled nobles in 1894 but not in 1909, province of West Prussia 4.12. Properties owned by untitled nobles in 1909 but not in 1894, province of West Prussia 4.13. Landholdings of all nobility and royalty, by province 4.14. Landholdings of non-noble persons, by province 4.15. Properties of 100 hectares or more in two districts of East Prussia 4.16. Numbers of ofﬁcers and non-ofﬁcers among land owners 4.17. Landholding by ofﬁcers and non-ofﬁcers 4.18. Average landholding by ofﬁcers and non-ofﬁcers 4.19. Ofﬁcer-owners by civil status 4.20. Landholdings of the Prussian State, by province 4.21. Landholdings of non-physical legal persons other than the Prussian State 4.22. Total area and total tax assessment by owner class 4.23. Number of owners and average area and tax value, by owner class 4.24. Ownership of forest land by major owner groups 5.1. Conrad’s twenty largest landowners by area c.1882 5.2. Reconciliation of discrepancy in holdings of Count Otto zu StolbergWernigerode, c.1882 5.3. Large landowners with more land than Grand Duchess of Saxony-Weimar but less than Prince Hatzfeld, c.1882 5.4. The ﬁfty largest landowners by total area owned, c.1882, excluding the Prussian State 5.5. The ﬁfty largest landowners by total area owned, c.1895, excluding the Prussian State 5.6. The ﬁfty largest landowners by total area owned, c.1907, excluding the Prussian State 5.7. The ﬁfty largest landowners by tax assessment, c.1882, excluding the Prussian State 5.8. The ﬁfty largest landowners by tax assessment, c.1895, excluding the Prussian State 5.9. The ﬁfty largest landowners by tax assessment, c.1907, excluding the Prussian State 5.10. The top ﬁfty landowners by socio-economic status 5.11. Share of each province in holdings of top ﬁfty landowners 5.12. Top ten owners by total area excluding the Prussian State, by province 5.13. Top ten owners excluding the Prussian State, by tax value, by province

xix 104 106 107 107 109 110 110 112 113 113 114 116 118 120 121 122 125 127 129 130 132 134 137 139 141 144 145 147 150

xx

List of Tables 6.1. Data from Rauer regarding numbers of knight’s estates

156

6.2. Comparison of Rauer’s data with Handbook data

157

6.3. Continuity and change of ownership of sample properties in the 1834 Matrikel also found in the 1882 database

160

6.4. Noble ownership of knight’s estates in East and West Prussia, 1834–1907/9

162

6.5. Comparison of numbers of knight’s estates in Rauer and in the Handbooks

163

6.6. Numbers of knight’s estates under 100 ha in the Handbooks

164

6.7. Changes in numbers and area of knight’s estates of 100 ha or more

165

6.8. Number of knight’s estates of 100 ha or more owned in various years, by owner category

166

6.9. Area of knight’s estates of 100 ha or more owned in various years, by owner category

166

6.10. Mean share of each land type in clusters

170

6.11. Share of each owner group in number of properties in cluster

171

6.12. Size category share for each owner group in number of properties, c.1882

171

7.1. Industrial establishments on large and small properties enumerated in the Handbooks (excluding dairy plants)

178

7.2. Enumerated industrial establishments (excluding dairy plants), by province, c.1882

180

7.3. Industrial establishments on principal properties of all sizes

181

7.4. City/rural ratio of employees per plant, selected industries, 1907: overall average for seven eastern provinces

185

7.5. Enumerated industrial establishments (excluding dairy plants) on agricultural estates of 100 hectares or more, by province, c.1882

186

7.6. Distribution of industrial establishments (excluding dairy plants) by size of estate on which they were found, c.1882

186

7.7. Industrial establishments (excluding dairy plants) per thousand hectares of estates of 100+ ha, by province and by size category of estate, c.1882

187

7.8. Number of establishments on estates of 100 hectares or more as share of total number enumerated in the census

188

7.9. Distribution of industrial establishments c.1882, by owner category

191

7.10. Share of each kind of industry in total number of industrial establishments located on the properties of each owner category c.1882

191

7.11. Owners of properties having twelve or more industrial establishments (excl. dairies) on larger estates in the seven eastern provinces of Prussia, c.1882

194

7.12. Owners of properties having twelve or more industrial establishments (excl. dairies), on larger estates in the seven eastern provinces of Prussia, c.1895

196

List of Tables 7.13. Owners of properties having twelve or more industrial establishments (excl. dairies), on larger estates in the seven eastern provinces of Prussia, c.1907 7.14. Owners of properties having twelve or more industrial establishments, in rank order 7.15. Comparison of simple regression results 7.16. Core regression results with provinces and size categories 7.17. Core regression results for regions and size categories 8.1. Number of districts in each province having land tax net yield average for all land in the following ranges, c.1862 8.2. Average tax assessments (marks/ha) of six land types in 259 districts of the seven eastern provinces, c.1862 8.3A. Number of districts in each province having land tax net yield for ploughland in the following ranges 8.3B. Share of districts in each province having land tax net yield for ploughland in the following ranges 8.3C. Share of each province in number of districts having land tax net yield for ploughland in the following ranges 8.4A. Number of districts in each province having land tax net yield for meadow in the following ranges 8.4B. Share of districts in each province having land tax net yield for meadow in the following ranges 8.4C. Share of each province in number of districts having land tax net yield for meadow in the following ranges 8.5A. Number of districts in each province having land tax net yield for pasture in the following ranges 8.5B. Share of districts in each province having land tax net yield for pasture in the following ranges 8.5C. Share of each province in number of districts having land tax net yield for pasture in the following ranges 8.6A. Number of districts in each province having land tax net yield for forest in the following ranges 8.6B. Share of districts in each province having land tax net yield for forest in the following ranges 8.6C. Share of each province in number of districts having land tax net yield for forest in the following ranges 8.7. Comparison of databases used for cluster analysis with those containing all principal properties of 100 ha or more 8.8. Land-use proﬁles of the six clusters: cluster centres and their standard deviations 8.9. Number of properties in six clusters in the three sample years

xxi

197 198 202 203 205 215 216 216 217 217 218 218 219 219 220 220 221 221 222 223 226 227

xxii

List of Tables

8.10. Total area and total tax value of properties in six clusters in the three sample years 8.11. Distribution of cluster member properties by province 8.12. Properties of 100 hectares or more in East Prussia: numbers and total area in 000 ha 8.13. Number of properties in cluster by owner group 8.14. Total area of properties in cluster by owner group 8.15. Tax value of properties in cluster by owner group 9.1. Regression estimation, tax value as dependent variable, simplest form 9.2. Regression estimation, tax value as dependent variable, simple dummy form 9.3. Regression estimation, tax value as dependent variable, all interactive dummy variables used A.1. Geographic location numbering scheme A.2. Owner codes for properties owned by physical persons A.3. Owner codes for religious bodies and states A.4. Owner codes for communities or business ﬁrms A.5. Values for the variable ‘Unﬁt’

229 230 231 233 233 234 245 246 247 251 252 253 254 255

List of Figures 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 3.7. 3.8. 3.9. 3.10. 3.11. 3.12. 4.1. 4.2. 4.3. 6.1. 6.2. 7.1. 7.2. 7.3. 7.4. 7.5. 7.6. 7.7. 8.1. 8.2. 8.3. 8.4.

Total area of large properties by size category and region (000 ha) Total area of large properties by size category and region (%) Property size vs. total area owned Distribution of total area and tax value of large properties by decile c.1882 Distribution of total area and tax value of large properties by area deciles c.1882 Distribution of total area and tax value of large properties by tax deciles, c.1882 Frequency distribution of logarithm of total area Frequency distribution of logarithm of land tax net yield Lorenz curves of total area in large properties Share of large properties by owner type, by province, c.1882 Area in large properties by owner type, by province, c.1882 Tax value in large properties by owner type, by province, c.1882 Owners by rank and category, excluding the State: area owned Owners by rank and category, excluding the State: tax value owned Ofﬁcer-owners: share of different ranks in numbers, total area, and total tax value Ownership of knight’s estates by number and area Share of each owner in total for cluster Number of plants by industry, seven East Elbian provinces Number of employed personnel by industry, seven East Elbian provinces Average number of workers per plant, seven East Elbian provinces Industrial establishments on properties of 100 ha or more by province or area Industrial establishments on larger landed properties: total number by size category Distribution of properties of 100 ha or more: number by size category Ownership of industrial establishments: number by owner category ‘Land yield index’ by province, early 1860s ‘Land yield index’ by administrative region, early 1860s Composition of clusters by land type, major clusters Composition of clusters by land type, minor clusters

70 71 73 75 75 76 77 78 83 83 84 85 90 90 115 167 172 183 184 184 189 189 190 193 212 213 227 228

Map: Central Europe in 1910, with seven East Elbian provinces outlined in white. Reprinted with permission of Paul Robert Magocsi, Historical Atlas of Central Europe, revised and expanded edition (Seattle: University of Washington Press, 2002).

1 Introduction 1.1. THE SETTING In German domestic and foreign affairs, the century leading up to the Great War was overwhelmingly the century of Prussia. After the end of the Napoleonic Wars and the initially rocky economic times that followed—brought on primarily by the loss of former export markets as a result of the enactment of the Corn Laws in Britain—Prussia’s emergence to pre-eminence in German affairs became ever more evident and overwhelming.¹ Three signal events punctuated this emergence: 1. The formation of the German Customs Union (Zollverein) in 1834, dominated by Prussia. She set the policy for both admission to, and tariffs of, the Zollverein, sometimes even resorting to strong-arm tactics to force recalcitrant principalities—those that failed to be seduced in by the generous revenue-sharing policies Prussia had introduced—to join.² She also blocked the entry of unwanted states into the Zollverein; at the top of this list stood Austria.³ 2. Prussia’s defeat of Austria in the Seven Weeks’ War in 1866. The trumpets of Prussian victory sounded the death knell of the German Confederation and ended Austria’s dreams of hegemony in the German-speaking world. Prussia now could, and did, carry out her ‘little German’ (kleindeutsch) plan to unify the German lands without the participation or membership of Austria, leading directly to 3. The formation of the German Empire in 1871, with the Prussian King as Emperor and the Prussian Minister-President as Chancellor performing their dual duties in Berlin, the dual capital. Prussian nobility dominated the ¹ Hans-Jürgen Puhle thinks that this position extended even beyond Germany, that Prussia had ‘emerged as the uncontested hegemonic power in Central Europe after Bismarck became Prussian Minister-President in 1862’. See his ‘Lords and Peasants in the Kaiserreich’, in Robert G. Moeller, ed., Peasants and Lords in Modern Germany: Recent Studies in Agricultural History (Boston: Allen & Unwin, 1986), 81. ² See W. O. Henderson, The Zollverein (London: Frank Cass, 1984), for details. ³ Ludwig Láng has chronicled the unsuccessful Austrian efforts to join the Zollverein in his Hundert Jahre Zollpolitik (Vienna & Leipzig: K.u.k. Hof-Buchdruckerei und Hof-VerlagsBuchhandlung, 1906), esp. 182–97.

2

Introduction foreign service and the military,⁴ and a former internal Prussian problem, the ‘Polish question’, soon became one of the most burning issues of German politics.⁵

Prussia’s emergence as the ever-more-dominant force in German political affairs was followed, from about mid-century onwards, by the parallel emergence of Germany to a position of industrial might and Great Power status in Europe, coming to rival France and surpassing Austria in both categories. The growth of industry touched off a debate about the direction of German society, the famous ‘industrial state or agrarian state’ debate that largely pitted industrial West vs. agrarian East.⁶ Although the debate continued right up to the outbreak of war, the practical resolution of it came through the famous ‘marriage of iron and rye’⁷ —protective tariffs for both industry and agriculture from 1879 onwards, but especially from the late 1890s. The agrarian camp included diverse groups and positions. There was, for example, a romantic wing that saw despair, degradation, and depravity in the cities and looked back on an idealized (if non-existent) agrarian past. Its followers, in the words of Fritz Stern, ‘sought to destroy the despised present in order to recapture [that] idealized past in an imaginary future’.⁸ This group typically also saw agriculture as the source of sturdy peasant lads to form the backbone of the army. A much more practical and far less romantic group considered the food supply to be a strategic essential in wartime, and agricultural self-sufﬁciency as, therefore, a paramount goal of public policy. The political hard core, but at the same time a serious weakness, of the agrarian camp was seen to be the landowning aristocrats of Prussia’s eastern provinces—the region dubbed ‘East Elbia’ in many accounts.⁹ These noblemen ⁴ An excellent quantitative study of this dominance is Nickolaus von Preradovich, Die Führungsschichten in Österreich und Preussen (1804–1918), Veröffentlichungen des Instituts für Europäische Geschichte Mainz, 11 (Wiesbaden: Franz Steiner Verlag, 1955). ⁵ Boleslaw Grzes, ‘Teoretyczno-propagandowe aspekty dyskryminacji Polakow w Poznanskiem na przelomie XIX i XX wieku’ [Theoretical-propaganda aspects of discrimination against the Poles in turn of the century Poznania], in Lech Trzeciakowski and Stanislaw Kubiak, eds, Rola Wielkopolski w dziejach narodu Polskiego [The role of Greater Poland (Poznania) in the history of the Polish nation] (Pozna´n: Wydawnictwo Naukowe Uniwersytetu im. Adama Mickiewicza w Poznaniu, 1979), 202. ⁶ A good short survey can be found in Herman Lebovics, ‘ ‘‘Agrarians’’ versus ‘‘Industrializers’’: Social Conservative Resistance to Industrialism and Capitalism in Late Nineteenth Century Germany’, International Review of Social History, 12 (1967), 31–65. ⁷ Alexander Gerschenkron, Bread and Democracy in Germany (Berkeley: University of California Press, 1943). It could be noted that this was followed by a similar ‘marriage of textiles and wheat’ in the Austro-Hungarian Empire. See Scott M. Eddie, ‘The Terms and Patterns of Hungarian Foreign Trade, 1882–1913’, Journal of Economic History, 37/2 ( June 1977), 351. ⁸ Fritz Stern, The Politics of Cultural Despair: A Study in the Rise of German Ideology (Berkeley: University of California Press, 1963), p. xvi. ⁹ Whether the two Mecklenburgs should be included in this grouping because of a similar landowning structure, even if they did not belong to Prussia, elicits varying opinions. Most scholars seem to lean toward keeping the group purely Prussian. This study will also adhere to that position, to focus on Prussia and exclude Mecklenburg.

Introduction

3

(and women, but only a tiny number of women actually owned land¹⁰) had enjoyed signiﬁcant privileges for a long period of time; they were prepared to ﬁght to keep those privileges, which they considered their due from simple birth as well as from a long history of military service and sacriﬁce for their monarch. Dubbed the Junkers,¹¹ they saw in themselves both the repository and the wellspring of the nation’s values and character, while others—including the pro-peasant wing of the agrarians—thought their elite status and aristocratic privileges to be not only anachronistic, but a drag on the agrarian cause and economic and social development in general. One of the most prominent of the agrarians, Max Sering, advocated subdividing some of the estates of the large landowners into family farms for peasants as a means of increasing agricultural productivity and not only stopping, but reversing, the ‘ﬂight from the land’.¹² Needless to say, this position was not popular among the large landowners.¹³ Land is of course the fundamental form of wealth in an agrarian society, so whatever the theoretical or practical merit of proposals such as those of Sering, the issue of who did, or who should, own the land ﬁgured in every variant of the agrarian position. In the absence of concrete information, of course, various folklores could grow and ﬂourish: if many favoured peasant agriculture—the quintessential ‘yeoman farmer’ so prominent in the British folklore—there were others who felt that large estates and leading owners not only were essential to the preservation of rural values and rural order, but were also the very group whose larger resources enabled it to lead technological advance in agriculture. The ‘peasant’ group, on the other hand, saw the great landlords as blocking access to land, sometimes even as virtual monopolists in landownership, despite the rather brisk market trafﬁc in large and small landholdings that had characterized most of the nineteenth century in Prussia. ¹⁰ c.1895, only 708 of the 11,033 owners of 100 ha or more could be identiﬁed as women, sometimes listed together with a husband or male relative as owner. These women owned just over 672,000 ha of land, about 6% of all land owned by those possessing 100 ha or more. ¹¹ The term Junker, whose etymology is apparently the phrase Jung Herr —‘young lord’, has come to be used in many different ways, from meaning the group of noble owners of knight’s estates (Rittergüter) to meaning all nobles from the eastern provinces to, at an extreme, all owners of large properties, irrespective of birth. In this study I will rather abjure use of the term, but my own preference is to consider that being a Junker was a combination of noble birth and state of mind regarding land and its position in the social order. This is not a quantiﬁable concept, so—in the sections of this study on the social distribution of landownership—I will stick to more concrete categories such as noble and non-noble persons, the former for some purposes also divided by rank. ¹² Sering’s most inﬂuential book was probably Die innere Kolonisation im östlichen Deutschland, Schriften des Vereins für Socialpolitik, 61 (Leipzig: Duncker & Humblot, 1893). ¹³ Although when the government actually instituted a policy of buying up large estates for subdivision into peasant farms in the provinces of Posen and West Prussia, landowners soon ﬂooded it with offers to sell. See e.g. Scott M. Eddie and Christa Kouschil, The Ethnopolitics of Land Ownership in Prussian Poland, 1886–1918: The Land Purchases of the Ansiedlungskommission, Trondheim Studies on East European Cultures & Societies, 9 ( Trondheim, Norway: NTNU, 2002).

4

Introduction

1 . 2 . W H O OW N E D T H E L A N D ? H OW D O W E K N OW ? In these circumstances, there was a widespread fascination with the question of who owned the land, not just in Germany, but in other countries of Europe: so great was the interest engendered by the publication in 1875 of the Return of Owners of Land,¹⁴ which listed the owner of each parcel of land of 1 acre or larger in England and Wales outside of London, that the copy obtained by one Tory club in London ‘was reduced to rags and tatters within a fortnight’, John Bateman tells us.¹⁵ The reason that various folklores regarding landownership could grow up side by side is that neither Germany nor Prussia ever collected and published any data on landownership. Therefore the methods of the German Historical School, with its emphasis on ﬁnding out information about real questions and institutions—which one might expect would have been applied with vigour to landownership data—could not be brought to bear on this issue without extreme, and mostly insurmountable, difﬁculties. There were certain partial compilations, such as the Matrikeln for each province—lists of knight’s estates (Rittergüter) and their owners who were qualiﬁed to be members of the First Estate (Ritterschaft) under the class-based allocation of seats in the Prussian provincial assemblies (Landtage) that was established in the 1820s¹⁶—but never any ofﬁcial or comprehensive statistics,¹⁷ even though everything necessary for carrying out such an inquiry had, in fact, been in place for decades by the turn of the century.¹⁸ Into this gap stepped some private publishers, who collected data on large properties for a variety of market reasons. The earliest of such attempts at presenting data on large landholdings were published almost simultaneously in Bohemia and Prussia: Paul Alois Klar’s Böhmens grosser Grundbesitz, wie dieser in der königl. Landtafel inneliegt in 1856,¹⁹ and Karl Friedrich Rauer’s Adressbuch der Rittergutsbesitzer und Rittergüter in den Preussischen Staaten in 1857.²⁰ Both drew their data from ofﬁcial government sources: Klar from the Landtafel, which recorded the estates of the nobility, and Rauer from the Matrikeln ¹⁴ Great Britain. Local Government Board, The Return of Owners of Land in England and Wales Exclusive of the Metropolis, Presented to both Houses of Parliament (London: HMSO, 1875). ¹⁵ Bateman published his famous book, which went into several editions, based on the Returns for England, Wales, Scotland, and Ireland. The quote is from The Great Landowners of Great Britain and Ireland (Leicester: Leicester University Press, 1971; repr. of 4th edn, 1883), preface, p. v. ¹⁶ Chapter 6 contains the results of analysis based in part on the 1834 Matrikel for East and West Prussia. ¹⁷ Sigrid Dillwitz avers that this was so because it was not really in the interest of the State to compile such data. See her ‘Quellen zur sozial-ökonomischen Struktur der Bauernschaft im Deutschen Reich nach 1871’, Jahrbuch für Wirtschaftsgeschichte, 2 (1977), 248. ¹⁸ Eduard Müller, Der Grossgrundbesitz in der Provinz Sachsen: Eine agrarstatistische Untersuchung, Sammlung nationalökonomischer und statistischer Abhandlungen des staatswirtschaftlichen Seminars zu Halle a.d.S., 67 ( Jena: Gustav Fischer, 1912), 1. ¹⁹ Prague: self-published, 1856. ²⁰ 2 vols, Berlin: R. Kühn.

Introduction

5

deﬁning membership in the First Estate, mentioned earlier. Both were necessarily incomplete, since there were large properties in both kingdoms that were not eligible to be included in these records. Rauer’s work was more analytical, and the actual address book portion²¹ covered only a subset of noble owners from the data that he had gathered. The ﬁrst Prussian address book of large properties to include economic data about the properties listed was Adolf Frantz’s General-Register in 1863,²² but—at only 117 pages for the entire province of Brandenburg—it was seriously incomplete. There followed a grandiose project to produce a series of address books for Germany and Austria-Hungary, but it failed almost before it got started.²³ In the end, there were two principal series of Prussian directories of landowners that managed to publish volumes on all seven of the eastern provinces. The ﬁrst and most extensive was the series from Paul Parey Verlag in Berlin, under the general editorship of Paul Ellerholz, entitled Handbuch des Grundbesitzes im Deutschen Reiche.²⁴ Begun in 1879,²⁵ it published individual volumes purporting to contain all properties of 100 hectares or more, or having at least 1500 marks of ‘land tax net yield’ (Grundsteuerreinertrag, hereafter GSRE), the basis for assessing the land tax in all of Prussia. A separate volume covered each of the seven provinces, and these were typically revised at ﬁve- to seven-year intervals. Later the series expanded to Mecklenburg and some other German states. The main competition for the Handbook series began later,²⁶ but continued longer, from the publishing house of Niekammer in Stettin (Szczecin),²⁷ Niekammers landwirtschaftliche Adressbücher. Niekammer’s series covered the eastern provinces of Prussia separately in individual volumes revised and carried ²¹ Published separately as Alphabetischer Nachweis (Adressbuch) des in den Preussischen Staaten mit Rittergütern angesessenen Adels (Berlin: self-published, 1857). ²² General-Register der Herrschaften, Ritter- und anderer Güter der preussichen Monarchie. Mit Angaben über Areal, Ertrag, Grundsteuer, Besitzer, Kauf- und Taxepreise, usw. (Berlin: Grellius, 1863). ²³ General-Adressbuch der Grossgrundbesitzer des Deutschen Reiches u. der Österr.-Ung. Monarchie. Vollständig in 40 Bänden. Although the title claims 40 volumes, so far as I have been able to determine, only the ﬁrst three volumes were ever published (Berlin: Kühn, 1878). ²⁴ An earlier series from the publisher F. Bürde in Berlin (Adressbuch des Grundbesitzes in : dem Areal nach von 500 Morgen aufwärts) began in the 1870s, but was apparently soon abandoned. I have seen only the volumes for Posen and Silesia (both published 1872). They are virtually identical in format to the volumes in the Ellerholz series, but the coverage is somewhat less complete, since 500 morgen = c.128 ha. ²⁵ Johannes Conrad and Ilona Buchsteiner (see Sect. 1.3 below) give the beginning date as 1875 and ‘middle of the 1870s’ respectively, but according to the catalogue of the Staatsbibliothek in Berlin, it began in 1879 as a publication of Verlag des Landwirthschaftlich-Statistischen Bureaus Lodemann together with Nicolai Verlag in Berlin. This can be conﬁrmed by looking at the illustration on the cover of Klaus Hess, Junker und b¨urgerliche Grossgrundbesitzer im Kaiserreich. Landwirtschaftlicher Grossbetrieb, Grossgrundbesitz und Familienﬁdeikommiss in Preussen (1867/71–1914) (Stuttgart: Steiner Verlag, 1990). ²⁶ Niekammer published only one directory for one province before 1900. Hess, Grossgrundbesitzer, 80 n. ²⁷ Later moved to Leipzig.

6

Introduction

on into the 1930s, whereas the Parey series appeared to cease in 1929. Both sources present data in tabular form, both purport to contain complete coverage of properties over 100 hectares in size, both were based on survey questionnaires compiled by the publishers and supplemented with government help, and there seems very little ground on which to prefer one over the other.²⁸ Other German publishers produced occasional similar address books, but the only one to appear regularly was Wilhelm Gottlob Korn’s Schlesische Güteradressbücher, which also covered properties assessed at 1500 marks or more of GSRE, irrespective of size, but specialized on the one province. By 1909 it was already its ninth edition.

1 . 3 . A D D R E S S B O O K S O F L A N D E D P RO PE RT I E S A S S O U RC E S F O R S C H O L A R S

1.3.1. The Pioneer: Johannes Conrad Agricultural address books, for whatever country or province they were produced, were usually published for commercial purposes. They were primarily intended as a source of information for potential buyers of agricultural products or land, and for sellers of products or services to agriculture.²⁹ Johannes Conrad, Professor of Political Economy at the University of Halle and editor of the inﬂuential journal Jahrbücher für Nationalökonomie und Statistik, was the ﬁrst to exploit the potential of these address books as a source of property statistics in a series of studies entitled ‘Agrarian-Statistical Investigations’, published in the journal that he edited. Conrad had access to the German government census of agriculture in 1882, but being aware that it did not contain any useful ownership data,³⁰ he turned instead to the Parey series of Handbooks.³¹ That Conrad’s choice was superior to using the census data was recognized by no less a personage than August Meitzen, who used Conrad’s data in his own monumental nine-volume ²⁸ René Schiller thinks that the Niekammer series, at least in its latest pre-war directory for Brandenburg, had more complete coverage of properties in the 100–200 ha range than did the Handbook. See his Vom Rittergut zum Grossgrundbesitz: Ökonomische und soziale Transformationsprozesse der ländlichen Eliten in Brandenburg im 19. Jahrhundert, Elitenwandel in der Moderne, 3 (Berlin: Akademie-Verlag, 2003), 182, 186. Hess, Grossgrundbesitzer, 95, ﬁnds the same phenomenon for the administrative region of Trier in the western part of Prussia. My own comparisons of the 1905 Niekammer directory with the 1910 Handbook for Pomerania lend further credence to this claim. (See discussion of Buchsteiner’s study of Pomerania below for details.) ²⁹ Nowadays, interestingly enough, their principal use seems to be in genealogical research. See e.g. the forums at http://list.genealogy.net/mailman/listinfo/adressbuecher. ³⁰ ‘The 1882 survey has at least oriented us with true precision concerning the agricultural operating units. . . . How many of those are united in one hand, however, the ofﬁcial statistics are unable to say.’ Johannes Conrad, ‘Der Grossgrundbesitz in Ostpreussen’, Jahrbücher für Nationalökonomie und Statistik, 3rd series, 2 (1891), 822. My translation. ³¹ For the Prussian provinces the work was generally in the second edition by the time Conrad employed it.

Introduction

7

work on land and agriculture in Prussia:³² ‘Conrad’s data are on the whole correct and more thorough than those of the ofﬁcial statistics; they are therefore used as the basis here.’³³ Conrad became concerned with the issue of landownership well before he began to examine the Prussian case. His interest was stirred by the data of the British Return of Owners of Land, on which he reported to the readership of the Jahrbücher in two separate articles already in 1876.³⁴ With regard to Prussia, Conrad was concerned with ﬁve major questions: 1. the extent of the latifundia (which he arbitrarily deﬁned as estates over 5000 ha), who owned them, and whether they were eating up smaller properties; 2. whether the nobility were still dominant or not in the ownership of large estates in any given province; 3. the amount of State property and the terms of its leasing; 4. the extent of entail; 5. the effect of the extent and character of large properties in restricting overall or agricultural population and in spurring emigration. Conrad began his series in the late 1880s;³⁵ in it he set out to examine the structure of landownership in the so-called core provinces of Prussia in the early 1880s.³⁶ The ﬁrst article (1888) dealt with the seven provinces as a unit;³⁷ there followed articles on individual provinces: East Prussia (1891), West Prussia (1892), Posen (1893), Pomerania (1895), and Silesia (1898). Although his stated goal was to cover all seven of the core provinces, he only completed these ﬁve, even though he remained as editor of the Jahrbücher for seventeen more years (until his death in 1915).³⁸ Conrad’s studies, while path-breaking, contain inconsistencies that make it very difﬁcult to compare provinces, even using his own data. His unit of analysis ³² August Meitzen, Der Boden und die landwirtschaftlichen Verhältnisse des Preussischen Staates nach dem Gebietsumfange vor 1866 , 9 vols (Berlin: Paul Parey, 1874–1906). ³³ ‘Die Conrad’schen Zahlen sind aber im Ganzen zutreffend und eingehender als die der ofﬁziellen Statistik; sie sind daher hier zu Grunde gelegt.’ August Meitzen and Friedrich Grossmann, Der Boden und die landwirtschaftlichen Verhältnisse des Preussischen Staates nach dem Gebietsumfange vor 1866, vi (Berlin: Paul Parey, 1901), 557 n. ³⁴ ‘Die Besitzverhältnisse an Grund und Boden in Schottland’, ns 26/1 (1876), 50–6; and ‘Die Grundbesitzverhältnisse im brittischen Reiche’, ns 27/5–6 (1876), 377–96. ³⁵ Jahrbücher für Nationalökonomie und Statistik, 1888–1898 passim. The ﬁrst four articles of the series dealt with technical questions of agriculture and agronomy and were published earlier; it is only with the ﬁfth article in 1888 that the series moves into its primary focus on landownership. ³⁶ Kernprovinzen. These were East Prussia, West Prussia, Posen, Silesia, province of Saxony, Brandenburg, and Pomerania. The current work covers exactly the same provinces. ³⁷ ‘Die Latifundien im Preussischen Osten’, Jahrbücher für Nationalökomie und Statistik, ns 16 (1888), 121–70. ³⁸ Conrad was the second editor of the journal, taking over from the founding editor, Bruno Hildebrand, in 1878.

8

Introduction

was not the individual property, but how much land an individual owner held in a given province. For this calculation, he apparently had his statistical assistant count all properties in each Handbook, even though some of the properties were smaller than 100 hectares. The Handbooks included these smaller properties, but not systematically, so that while they accounted for only trivial portions of the total area of properties in the Handbooks for Pomerania and Posen (0.23 and 0.57 per cent, respectively), the ﬁgure rose to 1.82 per cent in West Prussia and 2.20 per cent in Silesia. For Saxony (see below regarding Müller’s work with data supplied by Conrad), the ﬁgure was nearly 12 per cent. Conrad’s basic schema for the ownership data, to which—alas—he did not adhere consistently, was to present the following tables: 1. A summary table (usually called ‘Table A’) showing data on ownership by major classes of owners (nobility, ‘bourgeois’, communities, churches, etc.) for owners with less, and owners with more, than 1000 hectares. 2. A more detailed table (usually ‘Table B’) providing both more information about the owners and properties, and broken down into ﬁner size categories (100–200, 200–300 ha, etc.). Table A was thus simply an ‘executive summary’ of Table B. 3. A table of all owners, by name, who owned more than 5000 hectares in the given province (‘Table C’). Conrad did not always provide the same information, or use the same categories, in his ‘B’ tables, and some of the information he did provide was extremely suspect: for example, for the vast majority of owners listed in the Handbooks, no indication of their residence address was given, yet Conrad presented data for the number of owners who lived on their estates. He also divided the estates according to whether they were owner-operated, leased, or administered. In his tables, the sum of these three categories was exactly equal to the total number of estates, implying that these categories were mutually exclusive. They were not: many properties were listed with both one or more lessees and one or more administrators; moreover, there was never any indication in the Handbook whether these people rented or administered all or only part of the given estate. Since we know from qualitative evidence that cases where an owner ran part of the estate himself, and leased out part, were relatively common, this becomes a serious deﬁciency in the Handbook data. It would be better to ignore, rather than perpetuate, Conrad’s numbers on residence and mode of estate operation. Conrad also omitted the properties of the Prussian state from his purview,³⁹ although he did include the holdings of other German states and foreign states. ³⁹ Conrad did, however, publish a separate article on a subset of State properties in the provinces of East and West Prussia: ‘Die Domänenvorwerke in der Provinz Preussen’, Jahrbücher für Nationalökonomie und Statistik, 3rd series, 6 (1893), 27–59.

Introduction

9

The 100-hectare-and-over properties of the Prussian state accounted for from 4.5 per cent of the total area of the province in Pomerania to a high of 12.6 per cent in West Prussia,⁴⁰ so this was a signiﬁcant omission. Indeed, if we look at the share of State land in all land in properties of 100 hectares or more, it ran from a low of 10 per cent in Silesia to a high of nearly 35 per cent in East Prussia—and overall for the seven provinces nearly 21 per cent—c.1882. In one important sense, the omission can be justiﬁed: the Prussian state was far and away the biggest landowner in every one of the provinces here under review; including it in the data could let a single owner so overwhelm the distribution as to mask the situation of the other landowners in the province. In this book I will seek both to delineate more precisely the extent of Prussian state landownership and to cast light into the shadow it throws over other categories of landowners.

1.3.2. Conrad’s Students and Followers One of Conrad’s students soon followed his lead to publish his own ‘AgrarianStatistical Investigations’: Eduard Müller gave his book, Der Grossgrundbesitz in der Provinz Sachsen,⁴¹ the subtitle ‘Eine agrarstatistische Untersuchung’ to emphasize its Conradian lineage, as did Hanns-Joachim Richter later to his book on Silesia.⁴² In his introduction, Müller informs us that Conrad had collected the required materials but, for lack of time, was unable to bring the series to completion.⁴³ Thus Müller added the sixth of the seven provinces to the series, following Conrad’s methodology very closely, but still not in directly comparable form, but at least he reproduced a Conrad-style Table B for 1885.⁴⁴ Richter, on the other hand, chose the later date of 1891 to begin his study because Conrad had already published an article on Silesia in 1880,⁴⁵ as a part of his series. Richter’s book covered Silesia at three dates, 1891, 1917, and 1937, and used the previously mentioned Schlesische Güteradressbücher of Korn as his ⁴⁰ Because Conrad used the Handbook data ‘raw’, so have I in these calculations, which use the area of state properties before the adjustments to Prussian state property detailed in the next chapter. They therefore represent a lower bound to the State’s share of the total area of the province. ⁴¹ Sammlung nationalökomischer und statistischer Abhandlungen des staatswissenschaftlichen Seminars zu Halle a.d.S., 67 ( Jena: Gustav Fischer, 1912). ⁴² Die Entwicklung des Grossgrundbesitzes in Schlesien seit 1891 (Breslau: Wilhelm Gottlob Korn, 1938). ⁴³ Müller, Grossgrundbesitz in der Provinz Sachsen, 6. ⁴⁴ He also had access, he informs us in his foreword, to the original materials of the Prussian Statistical Ofﬁce. Müller was by and large very careful in his work, but on one point he seems to have made an error: he ﬁnds an increase in the area of large properties between Conrad’s data from the 1885 edition of the Handbook and his data from the 1907 edition. He claims that part of the increase arises as a result of leaving out properties with more than 100 ha of area but less than 1000 marks of land tax net yield in 1885 (ibid. 75). But in fact the 1907 edition contains fewer, not more, such properties: 77 in 1885, only 62 in 1907. ⁴⁵ Conrad, ‘Der Grossgrundbesitz in Schlesien’, Jahrbücher für Nationalökonomie und Statistik, 3rd series, 15 (1898), 705–29.

10

Introduction

source. Focusing on the 1937 territory of the province, he took out from the earlier address books all properties that fell to either Poland or Czechoslovakia after the Great War, so his data are not comparable with Conrad’s. In particular, this limitation of territory left the Prince of Pless with only around 11,000 hectares in Richter’s tables, depending on the year, since the principality of Pless wound up in Poland. In fact, the earliest study to follow in the Conradian tradition was a dissertation which referred to Hungary, published in 1893: Alfred Hirsch, Ungarns Grundbesitzverhältnisse: Agrarstatistische Untersuchungen.⁴⁶ The subtitle even made it into the Hungarian-language literature in Lajos Beck’s A magyar földbirtok megoszlása: Agrár-statisztikai tanulmány [The distribution of Hungarian landed property: an agrarian-statistical investigation].⁴⁷ This kind of work ceased after the publication of Richter’s book, and was not taken up again until recently.

1.3.3. Recent Work in the Conradian Tradition: Hess, Nabert, Buchsteiner, Schiller The four most representative and most-often-cited examples of new work based on Prussian address books are Klaus Hess’s Junker und bürgerliche Grossgrundbesitzer im Kaiserreich,⁴⁸ Thomas Nabert’s Der Grossgrundbesitz in der preussichen Provinz Sachsen, 1913–1933,⁴⁹ Ilona Buchsteiner’s Grossgrundbesitz in Pommern 1871–1914,⁵⁰ and René Schiller’s Vom Rittergut zum Grossgrundbesitz.⁵¹ Of the newer works, Hess’s is the only one to treat Prussia as a whole. He divides his study into three principal parts. The ﬁrst section evaluates and examines the principal sources of data about large landed properties: the agricultural enterprise statistics (agricultural censuses) of 1882, 1895, and 1907; statistics based on the land tax assessments of the early 1860s plus inquiries into the state of private properties in 1878, 1893, and 1902; and the address books of large properties beginning at the end of the 1870s. Hess’s extremely thorough evaluation of the ⁴⁶ Halle a.S.: Ehrhardt Karras. ⁴⁷ Budapest: Pallas, 1918. ⁴⁸ Junker und bürgerliche Grossgrundbesitzer im Kaiserreich. Landwirtschaftlicher Grossbetrieb, Grossgrundbesitz und Familienﬁdeikommiss in Preussen (1867/71–1914), Historische Forschungen, 16 (Stuttgart: Steiner Verlag, 1990). ⁴⁹ Der Grossgrundbesitz in der preussichen Provinz Sachsen, 1913–1933: Soziale Struktur, ökonomische Position und politische Rolle (Cologne-Weimar-Vienna: Böhlau Verlag, 1992). ⁵⁰ Grossgrundbesitz in Pommern 1871–1914: Ökonomische, soziale und politische Transformation der Grossgrundbesitzer (Berlin: Akademie Verlag, 1993). Shorter studies using similar data and techniques are my own ‘Junkers and Magnates: The Social Distribution of Landed Wealth in Prussia and Hungary; A Case Study of Pomerania and Transdanubia, 1893’, Österreichische Osthefte, 1 (1994), 109–31; ‘Grossgrundbesitz im ostelbischen Preussen: Datenbasis und methodologische Probleme’, in Heinz Reif, ed., Ostelbische Agrargesellschaft im Kaiserreich und in der Weimarer Republik (Berlin: Akademie Verlag, 1994), 141–55; or ‘Cluster-analízis a földbirtokszerkezet vizsgálatában’ [Cluster analysis in the examination of the structure of landed property], AETAS 3 (1994) 17–36. ⁵¹ Full citation in n. 28.

Introduction

11

strengths and weaknesses of these sources should be made required reading for anyone who wants to look at questions of landownership in Prussia. Here his most startling conclusion is that If one takes the indicator of land tax assessment, which is a much better measure of the economic strength of an enterprise, for judging the situation of landownership, an entirely different picture of the regional distribution of large landed properties emerges: in West Elbia, in relation to its overall area, there were more large landed properties than in East Elbia; even in regard to their economic results these did not stand behind their East Elbian counterparts.⁵²

Later in the present book I will also examine the distribution of properties by their land tax assessments, especially in Chapters 3 and 4, and consider who were the biggest landowners by tax value in Chapter 5. The second section of Hess’s book concerns entailed properties (Fideikommisse), which by 1914 covered about 7 per cent of the area of Prussia, and whose numbers had doubled, and area increased by half, since 1870, with most of the increase taking place in the eastern provinces.⁵³ Hess examines their legal position, attempts at reform of the laws governing such properties, and the extent and ownership of entails by social group, ﬁnishing with chapters on the social and economic importance of the entailed properties and on government policy toward entails. One of the more interesting contributions Hess makes is to extend the analysis, from unpublished sources, to data on ‘money entails’ (Geldﬁdeikommisse) that did not have any land in their makeup.⁵⁴ With regard to the economic and social signiﬁcance of the entails, Hess ﬁnds—contrary to popular opinion based on a misleading overall average tax value per hectare—that entails tended to encompass both agricultural and forest land of higher than average quality for the districts in which they were located. The lower overall average was simply the result of a composition effect: entails generally had a much higher than average proportion of forest land, forest land was assessed much lower than arable land, and therefore the overall average tax value per hectare came out lower, even though the per hectare tax value was higher for each separate type of land.⁵⁵ He also stands the conventional demographic wisdom on its head: although the rate of emigration from districts with heavy preponderance of entails was lower than from districts of mostly freely alienable private property, when one adds in the rate of natural increase—which was considerably lower in the districts with lots of entailed properties—the net ⁵² Hess, Grossgrundbesitzer, 313. Here Hess puts a quantitative dimension to a general observation of the Prussian State Statistical Ofﬁce from a 1921 publication, in which the ofﬁcial statisticians note that when the land tax net yield is used as the measure, the distribution of large properties in the Eastern and Western provinces hardly differs at all, especially if 1500 marks be taken as the lower limit. Prussia, Preussisches Statistisches Landesamt, ‘Grundbesitzverteilung in Preussen nach den Ergebnissen der ländlichen Verschuldungsstatistik,’ Zeitschrift des Preussischen Statistischen Landesamts, suppl. 42 (Berlin: PSL, 1921), 3. ⁵³ Hess, Grossgrundbesitzer, 142–5. ⁵⁴ Ibid. 157–9. ⁵⁵ Ibid. 168–75.

12

Introduction

result was that areas of mostly allodial property had overall greater population growth.⁵⁶ An interesting social effect of ownership of entails had to do with the marriage experience and number of children of owners and their brothers. Hess ﬁnds that of the brothers, fewer than two-thirds even entered into a marriage, and of the marriages contracted fully a quarter remained childless. Hess ﬁnds this to be evidence conﬁrming Brentano’s claim that establishment of entail tended toward the dying out of a family.⁵⁷ So far as careers outside of agriculture went, the owners and their brothers were occupied almost exclusively as military ofﬁcers or ﬁlled posts in the local, provincial, state and court administrations. Only these were seen to be beﬁtting their status.⁵⁸ The third and ﬁnal section concerns the proﬁtability of landed estates. Hess focuses particularly on the ‘crisis’ of the 1890s, in which falling grain prices supposedly led to a general threat of bankruptcy for agricultural enterprises. Here some simple calculations prove instructive: 1. The prices of 1891, from which the falling trend was measured, were the highest for many decades. If ten-year averages are used, the price fall in the 1890s disappears. 2. Most farms also had animal husbandry, and the prices of animal products did not fall, while production increased signiﬁcantly. 3. The period from the 1880s in particular witnessed an enormous intensiﬁcation of agriculture in Germany, with yields per hectare of some important crops doubling or even tripling, while fallow land decreased with new rotations. With such a large increase in output, the conclusion that lower prices led to less revenue is not easily supportable. 4. Except for wages, the costs of most inputs to agriculture were falling throughout the period, and large estate owners largely got around the wage problem by importing cheap foreign labour.⁵⁹ The rest of the ﬁnal section is a concerted and exhaustive attack on the common thesis, both contemporary and historical, of a lasting crisis in agriculture from the 1880s onward that brought especially the large landowners of the eastern provinces to the brink of ruin. Using data on yields and prices of agricultural products, government surveys of agricultural indebtedness along with annual ﬁgures for the inception and retiring of mortgages, reports from the Prussian statistical bureau on forced auctions of landed properties, statistics on change of ownership of land, and ﬁnally an evaluation of the course of land prices from 1895 onwards, Hess can ﬁnd no support from any of these data for the ⁵⁶ Hess, Grossgrundbesitzer, 176. ⁵⁷ Ibid. 181–2. The reference is to Lujo Brentano, Familienﬁdeikommisse und ihre Wirkungen (Berlin: Simion, 1911), 11 ff. ⁵⁸ Hess, Grossgrundbesitzer, 185. ⁵⁹ Ibid. 218–21.

Introduction

13

‘conventional wisdom’. In fact, everything pointed in the opposite direction: ‘Instead of speaking of a ‘‘structural’’, ‘‘permanent’’, or ‘‘general, as it were, creeping’’ ‘‘crisis’’ of East Elbian large-farm agriculture, it would rather appear appropriate to talk of an altogether continuous upswing that sometimes took on boom-like dimensions, and around 1900 turned into a lasting boom, with all of its negative accoutrements.’⁶⁰ How could such a false picture arise? Hess ascribes it to many historians’ ignorance of agriculture and ‘misinterpretation of important indicators’, a confusion between the relative decline of agriculture as industry grew in Germany with an absolute decline, along with an uncritical acceptance of the ‘lament of the representatives of agricultural interests’.⁶¹ All in all, then, using a most exhaustive statistical underpinning, Hess has stood the conventional wisdom about large-estate agriculture and ownership during the Second Empire completely on its head. Under his scrutiny, folklore has had to yield to data. As we shall see at many points later, the data of this book are completely consistent with Hess’s interpretation, but seldom with the conventional wisdom that he has so thoroughly undermined. Unlike Hess, Nabert, Buchsteiner, and Schiller deal with individual provinces, which seems to presage the current trend in such studies.⁶² Nabert takes off from Müller (and Conrad), but follows the same path as Buchsteiner did before,⁶³ and Schiller after him—not just to determine who owned the land, but to place these owners into their social and economic context. Nabert’s book is, in the end, disappointing from an empirical point of view, even though he calls attention to the importance of empirical work and emphasizes that he intends to produce a ‘recording (Erfassung) of the most exact possible structure of the large landholders from the ownership and property relations’.⁶⁴ His working up of the address book data constitutes one of the primary contributions to knowledge of this work, yet all he does with those data is to determine the number of properties and total area of land owned by four groups of landowners: nobility, bourgeois, state and public-sphere owners, and what he calls ‘private-social’ owners, a tiny residual category. There is no analysis of the size distribution of ownership, such as that found in Conrad and Müller, nor any indication of differentiation among the noble or other owners by amount of property owned, or by noble rank, distributions that both Conrad and Müller emphasized. Müller also provided a table of the forty largest landowners by the tax value of their land, not just simple ⁶⁰ Ibid. 310. Quotation marks in original. My translation. ⁶¹ Ibid. 315. ⁶² Including my own for the province of Burgenland in Austria: Scott M. Eddie, Historisches Verzeichnis der Grundbesitzer des Burgenlandes/Burgenland történelmi gazdacímtára 1893–1930, Burgenländische Forschungen, 79 (Eisenstadt, Austria: Verlag des burgenländischen Landesarchivs, 1999). ⁶³ Although her book was published after Nabert’s, the Habilitationsschrift upon which it was based was completed in 1988, whereas Nabert pursued his dissertation research between 1989 and 1991. ⁶⁴ Nabert, Provinz Sachsen, 1913–1933, 4.

14

Introduction

area,⁶⁵ but Nabert does no follow-up, even though the ‘land tax net yield’ is a much better measure of the value of a property than is its area in hectares, and these tax data are provided in the address books for each property listed. While Nabert has gone to considerable lengths to glean from other sources social and economic information such as marriage relationships, debt, membership in student societies, service in the provincial assembly or as Landrat (head of the district [Kreis] administration), and even military rank, of his large landowners, he has made a list of the biggest landowners for only one year (1929). He has not provided any comparison with 1913 or 1922, nor even with Müller’s or Conrad’s own lists of big owners. And the one table that Nabert does provide, entitled ‘the ten most signiﬁcant landowners in the province of Saxony, 1929’, ranked by total area owned, contains only nine entries.⁶⁶ Buchsteiner’s original work on Pomerania can be regarded as the study that revived application of the neo-Conradian methodology to a single province. Even before eventual publication as a book, scholars were citing from her Habilitationsschrift.⁶⁷ It has been often cited since publication,⁶⁸ and served, to some extent, as a model for Nabert’s work.⁶⁹ It is, for all that, a rather frustrating work to deal with, because the concepts employed have not been clearly deﬁned, and the interpretation of the data is sometimes inconsistent. For example, in the ﬁrst table of the book,⁷⁰ Buchsteiner compares numbers and areas of properties from the 1910 Handbook with those of farms from the 1907 census (she calls both of them ‘large farms’—Grossbetriebe). From the Handbook data she uses size categories according to total area of the property, from the censuses according to total agricultural area of the farm,⁷¹ and presents what she calls ‘total area’ (Gesamtﬂäche) for both. This is misleading—and unnecessary—for Buchsteiner recognizes the difference in the two concepts, but ‘a conversion to total area or total agricultural area, was . . . so complicated, that it had to be dispensed with’.⁷² ⁶⁵ Müller, Grossgrundbesitz in der Provinz Sachsen, 62. ⁶⁶ Nabert, Provinz Sachsen, 1913–1933, appendix 8, 161. ⁶⁷ Indeed, I myself cited Buchsteiner’s Habilitationsschrift (called a ‘dissertation B’ in the ofﬁcial parlance of East German academia of the time) in a paper presented at a conference in Berlin in 1992 (later published as ‘Grossgrundbesitz im ostelbischen Preussen’ in 1994, with the reference changed to cite her by-then published book). ⁶⁸ Schiller chides both Hess and Buchsteiner for their abstract, aggregative approach, so that their works ‘do not allow the people doing business to be recognized’. Schiller, Vom Rittergut zum Grossgrundbesitz, 19. ⁶⁹ Among the works ‘showing the way’ for his work, Nabert cites Buchsteiner’s dissertation B, in his Provinz Sachsen, 1913–1933, 3. ⁷⁰ Ch. 1, table 1, 30. ⁷¹ ‘Agricultural area’ (landwirtschaftliche Nutzﬂäche) was deﬁned in the censuses as ploughland, garden (except for ornamental gardens and parks), meadow, rich or cultivated pasture, orchards, and vineyards. The other two categories of land in the censuses were woodland and ‘other land’, the latter consisting of house and farmyard plots, ornamental gardens, uncultivated pasture, water, roads and paths, moor, and unproductive land. See, for example, Germany. Kaiserliches Statistisches Amt, Statistik des Deutschen Reichs, ns 5 (Berlin: Puttkammer & Mühlbrecht, 1885), 1∗ . ⁷² Buchsteiner, Pommern, 29.

Introduction

15

Table 1.1. Amending Buchsteiner’s comparison of farm census and Handbook ( HB) data, c.1907

Size categories of agricultural land (ha)

100–199 (number) (area: ha) 200–499 (number) (area: ha) 500–999 (number) (area: ha) 1,000+ (number) (area: ha) total number of units ‘total area’ (ha)

Buchsteiner 1 Total area according to HB

2 Agricultural area according to census

Additional calculations 3 4 Agr. area Total area according according to HB to census

5 Total area: HB, w/size categ. of agr. area

298 45,097 950 335,701 837 585,598 414 723,772

735 114,137 1,235 424,356 657 417,636 51 56,721a

367 55,093 1,166 405,177 771 519,134 94 126,361

735 159,121 1,235 574,069 657 592,134 51 89,749

367 100,375 1,166 615,752 771 723,483 94 187,627

2,499

2,678

2,398

2,678

2,398

1,690,168

1,021,850

1,105,764

1,415,073

1,627,238

a

Sic. Correct area from census: 65,721 ha. Sources: Buchsteiner, Pommern, table 1; Statistik des Deutschen Reichs, ns 212; Handbooks.

To repeat: the area data for farms from the census (column 2 in Table 1.1) show only the agricultural area, not the actual total area including all other types of land. With a computer, and taking the readily available total area data from the farm census, I can make the calculation that Buchsteiner did not make.⁷³ The table thus presents her numbers and adds the requisite calculations for total area from the agricultural censuses and agricultural area from the Handbooks. In German statistical sources size categories such as ‘100–200’ mean ‘at least 100 but less than 200’. Throughout this work, to avoid apparently overlapping size categories, I have changed them to ‘100–199’, ‘200–499’, and so forth; and for smaller spans ‘0–1.9’, ‘2–4.9’. These are still to be understood as meaning greater than or equal to the ﬁrst number but less than the ﬁrst number of the next span. The categories and the data in columns 1 and 2 present the entire contents of Buchsteiner’s table 1. In order to compare apples with apples or oranges with oranges, however, we need to compare either the data of columns 2 and 3, which show the agricultural area in the various size categories of agricultural land, or columns 4 and 5, which contain the total area in each size category of agricultural land. No comparable data to column 1 can be gleaned from the agricultural censuses, because none of their size tables are ordered according to ⁷³ These are the data that I present in column 4 of Table 1.1; they are taken from: Germany. Kaiserliches Statistisches Amt, Statistik des Deutschen Reichs, ns 212, pt 2b (Berlin: Puttkammer & Mühlbrecht, 1912), Übersicht 1, app. 2∗ .

16

Introduction

the total amount of land comprising the farm. On the other hand, comparable data to column 2 can be extracted from the Handbook (column 3), or else total area can be extracted from both sources (columns 4 and 5). Had Buchsteiner been able to do either of these calculations, the correct differences between the two sources would not have appeared so ‘astonishingly great’,⁷⁴ and could have been quite easily ascribed to two main causes: 1. Larger numbers and agricultural area in the Handbooks in size categories above 500 hectares of agricultural area, but smaller below that, along with a greater total number of farms than of properties, indicate that it was common to operate a single large property as more than one farm. 2. Larger total area from the Handbook in all size categories of agricultural area above 200 hectares stems from the greater inclusion of forest properties in the Handbook, compared to the farm census.⁷⁵ For her comparisons over time, Buchsteiner uses address books of landed properties from 1879, 1893, 1905, and 1910. The sources for 1893 and 1910 are the very same Handbooks used in this study; the source for 1879 is the predecessor of the Handbook series.⁷⁶ Since there was no new edition of the Handbook for Pomerania between 1893 and 1910, Buchsteiner inserts a new source for 1905, from the series Niekammers Landwirtschaftliche Güteradressbücher [Niekammer’s address books of agricultural properties].⁷⁷ This last causes problems never fully resolved in her book, and about which I will have more to say in Chapter 2. A check of the two sources we have in common (Handbooks for 1893 and 1910) reveals the differences between my calculations and Buchsteiner’s, summarized in Table 1.2. The differences in numbers of properties and total area of those properties in 1893 could easily be put down to differences in deciding ambiguous cases, or at least in part to my strategy of ﬁlling in data from other sources for properties listed without area data in the Handbook. The difference in number of owners could most likely be put down to Buchsteiner’s greater knowledge of the landowning families of Pomerania.⁷⁸ The larger differences between us in 1910 are more unsettling, and I have no explanation for them.⁷⁹ A small part of the area discrepancy could again be the result of my ﬁlling in data from other sources ⁷⁴ Buchsteiner, Pommern, 30. ⁷⁵ See further discussion of this point in detail in section 1.4.1 below on incomparabilities in the agricultural censuses. ⁷⁶ General-Adressbuch der Ritterguts- und Gutsbesitzer im Deutschen Reich, I. Das Königreich Preussen, 2. Lieferung, Die Provinz Pommern (Berlin: Verlag des Landwirthschaftlich-Statistischen Bureaus Lodemann,1879). ⁷⁷ Pommersches Güter-Adressbuch (Stettin: Niekammer, 1905). ⁷⁸ Other than Buchsteiner’s greater local knowledge, our strategies were the same—when in doubt, consider even identical names as different persons. ⁷⁹ Given the extent of my efforts to root out errors and to avoid double-counting of properties (see Ch. 2 for details), I am conﬁdent that the differences could not have arisen from double-counting or errors in recording the data in my dataset.

Introduction

17

Table 1.2. Summary differences between Buchsteiner’s and Eddie’s calculations from the same sets of data for Pomerania Property data

No. of properties Total area (000 ha) No. of owners Average size of ownership (ha)

1893 Buchsteiner

Eddiea

Differenceb

1910 Buch- Eddiea Diffsteiner erenceb

2,547 1,715 1,589 1,079

2,568 1,736 1,610 1,078

0.8 1.2 1.3 −0.1

2,499 1,690 1,542 1,096

2,546 1,735 1,613 1,076

1.9 2.7 4.6 −1.9

a

Data without adjustment of State properties (see Ch. 2), for exact comparability. As % of Buchsteiner’s totals. Sources: Buchsteiner, Pommern, appendix tables 1, 3, and 4: 315; 317–18; Handbooks, 1893 and 1910. b

for properties listed in the Handbook without their total area, but this cannot be a primary explanation. I will merely note the differences here, so that the reader may be aware of them. Schiller’s weighty work (587 pages) represents by far the most comprehensive of the studies focusing on a single province. Taking full advantage of the relatively rich data sources for the province of Brandenburg, Schiller is able to provide the only really long-term picture (1804–1914) of the development of the structure of large landownership, along with data on the productive power and changes in the technology of exploitation of these estates, the social positions and relations of their owners, career chances of the children of these owners, and more. This work is social history of a high order, and sets a standard for breadth of view and multiplicity of sources that any subsequent work will be extremely hard-pressed to emulate. That said, the book does present some challenges to the reader.⁸⁰ Because of the order of argument and, consequently, the arrangement of tables and graphs, it is difﬁcult to make some simple comparisons over time. For example, Schiller tells us that his databank includes 3241 properties that belong to his deﬁnition of Grossgrundbesitz (large property holding).⁸¹ Although he does not explain it, this number must mean all properties that have ever appeared at any time in any of his sources, even if only once, for the largest total number of properties for any one year appears to be 2732 in 1914.⁸² Later on he discusses and analyses the disappearance of properties from the data sources, mainly knight’s estates (Rittergüter) that have lost their status or other large properties that have been subdivided into smaller units. A simple table showing the total number of ⁸⁰ For the current study, the sections on the structure of landownership are most relevant, and it is on these that I will focus here. ⁸¹ Schiller, Vom Rittergut zum Grossgrundbesitz, 20. ⁸² Ibid. 186, table 33.

18

Introduction

properties in the dataset used for each year, along with the sum of total area of those properties, is missing, but would be most useful for orienting the reader. Schiller uses his own unique deﬁnition of large property: it either has 100 or more hectares of territory or, if fewer, it must be a knight’s estate. His databank divides into two parts at 1863. Before that date, his sources are various lists of knight’s estates; afterwards, beginning with 1885, they are address books of landowners that purport to include all properties of 100 hectares or more in size (or 1500 marks or more of taxable net yield), irrespective of whether or not they are knight’s estates. The dividing line is an address book from 1863 that, as I have pointed out elsewhere,⁸³ is seriously incomplete. Schiller tells us that he supplemented his databank for parts of the province with data from the land tax assessment of the 1860s,⁸⁴ so I presume—although Schiller does not say so directly—that this supplementation was to ﬁll in gaps from the address book in question.⁸⁵ With this combination of property types he appears to be aiming at deﬁning a large property as one that accords a certain social status to its owner, but the arbitrary size cut-off—while absolutely traditional in the German literature⁸⁶ —includes in his group a signiﬁcant number of properties belonging to peasant owners. These may have been highly regarded among their own social class, but no one at the time considered them to be a Grossgrundbesitzer (big landholder) in the wider society. This—and the availability of sources of data for the period before 1885—probably determines his very strong focus on the knight’s estates, and his rather peripheral treatment of what he calls the ‘unprivileged’ large properties. But the constant division of his dataset into these two categories sometimes makes it difﬁcult to follow overall developments over major subperiods in his century-plus of coverage. As it turns out, Schiller would have sacriﬁced little, in my opinion, if he had used the simple 100 hectare cut-off that nearly all other studies of landownership have used. In 1885, my database (which comes from exactly the same source as his database for that year) includes only twenty-seven knight’s estates under 100 hectares with a total area of 1788 hectares (only 0.1 per cent of the total area of all large properties as he deﬁnes them). While he does not report the number for 1885, the number ‘around 1900’ of such knight’s estates was twenty-six.⁸⁷ ⁸³ Eddie, Historisches Verzeichnis, 219, 224. ⁸⁴ Schiller, Vom Rittergut zum Grossgrundbesitz, 173. ⁸⁵ Adolf Frantz, General-Register der Herrschaften, Ritter- und anderer Güter der preussichen Monarchie. Mit Angaben über Areal, Ertrag, Grundsteuer, Besitzer, Kauf- und Taxepreise, usw. (Berlin: Grellius, 1863). ⁸⁶ Cf. Theodor Freiherr von der Goltz’s classic Vorlesungen über Agrarwesen und Agrarpolitik ( Jena: Gustav Fischer, 1904), 82, in which he identiﬁes ﬁve classes of landholding: large landholding (100 ha or more), large peasant holding (20–100 ha), medium peasant holding (5–20 ha), small peasant holding (2–5 ha), and parcel holding (under 2 ha). ⁸⁷ Schiller, Vom Rittergut zum Grossgrundbesitz, 210, table 45. I presume that ‘around 1900’ means 1903 in this case.

Introduction

19

He could still have used these properties in his analysis of knight’s estates—for his pre-1885 data are almost exclusively for knight’s estates—while producing numbers for the post-1885 years that would be exactly comparable with those of other scholars. Most of the analysis of the book uses the numbers of large properties as he has deﬁned them, with only relatively short sections concerning the structure or distribution by area or tax value. It is particularly interesting, and rather puzzling, that he does not use the tax value data from the address books of landowners that form his primary database for the years 1885, 1896, 1903, 1910, 1914, and 1921,⁸⁸ but only data from separate tax and debt statistics published by the Prussian government. He thus passes up the opportunity to compare numbers, area, and tax value of properties directly from his database for the post-1885 years. Given his earlier focus on the economic power of large estates—his chapter 2 is entirely devoted to increases in size, yields, animal husbandry, and mechanization, while his chapter 3 deals with credit, capital, prices, and economic crises facing large estates—this would seem to have been a natural use of the data in his databank. The foregoing criticisms notwithstanding, Schiller’s book provides much information, breadth and depth of details, and considerable food for thought. He also conﬁrms some of the ﬁndings of others. Like Hess, Schiller ﬁnds the ‘destitute farmer’ (notleidender Landwirt) of late nineteenth-century agrarian propaganda to be largely a myth with regard to large estate owners, both for Prussia as a whole and for Brandenburg as a province: four-ﬁfths of the noble owners of large estates in Brandenburg, even with incomes probably underestimated in the tax rolls, fell into the category of more than 6000 marks annual earnings, which was a very comfortable income indeed in those days. Schiller concludes that the ‘destitute farmer’—to the extent that this type existed at all—would much more likely have been found among smaller owners, where the average ratio of debt to income from land was considerably higher than for the big landowners.⁸⁹ Before continuing on in the tradition set by Conrad and his followers, the preceding discussion has shown that it is necessary to establish quite precisely what is to be measured, what sources are to be used, and how the questions are to be asked and examined. With regard to sources, it is necessary to explain why the German government censuses of agriculture cannot be used for this work. Conrad and his students knew this,⁹⁰ but—as Hess has pointed out in detail, and we have again seen above—this knowledge has not always carried over into later historical work. ⁸⁸ For all of these years save 1914, he uses the same Ellerholz Handbook series that I employ throughout this book (see Ch. 2). ⁸⁹ Schiller, Vom Rittergut zum Grossgrundbesitz, 235. ⁹⁰ As Richter put it, the data from the address books of large properties are ‘the only useful source for ownership statistics of Silesian large land ownership’ and they also are ‘incomparably more authentic’ than any of the ofﬁcial enterprise statistics. Hanns-Joachim Richter, Schlesien, 11, 13.

20

Introduction

There are three primary characteristics of the farm enterprise data that make the agricultural censuses of 1882, 1895, and 1907 unsuitable as a basis for a study of landownership: 1. they are based on a concept completely inappropriate for revealing any information about landownership, 2. they are not consistent with each other, and 3. such information as they do contain about ‘property relations’ (Besitzverhältnisse) only serves to emphasize the inappropriateness of the fundamental unit of analysis for any kind of study of ownership of land.⁹¹ The next section explains why this is so.

1 . 4 . FA R M S TAT I S T I C S A R E N OT P RO PE RT Y S TAT I S T I C S To repeat for emphasis: this book, like the studies mentioned in the previous section, is a study of landownership. It is not a study of production relations, nor of the size of enterprises in agriculture, forestry, or other primary production. It is therefore incumbent upon the author, in discussing the distribution of landownership, to use concepts that are ownership concepts (e.g. properties or total holdings) and not production-unit concepts (e.g. farms or forestry ﬁrms). Since one property may be operated as several farms or as part of one or more bigger farms, whereas one farm may consist of more than one property or parts of several properties,⁹² there is absolutely no necessary correspondence between the production entity (the farm: Betrieb) and the ownership entity (the property or holding: Gut or Besitz or Besitzung).⁹³ Because this simple, yet fundamental, distinction has too often been ignored or confused in the literature on Prussian landownership,⁹⁴ it is necessary to point out how these two kinds of data differ in fact, and that the data from the address books of properties (the Handbooks) are not merely superior to the data typically used when discussing the distribution ⁹¹ Schiller goes further, even into the question of intent: he points out quite explicitly that, with the data of the agricultural censuses, ‘the Prussian statisticians fulﬁlled their duty, but still largely avoided that their material could be used in the political debate over large landownership’ (Vom Rittergut zum Grossgundbesitz, 185, my translation). ⁹² These examples by no means exhaust the list of possibilities, but are sufﬁcient to make the point. ⁹³ It is instructive to note the ofﬁcial position of the Prussian State Statistical Ofﬁce on this point: ‘The data of the newest agricultural enterprise statistics of 1907 do not permit their use as landownership statistics, because the enterprise statistics take no notice of ownership or situational relations of landholding.’ Preussisches Statistisches Landesamt, ‘Grundbesitzverteilung’, 1. (My translation.) ⁹⁴ As Hess summarized the situation, ‘the literature, if it does not entirely forgo information about the distribution of landownership from the outset, contents itself as a rule with data from the agricultural enterprise statistics’ (landwirtschaftliche Betriebstatistik). Hess, Grossgrundbesitzer, 18–19. My translation.

Introduction

21

of landownership, but that the data typically used are in fact just plain wrong for the purpose. Some of the confusion arises, I think, because many of the writers on agriculture—unlike Johannes Conrad, for example, who came from a family of substantial landowners and farmers in West Prussia—do not have direct experience of agriculture or else come from a family-farmer background, leading them to have an implicit, perhaps even subconscious idea that the farm and the property are one and the same entity. Indeed, Buchsteiner’s use of the term Gutsbetrieb (literally ‘property operation’ or ‘estate enterprise’), along with her use of Besitz and Betrieb as synonyms,⁹⁵ is an example of this mentality. That this identity is not necessarily the case can be illustrated by the case of the farming operation of my grandfather, Nicolai Eddie, and his brother William. Nicolai and William owned and operated their land jointly and in common. When the Eddie brothers both died in 1949, their farm enterprise contained 1440 acres (c.560 ha) in two properties—so one operational unit, two properties. Nicolai was a widower, so my father and his sister each inherited half their father’s share of the land, and William’s widow Grace inherited his share. The farming operations were now split into two separate enterprises, but the land of my father and aunt became part of another, larger farming operation that included property of my uncle, so the very same land now comprised one and a half farms and three properties. At Grace’s death her two children inherited her land, but there was no change in the farming operation, so now there were one and a half Betriebe and four Besitze. Skipping a couple of steps, the end result was that my sister and I jointly owned our father’s share, our two cousins jointly owned his sister’s share, and a single heir owned the shares of Grace’s two daughters. Because the daughters had earlier sold their share of the Eddie brothers’ smaller second property, their remaining land was now a single property, whereas the share of Nicolai’s heirs was held as four properties. But now each of these properties was part of a larger farming enterprise that included land owned by someone else. So the Eddie brothers’ land now consisted of six properties, one no longer owned by a family member. Two of the ﬁve remaining in family hands were part of the same farming operation and the other three were parts of three separate farms—so altogether now six properties and ﬁve part-farms. When my sister and I sold our share of the land, that was incorporated into yet a different farming operation; now the six properties were each only one constituent part of six different farming operations, each made up of parts of several different properties. It should be abundantly clear from this example that there is no necessary correspondence at all between the property and the farm operation.⁹⁶ ⁹⁵ See e.g. Buchsteiner, Pommern, 28 and 30. ⁹⁶ Had I included information about which farmers were both landowners and lessees, the picture would have become even more complex. And this is not even to mention the complexities

22

Introduction

There is another source of confusion, failing to differentiate between the right of possession (the Besitzrecht) and the property (the Besitz) registered in the Grundbuch, or land register. In other words, there are two meanings to the word Besitz: a property in the real estate sense, and in common parlance ‘that which is possessed’. But that which is possessed—whether through ownership, leasehold, copyhold, usufruct, delegation, occupation, theft, or whatever means—may be one property, one or more parts of a single property, or more than one property or parts of more than one property. This confusion leads Nabert, for example, to use the agricultural census data as if they were property statistics. He argues that the property owner, when leasing his land, has turned over the right of possession to the lessee for the term of the lease.⁹⁷ Even though Nabert does realize that the property and the farm operation are sometimes but not always the same, pointing out that ‘in practice land owned and land possessed come together in peasant properties, but also in the East Elbian estate farms’,⁹⁸ he does not seem to realize (a) that when an owner turns over his right of possession of his or her property, it is not necessarily to a single lessee or farm operator; (b) that one tenant may lease property from more than one owner; and (c) that usually only some, but not all, of the rights of possession are turned over, and these are in most cases speciﬁcally enumerated in the contract between the lessor and lessee. Just because the landowner temporarily turns over physical possession of the land to another, this does not mean that any single farm enterprise is also a single property, nor does it become one through the act of signing over some part(s) of the right of possession, and therefore does not justify Nabert’s use of the agricultural enterprise statistics as if they were property statistics.⁹⁹ For a tenant farmer to refer to the land he has leased from the landowner as ‘my farm’ or ‘part of my farm’ might be quite normal, but he would never refer to it as ‘my property’. Finally, there may be some confusion that arises because the address books universally used the term Besitzer (a noun) for the owner—not the lessor or administrator (these were listed separately)—of the units that they described, which were named properties in the land register. This is the term that I have translated, for the sake of both comparability and convenience, as ‘owner’, but of accounting for the various income streams and the arrangements for paying and receiving rent for the land. ⁹⁷ Nabert, Provinz Sachsen, 1913–1933, 26. ⁹⁸ ‘In der Praxis ﬁelen Bodeneigentum und -besitz gerade in bäuerelichen Betrieben, aber auch auf den ostelbischen Gutswirtschaften zusammen’ (ibid. 27). The latter part of this statement is incorrect as a generalization, as the previous discussion of Buchsteiner’s table 1 makes clear. ⁹⁹ He unfortunately compounds the confusion by treating the property statistics of the address books as if they were farm operation statistics: his appendix 5 (158), a table supposedly of land owned by social group, has as its units the number and area of Betriebe, or farm operations (and this even though the data of the address books include explicit data on lessees, and even show when a property is rented in parcels). And how does he know that different properties owned by the same owner in a given district, but showing no lessees, are not operated as a single farm unit?

Introduction

23

which was chosen very carefully, I believe, by the compilers of the address books, because there were some forms of possession that were not exactly ownership in fee simple, as we now understand it. Some properties, even if they had been in the same family for hundreds of years, were technically Lehngüter (ﬁefs), properties given by, and at the pleasure of, the monarch, usually for services rendered or to be rendered. The monarch retained some ultimate right of repossession under speciﬁc circumstances. By the time of coverage of this book, nearly all of the existing Lehngüter had been transformed into allodial properties, so the ﬁefholder had become an owner de jure and not just de facto. Quantitatively much more important was the entailed property (Fideikommiss) to which the person who ‘owned’ it really had only a lifetime right of possession. The actual legal owner of the property was a juridical person, the foundation (Stiftung) that had set up the entail and speciﬁed the terms of inheritance of the use of its lands. All studies of landownership that I have seen have nevertheless treated the holder of a Lehngut, or the current possessor of a Fideikommiss, as the owner, and I shall do the same. Hence my translation of Besitzer as ‘owner’, even though the proper equivalent of owner in German is Eigentümer, whereas Grundbesitzer, in the case of ﬁefs or entails, should perhaps be more properly rendered as ‘landholder’. The address books did not always have a perfectly consistent deﬁnition of what constituted a property or estate, so their treatment was sometimes inconsistent: This was especially true for manors (Herrschaften), as will be discussed in more detail in Chapter 2: manors were sometimes treated as single properties and sometimes each of their constituent parts was listed as a property in its own right. But those manors that were entailed must be treated as single properties, since they had to be passed on intact; no parts of them could be sold off separately.

1.4.1. Incomparabilities in the Data of the Agricultural Censuses As mentioned earlier, neither Prussia nor Germany ever collected comprehensive statistics on landownership in the period under review, despite plans, exhortations, and even preparations to do so. Therefore scholars have often turned to the ready-made tables of the German agricultural censuses taken in 1882, 1895, and 1907.¹⁰⁰ This is on the face of it questionable, since these censuses took pains to point out explicitly that they were not censuses of ownership;¹⁰¹ moreover, they left out of their agricultural enterprise size tables entirely those operating ¹⁰⁰ Germany. Kaiserliches Statistisches Amt, Statistik des Deutschen Reichs (Berlin: Puttkammer & Mühlbrecht), various volumes: 1882: ns 5 (1885); 1895: ns 112 (1898); 1907: ns 212 (1909). ¹⁰¹ For example, ‘One should never lose sight of [the following]: ﬁrst, that areas of operation are treated here, not areas of ownership; second, that only the agricultural areas of operation, not the total area of the farms, come into the presentation.’ Germany: Kaiserliches Statistisches Amt, Landwirthschaftliche Betriebstatistik nach der allgemeinen Berufszählung vom 5. Juni 1882, ‘Statistik des Deutschen Reichs,’ ns 5 (Berlin: Puttkammer & Mühlbrecht, 1885), 9*. My translation.

24

Introduction

units—in particular forestry enterprises—that did not cultivate any agricultural area. Beyond that, the 1882 census left out much more forest area than did the other two, as we shall see presently. The consequences of these deﬁnitional decisions were manifold, but the three most important from the point of view of studies of landownership are that 1. the census tables of size distributions show only the sizes of farms (operating units) and not the sizes of holdings or properties (ownership units);¹⁰² 2. the size categories used in the censuses were based on the amount of agricultural land only,¹⁰³ so that the total amount of land of the enterprise was inevitably larger than the amount included in the census;¹⁰⁴ moreover, 3. if the forestry operations of the estate were separate from the farming operations, as was very often the case, the forest land was left out of the agricultural tables entirely. Thus often none, or only a part, of forest land would be included in the total land area data of the census, depending not upon whether it was part of a property that also included agricultural land, but upon whether it was part of an enterprise that also cultivated some agricultural land.¹⁰⁵ This last point is quantitatively important. The 1895 and 1907 censuses counted all forest land, but—as in the 1882 census—included only those forestry enterprises that also cultivated some agricultural land in the size distributions of agricultural operations, the data used by so many scholars. For the seven provinces of our concern, the amount of forest land so included was 3.40 million out of 5.39 million hectares of forest enumerated in 1895, 3.37 million of 5.35 million in 1907, or 63 per cent in both years. Thus almost 2 million hectares of forest land were left out of those census tables that many have used to show the distribution of ‘landownership’. Since the total land area of the seven eastern ¹⁰² In German, the distinction is between Betrieb (the enterprise or production unit, i.e. the farm), which is the concept employed in the censuses, and Besitz (holding), the unit of property ownership. Therefore the data of the agricultural censuses are a Betriebstatistik, not a Besitzstatistik. Even as a Betriebstatistik, however, the censuses are not perfect. They were in each case a part of an overall employment survey, using the household as the basic reporting unit. In the agricultural part of the survey, the forms to be ﬁlled out were very explicit that ‘for each agricultural household only one agricultural form is to be ﬁlled out. If there are several persons in the household who work an area of land, the entries concerning the area and animal husbandry of all these persons are to be combined together in one agricultural form.’ Statistik des Deutschen Reichs, ns 112, 3*. So, for example, the farms of three brothers farming independently, but living together in the parental household, would appear as only one farm in the census. ¹⁰³ For the ofﬁcial deﬁnition of ‘agricultural land’ (landwirtschaftliche Nutzﬂäche), see n. 71. ¹⁰⁴ More on this point below. ¹⁰⁵ Beginning with the 1895 census, forestry operations which worked no agricultural land were also enumerated, but the data of these operations were included only in the forestry tables. The agricultural tables contained only the forest area of those operations that also had agricultural land. Because of different methods of presentation and of information not included in the tables, these two sets of forest land data from the censuses cannot be combined.

Introduction

25

provinces comprised 22.7 million hectares,¹⁰⁶ the amount of forest land left out of the tables in these two years was a considerable fraction of the total area of the provinces—on the order of 9 per cent. But in 1882 the criteria for inclusion in the census tables on size distribution must have been different, for in that year—in our seven provinces—only 2.25 million hectares of forest land were included in the tables. This was only 42 per cent of the forest land in these provinces.¹⁰⁷ The forest land completely left out of the census in 1882 amounted to nearly 14 per cent of the total land area of the seven East Elbian provinces. The discrepancy means that, for the purposes of determining overall size of farm enterprises, the 1882 census is not consistent with those of 1895 and 1907. Let us examine more closely the discrepancy in forest land inclusion in the 1882 and 1895 censuses for large farms. Table 1.3 compares numbers and areas of woodland in farm enterprises with at least 100 hectares of land, according to the standard census size categories: Table 1.3. Woodland area of large farm enterprises in 1882 and 1895 censuses Agricultural land (ha) 100–199 200–499 500–999 1,000+ total

Number of farms 1882 1895

Ratioa

Woodland area (ha) 1882 1895

Ratioa

6,747 7,637 3,055 481

6,835 7,355 3,014 525

1.013 0.958 0.987 1.091

150,857 529,748 526,381 236,534

439,367 665,010 601,749 281,905

2.912 1.255 1.143 1.192

17,920

17,729

0.989

1,443,520

1,988,031

1.377

a Ratio 1895/1882. Sources: Statistik des Deutschen Reichs, ns 5, 112, 212.

Although the 1895 agricultural census counted 1 per cent fewer large farm enterprises, these large farms contained 437,000 more total hectares, but almost 545,000 more hectares of forest, compared to the 1882 tables.¹⁰⁸ That is 47 ¹⁰⁶ According to the Statistical Yearbook of the German Empire (Statistisches Jahrbuch des Deutschen Reichs), the total area of the seven provinces with Berlin ‘according to the latest surveys’ varied between 22.696 and 22.714 million hectares between 1880 and 1906 (passim). ¹⁰⁷ According to the 1883 enumeration of forest land, there were 5.34 million hectares of forest land in total in the seven East Elbian provinces. Statistisches Jahrbuch des Deutschen Reichs (hereafter ‘Statistical Yearbook’), 1885, 26. ¹⁰⁸ The difference between the two numbers, some 107,000 hectares, represents 0.6% of the total enumerated land of the large farms in 1882. If we factor out the woodland, enumerated agricultural and other areas declined by somewhat more than did the numbers of farms (1.6% vs. 1.1%), implying that the average agricultural enterprise cultivated slightly less land in 1895 than it did in 1882. The apparent increase in average size of a farm enterprise is more than accounted for by the difference in coverage of forest areas in the two censuses.

26

Introduction

per cent of the difference in the total enumerated forest area of agricultural enterprises between the two censuses, for all size categories. While large farms contained 45 per cent of all the land enumerated in 1882, they contained 64 per cent of all the enumerated forest. The comparable ﬁgures for 1895 are 44 per cent and 58 per cent. Because forest land was more strongly concentrated in large enterprises, these discrepancies reveal that not only were the censuses inconsistent with each other, but that the source of the discrepancy lay mostly in the smaller size categories in 1882, where apparently a larger fraction of the forest land was uncounted, compared to later years. Comparisons of size distribution of farm enterprises between 1882 and one of the later census years may therefore somewhat understate any reduction in the inequality of enterprise size, although this is not a mathematical certainty. Later in this chapter I introduce an explicit adjustment to make the 1882 census data comparable with those of the later years. Leaving out some kinds of land means that the size categories of the census—which measure the amount of agricultural land only—are not even size categories for the enterprise as a whole, let alone the property. The non-included lands cause the relative effect to differ as between size categories in a given census, as the 1882 census data clearly show: Table 1.4 reveals that, for farms with 100 or more hectares of agricultural land, the excluded land types become increasingly important as farm size increases. The total is already 25 per cent larger than the agricultural area for the 100–199 hectare category, growing to 44 per cent for farms with more than 1000 hectares of agricultural land, even though these data do not include the adjustment for under-enumeration of forest land in 1882 (see Section 1.4.4 on the adjustment of the 1882 census data below). Were that adjustment included, the discrepancy between agricultural area and total area of farms would be magniﬁed yet further. Table 1.4. Agricultural land area and total area per farm, all seven eastern provinces, 1882 Agricultural Agricultural area land (ha) No. of Total farms (000 ha) 100–199 200–499 500–999 1,000+ total over 100 total all sizes

6747 7637 3055 481 17920

Ha per farm

945.6 140.2 2486.2 325.6 2024.5 662.7 667.1 1,387.0 6123.4

341.7

1,572,402 14,589.8

9.3

Other areas (000 ha) Wood- Other Total land land land

Avg. total (ha)

Total/ agr. land ratio

150.9 529.7 526.4 236.5

84.8 234.8 178.6 58.5

1181.2 175.1 3,250.8 425.7 2,729.5 893.4 962.1 2,000.3

1443.5

556.7

8,123.6

453.3 1.33

2,249.2 1,271.2 18,110.1

11.5 1.24

Source: Calculated from data in 1882 agricultural census: Statistik des Deutschen Reichs, ns 5, 8–23.

1.25 1.31 1.35 1.44

Introduction

27

In sum, therefore, although scholars have often used the census data to show how unevenly the ownership of landed property was distributed, all they have really shown is how unevenly the sizes of operating farms were distributed—and that on the basis of data that are both incomplete and incomparable between years. The data are, moreover, even incomparable among the size categories in any given census itself. Thus we must come to the same conclusion as Klaus Hess, who very pointedly and very correctly noted that ‘because of the method of their collection, the ‘‘agricultural enterprise statistics’’¹⁰⁹ in no way convey the character of ownership statistics, particularly for large landholdings, contrary to the common opinion.’¹¹⁰

1.4.2. Owning, Renting, and Ownership Statistics The incomparability with ownership statistics arises not simply because (a) the censuses did not count some forest land in their tables of agricultural enterprises, and (b) they were inconsistent with each other in the land they left out; the problem has deeper roots. Although many writers of ﬁne literature have extolled the yeoman farmer or sturdy, independent peasant, in fact no farmer is limited to cultivating only, or even all of, his own land. He can rent in or rent out land, and hire in or hire out labour, depending on whether he or his family can earn more that way than by simply being a ‘family farmer’ in the strict sense, that is, cultivating only his own land with only the labour of his own family. The agricultural censuses provided tables showing the ‘property conditions’ (Besitzverhältnisse) of the agricultural enterprises listed. The 1895 census, for example, showed whether the enterprise operated with only owned land, only leased land, or some combination of the two, along with whether the enterprise operated wholly or partly with share-crop land, ‘allowance land’ (Deputatland¹¹¹), ‘service land’ (Dienstland ¹¹²), or community-owned land. Sharecropping and the other forms following it were very small in comparison to owned or leased land, so it is here sufﬁcient to concentrate only on owned and leased land to make the point once again that enterprise = property, which the illustrative data in Table 1.5, taken from the 1895 agricultural census, make clear. ¹⁰⁹ Landwirtschaftliche Betriebstatistik, i.e. the data from the agricultural census. ¹¹⁰ Hess, Grossgrundbesitzer, 97; emphasis added, my translation. Häbich, in his standard treatise on latifundia and more than sixty years before Hess, understood the difference equally clearly. Theodor Häbich, Deutsche Latifundien: Ein Beitrag zur Berichtigung unserer Vorstellung von der bestehenden Verteilung des ländlichen Grundeigentums (Königsberg: Gräfe & Unzer, 1929), 10. I do not understand how other scholars can ignore the explicit warning of the censuses themselves, and the clear statements of the conceptual differences in such standard works as those of Conrad, Häbich, and Hess, and still go ahead and use the census data as if they were data on ownership. ¹¹¹ Land use given to the holder of a particular ofﬁce or position as part of his/her compensation. ¹¹² Payment for use of this land made in some sort of labour service, rather than in money.

28

Introduction

Table 1.5. Farm enterprises with owned and leased land, all seven eastern provinces, 1895 Agricultural land (ha)

Number of farms

5–9 10–19 20–49 50–99 100–199 200–499 500–999 1,000+

156,304 119,783 95,046 20,901 6,655 6,980 2,890 507

Number of farms made up of Only Only Partly owned land rented land rented land 91,538 83,538 76,413 16,546 4,720 4,676 1,854 315

4,351 2,438 1,187 556 671 1,291 517 42

Only owned/ total (%)

Partly rented/ total (%)

62,760 32,074 15,463 3,270 1,186 990 502 151

58.6 69.7 80.4 79.2 70.9 67.0 64.2 62.1

40.2 26.8 16.3 15.6 17.8 14.2 17.4 29.8

total over 100

17,032

11,565

2,521

2,829

67.9

16.6

total all sizes

409,066

279,600

11,053

116,396

68.4

28.5

Source: Calculated from data appearing in 1895 Agricultural Census: Statistik des Deutschen Reichs, ns 112, table 2, 10–25.

For purposes of this comparison, I took 5 hectares as the lower limit of size of a ‘farm’, given that most farms under that size did not constitute an independent enterprise, that is, one that served as the main income source for its operator.¹¹³ We can readily see that the importance of supplementing one’s own land with some rented land was most important at the two extremes of the size distribution, and less important in the middle. On average, one out of six farms with 100 hectares or more of agricultural land operated with some leased land. But this is only part of one side of the story: the data of the table show us only the number of farms made up of some owned land and some leased land; what they do not show us is how many of those farms that were operated solely with either owned or leased land were themselves part of some larger property, or whether (a) farms made up of exclusively owned land consisted of more than one property or parts thereof, or (b) the farms using exclusively leased land were renting land from more than one property owner. That such land-use patterns were widespread in Prussia in the period under review is clear from the illustrative data for the province of Pomerania in Table 1.6, taken from the 1907 agricultural census.¹¹⁴ I have included in this ¹¹³ Even in the category of farms with 5–10 ha of agricultural land in 1907, for example, it was typical for about a quarter of these farms to be only a ‘supplementary operation’ (Nebenbetrieb) for their operators. See 1907 census, Statistik des Deutschen Reichs, ns 212/1, 9–51. ¹¹⁴ A single province was chosen for reasons of convenience and ease; any province would have done for this purpose. I chose Pomerania in part because of the confusion of the concepts of farm and property in Buchsteiner, Pommern. Buchsteiner seems to have been unaware of the kind of data contained in Table 1.4—even though reported directly in the census—and their implications. For more discussion on this point, see my ‘What Size and Kind of Agricultural Units? Reﬂections on the Study of Prussian Land Ownership before the Great War’, Begegnungen (forthcoming), esp. MS 10–12.

Introduction

29

Table 1.6. Census data concerning ownership relations on farms in Pomerania, 1907 Farms in Pomerania, 1907

Number using exclusively (%) Own Rented land land

Share of (%) Own land

Rented land

Other land

Agricultural land (ha)

Number of farms

Total land area (ha)

Under 2 2–4 5–19 20–49 50–99 100–199 200–499 500–999 1,000+

104,946 19,824 38,203 10,463 1,765 735 1,235 657 51

67,482 82,305 487,672 421,825 183,995 159,121 574,069 592,134 89,749

18.5 44.3 60.4 77.3 73.6 66.0 63.6 67.7 54.9

23.7 9.4 3.8 3.4 8.3 19.5 24.7 14.2 7.8

41.3 67.8 85.1 93.3 90.0 84.3 77.0 86.7 88.2

21.4 25.8 13.4 6.3 9.8 15.7 23.0 13.3 11.8

37.3 6.4 1.5 0.5 0.2 0.0 0.0 0.0 0.0

total

177,879

2,658,352

35.1

16.5

83.7

14.8

1.5

Source: Calculated from data appearing in 1907 agricultural census: Statistik des Deutschen Reichs, ns 212, part 1, 96–9.

table all size categories, in order to show more clearly the dependence of operators of the very smallest properties on land owned by others. That only 51.6 per cent of all farms in the table were made up exclusively of either owned or rented land implies that nearly half (the other 48.4 per cent) of all farms used both their own land and land owned by others. This ﬁgure is skewed by the very smallest farms (under 2 ha) where 57.8 per cent of farms operated this way (since only 18.5 per cent used exclusively owned land and only 23.7 per cent used exclusively rented land). In contrast to this, in every size category between 20 and 999 hectares, less than 20 per cent of farms were operated with such mixed tenures, with that ﬁgure rising again to over 37 per cent of farms larger than 1000 hectares, roughly the same share as for the 5 to 20 hectare size category. Thus, in Pomerania the farms most likely to operate with both owned land and rented land together were those at the bottom and the top of the size scale, just as we saw for the seven provinces as a whole. In other words, as our intuition surely would have suggested, it was the very smallest and the very largest agricultural enterprises that were least likely to be operated as the archetypal ‘family farm’. On the other hand, over a very wide range of sizes (5 to 999 hectares of agricultural land), typically some three-ﬁfths to three-quarters of all farms were owner-operated and did not rent in any land at all. Whether they rented out any land remains unknown; the larger the property, the more likely renting out would be. More interesting, perhaps, is the considerably wider variance in the number of farms operated only with rented land, which reached nearly a quarter of all farms cultivating between 200 and 500 hectares of agricultural land. Just as the census data did not reveal whether owner-operators rented out land, neither

30

Introduction

did they reveal whether these substantial tenant farmers rented from one or more than one landlord.¹¹⁵ In short, the data from which this table was derived reveal nothing whatever about the size distribution of landownership in the province of Pomerania. Other tables of size distribution in the censuses, which do not even so much as allude to ownership, of course can provide no more information.

1.4.3. Can Census Data Serve as a Proxy for Property Data? How much difference do all these considerations make in practice? If we cannot use the census data directly, perhaps we could still use them as a proxy to estimate trends or relative changes in landownership. Just as economists often use wholesale price indices as a proxy for retail price indices by assuming a ﬁxed vector of mark-ups which does not change over time—without having to specify how big those mark-ups might be—so perhaps could we use the data on farm sizes as a proxy for data on the sizes of properties by assuming some sort of relationship between the two.¹¹⁶ But until we can specify what that relationship is, we cannot use the farm size data directly as property size data in any given year; we could, however, use them to indicate changes over time if the relationship, even if not precisely known, were stable. Using actual ownership data gleaned from the basic underlying sources for this study,¹¹⁷ in what follows I intend to address both the differences at a given time and the trend over time, to see how good a proxy the farm data might be.

1.4.4. Adjustment of the 1882 Census Data Before making the comparisons, however, it is necessary to adjust the data of the 1882 census for the under-inclusion of forest land, compared to the other two censuses. If we did not do that, it would exaggerate the difference between the census and property data in 1882, compared to other years, and thus bias the comparison. The adjustment is a simple one: I have simply taken the ratio of woodland to total land from each of the size categories in 1895 and have changed woodland area in 1882 to produce an adjusted total that would result in these same ratios in 1882. Algebraically expressed, the adjustment was as follows: Let W1895 and W1882 = hectares of woodland, and T1895 and T1882 = hectares of total area in the 1895 and 1882 censuses, respectively. ¹¹⁵ For an interesting example from a Hungarian directory of farmers from 1895, in which every single one of the 42 farm operators listed on a particular page cultivated land from two or more owners, see Eddie, ‘What Size and Kind of Agricultural Units?’, esp. MS 12. ¹¹⁶ It should be pointed out, however, that I have yet to ﬁnd that assumption explicit in any of the studies that use the enterprise data as property data. ¹¹⁷ See Preview to Ch. 2 for a discussion of these sources.

Introduction

31

Table 1.7. Adjusted areas for large agricultural enterprises in seven eastern provinces, 1882 census (all absolute ﬁgures in 000 hectares) Size categories: Woodagricultural land in land (ha) 1895

Total area in 1895

Ratio Census wood to woodtotal land in 1882

Total census area in 1882

Adj. woodland in 1882

100–199 200–499 500–999 1,000+

439 665 602 282

1,467 3,234 2,777 1,083

0.461 0.280 0.298 0.378

151 530 526 237

1,181 3,251 2,729 962

440 704 609 255

1,471 3,425 2,812 981

1,988

8,561

0.326

1,444

8,124

2,020

8,701

2,010

8,690

total Sum of details

Adj. total area in 1882

Source: See Table 1.4. 1895 The ratio of woodland to total in 1895 is therefore R = W T1895 To make the ratio of woodland to total area in 1882 the same as in 1895,

adjust W1882 (and therefore T1882 ) by an amount X so that adjusted i.e.W1882

= W1882 + X and

adjusted T1882

= T1882 + X

W1882 +X T1882 +X

=R

The results are shown in Table 1.7.¹¹⁸ The last two rows of the table show that the sum of the details is nearly identical to the total adjustment calculated simply from the average ﬁgures; this in effect provides a check on the internal consistency of the estimation procedure. We are now ready to look at how the agricultural census data for farm enterprises compare to the Handbook data on agricultural properties. Consider the following magnitudes from the two sources: 1. 2. 3. 4.

number of units reported; agricultural area of reporting units; total area of units, including forest, moor, water, unproductive land, etc.; as a derivative of (3), the average size of reported units.

Tables 1.8 and 1.9 report these respective magnitudes. Taking the number of properties in each size category and comparing them to the number of farms in the same size category in the same year, we ﬁnd no uniformity whatever, except ¹¹⁸ As a check, I did a simpler adjustment by taking the ratio of woodland to agricultural area in 1895, then applying that ratio to the agricultural area of 1882 to get an adjusted woodland area in 1882. This resulted in almost exactly the same details and totals. Since the total by this alternative method was only about a quarter of a per cent less than by the other method, I present only the one calculation.

32

Introduction Table 1.8. Agricultural census data and Handbook data on number of farms and properties, seven eastern provinces Agricultural land (ha)

100–199 200–499 500–999 1,000+ total 100+

Numbers of farms by size of agricultural area 1882 1895 1907

Numbers of properties by size of agricultural area c.1882 c.1895 c.1907

6,747 7,637 3,055 481

6,835 7,355 3,014 525

6,607 7,242 2,593 335

4,116 7,170 3,188 576

4,179 7,240 3,129 570

3,993 6,769 2,877 581

17,920

17,729

16,777

15,050

15,118

14,220

Sources: Farm data: Statistik des Deutschen Reichs (1882: ns 5, 1895: ns 112, 1907: ns 212); property data: Handbooks.

Table 1.9. Ratio of number of properties to number of farms by size, seven eastern provinces (%) Agricultural land (ha)

c.1882

c.1895

c.1907

100–199 200–499 500–999 1,000+

61.0 93.9 104.4 119.8

61.1 98.4 103.8 108.6

60.4 93.5 111.0 173.4

84.0

85.3

84.8

total 100+ Sources: See Table 1.8.

that the ratio increases as the size of the agricultural area of the unit increases. This increase, however, is not stable across years. To some extent, the differences between 1895 and 1907, for example, may be the result of differences in the deﬁnition of what constitutes a property in each year. In particular, were we able to break the manors (Herrschaften) down into their constituent parts, the numbers of properties in the smaller size categories would grow, and the number in the largest size category shrink. This would affect all of the tables that follow, as well. But, as the ownership relations data from the preceding section have shown, this could not bring about a close correspondence between enterprises and properties, even if the change by subdividing manors were to be large. But in fact the number was large in only one province, Silesia. In 1909, manors there numbered 150, with 207 thousand hectares of agricultural land among their total of 592 thousand hectares. In the other provinces there were far fewer manors, typically no more than 15. And, since many of these were entailed estates (Fideikommisse), which had to be passed on intact and could not

Introduction

33

Table 1.10. Agricultural census data and Handbook data on agricultural land area of farms and properties, seven eastern provinces (000 ha) Agricultural land (ha)

Total agricultural area of farms 1882 1895 1907

Total agricultural area of properties c.1882 c.1895 c.1907

100–199 200–499 500–999 1,000+

946 2,486 2,024 667

953 2,375 2,017 747

939 2,337 1,708 457

607 2,337 2,142 930

617 2,363 2,103 916

593 2,214 1,928 918

total 100+

6,123

6,092

5,441

6,016

5,999

5,653

Sources: See Table 1.8.

be alienated in part, they must necessarily be regarded as single properties. Even given all this, however, it is still the case that the vast, overwhelming majority of properties listed in the Handbooks are unambiguously single properties. Thus any argument based on aggregation into manors or other larger units could be only a peripheral one at best, which means that this potential criticism of the property data in the tables cannot vitiate the arguments made on the basis of those tables. Let us now turn to the second magnitude, the agricultural area of the units reported from the two sources. Immediately apparent from Table 1.10 is the remarkably close correspondence in total agricultural area from the farm statistics and from the property Handbooks, in the ﬁrst two years at least. This correspondence argues strongly for the notion that the difference in farm numbers and property numbers is signiﬁcant, and considerably reinforces the contention that the farm statistics cannot be used as property or ownership statistics, especially for units with 100 hectares or more of agricultural land. The fall-off in agricultural area in the two largest size categories between the 1895 and 1907 censuses was quite dramatic, but was not repeated in the property statistics: The total agricultural area in the seven eastern provinces for all sizes of farms rose, however, from 14,590 thousand hectares in 1882 to 14,609 thousand in 1907. Therefore the drop-off from the larger size categories between 1895 and 1907 represents a shift of agricultural area to smaller units, but not an actual decline in the area of cultivated land. When we look at the ratios by size, as before, Table 1.11 shows us again a very mixed picture. Compared to the Handbook data for total agricultural area, we have already seen that the totals are quite close (reassuring, also, for the completeness of coverage of the Handbooks), but that only in the two middle size categories do the areas of agricultural land come close to each other in the two sources. Even this small correspondence disappears when we look at the total area of all land in the enterprise or property, as Table 1.12 shows. The ratios between the total area measures follow a similar

34

Introduction Table 1.11. Ratio of agricultural area of properties to agricultural area of farms by size, seven eastern provinces (%) Agricultural land (ha)

c.1882

c.1895

c.1907

100–199 200–499 500–999 1,000+

64.2 94.0 105.8 139.4

64.7 99.5 104.3 122.6

63.2 94.7 112.9 200.9

98.3

98.5

103.9

total 100+ Sources: See Table 1.8.

Table 1.12. Agricultural census data and Handbook data on total area of farms and properties, seven eastern provinces (000 ha) Agricultural land (ha)

Total area of farms 1882a 1895

1907

Total area of properties c.1882 c.1895 c.1907

100–199 200–499 500–999 1,000+

1,471 3,425 2,812 981

1,467 3,234 2,777 1,083

1,404 3,231 2,358 635

1,318 3,670 3,182 1,778

1,379 3,762 3,157 1,729

1,298 3,573 3,033 1,781

total 100+

8,701

8,561

7,628

9,948

10,027

9,685

a Adjusted. Sources: See Table 1.8.

Table 1.13. Ratio of total area of properties to total area of farms by size, seven eastern provinces (%) Agricultural land (ha)

c.1882

c.1895

c.1907

100–199 200–499 500–999 1,000+ total 100+

89.6 107.2 113.2 181.2 114.3

94.0 116.3 113.7 159.6 117.1

92.5 110.6 128.6 280.5 127.0

Sources: See Table 1.8.

pattern of rising by size category, but are unstable over time, just as we saw for the previous two measures (see Table 1.13). Unsurprisingly, then, the average size of unit—when all types of land are taken into consideration—among the properties of the Handbook is considerably

Introduction

35

Table 1.14. Agricultural census data and Handbook data on average total area of farms and properties, seven eastern provinces (ha) Agricultural land

Average total area of farms incl. non-agricultural area 1882 1895 1907

Average total area of properties, incl. non-agricultural area c.1882 c.1895 c.1907

100–199 200–499 500–999 1,000+

218 449 921 2,040

215 440 921 2,062

213 446 909 1,895

320 512 998 3,087

330 520 1,009 3,033

325 528 1,054 3,065

486

483

455

661

663

681

total 100+

Sources: See Table 1.8.

Table 1.15. Ratio of average total area of properties to average total area of farms by size, seven eastern provinces (%) Agricultural land (ha)

c.1882

c.1895

c.1907

100–199 200–499 500–999 1,000+

146.8 114.0 108.4 151.3

153.5 118.2 109.6 147.1

152.6 118.4 116.0 161.7

total 100+

136.0

137.3

149.7

Sources: See Table 1.8.

in excess of the average size of unit in the agricultural enterprise statistics, but with no ﬁxed or even approximately constant ratio between them, as the panels of Tables 1.14 and 1.15 show. In the category of units with between 100 and 200 hectares of agricultural land, the difference in average overall size is around 50 per cent, and it even exceeds 50 per cent in the category of units with 1000 or more hectares of agricultural land in two of the three years. Only in the 200–999 hectare categories are the average total sizes even approximately the same, with properties showing roughly 10 to 20 per cent more total area than farming units with the same amount of agricultural area. This comparison of census (farm) to Handbook (property) data shows—especially for the interval from 1895 to 1907—the general tendency of smaller farms to grow at the expense of larger ones.¹¹⁹ For larger farms or properties (those with at least 100 hectares of agricultural land) the ratio of number of farms to ¹¹⁹ See Hess, Grossgrundbesitzer, 29–36, for an excellent summary of the details of this tendency.

36

Introduction

number of properties changes less than the ratio of farms’ total area to the total area of properties. The data of the table do not support the assumption of a stable relationship of the distribution of farms to the distribution of properties over time—indeed, they show that the census data became a progressively worse representation of property data as time passed. So the size distributions of properties cannot be used as a proxy for size distributions of properties or landownership, not even in the limited sense of tracking changes over time. Once again, we see no stable relationship between size distributions of large farm enterprises and size distribution of large rural properties, so once again the use of census data as a proxy for ownership data fails. Given the unsuitability of the census data, both because they are inconsistent with each other and because they cannot even serve as an approximate proxy for the distribution of landownership by property unit, they are obviously even less suitable for drawing conclusions about the concentration of ownership of land. Those scholars who use the census data on farm enterprises in an attempt to indicate the distribution of landownership by size, or concentration of that ownership, are simply wasting their time and ours. In order to be able to say anything about the distribution of landownership, one must use ownership statistics, not enterprise statistics. That is the task of the following chapters.

2 Data Considerations 2 . 1 . P R EV I EW Data are like potatoes: you have to clean them before you can cook them. This chapter describes the principal data sources and the extensive effort made to ‘clean’ them, that is, to ensure—insofar as possible—that the data would be accurate, consistent, and complete. After describing in some detail the principal sources used, and what types of data were included in them, this chapter explains the choice of ‘snapshots’ of the data focusing on the years of the agricultural censuses of Germany (1882, 1895, and 1907) and the choice of a lower bound of 100 hectares for inclusion of a property into the analysis. I show that there is very little cost, and considerable gain in consistency, attendant upon the 100-hectare lower limit. There is also extensive discussion of several important data problems and what could be done about them: 1. Inconsistency in recording names of owners, even within a single source volume. Even after extensive checking, the number of individual owners is still no doubt overstated, especially for the bourgeoisie. 2. Changes in administrative boundaries within provinces, especially for the province of West Prussia in 1887. 3. Missing data, especially details for different types of land; details that did not add up to the reported total area; or the reporting of two or more land types (e.g. ploughland and meadow or water and moor) as a combined single ﬁgure. 4. Incomplete coverage of the reported data, speciﬁcally in the principality of Pless and in the ownership of forest land by the Prussian state.

2 . 2 . P R I N C I PA L S O U RC E S A N D U S E S O F OW N E R S H I P D ATA The principal sources used for the data on ownership of large properties are the multi-volume Handbuch des Grundbesitzes im Deutschen Reiche [Handbook

38

Data Considerations

of landholding in the German Empire, hereafter referred to as Handbook],¹ which purported to list all properties of 100 hectares or more (or of 1500 or more marks of taxable assessment, regardless of acreage).² Prepared with the active help of the Ministry of Agriculture, these Handbooks listed each large property according to the district (Kreis) in which it was located, and provided the following information about the property: 1. Its size in hectares, including the number of hectares of ploughland, meadow [hayﬁeld], pasture, woodland, water, and ‘moor’ (Oedland). ‘Moor’ was a catch-all category that included not only moorland per se, but also all unproductive land as well as untaxed land such as paths, roads, and farmyards. 2. The ‘land tax net yield’ (Grundsteuerreinertrag, the assessment basis for the land tax³) for the whole property, but not for any of its individual land types. 3. The name, and sometimes address, of the owner(s). 4. The existence of any non-agricultural or agricultural processing establishment on the property, such as a ﬂour mill, sawmill, brickworks, distillery or brewery, starch factory, inn (Gasthaus or Krug), electrical generating plant, etc. 5. Location information: nearest post ofﬁce, nearest railway station, telephone number if there was one, etc. 6. Miscellaneous notes, such as whether the estate had a herd of purebred cattle, or whether it was undergoing subdivision into smaller parcels. The publication of these Handbooks was intended to provide directories for the use of both suppliers of goods and services to agriculture and buyers of agricultural products or land. The Handbooks were to appear in several series, the ﬁrst and earliest being for the Kingdom of Prussia. Within this series, individual volumes covered single provinces, and these volumes were normally to be revised every four to seven years, although longer gaps did exist. The series began in 1879 with a volume for Brandenburg, and other Prussian provinces followed soon after. With the sole ¹ Paul Ellerholz, gen. ed., most editions published by Paul Parey Verlag in Berlin. ² Schiller (Vom Rittergut zum Grossgrundbesitz, 54) says that they also include all knight’s estates, regardless of size or tax assessment. Both Schiller and Hess believe these Handbooks to be short on the listing of properties in the 100–200 ha range (cf. Ch. 1). ³ The ‘land tax net yield’ (Grundsteuerreinertrag) was an estimate of how much net income a given piece of land would produce under ordinary management at local farm gate prices. The land tax, which was a ﬁxed percentage of this ‘net yield’ in all provinces of Prussia, was thus a tax on the potential, not the actual, income of the land. The assessment procedure was to evaluate the quality of a given piece of land, then refer to tables which gave the ofﬁcial net yield for that quality grade of land in that district or subdistrict. These tables divided most land types into eight quality grades, so the procedure was quite precise, at least in principle.

Data Considerations

39

exception of Silesia, where I use the ﬁrst edition from 1880, all the Handbooks were in at least their second edition by the early 1880s. Johannes Conrad was the ﬁrst to recognize the potential of these Handbooks for the study of landownership, as discussed in Chapter 1. The current study will follow Conrad’s lead in focusing on the seven ‘core provinces’ of East Elbian Prussia, namely East Prussia, West Prussia, Pomerania, Posen, Brandenburg, Silesia, and Saxony. Note that this last is not the Kingdom of Saxony, with its capital in Dresden, but the province of Saxony, comprising the three administrative regions (Regierungsbezirke) of Merseburg, Magdeburg, and Erfurt. The territory of this former province now encompasses approximately today’s state of Saxony-Anhalt and part of the state of Thuringia. The principal descriptions and analyses of this book will come from three ‘snapshots’ of landownership data from the Handbook for each province. The ‘snapshots’ were of the data from the edition of the Handbook published closest to each year in which Germany took an agricultural farm census: 1882, 1895, and 1907. Table 2.1 shows the actual year of publication of each Handbook used:⁴ Table 2.1. Dates of data ‘snapshots’ for the seven East Elbian provinces Province

Date of publication of Handbook nearest to 1882 1895 1907

East Prussia West Prussia Pomerania Posen Brandenburg Silesia Saxony

1884 1885 1884 1884 1885 1880 1885

1895 1894 1893 1896 1896 1892 1899

1907 1909 1910 1910 1910 1909 1907

The Handbooks were supposed to cover all properties of at least 100 hectares in size or at least 1500 marks of taxable net yield, but they also included some properties under 100 hectares, even if they had less than 1500 marks of taxable net yield—on an apparently random basis and with great differences in inclusion in different provinces. If one were to take all the data of each Handbook for comparison of one province with another, it would be quite problematic, as Table 2.2 illustrates. ⁴ The single exception to the use of volumes from the Handbook series came for Silesia c.1907. Because the Handbook did not publish a new edition for Silesia between 1892 and 1922, I chose to use a locally published directory only for the province of Silesia, Wilhelm Gottlob Korn’s Schlesisches Güteradressbuch (9th edn, Breslau: Korn, 1909).

40

Data Considerations

Table 2.2. Total area of province and of various categories of properties, c.1882 (unadjusted %) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Total area of province

Total area of properties included in the Handbook (000 ha) 100 ha 0–99 50–99 + ha +ha

0–99 ha as % of 100+ha

50–99 ha as % of 100+ha

3,699 2,552 3,983 3,012 2,897 4,031 2,525

1,292 1,216 1,764 1,629 1,690 1,983 766

17.7 22.5 18.7 4.1 9.6 44.5 100.6

15.5 21.3 15.9 3.6 8.5 39.5 60.7

1.37 1.85 1.06 0.25 0.57 2.24 13.13

1.20 1.75 0.90 0.22 0.50 1.99 7.92

22,699

10,340

217.7

164.9

2.11

1.59

While the Handbooks for Pomerania and Posen included only trivial numbers of properties smaller than 100 hectares, the Silesian one included considerably more, and that for the province of Saxony provided much greater coverage of properties of between 50 and 100 hectares.⁵ Because Conrad appeared to have included all properties from the Handbooks in his statistics on ownership, irrespective of their size, the discrepancies seen above are a source of incomparabilities between provinces in his results.⁶ For the sake of data consistency, therefore, I have chosen to follow the standard practice of limiting my purview to properties of 100 hectares or larger. In order to avoid incomparabilities among provinces, this must be a strict limitation: for example, when totalling up all ownership by any given owner, only properties larger than 100 hectares each will be included in the total, even if there is information on smaller properties belonging to that owner. Thus even though the 1896 Handbook for the province of Posen included the property Kuznika skakawa, owned by Prince Heinrich von Preussen, a member of the Hohenzollern family who also owned many other properties, it is not included in his ownership totals because it was only 91 hectares in size. Does this procedure simply create its own set of problems? Not really—for example, for all noble owners, including royalty, the total area of such under-100-hectare properties in the Handbooks for c.1895 represented less than 0.2 per cent of the area of their over-100-hectare properties. Consistency in the data, therefore, appears to come at very little cost. ⁵ Full coverage of the properties of 50 or more ha came only with the Handbook volume for Saxony in 1913. ⁶ In the various articles he also varied the format and data inclusion of his tables, another frustrating inconsistency. (See discussion in Sect. 1.3.1.)

Data Considerations

41

2 . 3 . C O R R E C T I N G A N D A D J U S T I N G T H E D ATA

2.3.1. Internal Consistency of the Data Because the data of the Handbooks were not always complete and consistent, they had to be ‘cleaned’ before they could be used; because they were so voluminous, they also had to be checked periodically to make sure nothing was lost in the cleaning process. The procedure used was as follows: 1. The records were examined page by page to make sure all the properties on each page of the source had in fact been entered into the computer ﬁle. 2. To prevent double-counting of properties, the records were then examined individually to be sure that they had each been correctly assigned as a ‘principal property’ or a ‘subunit’. A ‘subunit’ was a property whose data were already included in the data for a principal property. The most common case of this was the case of a Herrschaft (manor) which was listed as ‘bestehend aus’ (consisting of ) a list of individually named subunits. 3. Every property received an owner code, identifying the social and legal status of its owner (e.g. noble person, non-noble person, religious body, the State, etc.), as detailed in the Appendix. Later, after other changes and corrections had been made to the data, the owner codes were checked again, one by one, to ferret out mistakes in coding (e.g. I searched the database for instances where the property owner was coded as ‘non-noble’, yet had either a noble title or ‘von’ in his or her name). 4. Changes in administrative divisions affected one province in particular, West Prussia, and to a lesser extent, Posen. The Prussian government carried out a major realignment of district boundaries, which created many new districts, in 1887. All properties from West Prussia, 1884, had to be checked and assigned to their proper post-1887 district, through the use of maps, gazetteers, and reports of the Prussian government concerning the administrative realignment. A similar procedure was followed for Posen. 5. The single most vexing problem was inconsistency, even within a single volume, in the recording of names of owners. I had ﬁrst recourse to the name index of the volume (done in every case for every year and province), which often could resolve the differences because it showed a given owner with multiple properties in that province. The name of the owner could then be entered in a consistent and identical way for each such property. Unfortunately, the indexes for some provinces and years did not make any attempt to consolidate owners, so from the index there was no way to tell if two owners—even if they had identical names—who appeared on different pages of the Handbook were the same person or two different persons. By default, I had to assume that they were different

42

Data Considerations

persons, unless I could ﬁnd other clues to their identity. For example, if the name index did not help, sometimes one could deduce that two names listed somewhat differently were really one and the same person. It was relatively easy to ﬁgure out that the owner of the estate Faulheden in Kreis Lötzen in East Prussia in 1884, ‘Baron von Schenck zu Tautenburg, auf Partsch’ was Baron Georg von Schenk zu Tautenburg, listed as owner of Partsch in Kreis Rastenburg; and not Baron Wolfgang von Schenck zu Tautenburg, owner of the estate Doben in Kreis Angerburg. Even easier was the case of ‘L. MacLean, auf Roschau, Kreis Danzig’—owner of Kobierschin in Kreis Preussisch-Stargard in West Prussia in 1884—being the same person as Lachlan MacLean, owner of Roschau. But was ‘Pförtner von der Hölle’, owner of Steinborn in Kreis Freystadt, Silesia in 1880 really Pr.-Lieutenant Richard Pförtner von der Hölle, owner of Schmarker and Ellguth-Schmarker in Kreis Trebnitz? Perhaps, since there was no other Pförtner von der Hölle listed as an owner of property in any other district of Silesia (or in any other province, for that matter). But, since there was no evidence by which to establish or deduce the identity of the owner without a given name listed—who could easily have been father, son, brother, uncle, or cousin,⁷ to Richard—I have had to count these as two separate owners. 6. Because quite a few nobles, in particular, owned properties in more than one province, the name check for this group had to be repeated, insofar as possible, across provinces as well as within provinces. Noble owners each received a single four-digit owner number that did not refer to any single province. (Other owners, in particular the Prussian State, who owned property in more than one province also received a single four-digit owner number for all provinces.) The owner numbers are important for determining how many owners owned a given set of properties. As pointed out before, the doubts and lack of evidence must mean that the process of assigning owner numbers overstates the actual number of owners. For example, the owner of Garlepow in Kreis Rügen in Pomerania, 1884, is simply listed as ‘von Berg’. Is he perhaps the same person as another von Berg without any given name, owner of Dubkewitz and Lüssvitz in the same district, or the third von Berg in that district, owner of Renz, Silenz, and Venzvitz? And are any of these Max von Berg, who owned property in Silesia, or Fritz von Berg, landowner in East Prussia, or perhaps Count Carl Friedrich von Berg, from Brandenburg? And, since Fritz is a diminutive of Friedrich, could Fritz and Carl Friedrich be the same person? Because of lack of necessary information there had to be six separate owner numbers for this group of von Bergs, even though in reality there may have been only two, or three, or four, or ﬁve individual von Bergs here, not six. ⁷ The chances of this owner being a female relative are extremely remote: so far as I could tell, the Handbooks always identiﬁed females as such, either by use of the given name or a designation such as ‘Frau’ or ‘Witwe’ (widow) or the female version of a noble title. c.1895, for example, I found 708 female owners holding just over 692,000 ha of land. Compare this to a total of 11,033 owners with 10.7 million ha in properties of 100 ha or more. (See Ch. 1 n. 10.)

Data Considerations

43

7. Non-noble owners were even more difﬁcult to identify, since the information was often much scantier than for nobles. Insofar as indexes and deduction would permit, names for this group were also consolidated, but in the overwhelming majority of cases the non-noble owners received an owner number only for a given province. No cross-province check of such owners could be made, unless an ownership entry in one province made speciﬁc reference to that person’s ownership of a property or residence in another province.

2.3.2. Problems with Individual Properties General Other checks and corrections dealt with individual properties which exhibited problems identiﬁed by the computer: 1. Sometimes only the total area of the property appeared in the given volume of the Handbook, with no details about the number of hectares of ploughland, meadow, etc. If the property appeared in an earlier or later version of the Handbook,⁸ and there had the same total area with details of the individual land types, I used those details also for the year in which they were missing. Otherwise, the record was left as originally entered, without any details. 2. When details were given, but they did not add up to the total area given for the property, I examined the source to see if there were simply some transposed digits, wrongly recorded digits or (as happened in only two or three cases) the area of a particular land type was printed on the line below or the line above where it should have been in the Handbook. If this did not explain the difference between the sum and the total, I again referred to earlier or later editions of the Handbook to try to resolve the anomaly. If the anomaly still existed (it would sometimes be carried through to successive editions of the Handbooks), I had to leave it. After all these corrections and adjustments, the remaining discrepancies were very small: The sum of absolute values of remaining anomalies of all sorts, including properties where the anomaly was the entire area of the property because no details were given, amounted in 1882 to only 0.3 per cent of total area of properties in the database, 1.1 per cent in 1895, and 3.1 per cent in 1907. Taking out the properties for which no details were given, the remaining anomalies fell to 0.1 per cent, 0.2 per cent, and 0.7 per cent of the area of those properties for which details were given in the respective years. Thus the overwhelming source of discrepancies between the total and the sum of the details for the database properties is the set of cases where the sum of details = 0, and the discrepancy amounts to the entire area of the property. Other sources of remaining discrepancies were trivial.

⁸ In the search I included editions of the Handbook not used in my ‘snapshots’, e.g. Brandenburg 1903 or Saxony 1913.

44

Data Considerations

3. If the total area of a property was not listed, but details of the area of individual types of land were given, I ﬁrst checked previous and subsequent editions for the Handbook to see if, in the case of exactly the same details, a total was given. If this did not provide the answer, I took the sum of the details to be the total area of the property. 4. A more common occurrence was that the assessed net yield for the land tax (Grundsteuerreinertrag, or GSRE) of the property was missing, even though all the other data were given. Missing GSRE was particularly prevalent for properties owned by the Prussian ﬁscus, since state domains were exempt from the land tax.⁹ In such a case, if reference to previous or subsequent editions of the Handbook did not solve the problem,¹⁰ I took the taxable value to be the amount calculated by multiplying the average tax assessment per hectare for each land type in the district (Kreis) where the property was located, times the number of hectares of land of that type in the property. The sum of these then became the estimate of the GSRE for the property. In those cases where only a total area was available, the GSRE was taken to be the product of that area times the overall average assessment per hectare of all taxable land in the district. Normally such estimates had to be made for only two or three properties in any given district in any given year. In one case, where only the GSRE was given, but no data about area of land, the total area was estimated by dividing the total GSRE of the property by the average GSRE per hectare for all taxable land in the district. 5. Sometimes the area of different land types was given only as a combined total, for example, plough plus meadow, or water and ‘moor’. In such cases, if references to earlier or later Handbooks did not provide the necessary details, if it was a total for two land types, I simply divided it in half; if for three, then in thirds. These properties were marked with a special ‘unﬁt’ code to prevent their being used in any regressions or cluster analysis based on the areas of individual land types.

The Principality of Pless The Prince of Pless, the largest individual landowner in Prussia after the Kaiser himself—as will be detailed in Chapter 5—owned the manor of Pless in the district of the same name in Silesia; the manor had been designated a principality (Fürstentum). According to the Handbooks, the area of the manor of Pless was 49,797 hectares in 1880,¹¹ but only 34,755 in 1892; the Silesian address book ⁹ But all properties, even state-owned ones, had been assigned a land tax net yield, to allow for the eventuality that they might be sold to an owner who was not tax exempt. Thus the problem was to ﬁnd out or to estimate what had simply not been reported. ¹⁰ The taxable value, after being set in the early 1860s, was never changed. This was in part an incentive to owners to invest in improvements to their land with no fear of adverse tax consequences. See Wilhelm von Lesigang, ‘Grundsteuer’, in Johannes Conrad, ed., Handwörterbuch für Staatswissenschaft, 2nd edn, iv ( Jena: Gustav Fischer, 1900), 898. ¹¹ It is clear that Conrad took the Handbook data for Silesia at face value. He calculated the Prince of Pless’s total landholdings in the early 1880s as 70,139 hectares (my database produced an almost identical ﬁgure—70,122—from the same original data). Conrad, ‘Die Latifundien’, 159.

Data Considerations

45

put it at 40,437 hectares in 1909. Fluctuations of this magnitude were not possible, especially as no other owners appeared in Pless district of Oppeln region in 1892 who might have taken up the 15,000 hectares that appeared to have gone missing from the principality of Pless, or who later could have returned nearly 6000 hectares to the principality. Moreover, the ﬁgure for 1880 seems to be an overstatement, based upon the following reasoning: the manor of Pless was in large part forested. In the 1909 address book, information about the administration of the ‘upper’ and ‘lower’ forests is listed in the general description of the manor,¹² and it seems abundantly clear (but is not explicitly stated) that these areas of forest administration are included in the total area of the individual properties listed as comprising the manor. In the 1880 Handbook, however, the ‘upper forests’ and ‘lower forests’ are listed as if they were properties in their own right, and therefore making up a separate part of the total area of the manor. This appears to be a case of double counting, but because the list of the other properties is incomplete, taking this forest area (more than 23,000 hectares) out of the total would then produce a result almost laughably low. The data in the 1892 Handbook were also very incomplete, but the ‘upper forests’ and ‘lower forests’ played no role whatever in the list that year. Taking these considerations together, it seemed reasonable to assume that the ﬁrm publishing the Silesian address book had better information, especially since it had started this series in 1870 and was already in the ninth edition by 1909, whereas the publishers of the Handbooks had produced only two editions for Silesia, and could not manage a third edition until thirty years after publishing the second. In consequence, I have felt it necessary to modify the data for the principality of Pless: in the interests of greater accuracy, I have substituted the data from the third edition of the Silesian address book (1886) for the data of the 1880 Handbook. This made the total area 40,942 hectares with a taxable value of 225,652 marks. The fourth edition of the address book (1891) gave the ﬁgures as only 37,280 hectares, with a taxable value of 209,300 marks, so I used the data from 1905—40,029 hectares¹³ and 223,598 marks,¹⁴ respectively, in place of the original ﬁgures for 1892. ¹² Schlesisches Güteradressbook (9th edn, 1909), 520. ¹³ The close correspondence with Conrad’s own calculations is remarkable: in his article on Silesia, using the 1892 Handbook, Conrad put Prince Pless’s holdings in that province at 51,112 ha. The Handbook for 1892 shows Prince Pless’s other properties in that province totalling 11,029 hectares, so the difference is 40,083 hectares. Given that the data in the Handbook for individual properties within the principality of Pless added up to only 34,755 hectares and that the Silesian address book for 1891 reported 37,280, the near-identity of Conrad’s ﬁgures with those of the 1905 address book is even more remarkable. The taxable value calculations do show a more marked discrepancy, however: Conrad’s ﬁgures imply a taxable value for the principality of Pless of 217,406 marks, whereas the Silesian address book for 1891 has 209,300 and for 1905, 223,598. See Conrad, ‘Schlesien’, 713. ¹⁴ I am greatly indebted to my colleague and friend Dr Christa Kouschil for these data from the pre-1909 editions of the Silesian address books.

46

Data Considerations

2.3.3. Incomplete Coverage of the Data Calculations for the regional distribution of large landed properties revealed some serious shortcomings in the recording of Prussian State properties in three provinces. This began with the calculation shown in Table 2.3. From the data of the table, the provinces fall into three groups: Table 2.3. Share of area and tax value of properties of 100 hectares or more in area and tax value of all land outside of cities, by province (unadjusted %) Province

Share of total area Share of total tax value 1882 1895 1907 1882 1895 1907

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

36.2 49.8 48.5 58.0 61.4 50.9 33.9

35.8 50.3 48.8 61.8 61.8 51.9 33.8

45.6 46.3 48.9 61.8 52.0 52.6 36.6

38.5 41.8 44.5 62.7 61.1 42.8 31.5

37.0 41.0 43.9 64.8 61.0 43.9 33.7

44.3 37.6 43.3 61.7 50.6 45.3 36.7

total

48.4

49.2

49.5

44.2

44.8

44.8

1. One province in which large properties apparently gained rather signiﬁcant amounts of both area and tax value: East Prussia. 2. Four provinces in which large properties gained at least some area, if not always gaining tax value: Brandenburg, Pomerania, Silesia, and Saxony. 3. Two provinces in which large properties declined rather signiﬁcantly in both area and tax value: West Prussia and Posen. The declines are easily explained: West Prussia and Posen are the provinces in which the Prussian government’s Royal Settlement Commission was active in buying up and subdividing large estates after 1886, and especially after 1898, when it received a massive budget infusion that doubled its land-purchase fund.¹⁵ Among the other provinces, East Prussia stands out, and its apparently anomalous pattern after c.1895 requires explanation. The 2778 properties of 100 hectares or more in East Prussia in the 1907 Handbook had an area of 1,625,571 hectares, compared to the 1,278,261 hectares of the 2238 large properties in the 1895 Handbook. The difference of 347,310 hectares needs to be explained, since this was an increase of 27 per cent, compared to increases for the other provinces on the order of 2 per cent or less, except 8 per cent for Saxony. ¹⁵ See Eddie and Kouschil, Ethnopolitics of landownership, esp. 10–14.

Data Considerations

47

In order to explain this phenomenon, I inspected all individual properties of 100 hectares or more in East Prussia from the 1895 and 1907 databases, to determine which were pairs—that is, which named properties appeared in both databases—and which were not. Those large properties that appeared in the 1907 database, but did not appear in the 1895 database, then became the subject of further analysis: of the 2778 East Prussian properties of 100 hectares or more in 1907, 707 had no pair in the 1895 database. Of these 707, 12 did appear to have a pair in the 1882 database.¹⁶ I thus focused my analysis on the 695 properties which appeared not to have a pair in either of the earlier databases.¹⁷ The principal source of discrepancy lay in the properties of the Prussian state.¹⁸ In 1907 in East Prussia, the Prussian state owned 91 properties of 100 hectares or more, totalling 316.9 thousand hectares, that did not have a matching pair in 1895. In addition, it owned an additional 30 properties, comprising 23.6 thousand hectares, that had been owned by someone else in 1895, for a total of 340.5 thousand hectares. Subtract from this the properties of 100 hectares or more owned by the Prussian state in 1895 that did not have a corresponding match in the 1907 database (27 properties of 37.3 thousand hectares) and the two properties owned by the State in 1895, but by someone else in 1907 (only 760 hectares), for an apparent net increase of 302.4 thousand hectares. This large increase noted in Prussian state properties presents other serious difﬁculties as well: some of these properties were very large; 31 of the 91 properties of 1907 with no match in 1895 were each 5000 hectares or more (198,434 hectares in total). It is unreasonable to think that the State acquired these as a series of parcels, each of which was much smaller, and most of which would have to have been less than 100 hectares. The lands in question were 76 per cent forest, which does not grow overnight. According to the Statistical Yearbook for the German Empire, reporting data of forest surveys done by the government, there were 6724 sq. km of forest land in East Prussia in 1878, 6621 in 1883, 6477 in 1893, 6445 in 1900, and 6608 sq. km in 1913. If we interpolate ¹⁶ This apparent anomaly could occur if, for example, a property listed separately in 1882 was considered a part of a larger property complex (but not explicitly named) in 1895, and then reverted to being listed as a separate property again in 1907. It also could happen if a property had both a German and a Polish name, and the Polish name were used in one year and the German name in the other two, or vice versa. Matching names by a simple alphabetical sort, unless the Polish and German names were very similar, would not capture this possibility. ¹⁷ Properties that were located in more than one district (Kreis) could have been reported in different districts in the different editions of the Handbook. Such cases would show up as unpaired, even though they might in fact have had a pair. This should not have been a serious problem, however, since the estates were usually reported in the district in which the property seat or manor house was located, except especially for some State properties in East Prussia, as we will see below. ¹⁸ Nabert (Provinz Sachsen, 1913–1933, 7) noted that State properties of small size and low tax assessment were missing from the Handbooks, but—as we shall see in what follows—many large properties also seemed to have been left out.

48

Data Considerations

between 1900 and 1913, these data imply a forest area of roughly 6550 sq. km in 1907—an increase since 1893 of around 100 sq. km, or only approximately 10,000 hectares. It is clear from these data that the recorded increase in the area of forest land owned by the Prussian State could not have come from a massive program of afforestation. We shall therefore have to examine the properties of the Prussian State in detail, in order to resolve the anomalies noted, which is the main business of the next section.

2.3.4. Adjusting for the Anomalies in Reporting State-Owned Properties Introduction The apparently non-standard change in the area of large properties over time in the province of East Prussia has revealed a problem with the inclusion of State properties, especially forest lands, in the Handbooks. Comparing forest areas to ofﬁcial survey data for forest lands revealed further anomalies in the areas owned by the Prussian state in two other provinces—Pomerania (in two years) and Brandenburg (one year).¹⁹ In all other provinces and years, the Handbook totals for land owned by the State and Crown always exceeded 90 per cent of that reported by the surveys, my arbitrary cut-off point for triggering an adjustment to the data. Since each of the three provinces—Pomerania, Brandenburg, and East Prussia—exhibited a different problem or set of problems, I shall deal with them separately in turn, and in order of increasing difﬁculty of the problems presented.

Adjustments to State and Crown Properties for Pomerania, Brandenburg, and East Prussia Pomerania In Pomerania in 1884, the Handbook data showed State and Crown forest land to be only 35 per cent of that shown as owned by State and Crown in the ofﬁcial survey of forest land for 1883. The ﬁgure rose to 81.1 per cent in 1893, compared to the 1893 survey, and 98.6 per cent in 1910, compared to the ofﬁcial 1900 survey.²⁰ Here, interestingly enough, the problem seemed clearly a case of not including most of the larger State properties in the Handbook in the earlier years: in the original data for 1884 there were only 12 properties of 2000 hectares or more containing 51,781 hectares of forest vs. 161 under ¹⁹ Prussia took ofﬁcial surveys of the amount of forest land in the kingdom in the years 1878, 1883, 1893, 1900, and 1913. See Statistical Yearbook, 1880, 21; 1885, 26; 1895, 14; 1902, 16; and 1915, 47. ²⁰ It would have been preferable to compare the 1910 Handbook to the 1913 survey data, but because the survey for 1913, as reported in the Statistical Yearbook (1915, 47), did not include a breakdown by ownership, I could not use its data.

Data Considerations

49

2000 hectares with a mere 9,409 hectares of forest. By 1896, these numbers had changed to 34 with 141,364 hectares of forest vs. 162 with 8932. Moreover, only three of the 34 largest properties contained less than 80 per cent forest land, and only seven less than 90 per cent. The trend continued to 1910, where the numbers were 45 properties of 2000+ hectares with 174,745 hectares of forest vs. 176 under 2000, with 13,183 hectares of forest. Of the 45 largest properties, only 4 had less than 80 per cent forest land, and only 17 less than 90 per cent. We therefore see what appears to be a dramatic improvement in the coverage of the larger forest properties in the Handbook, rather than any programme of massive state purchases of forests, since—as previously noted—it is very unlikely that the gain in forest land could have been the result of buying up legions of adjoining small woodlands in properties under 100 hectares in size. This assumption then determines the method used for adjusting the data for Pomerania. Because the adjustment made for 1884 depended on the adjustment made for 1893, let me deal with the latter ﬁrst. In Pomerania there were no properties that appeared in the 1884 and 1910 editions of the Handbook, but not in the 1893 edition. Five properties appeared as state properties in 1884, but not in 1893; of these, four had different owners in 1893 and the ﬁfth had no forest land. I therefore did not add any of these properties to the 1893 data. Of the properties that appeared as State properties only in the 1910 directory, 19—containing 26,156 ha of forest land—did not appear anywhere in the 1893 directory. Assuming that the Prussian state also owned these properties in 1893, adding them to the existing data for State and Crown properties raised the Handbook share of forest owned by State and Crown to 94.6 per cent of the total counted in the 1893 forest survey. This ﬁgure represents an improvement over the original coverage of the Handbook, but is still not completely satisfactory because of its arbitrary nature. In similar fashion, assuming that State properties that appeared in both the 1893 and 1910 editions of the Handbook for Pomerania—but not in 1884—were nevertheless actually owned by the Prussian state in 1884, I added the data for these 18 properties (they included 82,072 hectares of forest land) to the 1884 Handbook data. This raised the area of forest land owned by State and Crown only to 82.0 per cent of the 1883 survey total. Because it was still well short of the area reported in the ofﬁcial forest surveys, I took the further step of adding the same properties to 1884 that I had added to 1893, with their additional 26,156 hectares of forest land, raising the ratio of State and Crown property in the Handbook of 1884 to 96.5 per cent of the ﬁgure in the ofﬁcial 1883 survey of forest land. Because doubly arbitrary, and because the percentage ﬁgure was most likely still short of what the true ﬁgure should have been, this solution to the problem is less than perfectly satisfactory, but at least it seems in accord with the apparent pattern of incomplete coverage of large properties.

50

Data Considerations

Brandenburg There was a similar, but more perplexing, problem with the province of Brandenburg. For this province, State and Crown forest land in the Handbooks came to 98.5 per cent of the ofﬁcial survey total in the ﬁrst year (1885 vs. 1883, respectively) and 96.5 per cent in the ﬁnal year (1910 compared to 1900). Given that the Handbooks did not include all properties under 100 hectares, and the distance in time between the Handbook and the survey in the second comparison, these ﬁgures can be regarded as essentially complete correspondence between the Handbook data and the survey data for State and Crown properties in the ﬁrst and last years. But in the middle year (1896/1893), the ﬁgure reached only 86.4 per cent. The problem again seemed to be one of failure to include large properties, but why should this happen in an edition published after one in which they were included? To examine the problem, we can start again with the breakdown of 2000+ and under-2000 hectare properties, as presented in Table 2.4. Applying the dictum that ‘nature does not make jumps’, properties containing 30–40,000 hectares of forest appear to be missing from 1896. But a look at area totals shows that these properties are not missing altogether: Table 2.5 contains the totals for Table 2.4. Brandenburg: Original data for State and Crown properties (before adjustment) Year and size of property (ha)

No. of properties

Area of forest (ha)

Forest as % of total area

1885, under 2000 1885, 2000+ 1896, under 2000 1896, 2000+ 1910, under 2000 1910, 2000+

146 81 121 47 150 90

13,537 403,501 5,207 365,773 18,696 409,779

16.3 91.8 9.8 88.7 21.4 89.9

Table 2.5. Forest land in all principal properties in the Handbooks, for all owners, Brandenburg and Pomerania (unadjusted data) Brandenburg Year

Ha forest

Pomerania Year Ha forest

1885 1896 1910

864,871 882,779 887,316

1884 1893 1910

378,400 468,836 525,966

Data Considerations

51

forest land for all owners of all principal properties in Brandenburg; it shows a quite normal progression in the total area of forests covered by the Handbooks for Brandenburg alongside the dramatic change noted earlier for Pomerania. For Brandenburg, we must therefore look for a solution different from the one chosen for Pomerania. Comparing all properties of the Prussian State and Crown in the province of Brandenburg in 1885 with those in 1910 revealed six State properties and two Crown properties that appeared in these two editions of the Handbook, but not as State or Crown property in 1896. Assuming continuity of ownership, I could have added these eight properties (1885 data) to the database for 1896, except that each of them was already listed in 1896, but with a different owner. Examining these properties individually, and comparing them also with the data from the 1903 Handbook for Brandenburg, revealed that for three of the biggest properties, the person listed as owner in 1896 was listed as lessee or administrator in 1903, and in a fourth case the person listed as owner in 1896 was a Royal Chief Forester (Königlicher Oberförster), so also likely the administrator of the estate, which was listed as the property of the state in 1885, 1903, and 1910. Moreover, nearly all of these properties were listed as being a ‘royal domain’ or ‘royal forest’ in the property description in the 1896 Handbook. Assuming that this confusion of lessee or administrator for owner was the case for all properties which were listed as being owned by either State or Crown in 1885 and 1910, I changed the ownership of these eight properties, thereby adding 27,384 hectares to the total forest area owned by State and Crown in 1896 (and of course subtracting the same amount from the total possessed by other owners). This brought the total for Crown and State forest land to 91.9 per cent of the 1893 survey ﬁgure. Within that total, the 40,186 hectares of Crown land come very close to the 1885 ﬁgure of 40,850, and the 1910 total of 39,653 hectares, so any remaining shortage most likely comes in category of State land (1885: 376,188 hectares; 1896: 354,657; 1910: 388,822). As before, the adjustment began with State properties in 1910 that did not appear in the 1896 Handbook for the province of Brandenburg. There were 21 of these in all, of which 9 did not appear in the 1903 Handbook either, plus a tenth property owned by someone else in 1903. Eliminating these latter 10 properties from further consideration, I assumed that property owned by the State in 1903 and 1910 was also owned in 1896, and so added the remaining 11 properties to the 1896 database. These 11 properties made the only net addition to the total forest land reported for Brandenburg (unlike the eight properties for which only the owner was changed) of 17,931 hectares of State forest, bringing the number to 368,135 hectares for the State. The ﬁnal adjusted total of 408,321 hectares of State and Crown forest reached 95.1 per cent of the total from the ofﬁcial survey of 1893. This ended the adjustments for Brandenburg.

52

Data Considerations

East Prussia According to Meitzen’s data from the land tax assessment of the 1860s, State forest properties in East Prussia amounted to 452,474 hectares—not all forest land, since some state forests included other types of land as well, such as water and even ploughland and meadow. But we can assume, I think, that this ﬁgure includes about as much forest land as shows up in the 1883 total of 370,572 hectares. There were no Crown properties listed in East Prussia, so here we can concentrate solely on properties owned by the Prussian State. Only three properties appeared in the 1884 and 1907 Handbooks, but not in that for 1895. Adding these three to 1895 added only 3796 hectares of forest in that year, using the 1884 data. There was no property that appeared only in the last two years, so nothing to add to 1884 from that viewpoint. That leaves only the State properties found in the 1907 Handbook, but not in either of the ﬁrst two. There were 89 principal properties that satisﬁed this criterion; they contained 250,266 hectares of forest land. This is about 100,000 hectares more than the shortfall in either 1884 or 1895. Under these conditions, to make the necessary adjustments for East Prussia, it became necessary ﬁrst to examine the data property by property and district by district. The ﬁrst anomaly appeared in one district (Goldap in the administrative region of Gumbinnen) that showed a considerably smaller area of state-owned forest properties in 1895, compared to both 1884 and 1907. The reason for this apparent decline was that, for two forest estates in 1895, Goldap and Warnen, only the total area was reported in the Handbook, but not the details. The data for the Goldap State forest were included with the estate of Warnen in 1884; the two were listed separately in 1895 and 1907, but without any details in 1895. The areas and tax values of Goldap and Warnen in 1895 added up to the total of Warnen in 1884, so I apportioned the details from 1884 between the two properties on the basis of their size relative to each other. This added 9784 hectares of forest to the reported total for Goldap district in 1895, removing the apparent anomaly entirely. In three districts (Gerdauen and Landkreis Königsberg in Königsberg administrative region along with Stallupönen in Gumbinnen region), large State forests seemed to have disappeared between 1895 and 1907. This turned out not to be the case, however, but was the result of relocating or combining the data for estates that lapped over the boundaries of more than one district: the estate of Klein-Astrawischken (with 1636 ha of forest) from Gerdauen underwent a remeasuring and was included in the data for Astrawischken in Insterburg district of Gumbinnen region in 1907; Fritzen forest (3432 ha), in the district of Königsberg in 1884 and 1895 reappeared as Preussisch-Eylau forest in the district of the same name in 1907; and from the district of Stallupönen in 1884 and 1895, another part of the estate of Warnen (with a further 1320 ha of forest) was consolidated into the estates of Warnen and Goldap in 1907 after having been separately reported earlier (see previous paragraph). In one instance,

Data Considerations

53

Table 2.6. 1907 properties added to 1884 and 1895, East Prussia Admin. regiona

District

Königsberg

Fischhausen Wehlau Heilsberg Allenstein Osterode i. O. Mohrungen Heydekrug Tilsit (Land) Pillkallen Insterburg Lyck Johannisburg

Gumbinnen

total a

No. of properties

Area in hectares Forest land Total Area

Tax value (marks)

3 3 2 6 5 4 2 1 2 4 1 6

15,197 9,136 5,445 20,443 19,566 8,806 7,425 3,049 7,957 16,416 3,654 29,797

17,474 10,656 5,732 24,787 24,170 13,734 11,011 4,647 11,469 20,437 4,091 37,532

32,733 33,338 12,083 33,704 46,661 47,301 28,630 11,085 24,258 87,926 6,396 35,219

39

146,891

185,740

399,335

Using 1895 administrative divisions.

however, an estate was apparently missed in 1907, most likely because the major portion of it lay in the district of Goldap, where we have noted earlier problems of a similar nature: Nassawen, with its 2470 hectares of forest in 1884 and 1895, was therefore added to the 1907 data for the district of Stallupönen. After dealing with these problems, it was time to deal with those districts which showed a sizeable increase in state-owned forest area between 1907 and the previous years. Using 1000 hectares difference as an arbitrary cut-off point, after eliminating two districts where the amount added would have very substantially exceeded the total difference in forest area between 1907 and 1884,²¹ I added properties out of the 1907 Handbook from 12 districts to both the 1884 and 1895 databases, in the same fashion as I had done earlier for the province of Pomerania. Table 2.6 summarizes these additions. The net result of these changes for the province of East Prussia was to raise the area of State forests from the Handbooks to 361,326 hectares in 1884 (97.5 per cent of the 1883 survey total) and to 361,314 hectares in 1895 (98.0 per cent of the survey total). With these adjustments in place, we are now ready to readdress the question of the regional distribution of large properties in the chapter that follows. Note that the foregoing adjustments used all the Handbook entries, regardless of size, whereas for all of our later analyses we return to the set of large properties comprising 100 or more hectares of land each. ²¹ The most likely explanation for this phenomenon is that much of the forest land that seemed ‘new’ in 1907 was in fact listed as part of other properties, even in other districts, in the earlier years, just as in the cases of Goldap, Warnen, and Fritzen.

3 The Geographical and Size Distribution of Large Landed Properties 3 . 1 . P R EV I EW This chapter is like the label on a food product: it tells us the provenance of the product (where the East Elbian large properties lay), the nutrition information (what the makeup of the properties was), and its calorie content (how big the properties were). It even has some serving suggestions (different ways to look at the data). With more than two-ﬁfths of the area of Germany, and with especially their noble landowners playing a leading role in the government, the seven provinces of our concern were a major element of the country. The chapter begins with a discussion of the land endowment of these seven provinces, introducing the concept of ‘land-use proﬁle’ and calculating that proﬁle for each district found in the land tax cadastre of the early 1860s. Putting these districts into six clusters according to their overall land-use proﬁles, I show how the different clusters were distributed by province. The result could be characterized as ‘regionally concentrated diversity’ with the land best suited to arable agriculture concentrated in the east central and southern provinces, to forestry mostly in Brandenburg and Silesia, whereas the Baltic provinces were best suited to widespread mixed land use. Properties of 100 hectares or more in just these seven provinces accounted for a ﬁfth of all the land in Germany. These large properties were quite evenly spread among the seven provinces, with the largest absolute area in Silesia, the largest relative area in Posen and Pomerania (until Posen fell off after 1895 as a result of the activities of the Royal Settlement Commission), and with their total area remaining remarkably stable over the period under review. In the 54 districts in which large properties covered at least 60 per cent of the land, 42 belonged to just three provinces: Posen, Pomerania, and Silesia. Of the 14 with under 20 per cent, 9 were from Saxony. The distribution of properties by size (area), for the seven provinces overall and severally, showed the greatest number of properties in the 200–499 hectare category, the greatest area in the 500–999 hectare category, for the provinces

The Distribution of Large Landed Properties

55

taken together. For East Prussia, Silesia, or Saxony, however, the greatest area was also to be found in the 200–499 hectare size category. In terms of the distribution of large properties by tax value, the data show that total area was more unequally distributed than was total tax value, a result mostly of the different composition of properties in different size categories. The distribution by tax value conformed very closely to a log-normal distribution, but that by area did not. As a lead-in to the succeeding chapter, this chapter shows how the size of properties was distributed by different classes of owners, and shows that the area of the properties was more unequally distributed by owner than by property, and that this varied considerably by province. Silesia was very clearly the bastion of the great landed magnates.

3 . 2 . I N T RO D U C T I O N The seven East Elbian provinces of our concern in this book covered an area of 227 thousand square kilometres, representing 65 per cent of the total area of Prussia (348.5 thousand sq. km). Prussia, in turn, accounted for 64 per cent of the total area of Germany (540,000 sq. km). These seven provinces, therefore, contained more than two-ﬁfths of the total territory of Germany.¹ Because these eastern provinces were more heavily forested, and contained more lakes and moorland, than the western parts of Prussia, and also because some areas were made up of relatively low-quality land,² these ‘core provinces’ accounted for only 56 per cent of the taxable value of land in Prussia.³

3 . 3 . T H E L A N D E N D OW M E N T O F T H E E A S T E R N P ROV I N C E S The data of the land tax cadastre of 1861–4 revealed the distribution of land by type—that is, ploughland, garden, meadow, pasture, forest, etc.—in each district of Prussia.⁴ For each of the 236 rural districts included in the data of the land tax cadastre for the seven eastern provinces, using the data on the total area ¹ 1894 data from Statistical Yearbook. ² On the other hand, the province of Saxony contained—in the area around Magdeburg—some of the richest soil in all of Germany. See sect. 3.3 and also Ch. 8 on land use proﬁles of large properties for a detailed discussion of the quality of the major types of land by province. ³ Because other German states each had their own land tax regimes, different from that of Prussia, the comparison cannot be extended to Germany as a whole. ⁴ Published in 25 volumes, one for each administrative region in the Kingdom: Prussia. Ministerium der Finanzen, Die Grundsteuerveranlagung. Ergebnisse f¨ur den Regierungsbezirk (Berlin: Ministerium der Finanzen, 1866– ).

56

The Distribution of Large Landed Properties Table 3.1. Cluster means of land-use proﬁles from land tax cadastre Land use Arable Animal husbandry Forestry Other Number of districts in the given cluster

Cluster number 1 2 0.473 0.246 0.244 0.037 37

0.379 0.439 0.155 0.026 7

3

4

5

6

0.640 0.161 0.184 0.015 74

0.369 0.140 0.476 0.015 29

0.529 0.125 0.329 0.017 44

0.798 0.098 0.095 0.009 45

Table 3.2. Number of districts in each cluster by province Province/Cluster no.

1

2

3

4

East Prussia West Prussia Pomerania Brandenburg Posen Silesia Saxony

15 6 8 2 1

4 1

1

2

15 9 12 3 12 14 9

total

37

7

74

5

5

6

Total

1 12 2 11 2

2 2 10 4 19 7

2 3 1 7 14 18

35 20 26 30 26 58 41

29

44

45

236

of each land type (ploughland, meadow, etc.), I calculated the land-use proﬁle of each district, deﬁned as the share of the total potentially taxable area devoted to arable agriculture (plough, garden), to animal husbandry (meadow, pasture), to forestry (woodland), and to other uses (moor, water).⁵ Using the k-means partitioning method of cluster analysis, I classiﬁed the districts into groups according to their land-use proﬁles. After examining the results for 4, 5, 6, 7, and 8 clusters, the results for 6 clusters seemed to provide the best differentiation without excessive detail.⁶ The results of this classiﬁcation are shown in Tables 3.1 and 3.2, which show, respectively, overall cluster means and cluster membership by province. Looking at the tables by cluster, we ﬁnd the following: 1. The 37 members of cluster number 1, consisting of districts exhibiting mean values of roughly half arable land and the other half split evenly between meadow and pasture on the one hand, and woodland on the other, ⁵ This excludes unproductive land, land untaxed because it was devoted to such public use as roads, and farmyards, but did include land owned by the Prussian State, even though State-owned lands were exempt from tax. The tax cadastre nevertheless assessed State lands, just as it did private lands, in case State property should ever pass into the hands of a non-exempt owner. ⁶ It was with six clusters that the ﬁrst notably small cluster appeared—the 7-member cluster 2 in the table. This cluster remained stable at 7 members also for division into 7 and 8 clusters.

The Distribution of Large Landed Properties

2.

3.

4.

5.

6.

57

were heavily concentrated along the Baltic in the north (29 of 37 districts), especially in East Prussia. Of the remaining eight districts, ﬁve were in the province of Saxony, only two in Brandenburg, one in Posen, and none at all in Silesia. This small group, with about two-ﬁfths of its land devoted to each of arable agriculture and animal husbandry on average, had members in only three provinces: four districts from East Prussia, one from West Prussia, and two from Brandenburg. Cluster 3, in which, at the mean, roughly two-thirds of the land was arable, and about one-sixth each was land for animal husbandry and forestry, was the largest cluster. It was the only cluster to contain districts from every province, although with only three districts in this cluster, Brandenburg was a very minor member. The fourth cluster, in which roughly half the land was woodland and another third or so devoted to arable agriculture, was geographically concentrated in Brandenburg and Silesia. Beyond that, other provinces hardly mattered, with only two districts each from Posen and Saxony, and a mere one each from East Prussia and Pomerania, and none at all from West Prussia. The ﬁfth cluster, containing districts with about half their land in arable and another third in forestry, consisted of 44 districts concentrated in Silesia (19 districts) and Brandenburg (10). Seven districts from Saxony, four from Posen, and two each from East Prussia and Pomerania, rounded out the membership in this cluster. The last cluster, with about 80 per cent arable land on average, was also heavily concentrated, this time in Saxony (18 districts) and Silesia (14). After the seven districts from Posen, the rest of the provinces hardly mattered to this cluster.

Looking at the tables by province, we ﬁnd the following: 1. Thirty of the thirty-ﬁve districts in East Prussia fell either into cluster 1 or cluster 3. These were clusters in which—at the mean—land for animal husbandry and land for forestry were found in roughly equal proportion, while arable land made up half to two-thirds of the total area. East Prussia also accounted for the majority of districts in cluster 2, districts in which arable and animal husbandry together made up about four-ﬁfths of the total land. Other cluster conﬁgurations were either not found (clusters 5 and 6, the clusters with the highest shares of arable land), or claimed only one district from East Prussia (cluster 4—forestry with some arable agriculture). 2. West Prussia showed 9 of her 20 districts in cluster 3, and 6 in cluster 1, so presented an overall land-use proﬁle similar to that of East Prussia. West Prussia had only two districts each in clusters 5 and 6, one in cluster 2, and none at all in cluster 4.

58

The Distribution of Large Landed Properties

3. Pomerania’s 26 districts were concentrated in clusters 1 and 3, just as East Prussia’s were. While the distribution of districts among the other clusters was different from that of either East or West Prussia, Pomerania’s overall land-use proﬁle proved very similar to that of the other two Baltic provinces. 4. Brandenburg’s districts, however, were concentrated in clusters 4 and 5, where at their respective means the land-use proﬁles showed over 80 per cent of the total land in arable agriculture and forestry taken together. The other Brandenburg districts were scattered thinly over the rest of the six clusters; indeed, Brandenburg was the only province that could show at least one district in each cluster. 5. Nineteen of Posen’s 26 districts were to be found in clusters 3 and 6, in which land for animal husbandry and forestry appeared in about equal shares, while land for arable agriculture was dominant—about 60% and 80% of total land at the respective cluster means. 6. Silesia’s 58 clusters were the least scattered of those of any province: no districts fell into clusters 1 and 2, and from 11 to 19 districts into each of clusters 3, 4, 5, and 6. So about half of Silesia exhibited land-use proﬁles very similar to those of Posen, while the other half looked much more like Brandenburg. 7. The province of Saxony exhibited more variety than any other province, save Brandenburg. Eighteen of her 41 districts appeared in the cluster that was most highly specialized in arable agriculture, group 6, with a cluster mean of 80 per cent arable land. Nine districts fell into cluster 3, and seven into cluster 5, clusters having, at their respective means, shares of arable and forestry together adding up to more than 80 per cent. Five districts in cluster 1, two in cluster 4, and none in cluster 2 completed the distribution. The greatest concentration of areas most especially suited for arable agriculture (cluster 6) was thus found in the east central and southern provinces. While Silesia and Saxony accounted for the largest numbers in this cluster, Posen’s relative contribution was larger than that of Silesia (27 per cent of all her districts vs. 24 per cent, respectively). Animal husbandry was the most important land use in the smallest cluster, number 2, with only seven districts. Otherwise it was scattered among all clusters, with a mean share of total land area never exceeding one-quarter. Of two clusters in which the mean share of land for animal husbandry reached at least one-quarter, the Baltic provinces accounted for 34 of the 43 districts found in those clusters. Forestry was especially notable in clusters 4 and 5; thus Brandenburg and Silesia dominated in these two clusters. The Baltic provinces, on the other hand, appeared especially ill-suited, from their natural gifts (or lack thereof), for any real specialization on this activity.

The Distribution of Large Landed Properties

59

The mean share of ‘other’ land in any cluster never reached 4 per cent, and was under 2 per cent for four of the six clusters, so no concentration of ‘other’ land could be identiﬁed at the district level. At the time of the land tax cadastre, therefore, these seven provinces showed a prevalence of land suited for arable agriculture especially in the south and in Posen, with forestry especially favoured in the west central and south-east, while the Baltic provinces appeared most suited to widespread mixed land use. In Chapter 8 I will take up again the classiﬁcation by land-use proﬁles, this time for individual large properties rather than administrative districts. But ﬁrst it would be well to look at the geographic distribution of the large properties that form the principal focus of our concern in this book.

3 . 4 . T H E R E G I O N A L D I S T R I BU T I O N O F L A RG E P RO PE RT I E S The large estates of East Elbia—taken to mean properties of at least 100 hectares in size, according to long-established German custom—encompassed 20 per cent of all the land in Germany, given that they made up 47 per cent of the total area of the seven eastern provinces c.1895. Because larger estates tended to have a higher share of lower-taxed land (forest in particular, but also pasture and moorland) than did smaller properties, they accounted for only 41 per cent of the taxable value of land in the seven provinces. If we consider only the total land area and taxable value outside of cities,⁷ the ﬁgures become 49 per cent and 45 per cent, respectively. Large properties were, in consequence, quantitatively very important in the provinces of our concern and, indeed, in Germany as a whole. It is not surprising, therefore, that—even aside from the social position of their owners—the large estates of East Elbia should have received considerable attention in both contemporary and historical literature. Table 3.3 shows the provincial distribution of such large properties by number and area at the years nearest each agricultural census, as reported in the Handbooks after the adjustment of their data detailed in Chapter 2. The stated numbers of properties cannot bear too much weight in analysis, given that what constitutes a property could be differently deﬁned in different years. The problem is most evident for the province of Silesia, where the total area of large properties continued to grow between the second and third ‘snapshots’, even though the number of properties appeared to have fallen by 15 per cent. As we saw in Chapter 2, Silesia was the only province for which we had to use ⁷ City–rural basic data from the land tax assessment of 1861–4 (cf. n. 4). As more and more towns became cities over time, the ‘rural’ area correspondingly shrank, so that the percentages shown are actually an understatement.

60

The Distribution of Large Landed Properties Table 3.3. Distribution of large properties by province, adjusted data Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Number of properties c.1882 c.1895

c.1907

Total area of properties (000 ha) c.1882 c.1895 c.1907

2,298 1,687 2,061 2,551 2,367 3,339 1,626

2,280 1,798 2,028 2,586 2,384 3,465 1,604

2,779 1,681 2,013 2,546 1,984 2,938 1,839

1,475 1,216 1,764 1,754 1,678 1,975 766

1,470 1,229 1,794 1,770 1,685 2,026 765

1,628 1,130 1,777 1,735 1,431 2,047 811

15,929

16,145

15,780

10,627

10,739

10,560

a source not in the Handbooks, and that for the last year. While the total area ﬁgures for 1895 and 1907 help to conﬁrm the consistency of coverage between the Handbooks and the Silesian address book for 1909, it also clearly points up the potential danger of relying on the data for number of properties for any kind of precise analysis. The adjustments made have reduced the gain in the area of large properties in East Prussia between c.1882 and c.1907 to just over 10 per cent, which is probably still an overstatement, but much more in line with what happened in the other provinces. The small decline in Pomerania is probably an understatement: in both provinces I was able to adjust only for under-coverage of State properties, but not for properties of other owners. Given that the data for c.1882 still likely understate the area of large estates in East Prussia and Pomerania, the two provinces taken together probably both witnessed a decline in the total area of these estates over the entire period, instead of the slight gain shown in the overall totals in the table for East Prussia and negligible decline for Pomerania. The safest general conclusion from these data would be that the area of large properties declined slightly over the period from the early 1880s to the eve of the Great War. Large properties covered the most area in absolute terms in Silesia, followed by Brandenburg, then Posen and Pomerania, East and then West Prussia, with the province of Saxony in last place. Perfect adjustment of the areas of East Prussia and Pomerania in the ﬁrst year would not have affected this rank ordering. The rank order of provinces did change after the 1890s, but only because of the reduction in area of large properties brought about by the actions of the Prussian government’s own Settlement Commission for Posen and West Prussia,⁸ as noted before. Absent this intervention in the land market, it is likely that there would have been no change at all in the rank order. ⁸ Note that the small overall decline in area in all seven provinces between c.1895 and c.1907 (about 123,000 ha) is well under a third of the Settlement Commission’s purchases of estates over its lifetime (430,000 ha), so that most of the reduction in large-estate land through the Settlement Commission was offset by growth in large estates in other provinces.

The Distribution of Large Landed Properties

61

Table 3.4. Total tax value and tax value per ha of large properties, by province, adjusted data Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Total tax value, million marks c.1882 c.1895 c.1907

Tax value per hectare (marks) c.1882 c.1895 c.1907

9.91 7.69 15.49 17.13 13.58 22.81 15.61

9.58 7.56 15.31 17.40 13.48 23.28 16.72

10.96 6.93 15.00 16.47 11.35 24.22 18.00

6.72 6.32 8.78 9.77 8.10 11.55 20.37

6.52 6.15 8.54 9.83 8.00 11.49 21.84

6.73 6.13 8.44 9.49 7.90 11.84 21.98

102.22

103.33

102.91

9.62

9.62

9.75

Table 3.4 shows the total tax value, in million marks, and average tax value (in marks per hectare) of the large properties in the seven eastern provinces. Since the land tax assessment for any given plot of land did not change during the period under review, where the value per hectare declines, that decline represents a change in composition of large properties toward a larger share of lower-taxed land (primarily forest, but also pasture, water, moorland, and untaxed land). Where the value rises, in Silesia and Saxony, the rise represents exactly the opposite change in composition, with an increase in the share of higher-valued land—ploughland and meadow—in larger properties. The adjustments to the data for East Prussia, Brandenburg, and Pomerania had very little effect on the distribution of tax value by province, and no signiﬁcant effect whatever on the provincial rank order of tax value per hectare of properties of 100 hectares or more in size. Silesia again leads the parade in terms of absolute value of land tax net yield in the large properties, but the rest of the order is changed, compared to that for simple area. Pomerania shows up in second place at the beginning, but Saxony overtakes her by c.1907. Brandenburg in fourth place, Posen in ﬁfth, East Prussia in sixth, and West Prussia in seventh do not change positions, despite the losses of land by large estates, especially in Posen. In terms of the average tax value of large estates, Saxony leads, with Silesia a distant second and Pomerania an even more distant third. Brandenburg, then Posen, then East Prussia, and ﬁnally West Prussia, bring up the rear. The differences among provinces observed in the table arise for two reasons: a difference in the quality of land and a difference in the average composition of properties (share of higher-taxed land such as ploughland and meadow compared to the share of lower-taxed land such as pasture, forest, water, and moor) across provinces. The province of Saxony stands out, with an average tax assessment per hectare more than double the average for all seven provinces taken together. In the case of the province of Saxony, the composition difference actually serves to mask,

62

The Distribution of Large Landed Properties

or offset, some of the difference from higher land quality, since larger properties in the province of Saxony contained a higher-than-average share of lower-assessed forest land (43 per cent of total area) compared to the seven-province average (37 per cent). In a previously published paper, I argued that—even a half-century after the land tax assessment was made—the assessed value for the land tax remained a remarkably good proxy for relative market value of land.⁹ At the time, I argued further that the proxy would be better, the smaller the territorial unit within which the comparison would be made. Two major inﬂuences could act to make the tax assessment a less perfect proxy: 1. Conversion of lower-taxed land (e.g. moor) into higher-taxed land (e.g. meadow) with no corresponding change in the tax assessment. This did happen to some extent, and the Prussian government gave an incentive for such land improvement in that it deliberately did not change the tax assessment when such improvements had been made, as was pointed out in Chapter 2 (n. 10). 2. Change in the relative prices of products produced on different types of land. For example, if the price of wood rose compared to the prices of ﬁeld crops, then the market value of forest land would rise, relative to the market price of ploughland. But if their relative tax assessments did not change, then using tax assessments as a proxy would tend to undervalue properties consisting primarily of forest land, compared to properties consisting primarily of ploughland. The land tax cadastre for the Prussian provinces in question was completed during the years 1861–4. By the early 1880s, then, approximately twenty years had passed, during which there was a general rise, then fall, of agricultural prices.¹⁰ Moreover, the price changes were more profound for some products (grains in particular) than for others (meat, dairy products), so that the regional impacts of the price changes as a result of existing geographical specialization in production were not uniform. For these reasons, one must always take the comparisons which follow with some reservations—with the proverbial grain of salt—since they compare regions which lay a considerable distance from

⁹ ‘The Distribution of Landed Properties by Value and Area: A Methodological Essay based on Prussian Data, 1886–1913’, Journal of Income Distribution, 3/1 (summer 1993), 101–40. Note that even though the ratio of market price to tax assessment might change over time (as e.g. with general inﬂation or with a change in the terms of trade of agriculture vis-`a-vis the rest of the economy), a property with a higher tax assessment should command a higher price than one with a lower tax assessment, whether at the beginning, end, or in the middle of the period. This is the notion of proxy for the relative market value. ¹⁰ This period consisted of, roughly, the second half of the secular rise in agricultural prices from c.1850 to c.1873, and the ﬁrst half of the great decline in farm prices from c.1873 to c.1896.

The Distribution of Large Landed Properties

63

Table 3.5. Number of properties of 100 or more hectares, their area and tax assessments, by administrative region, c.1882 Administrative region

Province

Merseburg Stralsund Magdeburg Erfurt Breslau Stettin Liegnitz Potsdam Oppeln Frankfurt/Oder Bromberg Posen K¨onigsberg Danzig Marienwerder K¨oslin Gumbinnen

Saxony Pomerania Saxony Saxony Silesia Pomerania Silesia Brandenburg Silesia Brandenburg Posen Posen East Prussia West Prussia West Prussia Pomerania East Prussia

all regions

Number of properties

Area (000 ha)

Tax value (000 marks)

Tax value per ha (marks)

749 651 672 205 1,460 845 1,059 1,037 820 1,024 1,051 1,316 1,490 638 1,049 1,055 808

319 309 366 81 682 599 627 872 666 892 667 1,011 938 368 848 846 537

7,382 6,124 6,882 1,342 10,798 6,450 6,272 7,960 5,741 7,528 5,445 8,139 7,033 2,479 5,211 4,553 2,880

23.15 19.93 18.80 16.53 15.82 10.77 10.00 9.13 8.63 8.44 8.17 8.05 7.50 6.74 6.15 5.38 5.36

15,929

10,628

102,219

9.62

one another and exhibited quite varied topography, land types, and agricultural products produced.¹¹ The seven provinces of our purview were divided into a total of seventeen administrative regions (Regierungsbezirke),¹² and if we look at the data by administrative regions, a much more varied picture emerges. In Table 3.5 I concentrate only on a single year, to reduce clutter. Since there was very little change over time—other than that brought about by the Royal Settlement Commission, especially in the region of Bromberg (Bydgoszcz)—one year is sufﬁcient to illustrate the points made in what follows. The greatest variance appears in the province of Pomerania, whose three administrative regions occupy second, sixth, and next-to-last place in the tax value per hectare ranking. The average tax value per hectare in the region of Merseburg is 2.4 times the all-region average, and 4.3 times that of the lowest region of Gumbinnen. The three regions of the province of Saxony hold three of the top four places in the ranking, and the bottom of the scale is accounted for almost exclusively by regions from the East and Northeast: the easternmost part of Pomerania, West and East Prussia, and Posen. There appears a general ¹¹ For an extended discussion of these issues, please see Ch. 8. ¹² These in turn were divided into different districts. Large properties could be found in all 259 of the rural districts c.1895 (see n. 13 below).

64

The Distribution of Large Landed Properties

gradient in tax value per hectare of large properties that runs from the south-west and south (Saxony and Silesia) to the north and north-east, which tends to follow the gradient for all properties found above. The two regions of Brandenburg (Potsdam and Frankfurt/Oder) ﬁt approximately in the middle of the pack, both geographically and in ranking terms. It should be noted that, because it has the highest share of relatively low-taxed forest land of any province (49.0 per cent), the regions of Brandenburg therefore rank lower in the table than they would if we could do the comparisons by, say, the tax value per hectare of arable land. Unfortunately, the data do not permit such a comparison for our set of large properties. As one would expect, when the data are examined by district (the table is not reproduced here, but can be found at EH.Net: www.eh.net), the range of variance becomes even more pronounced. Among the 259 rural districts,¹³ the top 10, and 15 of the top 25, positions with regard to average tax assessment per hectare are occupied by districts from the province of Saxony. With only two exceptions, these districts belong to the regions of Magdeburg and Merseburg. Of the rest of the top 25 districts, 9 are from Silesia, and one—the district of Marienburg in the Danzig administrative region—from, of all places, West Prussia, the province with the lowest average tax assessment per hectare in our provincial comparison. West Prussia’s two administrative regions, as we have seen, took up two of the last four places in the regional table. The district of Marienburg in West Prussia was unusual in that its large properties were almost entirely made up of ploughland and meadow, the two most highly taxed land types in these provinces, and they had very little forest—only nine properties had any woodland at all—or pasture, and almost no moorland or water.¹⁴ Moreover, the biggest of these 115 properties was only 401 hectares, and only 13 of them were 200 hectares or more; thus Marienburg looked far more like one’s stereotypical picture of a district in the province of Saxony than in West Prussia. This is yet another reminder that conditions in rural areas can vary a great deal within relatively short distances, and that comparisons using large geographic units such as provinces can as easily obscure the picture as illuminate it. The top ten districts have an average tax assessment per hectare that ranges from 3.72 to 4.75 times the seven-province average, while the bottom ten range from a mere 0.23 to a meagre 0.35, when compared to the seven-province ¹³ In the 1895 administrative structure of Prussia, which is the baseline for the present book, there were 259 rural districts in the seven provinces, up from 236 at the time of the land tax cadastre because of creation of new districts by splitting up old districts, sometimes combining parts from two or even three former districts. The biggest part of this change took place in the province of West Prussia in 1887. There is an excellent online source in the German-language Wikipedia that shows each district of Prussia, and—if relevant—when it was created and from what former district(s): http://de.wikipedia.org/wiki/Liste der Landkreise Preu%C3%9Fens. ¹⁴ The data show 401 ha of forest (366 of these ha in just two properties), 343 of pasture, 51 of moor, and 21 of water out of a total of 17,315 ha in the 115 properties of 100 ha or more in Marienburg.

The Distribution of Large Landed Properties

65

average. So it is fair to say that the district range runs from less than a third to well over four times the average for the seven provinces as a whole. In the bottom ten districts are found four from West Prussia, two from Posen, three from East Prussia, and one from Pomerania. The highest average land tax net yield of 46.71 marks was found in the district of Wanzleben in Magdeburg region, and the lowest—1.80 marks per hectare—came from Johannisburg district in the region of Gumbinnen.¹⁵ Thus the highest average was nearly 26 times the lowest, and close to 5 times the all-district average. In Wanzleben there were 38 properties of 100 hectares or more, encompassing 16,471 hectares, of which at least 14,596, or 89 per cent, were ploughland or meadow. The average size of these properties was 433 hectares. Contrast this to Johannisburg, where 37 properties of over 100 hectares covered almost exactly 60,000 hectares (an average size of 1622 ha), but consisted of at least 38,295 hectares of forest (64%) and only about 14,300 hectares of ploughland and meadow (24%). Everything was different about these two regions: size and land composition of properties, quality of the land, even the kind of owners: the Prussian State owned 44,600 of the 60,000 hectares in Johannisburg’s large properties—and non-noble persons owned all of the rest—while in Wanzleben nobles owned 18 per cent of the land in large properties, ‘bourgeois’ owners held 34 per cent, and the Prussian State 31 per cent. A business ﬁrm owned two of the 38 properties, the duchy of Brunswick three, and a nun’s cloister, one. An average large property in Wanzleben, despite being only slightly more than a quarter of the size of its counterpart in Johannisburg, had a tax yield nearly seven times as high. From the foregoing examples it is easy to see why regarding East Elbia as a uniform region could be fraught with peril. When one examines the share of the total land area of the district accounted for by large properties of 100 hectares or more, the situation looks very different. About 1895 this share varied from nearly 92 per cent,¹⁶ in Franzburg district, Stralsund administrative region in the province of Pomerania, to just above 8 per cent in M¨uhlhausen, Erfurt, province of Saxony. Of the 54 districts in which large properties covered over 60 per cent of the total area of the district, 42 were found in just three provinces: Posen (19), Pomerania (12), and Silesia (11). Not a single district from Saxony was to be found in this group; on the other hand, of the 14 districts with a share of large properties under 20 per cent, 9 were from the province of Saxony. Much of the previous discussion has focused on absolute numbers. Since provinces were of different sizes—Silesia was the biggest of the seven provinces, ¹⁵ Johannisburg was assigned to the new administrative region of Allenstein in the revision of 1907, which I do not use in this book, remaining with the 1895 schema. ¹⁶ The second-highest ratio was only 79%, however (Rummelsburg district, K¨oslin region, also in Pomerania).

66

The Distribution of Large Landed Properties Table 3.6. Share of area and tax value of properties of 100 hectares or more in area and tax value of all land outside of cities, by province, adjusted data (%) Province

Share of total area c.1882 c.1895

c.1907

Share of total tax value c.1882 c.1895 c.1907

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

41.4 49.8 48.5 58.0 61.0 50.7 33.9

41.2 50.3 48.9 61.8 61.2 52.0 33.8

45.7 46.3 48.9 61.8 52.0 52.6 35.8

40.1 41.8 44.7 62.8 60.8 42.7 31.5

38.7 41.1 43.9 64.9 60.4 43.7 33.8

44.3 37.6 43.2 61.8 50.8 45.3 36.4

total

49.2

50.0

49.4

44.3

44.9

44.8

covering 40.3 thousand square kilometres, with Brandenburg close behind at 39.8; Saxony and West Prussia were the smallest, with 25.2 and 25.5 thousand, respectively—it is necessary also to look at relative numbers, to ﬁnd out where large estates were most or least important. Table 3.6 shows the share of large properties in both area and tax value of rural land, that is, land outside of cities. By the measures of both area and tax value, Posen and Pomerania were the bastions of the large estates among the seven provinces, although the large estates of Posen lost ground (literally) by the end of the period. In terms of share of these estates in total rural area, West Prussia and Brandenburg were the average provinces, but in terms of share in tax value, the average province would be either Brandenburg or Silesia. Interestingly enough, the province of Saxony brought up the rear, and by a good margin, in both categories—a result primarily of the much smaller share of total area to be found in large estates in that province. So, while Silesian large estates were most numerous, and had the largest area and the greatest overall tax assessment, in relative terms the real bastions of large estates in the eastern provinces by the above measures were Posen and Pomerania, but Posen’s position weakened by the end of the period. Silesia, the home of the great landed magnates, in fact differed little from the average when the share of large estates in area and tax value is considered, irrespective of the year in question. East and West Prussia were at or below average in both categories in all three years, and Saxony occupied a position very different—and very distant—from all the others. In the foregoing we have seen that the seven East Elbian provinces of Prussia present a varied picture, but one that is quite stable over time. This stability would have been even more evident, but for the major intervention of the Prussian government in the land market, especially in the province of Posen, and to a lesser extent in West Prussia. This theme of cross-sectional variety but stability over time is one that will recur many times in what follows in this book.

The Distribution of Large Landed Properties

67

3 . 5 . T H E S I Z E O F L A RG E P RO PE RT I E S

3.5.1. Traditional View 1: Distribution by Total Area of Properties Area in hectares, despite the inherent internal inconsistencies in such a unit,¹⁷ has been the traditional measure used for the size distribution of landholdings. Table 3.7 shows the distribution by that measure, using the traditional size categories, for our three ‘snapshot’ years. Given the reservations stated above about possible differences in the deﬁnition of what constituted a property, and especially given the probably incomplete nature of the additions of Prussian State forests to the databases, Table 3.7 offers little more than evidence of a considerable stability in the area of large properties in the seven core provinces of Prussia. Within this overall stability, the apparent gains in area of properties between 2000 and 10,000 hectares in size at the expense of all smaller categories except the 100–199 hectare group could be simply the result of deﬁnitional differences in different editions of the sources; on the other hand, the apparent gains came exactly in those size categories in which the Prussian state was the principal landowner, so could also be the result of early under-enumeration of State forests. Across provinces, the differences in size distribution are marked enough to allow some conclusions. Tables 3.8, 3.9, and 3.10 present the provincial distributions, arranged by region. It is easier to see the differences among the Table 3.7. Size distribution of properties 100 hectares or larger, all seven provinces (areas in 000 ha) Size in hectares

100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 + total

c.1882 Number

Area

c.1895 Number

Area

c.1907 Number

Area

3,365 6,562 3,941 1,375 428 216 42

495 2,178 2,734 1,822 1,337 1,444 617

3,434 6,605 3,997 1,402 438 230 39

504 2,188 2,766 1,857 1,348 1,518 558

3,761 6,285 3,655 1,307 500 240 32

553 2,071 2,524 1,734 1,610 1,563 505

15,929

10,628

16,145

10,739

15,780

10,560

¹⁷ A hectare of swamp is far less of an economic asset than a hectare of vineyard, to cite an extreme example. Since most comparisons of size of property are motivated by a concern for concentration or inequality in the distribution of wealth and power, some measure of value of the properties in question would be more appropriate. I deal with this question later in Sect. 3.5.3.

68

The Distribution of Large Landed Properties

Table 3.8. Size distribution of properties 100 hectares or larger, Baltic provinces (areas in 000 ha) Size in hectares

c.1882 Number

Area

c.1895 Number

Area

c.1907 Number

Area

East Prussia 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

561 1,004 473 154 59 39 8

82.6 323.5 321.2 204.7 193.7 258.3 90.9

564 994 460 154 59 42 7

83.2 321.2 312.4 204 192.9 276.5 80.1

772 1,281 451 145 73 56 1

121.3 405.2 305.3 188.1 244.8 353.4 10.2

total

2,298

1,474.8

2,280

1,470.3

2,779

1,628.4

349 659 462 132 43 39 3

51.0 224.0 321.1 175.8 133.2 268.8 42.1

428 684 464 133 48 38 3

62.1 228.6 320.4 179.7 140.9 262.6 34.4

505 592 384 109 52 37 2

72.5 196.0 262.5 150.2 169.3 256.4 23.5

1,687

1,215.9

1,798

1,228.7

1,681

1,130.5

352 971 688 243 78 28 7

52.5 328.3 479 320 229 182 87.3

274 994 889 345 68 15 1

42.4 349.2 622.7 446.6 210.7 86.3 11.9

300 973 859 324 75 15 0

45.2 339.4 599.1 422.2 241.5 87.7 0

2,367

1,678.1

2,586

1,769.8

2,546

1,735.2

West Prussia 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more total Pomerania 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more total

regions when the data of these tables are presented in graphical form, for which please refer to Figs. 3.1 and 3.2. In absolute terms, there were both larger numbers and more area of large properties in the Baltic provinces, but largely because that region contained three provinces, and the others only two. Similarly, in absolute terms the lowest numbers and the least area were to be found in the central-south-west region, in similar fashion mainly because the province of Saxony was the smallest of the seven provinces.

The Distribution of Large Landed Properties

69

Table 3.9. Size distribution of properties 100 hectares or larger, central and south-west provinces (areas in 000 ha) Size in hectares

Brandenburg 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000+ total Province of Saxony 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000+ total

c.1882 Number

Area

c.1895 Number

Area

c.1907 Number

Area

278 763 618 260 87 49 6

40.5 258.6 439 347.9 272.7 325.8 79.2

251 739 635 256 86 54 7

36.8 251.1 451.4 344.4 273.2 345.1 91.9

272 712 626 254 85 58 6

40.5 242 443.8 342.2 267.7 358.8 82.3

2,061

1,763.7

2,028

1,794

2,013

1,777.5

647 606 225 100 38 9 1

90.3 193.1 153.6 136.4 120.6 57.9 14.1

600 624 235 97 38 9 1

84.9 199.9 159.4 132.2 119.8 58.1 11

789 659 239 108 33 10 1

109.5 210.9 163.3 144.2 105.4 63.8 14.1

1,626

766.1

1,604

765.3

1,839

811.1

In this case, the relative distributions by region are more interesting: the two biggest size categories rose in relative importance as one moved south and southeast through these provinces, while at the same time the Baltic provinces and the east-south-east provinces had similar shares of area accounted for by properties in the 200–999 hectare size group. This share declined from c.1895 to c.1907 in the east-south-east, but we have seen earlier that this apparent decline is more a result of differing deﬁnitions of a property as between the Handbooks and the Silesian property address book used as the source for 1909. The markedly larger relative share of the very largest properties in the east-south-east region is, of course, the result of Silesia’s being the home of most of the great landed magnates in Prussia.

3.5.2. Traditional View 2: Distribution by Total Area Owned One of the points made in Chapter 1 is that the very ﬁrst scholar to use the Handbooks as a source, Johannes Conrad, recognized the problem of relying on the property as the unit of analysis, and hence chose to present all of his size distributions in terms of the total area owned by a given owner. By using this

70

The Distribution of Large Landed Properties

Table 3.10. Size distribution of properties 100 hectares or larger, east and south-east provinces (areas in 000 ha) Size in hectares

Posen 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

c.1882 Number

Area

c.1895 Number

Area

c.1907 Number

Area

352 971 688 243 78 28 7

52.5 328.3 479.0 320.0 229.0 182.0 87.3

371 967 678 253 75 36 4

54.9 326.3 469.4 333.1 214.6 237.7 49.3

316 793 565 207 68 31 4

47.5 267.8 390.2 271.7 200.7 203.4 49.4

total

2,367

1,678.1

2,384

1,685.3

1,984

1,430.8

Silesia 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

911 1,572 601 146 57 36 16

136.4 504.1 407.5 194.8 183.6 256.6 292.1

946 1,603 636 164 64 36 16

139.5 512.1 430.1 216.9 196.3 251.5 279.3

807 1,275 531 160 114 33 18

116.3 409.8 359.9 215.1 380.9 239.4 325.2

total

3,339

1,975.2

3,465

2,025.7

2,938

2,046.5

Fig. 3.1. Total area of large properties by size category and region (000 ha).

The Distribution of Large Landed Properties

71

Fig. 3.2. Total area of large properites by size category and region (%).

unit (Gesamteigentum in German), Conrad avoided almost all problems arising from inconsistencies in his sources in deﬁning what constituted a property.¹⁸ Using the same size categories as before, the distributions look as presented in Table 3.11. In Panel A the Prussian State is included as owner, in Panel B, excluded. Although the number of owners shown is certainly greater than the actual number in the Handbooks because of the identiﬁcation problem detailed in Chapter 2, there is no reason to believe that the problem got either better or worse over time. Consequently, trends observed in the data of Table 3.11 are likely to be actual trends, except for the probability that the increase in number of owners in the 100–199 hectare range may be mostly accounted for by an improvement in the Handbooks’ coverage of smaller properties in the later years. The table shows that the very largest owners gained in area in each period; the gains of the largest owner of all, the Prussian state (which, as we saw earlier, may be overstated because its early totals may be understated) accounted for just over half of the gain in area between c.1882 and c.1895, and more than all of the gain in Panel A between c.1895 and c.1907, since Panel B shows a slight loss of area by non-State owners: the other owners in the top group gained about 106,000 hectares between the ﬁrst two years, but lost over 50,000 between the last two years. The ‘middling’ owners, in the 500 to 1999 hectare range, were overall losers, partly—but only partly—because the Royal Settlement Commission and other land parcelling agencies purchased primarily in this size range. ¹⁸ It is still theoretically possible that an ownership unit made up, say, of 200 ha in a 120-ha and an 80-ha block might be counted in one source as one property, but in another as two. In this case, the 80 ha would not enter into the calculation because under 100, so the choice of total ownership of properties over 100 ha as the unit of analysis still does not avoid a problem such as this at the lower margin.

72

The Distribution of Large Landed Properties

Table 3.11. Size distribution of total ownership area in properties 100 hectares or larger, all seven provinces (areas in 000 ha) Area owned (ha)

c.1882 Owners

Area

c.1895 Owners

A: including properties owned by the Prussian State 100–199 2,417 351 2,473 200–499 4,026 1,334 4,125 500–999 2,596 1,819 2,563 1000–1999 1,217 1,666 1,192 2000–4999 499 1,466 525 5000–9999 121 799 102 10,000 or more 47 3,193 53

Area

c.1907 Owners

Area

358 1,360 1,796 1,613 1,533 672 3,407

3,027 4,059 2,238 1,083 479 109 47

441 1,317 1,565 1,473 1,430 731 3,603

11,033

10,739

11,042

10,560

B: excluding properties owned by the Prussian State 100–199 2,417 351 2,473 200–499 4,026 1,334 4,125 500–999 2,596 1,819 2,563 1000–1999 1,217 1,666 1,192 2000–4999 499 1,466 525 5000–9999 121 799 102 10,000 or more 46 1,004 52

358 1,360 1,796 1,613 1,533 672 1,110

3,027 4,059 2,238 1,083 479 109 46

441 1,317 1,565 1,473 1,430 731 1,057

total

8,442

11,041

8,014

total

10,923

10,922

10,628

8,439

11,032

The size distribution by area owned (Table 3.11) differs considerably from the size distribution of properties (Table 3.7), even though—of course—they both add up to the same amount of land. Fig. 3.3 shows the differences. The size distribution in terms of area owned is considerably more ‘top heavy’ than that of properties. There is not much difference in the 100–199 hectare range, but in the 200–499 hectare range the area of properties is considerably larger than the area of ownership of 200–499 hectares, indicating that properties in this size range often made up a part of a larger ownership complex. This phenomenon is even more striking in the 500–999 hectare range. In the categories above 5000 hectares, the top range of the ownership distribution, and the next-to-top of the property size distribution, are dominated by properties of the Prussian State. Hence the very large, and opposite, differences in the 5000–9999 hectare and the over-10,000 hectare size categories. Had the limitations of the Handbooks not forced me to deﬁne the manor as a property, but allowed me in all cases to use the constituent parts of the manors as principal properties in themselves,¹⁹ rather than subunits, the difference between ¹⁹ Unfortunately there were far too many cases for which the details of the constituent parts were not given, and many in which even the list of names of properties making up the manor was incomplete or missing entirely.

The Distribution of Large Landed Properties

73

Fig. 3.3. Property size vs. total area owned.

ownership and property data would have been even more dramatic. It would have shown that properties in the 200 to 999 hectare size range were to an even greater degree parts of larger ownership complexes. The presence or absence of the Prussian State in the ownership statistics makes an enormous difference, as Table 3.12 demonstrates. It is easy to see from these data why Conrad chose to leave the Prussian State out of his calculations, based on total landownership by each owner. I shall do the same in most of the next two chapters, which look much more closely at groups of owners and even individual owners of large properties. This section can therefore serve as an introduction to Chapters 4 and 5.

3.5.3. A Non-Traditional View 1: Distribution by Tax Value Already in the ﬁrst chapter we found that one of the most startling results of Klaus Hess’s research into landholding in Prussia was that, when the measure of comparison was the value of the land and not merely its geographic extent, the western Prussian landlords were fully the equal of their eastern Prussian counterparts. Since the taxable value of a property is a much better measure of its economic importance than its mere area, it behoves us here to follow Hess’s lead and examine the distribution of large properties according to their taxable value. We have seen above that the average tax value per hectare for the large properties of the eastern provinces was very close to 10 marks. Using this as a comparison benchmark, let us look at the size distribution of properties according to their tax assessment (GSRE), with a comparison to their distribution by total

74

The Distribution of Large Landed Properties Table 3.12. Effect of including or excluding Prussian State as landowner Area measure

Total for all large owners (000 ha)

Average per owner (ha)

Total for owners of 10,000+ ha (000 ha) Average for owners of 10,000 + ha (ha)

Year

With Prussian state Included Excluded

1882 1895 1907 1882 1895 1907 1882 1895 1907 1882 1895 1907

10,628 10,739 10,560 973 973 956 3,193 3,407 3,603 67,928 64,286 76,652

8,439 8,442 8,013 773 765 726 1,004 1,110 1,057 21,831 21,352 22,968

area, in Table 3.13. Continuing to use the 1882 database as reference point, we can examine the distribution of total area and total tax value in the large properties, in the ﬁrst instance (in Fig. 3.4) employing the standard technique of arranging the data into deciles, each decile containing 10 per cent of the total number of properties.²⁰ Fig. 3.4 shows that the total area of large properties was more unequally distributed than was their total tax value. The second through ninth area deciles each contain a smaller share of total area than the second through ninth tax value deciles do of total tax value, while the respective tenth deciles contain 44 per cent of total area but ‘only’ 35 per cent of total tax value. The differences in Table 3.13. Size distribution comparison for large landed properties, c.1882 Size category (ha)

100–199 200–499 500–999 1,000–1,999 2,000–4,999 5,000–9,999 10,000 or more total

Properties

Total area (000 ha)

3,365 6,562 3,941 1,375 428 216 42

495 2,178 2,734 1,822 1,337 1,444 617

15,929

10,628

Size category (marks) Under 1,000 1,000–1,999 2,000–4,999 5,000–9,999 10,000–19,999 20,000–49,999 50,000–99.999 100,000 or more total

Properties

Tax value (000 marks)

916 2,119 5,813 4,487 2,004 529 49 12

604 3,170 19,834 31,597 26,930 14,749 3,195 2,103

15,929

102,218

²⁰ Because the number of properties was not evenly divisible by 10, the ﬁrst nine deciles contain 1593 properties, the last one, 1592.

The Distribution of Large Landed Properties

75

Fig. 3.4. Distribution of total area and tax value of large properties by decile, c.1882.

the two distributions arise primarily from the different composition of properties in different size categories, with the shares of lower-taxed land (pasture, moor, water, but particularly forest) in the total area of properties increasing with increasing total size of property. Two charts—Fig. 3. 5 showing ﬁgures for area and tax value in the area deciles, Fig. 3.6 showing the numbers in the tax value deciles—show the differences in distributions even more clearly. The smallest property in the tenth area decile contained 1160 hectares with 9495 marks of land tax net yield; the lowest property in the tenth tax value decile had a tax value of 12,859 marks and covered 965 hectares. There were many properties in this decile with even less total area. The least was found in Adalbert

Fig. 3.5. Distribution of total area and tax value of large properties, by area deciles, c.1882

76

The Distribution of Large Landed Properties

Fig. 3.6. Distribution of total area and tax value of large properties by tax deciles, c.1882.

Prinz’s property in Schwittersdorf in the Mansfelder Lake District, Merseburg region, province of Saxony, with only 192 hectares (all ploughland) and 14,998 marks of taxable net yield. Of the 1592 properties in each respective tenth decile, fewer than half (739) appeared in both the tenth decile of area and the tenth decile of tax value. Of these 739 properties 49 per cent of the area, and 47 per cent of the tax value, were found in just two provinces, Brandenburg and Silesia. The Prussian State owned 254 of these properties, accounting for 42 per cent of their total area and 33 per cent of their total land tax net yield. The smallest of these 739 properties by area was 1160 hectares (with 19,140 marks of land tax net yield) and the smallest by tax value had 12,861 marks land tax net yield (for 2347 ha).²¹

3.5.4. A Non-Traditional View 2: Logarithmic Distributions One of the problems with the traditional view of size distribution is that the intervals along which the distribution is measured are not in any way equal, and therefore do not provide a picture of the frequency distribution in a statistical sense. Equal arithmetic intervals (e.g. 100 ha) might make sense in some cases, but in our data show a frequency distribution with its maximum value at the far left, declining steadily from there. Indeed, some of the intervals at the top end are empty, and the non-empty ones mostly have a frequency of one. Shortly to the right of the mean value, on any reasonable size sheet of paper, the graph becomes indistinguishable from the horizontal axis. A frequency distribution based on equal arithmetic intervals is, to all intents and purposes, useless. ²¹ The two properties were, respectively, Rohrbeck in K¨onigsberg (Neumark) district of Brandenburg, owned by the ‘heirs of von Gerlach’, and Heinrichsdorf in Neustettin district of Pomerania, owned by Heinrich Leopold von Arnim, resident on the property.

The Distribution of Large Landed Properties

77

800

Frequency

600

400

200 Mean = 2.61 Std. Dev. = 0.369 N = 15,908

0 2.00

2.50

3.00

3.50

4.00

4.50

5.00

Log Area

Fig. 3.7. Frequency distribution of logarithm of total area.

Equal logarithmic intervals, however, do make good sense, since they maintain a given ratio of values of the variable in question: If we used logarithms to the base 10,²² and an interval of one, for example, then the interval boundaries would occur at values of 0, 1, 2, 3, etc. Each interval boundary would therefore be 10 times the value of the previous boundary (100 = 1, 101 = 10, 102 = 100, 103 = 1000, and so forth). Let us look ﬁrst at the frequency distribution of large properties according to the logarithm of their total area in hectares (Fig. 3.7). The curve of the normal distribution is added to the chart for reference. Above the mean, the logarithmic distribution of area comes close to a normal distribution, with some ‘heaping’ at the very highest values, but below the mean it deviates more and more from the normal distribution as the size of property decreases. Because the distribution is truncated artiﬁcially at 100 (= 102 ) hectares, we cannot know how the distribution might look below that value. Regardless of the nearness to, or distance from, a normal curve, the logarithmic frequency distribution gives an easily-interpretable picture of the way in which these properties were distributed c.1882. The frequency distribution of tax values (GSRE) presents a more interesting picture, since the units in which the variable is measured (marks) are exactly ²² We could also use natural logarithms. For the datasets of large properties, I did all the calculations with both kinds of logarithms; the shapes of the frequency distributions were the same whether natural or base: 10 logarithms were used, so I present the base-10 logarithm results here because I ﬁnd them simpler to understand and to explain.

78

The Distribution of Large Landed Properties 1200

Frequency

1000

800

600

400

200

Mean = 3.63 Std. Dev. = 0.389 N = 15,908

0 1.00

2.00

3.00 4.00 Log GSRE

5.00

6.00

Fig. 3.8. Frequency distribution of logarithm of land tax net yield.

comparable; moreover, in principle at least, a mark of land tax net yield was to be the same, regardless of geographic location or size of the property in question.²³ Thus we are not comparing hectares of swampland with hectares of rich ploughland, or hectares of forest with hectares of hayﬁeld, as if these were somehow equivalent. The total area in hectares measure is therefore full of non-comparable apples and oranges, but the land tax net yield comes very close to being a uni-fruit. Fig. 3.8 shows the frequency distribution of land tax net yield by logarithmic size c.1882. In the diagram there is an extremely close correspondence of our logarithmic data to the normal curve, so that the land tax net yield exhibits a near-perfect lognormal distribution.

3.5.5. Summary Around 1882, nearly 16,000 properties 100 hectares or larger in size covered 10.6 million hectares, or 47 per cent of the total area of the seven eastern provinces of Prussia. Of the area in these large properties, physical persons owned 8.1 million hectares, or 78 per cent, and the Prussian State owned at least a further 2.2 million hectares (20+ per cent of the total). These physical persons owned 83 per cent of the tax value of these large properties, and the Prussian ²³ In Ch. 9, Sect. 9.3, I address speciﬁcally the question of whether the land tax assessment was biased to favour one or more types of land or groups of owners over other land types or owners.

The Distribution of Large Landed Properties

79

Table 3.14. Distribution of large properties by size and owner (numbers) Size category (ha)

Numbers of properties owned c.1882 by Nobility and royalty

Nonnoble persons

Prussian State

All other owners

100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

751 2,583 1,998 898 214 47 29

2,392 3,408 1,563 354 51 2 0

72 325 262 89 151 165 12

150 246 118 34 12 2 1

total

6,520

7,770

1,076

563

Share owned by All owners

Physical persons (%)

Prussian State (%)

3,365 6,562 3,941 1,375 428 216 42

93.4 91.3 90.4 91.1 61.9 22.7 69.0

2.1 5.0 6.6 6.5 35.3 76.4 28.6

15,929

89.7

6.8

State 13 per cent. The discrepancy between the area share and the tax value share of the State was the result of the State’s owning primarily forest land, with a smaller tax assessment per hectare than ploughland or meadow, the principal components of the other properties. The holdings of the next biggest category of owner, communities and communal groups,²⁴ did not even reach 1.5 per cent of either area or tax value; beyond that, the other owner categories held only trivial shares of the land. Indeed, all other categories of landowner besides physical persons and the State, when taken together, owned under 3 per cent of the total area, and almost exactly 4 per cent of the tax value, of large properties. Personal landownership was therefore overwhelmingly dominant in the case of large properties, with the Prussian State becoming a ‘big player’ only when the size of property reached at least 2000 hectares. Tables 3.14, 3.15, and 3.16 provide details of the above-described distribution of types of owners by size of property for the seven eastern provinces taken as a whole c.1882. Among individual landowners, the nobility²⁵ was predominant: 3620 noble or royal owners²⁶ owned nearly half of the land and almost exactly half of ²⁴ This category includes not only communal groups per se, but also all lower levels of government administration, from villages on up through provinces. ²⁵ Unless otherwise qualiﬁed, I use the term ‘nobility’ also to include royalty. ²⁶ Each unique owner or unique combination of owners counted as one owner (see discussion in Sect. 2.3.1). Thus if two brothers owned property separately, but also owned at least one property together, they would constitute three owners. An owner that was a group of persons, such as ‘heirs of Benecke von Gr¨oditzberg’, also counted as one owner. Further, when two owners appeared to be the same person, but could not deﬁnitely be identiﬁed as such, I had to treat them as two different owners; each received a separate owner number. Thus whether the ﬁgure of 3620 noble owners might be an over- or underestimate of the actual number of noble persons who owned land is indeterminate.

80

The Distribution of Large Landed Properties

Table 3.15. Distribution of large properties by size and owner (total area) Size category (ha)

Total area (in 000 ha) of properties c.1882 owned by Nobility NonPrussian All other All and noble State owners owners royalty persons

Share owned by Physical Prussian persons State (%) (%)

100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

117.3 882.7 1409.3 1186.5 611.3 347.9 449.7

344.8 1102.4 1065.7 467.3 141.0 14.0 0.0

11.4 114.5 177.8 124.2 550.4 1070.2 140

21.5 78.5 81.0 43.8 34.7 12.1 27.8

93.4 91.1 90.5 90.8 56.3 25.1 72.8

2.3 5.3 6.5 6.8 41.2 74.1 22.7

total

5004.6

3135.2

2188.4

299.3

10,627.5 76.6

20.6

494.9 2178.1 2733.7 1821.8 1337.3 1444.3 617.4

the tax value in large properties, while 7017 non-noble persons²⁷ owned just under a third of the area and just over a third of the tax value. So the picture of East Elbia as the land of the Junker landlords is true in the sense that these persons owned more land than any other category of owner, but non-noble landowners taken together—physical and juridical persons—owned a greater share of large estates than did the Junkers. And when we take into account that the nobility tended to own very few properties smaller Table 3.16. Distribution of large properties by size and owner (total tax value) Size category Total tax value of properties (000 marks) c.1882 owned by Share owned by (ha) Nobility NonPrussian All other All Physical Prussian and noble State owners owners persons State royalty persons (%) (%) 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

2,187 13,374 15,886 9,769 4,097 2,184 2,492

6,540 14,091 9,910 3,166 742 74 0

223 2,174 3,148 1,392 2,796 3,442 363

493 1,584 1,228 344 337 62 120

9,444 31,223 30,172 14,671 7,972 5,762 2,975

92.4 88.0 85.5 88.2 60.7 39.2 83.8

2.4 7.0 10.4 9.5 35.1 59.7 12.2

total

49,989

34,523

13,538

4,169

102,219

82.7

13.2

²⁷ Here the number of owners is almost certainly an overestimate of the actual number of non-noble persons who owned 100 ha or more of land. Because, typically, the sources contained much less information about these owners (often only a surname), it must be the case that many people who owned multiple properties have been counted multiple times as separate owners, simply because they could not be identiﬁed as one and the same person.

The Distribution of Large Landed Properties

81

than 100 hectares,²⁸ the picture of their dominance in the overall landownership structure begins to fade considerably. It would not be unreasonable to assume that their holdings of properties under 100 hectares in size were negligible, which implies that they owned somewhat less than a quarter of all the land in these seven provinces. While that is still a signiﬁcant amount, it by no means approaches the level that the predominant folklore assigns to them. If we examine the three main categories of landowners, we notice opposite tendencies in the structure of ownership for the two groups of physical persons: as the size of properties becomes larger, non-noble individuals become ever rarer as their owners, while nobility and royalty become more important. The aristocrats’ trend is broken in the 2000–9999 hectare range by the emergence of the Prussian state as the most important owner in this size category. Note (in Table 3.17) that by the time we cross the 5000 hectare frontier—into the area of the commonly accepted deﬁnition of a latifundium²⁹—nobles and the State are almost the only owners³⁰ to be found. To some extent, of course, the distribution of land by size of property will vary with varying deﬁnitions of what constitutes a property.³¹ A less arbitrary distribution results when we look at the distribution of land by the amount owned by each owner.³² When we examine this concentration of ownership, the picture intensiﬁes, as Table 3.18 reveals. In Table 3.18 it is clear that the non-noble landowners of 100 hectares or more were a smaller-scale lot than the others: they tended very strongly to own under 1000 hectares each—93 per cent of their number and 72 per cent of their land fell into the 100- to 999-hectare range. Contrast ²⁸ Of the 3906 principal properties under 100 hectares in size listed in the Handbooks c.1882 (clearly not a random sample, most likely heavily biased toward the inclusion of noble-owned properties), only 326 were owned by nobility or royalty, with a total area of 19,436 ha. These numbers represented 8.3% of the number, and 8.9% of the total area of all 3906 properties, far smaller than the share of nobility and royalty in the number and area of properties of 100 ha or more (41 and 48% respectively). ²⁹ Both Conrad and H¨abich chose this number. Conrad, ‘Die Latifundien’, H¨abich, Deutsche Latifundien. While almost completely arbitrary, it nevertheless provides for comparison to other works in the literature. ³⁰ Only ﬁve of the 153 properties of this size did not belong to either the nobility or the State: three of them belonged to cities: G¨orlitz and Guben in Brandenburg owned 27,824 and 5369 ha respectively—nearly all forest—as did Sprottau in Silesia (6729 ha). The Jaff´e brothers owned the 6373-ha estate of Hammerstein (5751 ha of forest) in West Prussia and August Rehder and Hermann Schwenke owned the estate of Stolzenburg in Pomerania (7676 ha, of which 4716 were forest). ³¹ Cf. Ch. 1, pp. 16–17; Ch. 2, p. 41; and Ch. 5, p. 126. ³² As noted before, this is still subject to problems arising from the identiﬁcation of whether two owners differently listed are the same person, or whether two owners with the same name are the same or different persons.

82

The Distribution of Large Landed Properties

Table 3.17. Size of property and the share of the three largest owner categories in total ownership c.1882 (%) Size of property (ha)

Nobility and royalty No. Area Tax value

Non-noble persons No. Area Tax value

Prussian State No. Area

100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

22.3 39.4 50.7 65.3 50.0 21.8 69.0

23.7 40.5 51.6 65.1 45.7 24.1 72.8

23.2 42.8 52.7 66.6 51.4 37.9 83.8

71.1 51.9 39.7 25.7 11.9 0.9 0.0

69.7 50.6 39.0 25.6 10.5 1.0 0.0

69.3 45.1 32.8 21.6 9.3 1.3 0.0

2.1 5.0 6.6 6.5 35.3 76.4 28.6

2.3 5.3 6.5 6.8 41.2 74.1 22.7

2.4 7.0 10.4 9.5 35.1 59.7 12.2

overall

40.9

47.1

48.9

48.8

29.5

33.8

6.8

20.6

13.2

Tax value

Table 3.18. Concentration of ownership: total area owned by owner category, c.1882 (areas in 000 ha) Hectares per owner

Noble/royal No. of owners Area

‘Bourgeois’ No. of owners Area

Prussian State No. of owners Area

All other owners No. of owners Area

100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000 or more

243 1,011 1,067 764 387 104 44

38 355 766 1,061 1,138 683 963

2,112 2,917 1,480 413 85 9 1

304 948 1,020 547 243 63 11

0 0 0 0 0 0 1

0 0 0 0 0 0 2,188

62 98 49 40 27 8 1

9 31 33 58 85 53 29

total

3,620

5,005

7,017

3,135

1

2,188

285

299

this to the nobles and royals, only 23 per cent of whose land was accounted for by people owning under 1000 hectares. Comparing the two kinds of distribution of ownership of large properties presented in the previous tables, the Lorenz curves of Fig. 3.9 show what is to be expected: because some owners owned more than one property, the distribution of area owned by owner is signiﬁcantly more unequal than that of area by property size. Again, remember that the ‘owners’ curve understates the degree of inequality because the number of owners is certainly overstated, as pointed out above. The two curves should lie even farther apart than they do.

The Distribution of Large Landed Properties

83

Fig. 3.9. Lorenz curves of total area in large properties, by total number of properties or of owners.

3.5.6. Distribution in the Provinces The overall picture masks considerable variation among the seven provinces. The relative picture from Fig. 3.10 shows the nobility to have been most predominant among landowners in the provinces of Pomerania, Posen, and Silesia, where they held around 60 per cent of both the area and the taxable value of large properties. In the other four provinces they held half or less—in West Prussia under a third—of the area and tax value of the larger properties.

Fig. 3.10. Share of large properties by owner type by province, c.1882. A = area; T = tax.

84

The Distribution of Large Landed Properties

In this chart and those that follow, I have arranged the provinces in three bands from west to east: Pomerania, West Prussia, and East Prussia along the Baltic coast in the north; province of Saxony, Brandenburg, and Posen in the middle; and Silesia in the south-east. The absolute amounts of land held by the large landlords also varied considerably by province. Here both the size of the province and the share of landholding played a role, and we see in Fig. 3.11 that in terms of the numbers of hectares in large properties, Silesia led the way, with Posen and Brandenburg not too far behind. The province of Saxony showed by far the least land in large properties, less than half that found in Pomerania, Brandenburg, Posen, or Silesia. But when we compare the tax value, this disadvantage almost completely disappears. Compared to the tax value for the province of Saxony, only Silesia had signiﬁcantly more value to be found in large properties, and Pomerania somewhat more, as Fig. 3.12 shows. The difference seen there is, if anything, muted, since the province of Saxony—although blessed with some of the richest land in all of Germany—showed a somewhat smaller share of ploughland and meadow in total area of her large properties than the all-province average. The foregoing charts make it clear that Silesia was the bastion of the great noble landlords in the seven ‘core provinces’. The nobility owned more land there than anywhere else, and land in Silesia was in general more valuable than the land anywhere except in the province of Saxony.³³ If there can be said to be

Fig. 3.11. Area in large properties by owner type, by province, c.1882. ³³ This does not take into account the vast mineral wealth of some of the Silesian properties, the basis of the great fortunes of such families as the Henckels von Donnersmarck and the Counts Schaffgotsch.

The Distribution of Large Landed Properties

85

Fig. 3.12. Tax value in large properties by owner type, by province, c.1882.

a bastion of the bourgeois large landowner, it was—surprisingly enough—in the provinces of East and West Prussia, where non-noble persons had more, and more valuable, land than any other category of owner, including the nobility. The Prussian State owned more land in Brandenburg than in any other province, but only in Saxony and Brandenburg did it own more hectares than non-noble persons did. The value of the State’s landholdings, however, was less than that of either of the two classes of physical persons in all provinces, again because the state’s land was mostly forest (at least 1.64 million of its 2.19 million hectares, or 75 per cent). These considerations of the regional and size distribution of various groups of owners c.1882 can serve as the introduction to the next chapter, which discusses ownership by different groups, and even some individuals in those groups, in much more detail.

4 Ownership of Land by Rank or Class of Owner 4 . 1 . P R EV I EW Rank has its privileges, but tax money buys more land: in this chapter I divide landowners into classes of physical and juridical persons (inanimate entities such as corporations, churches, or banks), and differentiate the physical persons by rank, from non-noble up to royal. Physical persons owned about 80 per cent of the land in large properties, and the Prussian State nearly all the rest, leaving only some 3 or 4 per cent in the hands of other types of legal persons. The amount of land in large properties owned by physical persons (of which persons of noble or royal rank owned about three-ﬁfths) shrank slightly over the period. There appears to be little or no evidence of diversion of industrial or commercial wealth into the purchase of land in East Elbia, despite the rise in land prices over the period. Generally, the higher the rank of the person, the larger the average landholding, although most of the apparent absolute gain in land held by princes and dukes was the result of the elevation of a few counts with large holdings to the rank of prince. Most of the ‘ruling houses’ of German sovereigns owned land in East Elbia, but their importance shrank over the period, and was generally outshone by the holdings of the Hohenzollern family. Although the landholdings of counts as a group were artiﬁcially reduced simply by the promotion of some counts to higher rank, the lower nobility—barons and untitled nobles—suffered real losses in landholdings, especially in the latter half of the period, representing a net sell-off. Non-noble persons also saw shrinkage in their holdings of large estates, and the data used in this chapter actually understate their losses. Among the nobility, the high nobility quite literally ‘held their ground’ as the period progressed, but the lesser nobility were signiﬁcant losers in terms of overall landownership. The gap between the higher and lower nobility in landownership was much wider than the gap between the nobility and the bourgeoisie. Military ofﬁcers were important land owners, especially among the nobility, where ofﬁcers owned from one-quarter to almost one-third of all large properties owned by the group, depending on the measure used and the year examined. Noble ofﬁcers owned about three times as much land, on average, as their

Land Ownership by Rank or Class of Owner

87

non-noble comrades, and the gap between the two groups grew in both relative and absolute terms as the period progressed. Part and parcel of this was an increase in the relative position of the high nobility in the ranks of ofﬁcer-owners. Among legal persons, the Prussian State appeared to have gained considerable land over the period, even after the adjustments made to its holdings in Chapter 2. A large share of this gain was the result of purchase from physical persons, mostly non-noble owners, but on the other hand an indeterminable share appeared also to be simply an improvement in coverage of State properties in the underlying data sources. The relative size of these two types of gain varied considerably by province. Other German and foreign states held only minuscule amounts of land, relative to the Prussian State. Among the other (quite unimportant) legal persons as owners of larger properties, communities and business ﬁrms stood out, with the latter posting large relative, but small absolute, gains over the period. The majority of the land owned by religious bodies was owned by monasteries and convents; the majority of land in the hands of all other categories of owners was accounted for by hospitals, schools, and the University of Greifswald.

4 . 2 . I N T RO D U C T I O N The Prussian State was overwhelmingly the largest landowner in Prussia, and it dominates any distribution of landownership by category of owner—to the extent that including the State in such distributions tends to obscure what was happening with other owner groups. In consequence, for all but the ﬁrst introduction to this topic, the Prussian State will be treated separately from other owners of land. Table 4.1 shows the broad breakdown of ownership of properties of at least 100 hectares in size, by physical and non-physical juridicial persons (as a reminder again, these are the adjusted data as reported in Chapter 2). Table 4.1. Ownership of large landed properties by physical and juridical persons (area in 000 ha, tax value in 000 marks) Owner category

Natural persons Prussian State All others, unknowna total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

8,140 2,188

84,511 13,538

8,087 2,297

83,746 14,751

7,553 2,546

82,542 14,822

299

4,169

356

4,829

461

5,550

10,628

102,219

10,739

103,327

10,560

102,913

a Only 12 properties of 100 ha or more, totalling 5150 ha, had unknown owners c.1882. There were 8 for 3323 ha c.1895, and 7 for 2716 ha c.1907.

88

Land Ownership by Rank or Class of Owner

Despite the vastness of State landholdings in Prussia, physical persons were still overwhelmingly dominant in the distribution of large properties, owning around 80 per cent of both area and tax value in the early 1880s. This share sank slightly to 75 per cent of area, but remained above 80 per cent of tax value, by about 1907. Thus the ownership of large properties in Prussia was very much a personal thing throughout the period under review. At this point, we now set aside consideration of state-owned properties and focus on the rest of the owner categories. The principal division among physical persons was between the aristocracy (royalty and nobility) and the bourgeoisie (non-noble persons). Between these two subgroups, the aristocracy were the more important landowning class. Although both groups declined slightly, in terms of overall area and tax value owned in large properties, the decline was somewhat greater for the bourgeoisie in terms of area, but for the aristocracy in terms of tax value. Royalty and nobility owned 1.60 times as much land, and 1.45 times as much tax value in large properties as did non-noble persons in the early 1880s, but these multiples had changed to 1.69 and 1.40, respectively, around 1907 (see Table 4.2). Even as the total area owned was declining, non-noble persons were concentrating their ownership on more valuable types of land (ploughland and meadow, in the main) as compared to the less valuable land types (particularly pasture and forest), while the aristocracy showed little of this trend: the average tax value of non-noble properties rose from 11.01 to 12.23 marks per hectare over the period, while the rise for the aristocracy was only from 9.99 to 10.16.¹ The absolute difference in the averages per hectare between the two subgroups of owners is a reﬂection of the composition of properties, not their quality:² royalty Table 4.2. Large landed properties owned by non-state owners (area in 000 ha, tax value in 000 marks) Owner category

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

Aristocracy Non-noble persons All others, unknown

5,005

49,989

4,940

48,822

4,744

48,177

3,135

34,523

3,147

34,924

2,809

34,364

299

4,169

356

4,829

461

5,550

total

8,439

88,681

8,442

88,575

8,013

88,091

¹ Note that the assessments from the 1860s were never changed from their original values, so none of the increase in average assessment per ha is the result of increased assessments; it all stems from a shift in composition toward the kinds of land that were assessed more highly. ² The land quality issue forms one of the central foci of Ch. 3, and will arise again in the discussion of potential bias in the land tax assessments in the ﬁnal chapter.

Land Ownership by Rank or Class of Owner

89

and nobility tended to own much higher shares of forest land in their properties than did non-noble persons; this higher forest share dragged down the average tax assessment per hectare, since forest land had only around a third to a ﬁfth or even less of the land tax net yield per hectare, compared to ploughland or meadow. The overall result, however, is still rather surprising: during a time of rapid growth in industrial and commercial wealth in Germany the bourgeoisie in the seven eastern provinces, on a net basis, divested themselves of large properties, giving up not only relatively, but even absolutely, more land from large properties than did the aristocracy. Both the agricultural censuses and the property statistics have shown a slight trend toward smaller farms and properties, respectively, in East Elbia as agricultural operations intensiﬁed and became more capitalintensive. Despite this trend, the east does not appear to have been an attractive market for estate purchases by industrialists and merchants of the west, and the secular rise in land prices in the eastern provinces³ does not appear to have been at all the result of a diversion of non-agricultural wealth into purchase of landed estates. There existed considerable differences among the aristocracy as landowners, both with regard to rank as well as regionally within East Elbia. Figs. 4.1 and 4.2 show the overall distribution of ownership of large properties by rank⁴ within the category of non-state land owners. Figs. 4.1 and 4.2 show us, ﬁrst of all, that the untitled nobility were the largest single category of landowner among the aristocracy as I have deﬁned it (royalty plus nobility). Among the titled nobility, those with the rank of count (Graf ) owned about as much as the rest of the titled nobility taken together, but among all the ranks, only the princes and dukes as a group managed to increase both the area and tax value of their large landholdings over the period under review. Among royalty, the Prussian house of Hohenzollern—all branches taken together, not just the immediate family of the king—increased its holdings, ³ See my ‘The Price of Land in Eastern Prussia: Data from Capital Gains Tax Records, 1891–1907’, Historical Social Research/Historische Sozialforschung, 22/2 (1997), 195–216 for a discussion of the trends in land prices. ⁴ In coding properties for the rank of the owner, I always used the highest rank listed for the year of publication of the Handbook: So, if a woman born a princess married a count, she was still a princess for rank-coding purposes; if a countess married a prince and thus became a princess, her properties were coded for ownership by a princess. If a royal woman married, say, a non-royal prince, then her properties were still considered royal. If her child then inherited, that child would be a non-royal prince or princess, so the coding for owner rank would reﬂect the change in status for the year in which the heir appeared as owner in the Handbook. If a person was promoted in rank between the publication years of two Handbooks, his properties would be coded for the rank held in the year of publication of the Handbook in question. Thus, for example, Count Otto zu Stolberg-Wernigerode was elevated to the rank of prince in 1890, so in the Handbooks before that year, his properties were coded as being owned by a count, but after that year as being owned by a prince. As we will see later, much of what appeared to be passing of properties to persons of higher rank was merely the increase in rank of their existing owner (and, usually, some or all of his progeny).

90

Land Ownership by Rank or Class of Owner

Fig. 4.1. Owners by rank and category, excluding the State: area owned in the seven eastern provinces.

Fig. 4.2. Owners by rank and category, excluding the State: tax value owned in the seven eastern provinces.

but the trend for the holdings of other royalty in eastern Prussia was clearly downward. We examine these developments in some more detail after ﬁrst dealing with the ownership by rank and type of owner in general in the next section.

Land Ownership by Rank or Class of Owner

91

4 . 3 . OW N E R S H I P B Y R A N K A N D T Y PE O F OW N E R If we subdivide the category of aristocrats, an interesting—but not unexpected—pattern appears in Tables 4.3A through 4.3C, which—to remind the reader once again—count only properties of 100 hectares or more. The ﬁrst thing to note among physical persons is that the average area owned by an aristocrat was over three times that of a non-noble person, and this ratio grew to nearly four by about 1907. Among the aristocracy, increase in average area owned accompanied increase in rank, except for the ‘sovereign houses’—but since all of these were of princely or royal rank (see Table 4.5), had we lumped the sovereign princes with Prussian princes, and the other kings with the Prussian king, the steady progression of size with rank would probably have been unbroken. Among the aristocracy as a whole, the area owned per owner also increased over the period encompassed by Table 4.3. When we look at the tax value per property owned, we see it rise smoothly with rank, except at the very top: the Hohenzollern family’s extensive forest Table 4.3A. Distribution of ownership by category and rank of owner, all seven provinces, c.1882 Owner category

Properties per owner

Area per property (ha)

Area per owner (ha)

Tax value per property (marks)

Tax value per owner (marks)

Aristocracy Untitled nobility Barons Counts Princes and dukes ‘Sovereign houses’ Hohenzollerns ‘Bourgeois’

1.80 1.47 1.75 2.82 6.21 4.44 15.86 1.11

767 641 592 897 1,581 1,639 1,691 403

1,384 940 1,039 2,525 9,826 7,284 26,812 447

7,662 6,521 7,666 9,105 11,948 15,086 8,828 4,443

13,822 9,558 13,440 25,640 74,246 67,048 139,982 4,920

Physical persons

1.34

570

766

5,913

7,949

Religious bodies Prussian State Other states Communal owners Banks & business ﬁrms Other owners

1.83 1,076.00 3.40 2.11

345 2,034 477 707

630 2,188,391 1,620 1,495

5,271 12,582 19,646 7,290

9,632 13,538,292 66,795 15,415

1.68 2.20

503 507

847 1,116

9,376 7,013

15,791 15,429

total: all owners

1.46

667

973

6,417

9,358

92

Land Ownership by Rank or Class of Owner

Table 4.3B. Distribution of ownership by category and rank of owner, all seven provinces, c.1895 Owner category

Properties per owner

Area per property (ha)

Area per owner (ha)

Tax value Tax value per property per owner (marks) (marks)

Aristocracy Untitled nobility Barons Counts Princes and dukes ‘Sovereign houses’ Hohenzollerns ‘Bourgeois’

1.81 1.47 1.65 2.70 6.10 3.92 17.78 1.11

780 659 596 873 1,648 1,271 1,443 396

1,411 969 983 2,353 10,047 4,981 25,645 439

7,734 6,736 7,773 8,475 12,354 16,120 8,911 4,407

13,998 9,899 12,814 22,848 75,338 63,190 158,411 4,877

Physical persons

1.34

566

758

5,883

7,869

Religious bodies Prussian State Other states Communal owners Banks & business ﬁrms Other owners

1.7 1184.00 3.2 2.03

430 1,897 509 656

730 2,246,378 1,628 1,335

5,839 12,289 20,915 6,936

9,899 14,550,423 66,927 14,113

1.41 2.22

480 547

679 1,212

8,285 7,709

11,720 17,081

total: all owners

1.46

663

969

6,414

9,368

holdings helped to keep their average tax value per property relatively low, even though the average size of a Hohenzollern property was the largest of any category of physical persons. And, because their holdings were so extensive, the average tax value per Hohenzollern owner (140,000 marks ± about 5%) was nearly double that of the next greatest category of owner in all three years. The gap in landed wealth between barons and counts on the one hand and princes on the other, while expected, is remarkably large: the average tax value of land owned by a prince or duke was triple that of the land owned by an average count and between ﬁve and six times that of an average baron. In other words, the landed wealth disparities within the ranks of the Prussian aristocracy—even leaving out royalty entirely—were much larger than the difference between the aristocracy in general and non-noble land owners.⁵ With this in mind, we can now turn to more detailed consideration of aristocratic rank and its relation to landownership. ⁵ This was nothing new. Hanna Schissler has noted that ‘the discrepancies in property-holdings within the nobility were enormous’ already in the 18th century. See her ‘The Junkers: Notes on the Social and Historical Signiﬁcance of the Agrarian Elite in Prussia,’ in Moeller, ed., Peasants and Lords, 29.

Land Ownership by Rank or Class of Owner

93

Table 4.3C. Distribution of ownership by category and rank of owner, all seven provinces, c.1907 Owner category

Properties per owner

Area per property (ha)

Area per owner (ha)

Tax value per property (marks)

Tax value per owner (marks)

Aristocracy Untitled nobility Barons Counts Princes and dukes ‘Sovereign houses’ Hohenzollerns ‘Bourgeois’

1.73 1.44 1.64 2.44 4.39 5.83 15.30 1.07

843 673 651 1,026 2,348 1,102 1,462 356

1,461 971 1,069 2,500 10,311 6,428 22,367 380

8,564 7,072 8,700 10,231 18,321 12,985 9,654 4,360

14,840 10,207 14,283 24,934 80,470 75,746 147,712 4,654

Physical persons

1.27

559

710

6,111

7,764

Religious bodies Prussian State Other states Communal owners Banks & business ﬁrms Other owners

1.83 1,444.00 1.60 2.13

417 1,763 681 659

761 2,546,450 1,089 1,404

5,456 10,264 19,192 6,120

9,958 14,821,694 30,707 13,043

1.92 2.40

532 549

1,020 1,317

7,396 8,234

14,198 19,761

total: all owners

2.80

533

1,492

11,788

33,007

4.3.1. Royalty After uniﬁcation in 1871, the German Empire recognized 22 sovereign houses⁶ on its territory. I have, accordingly, recognized this politically superior rank in my coding system for rank or type of owners.⁷ In presenting the data on ownership of large properties by rank or type of owner, I have put one of these houses, the Hohenzollerns (as both the royal family of Prussia and simultaneously the Imperial family of Germany), in a separate category in the tables in this section. In this category are included all branches of the Hohenzollern family who owned land in the seven eastern provinces. Although East Prussia shows only zeros in all years in Table 4.4,⁸ the Hohenzollerns actually did own an estate there: in 1884 King Wilhelm I owned ⁶ Or ‘ruling houses’ (regierende H¨auser) in the ofﬁcial terminology. I follow Conrad in calling them ‘sovereign houses’ (souveraine H¨auser). ⁷ See Appendix on coding. ⁸ In the early 1880s, the database includes 119 principal properties owned by seven members of the Hohenzollern family, including one juridical person, the Royal Court Chamber (K¨onigliche Hofkammer). This appears to be simply a matter of terminology, for in the last database from c.1907, all of its properties are listed as entailed properties of the king. I have accordingly considered these to be properties of the king. Of the 119 Hohenzollern properties, the king owned 52 per cent of the total area and 66 per cent of the total tax value. Prince Carl-Anton von Hohenzollern-Sigmaringen

94

Land Ownership by Rank or Class of Owner

Table 4.4. Land ownership of the Hohenzollern family in East Elbia (area in 000 ha, tax value in 000 marks) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

0 26 71 28 36 22 5

0 95 330 98 123 238 96

0 25 76 38 36 51 5

0 116 345 175 190 504 96

0 34 80 34 13 57 5

0 165 402 147 136 549 80

188

980

231

1,426

224

1,477

Louisenhof, a property of 63 hectares in the district of Stallup¨onen, but by 1895 it had been acquired by a Mr Herzog, and the databases include no other Hohenzollern properties in any year in East Prussia. Conﬁning our attention only to properties of 100 hectares or more does not leave out very much in general. In all provinces in 1882, the Handbooks listed only 8 Hohenzollern properties under 100 hectares each, totalling 532 hectares and 12,125 marks of taxable net yield; in 1895, 7 with 574 hectares and 9607 marks of tax value; and in 1907 only two: 148 hectares and 897 marks. Between the early 1880s and mid-1890s, the Hohenzollern family added nearly 30,000 hectares of land to its total holdings. Massive gains in Pomerania and especially Silesia, adding up to over 39,000 hectares, were partly offset by small reductions in holdings in the other provinces. In Pomerania, the king’s holdings gained over 9400 hectares, half of which were accounted for by his acquisition of four properties (totalling just over 4700 hectares) in the district of Neustettin, all of which had been owned in 1884 by Kreisdirector Baron Carl Senfft von Pilsach. The other half was a forest property of 4725 hectares in the district of Greifenhagen, which appears to have been a part of the estate of Stolzenburg listed in the neighbouring district of Randow in 1884. This estate, owned in 1884 by August Rehder and Herman Schwenke, apparently had been broken up into two parts, with the bulk of the forest land and some arable acquired by the king, and the rest acquired by a certain A. Diestel, who owned no other property listed in the c.1895 database.⁹ In Silesia, Prince Albrecht, who became regent of the state of Brunswick (Braunschweig) after the ruling house died out in 1884, acquired four large owned a further 24 per cent of the total area and 14 per cent of the total tax value, so that just these two individuals owned over three-quarters of all land owned by Hohenzollerns, along with over 80 per cent of the taxable value. ⁹ This name does not appear at all in either the c.1882 or the c.1907 database. By 1910, the owner of Stolzenburg was a Lt Lenz of the Landwehr (Home Guard).

Land Ownership by Rank or Class of Owner

95

estates that had been the property of his late wife, Princess Marianne of the Netherlands, who died in 1883: Schnallenstein and Seitenberg, comprising almost 11,000 hectares in the district of Habelschwerdt, and Camenz and Plottnitz in the district of Frankenstein, for another 5000 hectares. Crown Prince Wilhelm, a minor at the time, received title to over 9000 hectares from the 23,000-hectare estate of Oels, formerly the property of the Duke of Brunswick.¹⁰ Most of the other sovereign houses were represented among the large landowners of Prussia, as Table 4.5 demonstrates. In addition to these rulers, two foreign sovereigns¹¹ owned land in East Elbia—both in Silesia only: the royal family of the Netherlands appeared in the Handbooks as owners up until about the turn of the century, but did not appear anywhere by the time of my 1907 ‘snapshot’. They had extensive holdings in Silesia, in particular the manor of Muskau (nearly Table 4.5. German sovereigns and their ownership of land in Prussia Sovereigns within the German Empire

Did a family member own at least one large property in the seven eastern provinces of Prussia? c.1882 c.1895 c.1907

King of Prussia King of Bavaria King of W¨urttemberg King of Saxony Grand Duke of Saxony-Weimar Grand Duke of Mecklenburg-Schwerin Grand Duke of Mecklenburg-Strelitz Grand Duke of Oldenburg Grand Duke of Baden Grand Duke of Hesse-Darmstadt Duke of Anhalt-Dessau Duke of Saxony-Coburg-Gotha Duke of Saxony-Meiningen Duke of Saxony-Altenburg Duke of Brunswick-Wolfenb¨uttela Prince of Waldeck Prince of Lippe-Detmold Prince of Schaumburg-Lippe Prince of Schwarzburg-Rudolstadt Prince of Schwarzburg-Sonderhausen Prince of Reuss-Greiz (¨a. L.) Prince of Reuss-Schleiz (j. L.)

yes yes yes no yes no no yes yes yes yes yes yes yes yes no no yes yes yes yes yes

a

yes no yes yes yes no no no no yes yes yes yes yes no no no yes yes yes yes yes

yes no yes yes yes no no no no yes yes yes no yes no no no no yes yes yes yes

House became extinct in 1884 (Hanover branch allowed to inherit in 1913).

¹⁰ The rest of Oels went to the King of Saxony (see section on other royals below). ¹¹ Beyond this, two properties in Pomerania in 1884 were properties of the Swedish crown. I took these to be State properties, not properties of the Swedish royal family per se.

96

Land Ownership by Rank or Class of Owner

Table 4.6. Land ownership of foreign and other German royal familiesa (area in 000 ha, tax value in 000 marks) Province

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

8 14 9 0 40 115 10

77 61 105 0 308 993 266

8 14 9 0 36 48 10

77 61 99 0 270 813 260

11 12 10 0 29 43 11

92 68 121 0 229 580 275

total

197

1,810

125

1,580

116

1,363

a

Excludes all properties smaller than 100 hectares.

30,000 ha, about 90 per cent forest¹²), owned by Prince Frederik in 1880, but by the end of the 1890s only his daughter Marie, wife of Prince Friedrich zu Wied, appeared in the Handbook as the owner of three properties of less than 3000 hectares in total. By 1909 Princess Marie held only two of these, apparently having sold the estate of Schildau (313 ha) to a German commoner, Reserve Lt Karl Krieg. Other than that, the Prince of Liechtenstein owned one single property of only 157 hectares in Silesia, which he retained throughout the period. Taken together, members of foreign royalty and the German sovereign houses owned roughly the same amount of land as the Hohenzollern family in the seven eastern provinces of Prussia c.1882, but their properties were much more valuable, with nearly double the taxable yield of the Hohenzollern properties. Their holdings declined rapidly, so that by 1895 these other monarchs held little more than half the area owned by the Hohenzollerns, but about the same amount of taxable value. This changed little up to 1907. The reason for the area-tax value discrepancy was not that the non-Prussian royalty owned less lower-valued forest land (compared to ploughland and meadow), but because they owned more land than the Hohenzollerns in the relatively high-value provinces of Silesia and Saxony. The Hohenzollern properties, on the other hand, were concentrated in the relatively low-value provinces of Brandenburg and Posen. Like the Dutch royal family, some of the German sovereign families also disappeared from the list of East Elbian landowners as the period progressed: The widowed Queen Marie of Bavaria was co-owner of a 784-hectare estate (Fischbach) in Silesia in 1880 with her sister, Princess Elisabeth,¹³ widow of ¹² By 1892 it had passed into the hands of Count Traugott Herrmann von Arnim and served as his family seat. ¹³ The sisters were born Princesses Marie and Elisabeth of Prussia, so were really Hohenzollerns, not Wittelsbachs or Battenbergs. As a principle, within the same rank (in this case royalty) I coded

Land Ownership by Rank or Class of Owner

97

the Grand Duke of Hesse. Elisabeth died in 1885 and Marie in 1889, so by 1892 this property had passed into the exclusive ownership of Elisabeth’s son, Grand Duke Ernst Ludwig IV of Hesse. After the late 1880s, Bavarian royalty was no longer to be found among the owners of large estates in the seven eastern provinces of Prussia. Grand Duke Nicolaus Peter of Oldenburg owned six properties in Silesia in 1880, totalling over 2100 hectares. By 1892, these properties found themselves in the hands of at least three different owners, and the Oldenburgs were no longer landowners in eastern Prussia. Similarly, Grand Duke Friedrich of Baden owned more than 4000 hectares in Posen, all in the district of Schildberg (later to become Kempen), in 1884, but by 1896 these had all come into the possession of Prince Heinrich of Prussia, so the Grand Dukes of Baden also disappeared from the ranks of big landowners in Posen. Wilhelm, Duke of Brunswick was one of the biggest landowners in Silesia in 1880, with over 32,000 hectares in two large estates. This house died out in 1884, and by 1892 the Brunswicks’ 8000-hectare manor of Guttentag was listed as being in the hands of King Friedrich August of Saxony,¹⁴ as was over 13,000 hectares of the manor of Oels. The balance of Oels came into the possession of Crown Prince Wilhelm of Prussia, as noted above. The families of the Duke of Saxony-Meiningen and the Prince of SchaumburgLippe held on a bit longer. Both disappeared only after the mid-1890s: the SaxonyMeiningens owned two properties in Posen totalling almost 3000 hectares, which must have been sold off and subdivided, since there is no trace of either of them in Posen in 1910. Prince Adolph George of Schaumburg-Lippe owned two estates in East Prussia in 1884, 639 hectares in total; by 1896, he had added a third to bring the total to 780 hectares, but by 1907 two had passed into the hands of a commoner and the third had disappeared entirely from the Handbook, again probably because of subdivision. Of the twenty-one German sovereign houses other than the Hohenzollerns, therefore, sixteen were to be found among the ranks of East Elbian landowners in the early 1880s. Of this number, only twelve remained in the mid-1890s, although the group grew to thirteen when the King of Saxony acquired the aforementioned estates of the deceased Duke of Brunswick. By about 1907, only eleven were left, after the Dukes of Saxony-Meiningen and the Princes of the properties of wives according to the family they married into, not the family into which they were born. If the ranks differed, however, I always assigned their properties in my coding scheme to whichever rank was higher, birth rank or married rank. As we will see later in the case of a Reuss princess who married a Bentheim-Tecklenburg prince, that meant her property was counted as belonging to ‘other German royalty’ even though she was now part of a family that were ‘only’ non-royal princes. ¹⁴ This may have been an error of transcription in the text, since the King of Saxony was Friedrich Georg Albrecht I from 1873 until his death in 1902. He was succeeded by his brother, Georg I (Friedrich August Georg), succeeded by his son, Friedrich August III when he in turn died in 1904.

98

Land Ownership by Rank or Class of Owner

Schaumburg-Lippe no longer held landed estates in East Elbia. The governments of Brunswick, Saxony-Anhalt, and Saxony-Meiningen also owned property in East Elbia in all three periods, but these will be covered in a later section on ownership by states.

4.3.2. The Nobility We turn now to the nobility in descending order of rank. For all ranks, although only the male titles are used, it is to be understood that the equivalent female ranks are also included.

Princes and Dukes In terms of both area and tax value, the landholdings of princes and dukes were heavily concentrated in Silesia, where over 60 per cent of the total area that they owned, and about half of the tax value, were to be found. Their most valuable properties were to be found in the province of Saxony, where—even though princes and dukes owned only about 12–15 per cent of the area there, compared to what they owned in Silesia—the land was assessed at one-third to 40 per cent of the value of their lands in Silesia. Immediately suspicious in Table 4.7 is the case of East Prussia—only a single princely property in both 1884 and 1895,¹⁵ but 16 properties with much larger area and tax value in 1907. In 1907, those other 15 were all owned by Prince Richard Wilhelm Ludwig zu Dohna-Schlobitten (born 1843), who had been elevated to the rank of prince in 1900. Of these 15 properties, six were owned in 1895 by Count Richard Friedrich zu Dohna-Waldburg, father of the then Count Richard Wilhelm Ludwig, who himself (as count) already Table 4.7. Landholdings of princes and dukes, by province (area in 000 ha, tax value in 000 marks) Province

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

0.3 8 8 30 74 279 13

1.3 70 43 392 554 1,852 207

0.3 8 20 30 94 324 36

1.3 70 93 381 719 1,930 649

9 6 26 23 83 341 37

117 66 148 396 669 2,059 648

total

413

3,118

512

3,842

526

4,104

¹⁵ The one property, Eyssehnen in Memel district (later to become a part of Lithuania), remained in the hands of Princess Oginska throughout the period.

Land Ownership by Rank or Class of Owner

99

owned six of these properties in 1895. One other property, Pfeilings (360 ha, in Mohrungen district of K¨onigsberg region) was owned in 1895 by Count Adolf zu Dohna-Schlodien. Between 1895 and 1907, Prince Richard Wilhelm also acquired most of the estate of Sumpf (in Preussisch-Holland district), which had been owned in 1895 by a G. Mittmann. One of his 1907 properties, St¨open (260 ha, also in Preussisch-Holland district), was not to be found in either the 1895 or the 1884 Handbook. So the change in East Prussia is really no change at all, but almost entirely the result of one person’s change in rank. West Prussia shows very little change. Apart from changes in the reported area of some properties, the passage of two properties (Marienfelde and Sternbach, in Schwetz district of Marienwerder region, together 2069 ha in 1909) from Duke Dino-Talleyrand (resident in Silesia as Duke zu Sagan) into the hands of the Prussian State accounts for the entire difference between the last year and the other two years. The area of properties owned by princes and dukes, as recorded in the Handbooks, appeared to have more than tripled in the province of Brandenburg between 1885 and 1910. As in East Prussia, nearly the entire ‘increase’ was merely an increase in rank for two counts: 1. Kaiser Friedrich III raised Count Friedrich Hermann Adolf zu SolmsBaruth to the rank of prince in 1888, and made the title hereditary to the ﬁrst-born son. The manor of Baruth, in J¨uterbog-Luckenwalde district of Potsdam region, covered about 12,000 hectares; other princely Solms-Baruth estates raised the total to 15,429 hectares in 1910. 2. Philipp, Count zu Eulenburg-Hertefels, became a prince in 1900. His estates of Hasen and Liebenberg (the family seat) totalled 3960 hectares in both 1896 and 1910.¹⁶ The rest of the increase between 1896 and 1910 was the result of some aggrandizement of existing estates. The case of Pomerania was only slightly different. The decrease between 1893 and 1910 in princely properties was accounted for by two events: 1. parcellization of the estate of Gothen (1187 ha in 1893), owned then by Heinrich, Prince of Sch¨onburg-Waldenburg, into small properties, presumably for peasant ownership; 2. the death in 1898 of Otto von Bismarck, whose properties in Pomerania had amounted to just over 9000 hectares in 1893. The heirs—except for his ﬁrst-born son Herbert, who received the knight’s estate Reinfeld B (1336 ha) in Rummelsburg district—did not inherit the princely title; they remained count or countess only. ¹⁶ Prince Philipp’s father, Count Friedrich, was listed as owner of Hasen (1915 ha) in 1896.

100

Land Ownership by Rank or Class of Owner

Again, what appeared to be a signiﬁcant change in ownership turned out primarily to be another statistical artefact, this time a reduction in rank of the owners because only one of them inherited his father’s princely rank. Posen presents a somewhat different case, in that the amount of property owned by princes and dukes rose between 1884 and 1896, then fell back again in 1910. This ‘movement’ was the result of the following: 1. Count Hugo von Radolinski became Prince Hugo von Radolin in 1888. His properties in Posen amounted to just under 5300 hectares in all three years. 2. The raising of Count Otto zu Stolberg-Wernigerode to the rank of prince in 1890. His properties in Posen amounted to 10,732 hectares in 1884 and 7988 ha in 1896. His heirs, Princes Hermann and Wilhelm, held 8137 hectares between them in 1910. 3. In 1910, Imperial Count Wilhelm von Hochberg was listed as owner of the Prince Pless estates of Krucz (8200+ ha) and Goray (416 ha) in Czarnikow district of Bromberg region, which he apparently inherited from his father, Hans Heinrich XI, Prince of Pless, Imperial Count of Hochberg and Baron of F¨urstenstein. The Prince of Pless in 1910, however, was Hans Heinrich XV, who must have been Count Wilhelm’s older brother. 4. Countess Constantia Mielzynska acquired the estates of Borzejewo and Wyslawice in Schroda district, each about 450 hectares, which had belonged to Prince Wladislaus Czartoryski in 1896. 5. Countess Hedwig Dzieduszycka, daughter of Count Wladimir Dzieduszycki, became a princess by marrying Prince Witold Czartoryski, thereby adding her 6136 ha to princely properties in 1910. 6. The Royal Settlement Commission had acquired the former estates of Prince Gustav Biron von Curland in Schmiegel district, about 4000 hectares, in 1903, as well as Grabowo and Kaiserswalde in Wirstiz district (together some 2300 hectares), bought from Prince HohenloheSchillingsf¨urst in 1908. Some other minor changes resulted from sale of smaller properties from princes to lower-rank nobility or commoners after 1896. In Silesia, most of the increases noted over the period were also accounted for by a rise in rank for an existing individual, although the sale and purchase of properties between owners of different rank were also signiﬁcant: 1. Count Guido Henckel von Donnersmarck became a prince in 1901; he owned nearly 22,000 hectares of land in Silesia in 1909. 2. About 313 hectares were lost when Marie, Princess of the Netherlands (wife of the Prince zu Wied) apparently sold Schildau in Schönau district

Land Ownership by Rank or Class of Owner

3.

4.

5.

6.

7.

8.

9.

10.

101

in Liegnitz region to a commoner, Lt Karl Krieg, sometime between 1905 and 1909. The Prince zu Putbus owned the manor of Deutsch-Lissa in Neumark district of Breslau region in 1892 (c.1600 hectares); in 1909 Asta von Riepenhausen, née Imperial Countess von Wyttich und Lochum, owned Deutsch-Lissa, which thereby passed out of the category of princely properties. The area owned by the Duke of Ratibor increased by nearly 4400 hectares between 1880 and 1892, then decreased by 4800 hectares between 1892 and 1909. Most of the early increase was accounted for by an addition of 3800 hectares of woodland to his manor of Zembowitz in Rosenberg district, Oppeln region. It is not clear where this forest land came from. The later decrease also was mostly accounted for by a loss of over 4000 hectares of woodland from the same estate; it appeared that this land had been acquired by the King of Prussia and incorporated into his forestry operation at Karmunkau in the same district. A different king, Friedrich August of Saxony, acquired another piece from the same manor (the estate of Thurzy, 341 ha) sometime between 1892 and 1909. The manor of Dyhernfurth, Wohlau district, Breslau region, comprised about 2800 hectares in 1880 and just over 2100 in 1892 and 1909. Sometime between 1880 and 1892, it passed from the hands of Countess Antoinette von Lazareff-Hoym, n´ee Princess Biron von Curland, to Countess Marie Antoinette von Saurma-Jeltsch, n´ee Countess d’AbzacHoym—presumably by inheritance—thus removing it from the ranks of princely properties. Two estates of Prince Calixt Biron von Curland in Sprottau district of Liegnitz region (some 1162 ha in 1880) had, by 1892, become the property of a commoner, a Lt Heyme. Prince Rudolph zur Lippe had acquired the 412-hectare estate of Drogelwitz in Glogau district of Liegnitz region from a commoner, Lt Oscar Ruckert. Between 1892 and 1909, Prince Christian Kraft zu Hohenlohe-Oehringen acquired two estates in Lauban district of Liegnitz region (1059 ha in 1909) from two different non-noble owners. Ownership of the manor of Guteborn (4365 ha in 1909) passed from the category of ‘other German royalty’ after 1892 to the category of ‘princes and dukes’ by 1909 when Prince Ulrich von Sch¨onburgWaldenburg inherited it from his grandmother, Anna, Princess zu Bentheim-Tecklenburg, n´ee Princess Reuss j.L. Sometime between 1880 and 1892, the manor of Lasisk (Gross-Strehlitz district of Oppeln region) passed from the hands of Wilhelmine, Countess

102

Land Ownership by Rank or Class of Owner

zu Solms-Roesa, to Prince Otto zu Stolberg-Wernigerode, thus transferring 17,413 hectares from the ownership category ‘counts’ to the category ‘princes and dukes’. 11. The estate of Klein-Droniowitz (some 2000 ha in Lublinitz district of Oppeln region) was the property of a count (Roger von Seherr-Thoss) in 1909, whereas it had been the property of a prince, Friedrich Wilhelm zu Hohenlohe-Ingelﬁngen, in 1892. 12. The ubiquitous Prince Otto zu Stolberg-Wernigerode owned some 8300 hectares of land in Lublinitz district in 1892, at least 7000 of which had been the property of the ‘Heirs of Count Johannes Renard’ in 1880. For at least two generations back, I could ﬁnd no family connection between Prince Otto and the Renards. In the province of Saxony, the increase in princely properties between 1885 and 1899 was entirely the result of the raising of Count zu Stolberg-Wernigerode to the rank of prince in 1890, already mentioned above in regard to Posen. The provincial values for 1899 and 1907 were nearly the same: except for some small differences in the reported area of existing properties, Prince Ernst zu Lynar had acquired the estate of Grosskmehlen (only 234 ha) in Liebenwerda district, Merseburg region, from a Major von Rothkirch. From the above it should be abundantly clear that most of the changes in the amount and value of land owned by the highest-ranking nobility were simply the result of a change in rank of the owner—through ‘promotion’ by the Kaiser, by marriage to a higher-ranking person, or via inheritance from a higher- or lower-ranking relative—and only seldom through any major investments or divestitures on the part of the high nobility. Changes took place, so to speak, almost entirely ‘within the family’. Conrad’s lumping of the high nobility into a single category—rank of count and above—clearly made sense.¹⁷ In my opinion, it also makes sense in many cases to treat the aristocracy as a single broad category, encompassing all owners from untitled nobility on up to the Kaiser himself. As a consequence, and to avoid even more tedious details, the three sections that follow, on counts, barons, and untitled nobility, will present summaries only.

Counts and Countesses Landholdings of counts and countesses are shown in Table 4.8. From the table, three movements stand out: the decline and small rise in West Prussia, the rise and decline in Pomerania, and the precipitous fall in Saxony between the early 1880s and mid-1890s. The last case is the easiest, since the change in the province of Saxony can be mostly accounted for by the elevation in rank to prince of two counts—the ¹⁷ See any of his provincial studies in the series Agrarian-Statistical Investigations. Cf. Ch. 1.

Land Ownership by Rank or Class of Owner

103

Table 4.8. Landholdings of count(esse)s, by province (area in 000 ha, tax value in 000 marks) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

117 69 216 126 281 439 83

1,081 410 2,017 1,599 2,340 4,643 1,412

118 57 211 141 224 458 55

1,021 352 1,794 1,632 1,820 4,746 931

116 61 202 131 201 477 59

1,128 383 1,756 1,348 1,631 5,123 1,057

1,330

13,502

1,264

12,297

1,246

12,425

Tax value

previously-mentioned Count zu Stolberg-Wernigerode (owner of 14,000 ha in Saxony in 1899), along with Count zu Stolberg-Rossla in 1893 (owner of 9000 ha). Most of the rest appears to have been the result of sales of individual estates to non-noble persons. In West Prussia, Leo Count Sk´orzewski’s properties in Schwetz district of Marienwerder region (1845 ha in 1885) passed into private non-noble hands or those of the Prussian State. Because no Sk´orzewskis appeared in the 1894 Handbook for West Prussia, we have to assume that Count Leo sold out his West Prussian properties, since he died only in 1901.¹⁸ The ﬁscus also gained ownership of 7000 hectares in Briesen and Strasburg districts of Marienwerder region from two counts (von Mielczynski and von Keyserling), and another 1000+ in Konitz district (Alt-Laska from Count Hans Karl Albert von K¨onigsmarck). Unlike the case in Saxony, therefore, the changes in West Prussia between 1884 and 1896 were primarily real changes in ownership, and most of that a transfer from the hands of a count to those of the Prussian State. The small increase in area owned by counts between 1896 and 1909 in West Prussia could be accounted for mostly by the promotion of Roland von Br¨unneck, who owned two properties in Rosenberg district of Marienwerder region totalling some 2600 hectares, to the rank of count in 1900. In addition, Count Finck von Finckenstein had acquired the estate Garden (1150 ha, also in Rosenberg district) from a non-noble owner, Victor Schmidt.

Barons and Baronesses Table 4.9, showing land owned by barons and baronesses,¹⁹ shows one general trend—a loss of property by barons between the mid-1890s and c.1907 in every ¹⁸ Berlin-Brandenburgische Akademie der Wissenschaften, Das digitale W¨orterbuch der deutschen Sprache des 20. Jahrhunderts, on the internet at http://www.dwds.de/?qu=polnisch&corpus=1&lm =500&kw=on. ¹⁹ I have taken this rank to include the male and female versions of Baron, Freiherr, and Edler.

104

Land Ownership by Rank or Class of Owner

province except Saxony—and some provincial peculiarities, especially the rather strong gains in Brandenburg, and the major losses in Posen and Saxony in the earlier period. Major changes shown in Table 4.9 occurred as follows: Table 4.9. Landholdings of baron(esse)s, by province (area in 000 ha, tax value in 000 marks) Province

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

49 24 83 75 42 125 72

442 196 719 972 418 1,951 1,435

41 28 112 70 26 120 57

365 140 1,108 926 282 1,807 1,253

50 20 68 58 22 111 58

460 107 582 828 204 1,788 1,292

total

471

6,133

454

5,882

388

5,260

1. In East Prussia, nearly the entire decrease between 1884 and 1895 can be accounted for by transfers of 14 properties (7866 ha c.1882) from the hands of barons to the hands of higher- or lower-ranking persons, including—as a majority of recipients—non-noble persons. Between 1895 and 1907, baronial holdings increased by about 8600 hectares. Of this total, barons were net buyers of 4153 hectares from lower-ranked nobles, and net sellers of 228 hectares to non-noble persons. Properties found in one Handbook and not in the other accounted for 2747 hectares net, and two properties that remained in the same family, but with owners of lower rank in 1895, accounted for a further 1111 hectares of the total gain. Barons also lost one property of 433 hectares to the Landbank A.G. in Berlin. These changes explain about 7350 of the 8600 hectares of net gain of land by barons between 1895 and 1907 in East Prussia. 2. In West Prussia, four properties (nearly 6000 ha) that passed into lowerranking hands accounted for most of the decrease between 1894 and 1909. A ﬁfth (Waldau, Flatow district of Marienwerder region,1109 ha) went to the Royal Settlement Commission for parcellization. The decrease from 1885 to 1894, however, was much more problematic: I could identify six properties of 3349 total hectares that clearly had passed into the hands of lower-ranking owners; two others (647 hectares total) disappeared from the Handbooks, but not into the hands of the Royal Settlement Commission. This would appear to account for nearly all of the change, except that I also found four properties totalling 3097 hectares that appeared to pass to lower-ranking owners. Since, however, the later owner had the same surname as the earlier baronial owner, this may be the result of the compilers of the Handbook in 1894 failing to record the proper title for

Land Ownership by Rank or Class of Owner

105

the new owner. These latter could not be resolved. Two were still in the hands of the 1894 owner in 1909, and two had passed into the hands of other owners (one an untitled nobleman, the other the Prussian State). As offsets, there were 2579 hectares in three properties that passed from non-baronial into baronial hands between 1885 and 1894. 3. In Brandenburg, nearly all of the reduction of just under 9000 hectares in baronial ownership between 1896 and 1910 could be explained by the passing of 12 properties (8622 ha) into ownership of higher- or lower-ranking owners or into ownership by a non-physical juridical person. 4. Most of the reduction in baronial ownership in Pomerania between 1894 and 1909 (about 7900 of about 8500 ha) can be accounted for by offsetting changes in ownership: nine properties totalling 11,000 hectares passed out of the hands of barons to higher or lower ranks,²⁰ with seven properties of 3100 total hectares coming into the hands of barons from other ranks. The reduction between 1893 and 1910 (about 6800 hectares) was also almost entirely the result of the same process: the properties passing from the hands of barons to other ranks exceeded those coming into the hands of barons from other ranks. 5. Posen showed a steep decline in baronial ownership between 1884 and 1896, almost exactly 15,000 hectares. Of that, I can account for about 12,500 hectares by changes in individual properties. Over 7,500 hectares of land in ten properties changed owners to one with a different rank; four properties (3500 hectares) disappeared from the Handbooks —none of these could be found among the properties purchased by the Royal Settlement Commission—and three properties of nearly 1400 hectares changed owners within the same family, but the new owner appeared not to have inherited the baronial title. Between 1896 and 1910, barons appear to have sold off over 10,000 hectares to other ranks, non-noble persons, or juridical persons,²¹ even though the net change in baronial ownership was only about 5200 hectares. In addition to this, four properties of the manor of Dzialyn in Gnesen district of Bromberg, totalling nearly 2800 hectares, passed from ‘Baron Carl Sprenger’s heirs’ to Heinrich von Sprenger, who was not listed as a baron. One property of 618 hectares disappeared from the Handbook’s next edition, and the Royal Settlement Commission acquired the estate of Recz (419 ha in Znin district of Bromberg region) from Baron von Seherr-Thoss in 1898. Barons must therefore have acquired roughly 8500–8600 hectares from other ranks over the same interval. It is worth emphasizing here that although the holdings of barons appeared to have been cut nearly in half ²⁰ Of these, the King of Prussia acquired four properties totalling 4709 ha in the district of Neustettin, K¨oslin region, that had been owned in 1884 by Baron Carl Senfft von Pilsach. ²¹ Among these was the Posen Provincial Association against Itinerant Beggary (Posener Provinzial-Verein gegen die Wanderbettelei), which acquired the 314-ha estate of Alt-Latzig in Filehne district of Bromberg region.

106

Land Ownership by Rank or Class of Owner

Table 4.10. Landholdings of untitled nobility, by province (area in 000 ha, tax value in 000 marks) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

213 207 449 627 508 280 121

1,864 1,601 4,426 5,613 4,629 4,164 2,162

207 191 438 613 489 293 126

1,673 1,459 4,261 5,596 4,320 4,294 2,471

195 147 432 576 443 296 136

1,676 1,130 4,144 5,096 3,955 4,711 2,679

2,405

24,459

2,357

24,073

2,226

23,393

between 1884 and 1910, almost none of this was the result of purchases by the Settlement Commission, even though Posen was the province in which it was most active. Barons appeared to have been energetically buying and selling land, but mostly selling. 6. Silesia seemed to follow the general trend, and showed no dramatic change, so I did not examine changes there on an individual-property basis. 7. Properties owned by barons in the province of Saxony decreased by about 8600 hectares between 1885 and 1899; about 7200 hectares (in 15 properties) of that total can be accounted for by a change of hands that involved a change of rank of owner. Although there remain uncertainties for a small number of properties regarding whether heirs of barons also inherited the title without the compilers of the Handbooks having noted it, the changes in landholdings of barons—unlike those of counts and princes—did not occur as a result of promotion in rank; we ﬁnd no case of a baron’s being named a count.²² The great bulk of changes in baronial landholdings occurred through what appear to have been normal market processes. Why barons should have been divesting themselves of land on a net basis remains a mystery, however. Perhaps they were much more like the untitled nobility than the higher nobility, since their behaviour with regard to landholding was so similar, as the next section shows.

Untitled Nobility Table 4.10 shows the trend in the landholdings of untitled nobility in the seven eastern provinces. The table shows a steady erosion of the holdings of untitled nobility over the period, except in the provinces of Silesia and Saxony, and a quite precipitous decline in West Prussia between 1894 and 1909, on a ²² There are a few cases of baronesses marrying up in rank, but these account for only a trivial fraction of the total change in landholdings of the baronial group.

Land Ownership by Rank or Class of Owner

107

Table 4.11. Properties owned by untitled nobles in 1894 but not in 1909, province of West Prussia New owner/status of property Royal Settlement Commission Private parcellization Bank Prussian State Count or countess Non-noble person Not found in 1909 Handbook total

Properties (no.)

Total area (ha)

15 3 1 26 12 44 26

7,872 1,507 746 12,348 12,682 20,018 12,348

127

67,521

net basis 77 properties and nearly 44,000 hectares. Examining the individual property records, I found 127 properties covering 67,500 hectares that appeared as property of untitled nobles in West Prussia in 1894 but not in 1909. These were distributed as shown in Table 4.11. With regard to the row in Table 4.11 showing properties owned by a count or countess in 1909, but an untitled noble in 1884: the former Martha von Sch¨onborn, wife of Albrecht von Alvensleben—who was made a count in 1888—is not listed as a countess in either the 1884 or 1894 Handbooks. Marie Mathilde von Alvensleben-Sch¨onborn, presumably the daughter of Martha and Albrecht and owner of Martha’s properties in 1909, was listed as a countess. Martha owned ﬁve such properties amounting to 6025 hectares in 1894; these were listed as seven properties in 1909. In addition, Countess Marie Mathilde owned two more properties in West Prussia in 1909, totalling 1058 hectares. One had been the property of Lt. Albert von Alvensleben in 1894, and the other the joint property of Martha von Sch¨onborn and her apparent sister-in-law, a Mrs. von Sch¨onborn, n´ee von Schlichten. Similarly, properties owned by untitled nobles in 1909, but not in 1894, were distributed as shown in Table 4.12. The plus and minus ﬁgures from the two tables do not account for all of the net change: They sum to −75 properties (very close to the net number) and about Table 4.12. Properties owned by untitled nobles in 1909 but not in 1894, province of West Prussia Previous owner/status of property

Properties (no.)

Total area (ha)

Non-noble person Bank Titled nobility Not found in 1894 Handbook

40 1 6 5

23,358 1,204 1,590 4,867

total

52

31,019

108

Land Ownership by Rank or Class of Owner

−36,500 hectares (about ﬁve-sixths of the net number). The discrepancies are to be found in the amalgamation, or sometimes division, of some properties, and also in unexplained changes in area of properties between Handbook editions, usually—one presumes—the result of purchase to round out a property or sale of a part of a property.²³ What is clear, however, is that the change observed in West Prussia must have been primarily the result of market transactions, usually sale to, or purchase from, a person of different status, including juridical persons. Among the latter, the Prussian ﬁscus was clearly a net buyer; it appeared to have sold no properties at all to untitled nobles, while purchasing several. From the data in the tables, untitled nobility in West Prussia appeared to have been net sellers to every category of owner except banks.²⁴ The increase in ownership by untitled nobles between 1899 and 1907 in Saxony amounted to about 10,000 hectares. Over half of this could be accounted for by the net change in ‘same family’ holdings (those properties for which the listed owner was an untitled noble person in one year, but either a non-noble person or a titled noble with the same surname in the other year²⁵—3186 ha) and properties found in one year but not in the other (2442 ha). The net increase in holdings acquired from higher-ranking nobles was approximately 3400 hectares between the two years, and the net decrease in holdings acquired by non-noble persons was over 2500 hectares, while the net gain from non-physical legal persons came to a scant 100 hectares. Altogether these net changes can explain only about 6600 of the 10,000-hectare gain, so less than two-thirds. The rest must have come in large part from the expansion of existing holdings by accretion of properties not listed in the 1899 Handbook because too small.

Summary Having examined the trees in rather excruciating detail, it is now time to refocus on the forest. Table 4.13 shows the distribution by province of large landholdings for the entire aristocracy, deﬁned here as all nobles and royals taken together. In Table 4.13 all of the transfers of property between owners of different ranks are, by deﬁnition, netted out, as are—of course—elevations of rank within the ²³ Properties listed as having disappeared could not be found in the 1909 Handbook nor in the list of properties purchased by the Royal Settlement Commission. They could have been properties amalgamated into others, or properties that were parts of a larger manor, individually identiﬁed in one Handbook but not in the other, for example. A name discrepancy could also have prevented identiﬁcation: some properties had both a German and a Polish name that were not similar; the 1909 Handbook tended to use the Polish name more often. Sometimes a property even had two different German names, but both of them were not listed in every edition of the Handbook, making identiﬁcation across years difﬁcult, and in some cases impossible. ²⁴ Even if we do not count the two properties of Roland von Br¨unneck (made a count in 1900), three other properties acquired by Count Oswald von Potocki and Countess von Myczielska from untitled nobles between 1894 and 1909 totalled over 2400 ha, compared to the 1590 ha acquired by untitled nobles from higher-ranking nobles between those same dates ( Table 4.11). ²⁵ It is quite possible that some of these had been either ennobled, or given a title, in the period between 1899 and 1907.

Land Ownership by Rank or Class of Owner

109

Table 4.13. Landholdings of all nobility and royalty, by province (area in 000 ha, tax value in 000 marks) Province

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

389 347 836 890 982 1,260 305

3,466 2,431 7,650 8,680 8,372 13,841 5,596

378 322 837 901 906 1,299 298

3,171 2,198 7,285 8,767 7,610 14,200 5,776

382 281 822 832 793 1,325 308

3,468 1,930 7,177 7,927 6,849 14,780 6,041

total

5,009

50,036

4,940

49,008

4,743

48,171

nobility. The general trend of stable to slightly declining total landholdings is evident in the table. The sharp declines in West Prussia and Posen have already been explained by unusual policy (the Settlement Commission), leaving only Silesia bucking the trend. As we saw in the preceding sections, this was the result of gains by all classes of the aristocracy as here deﬁned, except for ‘other royalty’ and barons.

4.3.3. Non-Noble Persons (‘Bourgeoisie’) The ‘bourgeoisie’—taken here to be all physical persons not belonging to either the nobility or royalty²⁶—were one of the biggest landowning groups in East Elbia. They owned a larger number of properties, but less land or tax value, than did the aristocracy. On a per-hectare basis, their properties were more valuable than those of the nobles and royals—not because they owned higher-quality land, but because they owned properties with higher concentrations of higher-valued land, namely ploughland and meadow. Table 4.14 shows the distribution of their holdings by province in the three reference periods. In the table the general trend of non-noble persons to reduce their holdings of large properties, noted above, holds across all provinces except East Prussia and Saxony, and—primarily as a result of the land purchases of the Royal Settlement Commission—was particularly pronounced in West Prussia and Posen. Examining the unusual result for East Prussia in more detail reveals the following: of a total of 35 districts in the province, 13 actually followed the general trend and showed a net loss of land by non-noble persons between 1895 and 1907. On the other hand, 14 districts showed a net gain of more than 2500 hectares, 6 of these registering more than 5000 hectares. To try to get an idea ²⁶ This category therefore also includes an indeterminate number of Grossbauern, large peasant owners, whom no one at the time would have considered to belong to the bourgeoisie.

110

Land Ownership by Rank or Class of Owner

Table 4.14. Landholdings of non-noble persons, by province (area in 000 ha, tax value in 000 marks) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

564 527 411 542 493 415 184

4,684 4,364 4,185 5,319 4,326 6,494 5,151

570 524 433 540 477 416 186

4,659 4,291 4,395 5,381 4,315 6,693 5,250

630 368 386 507 326 358 235

5,428 3,431 4,072 5,077 2,990 6,307 7,064

3,135

34,523

3,147

34,984

2,809

34,370

of the source of this change, I examined the district with the largest total area increase in each of the two administrative regions: K¨onigsberg (Land) in the region of the same name, and Sensburg in Gumbinnen. In K¨onigsberg (Land) I found 53 properties that appeared as properties of nonnoble persons in the 1907 Handbook but not in the 1895 edition, in Sensburg, 33. Of these, 44 and 26, respectively, were simply not there—new additions to the list of large properties. In K¨onigsberg, the remaining properties broke down as follows: seven had been acquired from nobles, one from a business ﬁrm, and one appeared to be engrossment of an existing property. The corresponding numbers for Sensburg were two, four, and one which appeared to have been a part of an existing property that had been sold off separately. The gains were therefore overwhelmingly represented by properties new to the list. Were these simply the result of improvement in the coverage of properties in the Handbook, or were they perhaps mostly parcels that had been severed and sold off from larger units?²⁷ Table 4.15. Properties of 100 hectares or more in two districts of East Prussia District

Total number

No. owned by non-nobles

Total area (ha)

Area owned by non-nobles

K¨onigsberg 1895 K¨onigsberg 1907 Difference

126 163 37

82 124 42

52,240 61,186 8,946

26,025 35,801 9,776

54 83 29

36 62 26

51,267 60,972 9,705

11,634 18,041 6,407

Sensburg 1895 Sensburg 1907 Difference

²⁷ They could, of course, also have been properties formed by the amalgamation of several small holdings. I consider it unlikely that this would have happened on the scale observed in these districts.

Land Ownership by Rank or Class of Owner

111

In both districts there was a substantial increase in the area found in properties of 100 hectares or more between the two reference years, in the case of Sensburg district accounting for more than the entire increase in landholdings of non-noble persons. While these ﬁgures are not conclusive, they strongly suggest that the increase in area owned by non-nobles was primarily an improvement in the identiﬁcation of large properties to be listed in the Handbook,²⁸ not a transfer of area from one class of large holdings to another. If this be representative for East Prussia as a whole, then the overall reduction in holdings of large properties by non-noble persons found earlier would be an understatement of what had actually taken place.

4.3.4. Another Look at Rank: Military Ofﬁcers In his inaugural lecture at the Humboldt University in 1992, Hartmut Harnisch reminded his audience that ‘The families of the landed nobility in East Elbian Prussia had demonstrated, more impressively than their counterparts in the middle-sized and small German states, their usefulness in State service under the absolutist rulers, especially as ofﬁcers.’²⁹ and that this allowed these nobles to be both a ‘functional elite’ at the state level as well as a ‘local and propertied elite’.³⁰ Just as Harnisch’s lecture reminded his audience of the Junkers’ role as ofﬁcers, so do the novels of Theodor Fontana and others remind us that an independent income from one’s agricultural estates was often necessary to maintain the standard of living expected of an ofﬁcer, at least in the elite regiments. But non-nobles could also become ofﬁcers. Among the owners of large properties, then, were there indeed large numbers of ofﬁcers? Did military rank correlate with noble rank? Did ofﬁcers tend to own a certain size or type of property? The Handbooks often pointed out a land owner’s rank as military ofﬁcer (no enlisted ranks were ever identiﬁed in the Handbooks),³¹ even if the ofﬁcer in question were not currently on active duty. I have interpreted the term ‘ofﬁcer’ liberally, to include ofﬁcers of all ranks from lieutenant to ﬁeld marshal, and ofﬁcers in the Home Guard (Landwehr) as well as in the more central branches of service such as cavalry, artillery, or infantry. Occasionally, a female owner was listed as ‘Frau General(in)’ (but only seldom as wife of any rank below major); I ²⁸ This is also consistent with the problems of incomplete coverage of State forests in East Prussia, as detailed in Ch. 2 (alas). ²⁹ Hartmut Harnisch, Adel und Grossgrundbesitz im ostelbischen Preussen 1800–1914, inaugural lecture, Faculty of Philosophy and Historical Sciences, Humboldt University in Berlin, 16 June 1992, 9. My translation. Available on the internet at http://edoc.hu-berlin.de/humboldt-vl/harnischhartmut/PDF/Harnisch.pdf. ³⁰ Ibid. ³¹ I cannot claim to know that they always did so, since no mention was ever made of any criteria for inclusion of data about an owner, but the mentions struck me as frequent enough to indicate at least a striving to note whether the owner was, or had been, a military ofﬁcer.

112

Land Ownership by Rank or Class of Owner Table 4.16. Numbers of ofﬁcers and non-ofﬁcers among land owners Non-aristocrats c.1882 c.1895

c.1907

Aristocrats c.1882 c.1895

c.1907

Ofﬁcers Non-ofﬁcers

707 6,310

585 6,587

577 6,807

961 2,659

904 2,598

941 2,305

total

7,017

7,172

7,384

3,620

3,502

3,246

included these women in my ‘ofﬁcer’ group of owners on the assumption that their properties were either a widow’s inheritance or a contributor to a husband’s existence while in service. I made no distinction between ofﬁcers on active duty and those retired, in the reserve, or out of service (ausser Dienst), on the grounds that it was the ofﬁcer culture and its connection with landholding that was important—(indeed, why identify the military ranks of landowners who were no longer in service, if this were not so?)—not whether a particular landowner was currently on active duty. Among natural persons, I counted 10,637 owners c.1882; 10,674 c.1895; and 10,630 c.1907; these ﬁgures amount to no change in the total number of owners over the years.³² Among these owners, ofﬁcers and non-ofﬁcers were distributed among the aristocrats (royal or noble) and non-aristocrats (bourgeois) as shown in Table 4.16. While the number of noble owners who were ofﬁcers held quite steady, the proportion of ofﬁcers actually grew within noble ranks, as the total number of noble owners shrank over the period. During the same time, the numbers of non-noble ofﬁcers declined both absolutely and as a proportion of all non-noble owners. A similar but slightly more varied picture emerges from Table 4.17, which makes the same comparisons by the number, total area, and total land tax value of the properties they owned. While both total area and total land tax net yield for the non-noble ofﬁcer group declined slightly, the corresponding values for noble ofﬁcers increased. Ofﬁcers became more important among noble land owners, less important among the bourgeois. Whatever measure is used, more than one-quarter of noble land was owned by ofﬁcers—approaching one-third c.1907—but for non-noble persons the fraction was more like one-tenth. The gap between the two groups of ofﬁcers grew over time. The average noble ofﬁcer-landowner owned about three times as much land with about two-and-a-half times the land tax net yield, compared to his non-noble counterpart, as Table 4.18 shows. Ofﬁcers by and large owned slightly more than average amounts of land for their group, and this land was more valuable than average. But from the ³² As pointed out in Ch. 2, these numbers represent upper bounds to the number of owners because of problems of identiﬁcation.

Land Ownership by Rank or Class of Owner

113

Table 4.17. Landholding by ofﬁcers and non-ofﬁcers Non-aristocrats c.1882 c.1895

c.1907

Aristocrats c.1882 c.1895

c.1907

Number of properties Ofﬁcers 788 Non-ofﬁcers 6,984

676 7,262

655 7,227

1,683 4,835

1,652 4,686

1,603 4,023

total

7,772

7,938

7,882

6,518

6,338

5,626

Total area (000 ha) Ofﬁcers 342 Non-ofﬁcers 2,794

288 2,858

288 25,231

1,282 3,722

1,361 3,580

1,533 3,210

total

3,146

2,809

5,004

4,940

4,744

Land tax net yield (000 marks) Ofﬁcers 3,876 3,591 Non-ofﬁcers 30,650 31,330

3,596 30,768

12,645 37,340

13,044 35,781

15,836 32,341

total

34,364

49,985

48,825

48,177

3,136

34,526

34,921

Table 4.18. Average landholding by ofﬁcers and non-ofﬁcers Non-aristocrats c.1882 c.1895

c.1907

Aristocrats c.1882 c.1895

c.1907

Properties per owner (no.) Ofﬁcers 1.11 Non-ofﬁcers 1.11

1.16 1.10

1.14 1.06

1.75 1.82

1.83 1.80

1.70 1.75

total

1.11

1.07

1.80

1.81

1.73

1.11

Total area per owner (ha) Ofﬁcers 483 Non-ofﬁcers 443

492 434

499 370

1,335 1,400

1,505 1,378

1,630 1,393

total

439

380

1,382

1,411

1,461

Total tax value per owner (marks) Ofﬁcers 5,477 6,138 Non-ofﬁcers 4,857 4,756

6,231 4,520

13,162 14,043

14,429 13,772

16,829 14,031

total

4,654

13,809

13,942

14,842

447

4,920

4,869

difference in size and value between ofﬁcers of noble or non-noble origin, it is clear that income from his estates would have gone much further in helping a noble ofﬁcer to maintain the expected lifestyle than it would have for a nonnoble ofﬁcer. Ranking ofﬁcer-owners by the tax value of their properties, only six non-nobles made it into the top 100 ofﬁcer-owners c.1882. The two biggest of these were Gustav Wendenburg, a cavalry captain from Saxony, and Eduard Lindner, reserve lieutenant from Silesia, with just over 40,000 marks each of land tax net yield, putting them in 50th and 51st place, respectively; each had

114

Land Ownership by Rank or Class of Owner Table 4.19. Ofﬁcer-owners by civil status Civil status

c.1882 Untitled noble Baron Count Prince, royalty Non-noble total c.1895 Untitled noble Baron Count Prince, royalty Non-noble total c.1907 Untitled noble Baron Count Prince, royalty Non-noble total

Owners (no.)

Area (000 ha)

Tax value (000 marks)

696 121 131 13 707

663 110 278 231 342

7,124 1,577 2,646 1,302 3,872

1,668

1,624

16,521

652 112 121 19 585

675 91 267 327 288

7,049 1,473 2,475 2,048 3,591

1,489

1,649

16,635

639 118 167 17 577

662 124 436 311 288

7,221 1,832 4,458 2,324 3,596

1,518

1,821

19,431

three properties, totalling 977 and 2411 hectares, respectively. About 1895 there were ﬁve non-nobles among the top 100, again led by Capt. Wendenburg, now in 17th place with three properties totalling 1728 hectares and nearly 65,000 marks of tax value. Next after him was retired Lieutenant Paul Naehrich in 53rd place; Naehrich had seven properties, 1220 hectares, and 40,771 marks of tax assessment. By about 1907 there were only two in the top 100: Bernhard Nette, a captain of cavalry from Saxony, whose one estate of 1112 hectares carried 68,524 marks of tax value, at position number 35, and reserve lieutenant Karl Bach, also from Saxony, whose two properties totalled 714 hectares and 34,705 marks of tax assessment.³³ Big-time property ownership by ofﬁcers was clearly an almost exclusively noble terrain. Table 4.19 shows the numbers of ofﬁcer-owners by civil status. The big jump in ownership by counts between 1895 and 1907 is suspicious. Checking each individual count, I found 85 that were listed as ofﬁcers in 1907 but not in ³³ Capt. Wendenburg had apparently died by this time: two of his properties were owned by Erich Wendenburg, probably his son but not an ofﬁcer. The other property had disappeared, probably subdivided into smaller holdings.

Land Ownership by Rank or Class of Owner

115

Fig. 4.3. Ofﬁcer-owners: share of different ranks in numbers, total area, and total tax value, c.1882, c.1895, c.1907.

1895.³⁴ Of those, 54 simply were not listed at all in the 1895 Handbooks. The rest of the cases were questionable, in that the owner could not be identiﬁed by ﬁrst name in 1895, or was listed in 1895, but not as an ofﬁcer. Since it easily could have been the case that someone became an ofﬁcer in between the two dates, I did not feel justiﬁed in assuming that if someone were an ofﬁcer c.1907, he also was an ofﬁcer c.1895. Moreover, sons who were ofﬁcers would only appear in the list when they actually inherited their fathers’ estates. Even with all these rationalizations, however, there still remains the nagging suspicion that maybe the recording of military ranks for landowners improved over time; it would therefore be well to consider the calculated increases in the holdings of counts as an upper bound. With this caveat in mind, we can note that the most important category of ofﬁcer-owner was the untitled nobility; if we lump the barons with them into ‘lesser nobility’, we ﬁnd that group with about double the tax value of land of the higher nobility, except c.1907, and that the untitled noble ofﬁcers alone had more than twice the land or tax value held by non-noble ofﬁcers, even though they exceeded their numbers by only about 10 per cent in 1907 and were actually fewer c.1882. The relative positions of these groups of ofﬁcer-owners may perhaps be more easily seen by means of a graph. The shrinkage in the relative position of nonnoble ofﬁcers as landowners shows up very clearly in Fig. 4.3, as does the growth in the relative position of the high nobility. Untitled noble ofﬁcers may have been the most numerous, but their relative position declined. The high nobility ³⁴ There were also counts listed as ofﬁcers in 1895, but not in 1907; therefore this number is bigger than the net difference in number of counts as owners.

116

Land Ownership by Rank or Class of Owner

Table 4.20. Landholdings of the Prussian State, by province (area in 000 ha, tax value in 000 marks)a Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total a

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

510 322 469 261 189 199 238

1,683 767 3,232 2,067 767 1,417 3,605

507 361 458 270 264 198 239

1,669 902 3,023 2,114 1,221 1,477 4,345

589 462 489 288 279 211 229

1,900 1,388 3,039 2,071 1,226 1,636 3,562

2,188

13,538

2,297

14,751

2,546

14,822

Adjusted data from Chapter 2.

became more important as ofﬁcer-owners as time passed; this would still be quite dramatic even if we discounted the growth between the last two years for some degree of exaggeration.

4.3.5. The Prussian State This is not the place to repeat the analysis of State landholding data from Chapter 2, but it is worthwhile to remind the reader that the data of Table 4.20, based on the adjustments to State landholdings made in Chapter 2, are not robust enough to bear the weight of very detailed analysis. The ﬁgures in the table can, nevertheless, suggest some general observations. Even after adjustments, these ﬁgures represent a lower bound to the amount of land owned by the Prussian State in these seven provinces, as Chapter 2 shows in some detail. Moreover, where Table 4.20 shows apparent increases in landholdings of the State, these may be more illusion than fact, especially between c.1882 and c.1895 and especially in those provinces where large adjustments to the data were necessary (Brandenburg, East Prussia, and Pomerania). Taking these warnings into consideration, there still appear to be major increases in State landholdings in West Prussia and Posen that require a more detailed look at the data. In West Prussia, the State holdings of 361,000 hectares in 1894 comprised 139 properties, while the 462,000 in 1909 were found in 289 properties, so the number of properties more than doubled, and their area increased by 28 per cent. Only 9240 of the 101,000 hectares of increase can be explained by the 19 properties acquired from private individuals but not yet subdivided by the Royal Settlement Commission. Examining the data by district, there were eight districts in which the net gain of land between 1894 and 1909 appeared to be greater than 6000 hectares. In those eight districts, there were twelve

Land Ownership by Rank or Class of Owner

117

properties comprising 51,572 hectares in 1909 that could not be found in the 1894 Handbook. At an average size of about 4300 hectares—the biggest was 10,270 hectares—it is highly unlikely that these properties had been assembled by the State through the purchase of numerous smaller but adjacent properties. A further 62 State properties totalling 45,220 hectares had been owned by private individuals in 1894. Based on this sample, we could speculate that the apparent gain in landholdings by the Prussian State in West Prussia was roughly half real gain and half simply an improvement in the coverage of the Handbook. In Posen, the situation was somewhat different. First of all, there were nine districts in which the Prussian State owned properties in 1896, but apparently not in 1884. Of the 28 State properties in these districts, six (totalling 4424 ha) were not to be found at all in the 1884 Handbook; the other 22 (totalling 12,733 ha) had been owned by private individuals—seven by non-nobles (2337 ha) and 15 by nobles (10,396 ha). Secondly, four districts showed apparent gains of more than 6000 hectares. In those four districts, there were three properties (15,775 hectares) that were listed in a different district in 1884, so represented no net gain at all. There was only one property, of 6399 hectares, that did not appear at all in the 1884 Handbook. All the rest of the properties in those four districts (some 21 covering 19,579 ha) had been owned by private individuals (all but three by nobles) in 1884. The evidence indicates that the gains by the Prussian State in Posen were for the most part real gains, and—in contrast to West Prussia later—mostly at the expense of the nobility. It appears that the State was constantly on the lookout for properties to buy,³⁵ with the possible exception of the province of Saxony. Whether it had a policy to try to buy from nobles, or merely that nobles had better access to State ofﬁcials (or owned larger properties, which the State favoured for purchase) when they wished to sell their properties, cannot be determined from the data at hand.

4.3.6. Non-Physical Juridical Persons (Other than the Prussian State) Table 4.1 showed that, after physical persons and the Prussian State, all other juridical persons constituted a rather minor fraction of owners of large properties in the seven provinces of our concern; they possessed around 3 to 4 per cent of the area of large properties and 4 to 5 per cent of their taxable value. Most important among these minor players were communities and businesses, according to Table 4.21. Chief among the corporate owners were the cities and business ﬁrms. In the 1907 database, 287 properties (out of a total of 309) totalling 196,000 ³⁵ Only secondarily for subdivision by the Royal Settlement Commission, at least early on: none of the 28 properties acquired in Posen between 1884 and 1896 (over 19,000 ha) was in the hands of the Settlement Commission in 1896.

118

Land Ownership by Rank or Class of Owner

Table 4.21. Landholdings of non-physical legal persons other than the Prussian State (area in 000 ha, tax value in 000 marks) Owner group

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

Religious bodies Other states Communitiesa Banks and business ﬁrms Others, unknown

55 8 144

838 334 1,480

77 8 153

1,039 335 1,623

78 5 204

1,026 154 1,891

48 45

900 617

56 62

961 871

114 59

1,590 889

total

299

4,169

356

4,829

461

5,550

a

Deﬁned as communal groups plus all political divisions, province or below.

hectares belonged to cities;³⁶ these contained over 96 per cent of the total area of communally owned properties in that year. The heaviest concentration of communally owned properties was in Silesia, where 109 of the 309 properties, covering 94,000 hectares, were located, followed by Pomerania, with 92 for 52,000. Thus these two provinces alone contained almost 72 per cent of the total area of community properties. Posen had the least—only two properties totalling 311 hectares: one belonged to a peasant colony, the other to the province. Properties owned by banks and business ﬁrms were primarily located in Silesia and the province of Saxony: in the former in 1909, banks and business ﬁrms owned 73 properties with over 518 thousand hectares total area; in the province of Saxony the properties were fewer but bigger, so 42 properties covered 673,000 hectares. Thus the nearly 1.2 million hectares owned by banks and business ﬁrms in these two provinces accounted for three-quarters of all the land held by such owners in all of the seven East Elbian provinces. While one might expect that the business ﬁrms owning large properties in Saxony might be mostly sugar factories, that was not the case. Indeed, the activities of ﬁrms owning large estates in Saxony were very varied, everything from plant breeding to producing copper slates to various kinds of mining, even a chemical factory. From the ﬁrm name alone the principal activity of the ﬁrm often could not be determined, but sugar producers owned at least nine of the 42 properties owned by business ﬁrms, and companies engaged in some form of mining at least seven. Interestingly, only one of the 42 properties was owned by a ﬁnancial institution, a savings association. ³⁶ This is still a small number; apparently many cities had long since sold off their lands to cover debts arising from ‘many wars’. See Werner Herde, Der Grundbesitz der Stadt Greifswald: Eine Untersuchung u¨ ber die volkwirtschaftliche Bedeutung st¨adtischen Grundbesitzes, Dissertation: Staatswissenschaftliche Fakult¨at, Universit¨at Greifswald (Greifswald: Julius Abel, 1921), 9.

Land Ownership by Rank or Class of Owner

119

In Silesia, at least 18 of the 73 properties owned by banks and business ﬁrms were owned by sugar factories, and at least 11 by ﬁrms engaged in mining. The Landbank Corporation in Berlin owned 9 properties in Silesia,³⁷ and a further ﬁve were owned by ﬁrms engaged in ‘land acquisition’ or ‘parcellization’. Again, the ranks of business ﬁrms owning land in Silesia exhibited considerable variety, perhaps even greater variety than in Saxony. Lands owned by religious bodies in the seven provinces were primarily the properties of monasteries and convents (68% of total area), secondarily of local parish churches (24% of total area). Nearly three-quarters (74%) of the area of properties owned by religious bodies was found in the three provinces of Brandenburg, Posen, and Pomerania in 1910. The character of ownership in the three provinces was very different, however: In Pomerania, all but three of the 52 church properties were to be found in the administrative region of Stralsund, and there almost all were owned by one or another of ﬁve monasteries/convents. In Posen, on the other hand, there were some properties owned by monasteries or convents, but the overwhelming bulk was owned by local parishes. The second largest province from this point of view, Brandenburg, was again different from the other two: only ten properties covered over 20,000 hectares;³⁸ two were owned by the Neuzelle Monastery, six by the Brandenburg Cathedral, another by the Cathedral Collegium, and only one by a parish church. In the ‘other’ category the principal owners were hospitals, schools, and the University of Greifswald, which by itself owned over 15,000 hectares of land with 267,000 marks of taxable net yield, all in the district of Greifswald, administrative region of Stralsund in Pomerania.

4 . 4 . S U M M A RY Table 4.22 presents summary ﬁgures for total area and total tax value owned by the various owner groups. Given that the properties of the high nobility (including royalty) were bigger, more obvious, and probably better known than those of the other groups of owners, I think it is fair to assume quite complete coverage of these properties in the Handbooks already from the 1880s. Therefore we should be able to conclude with some conﬁdence that the high nobility held its own as a landowning group in the seven eastern provinces from the early 1880s up to the eve of the Great War. By the same token, coverage of properties owned by the lesser nobility and the bourgeoisie in the Handbooks probably improved over the period, so the decline shown in the ownership of large properties by both groups probably represents a lower bound to their losses. We have seen that ³⁷ All told, the Landbank Corporation owned 40 properties of 100 ha or more in the seven provinces, with a total area of 23,705 ha c.1907. ³⁸ Almost 16% of the total area owned by religious bodies in the seven provinces.

120

Land Ownership by Rank or Class of Owner

Table 4.22. Total area (in 000 ha) and total tax assessment (in 000 marks) by owner class Owner class High nobilitya Lesser nobilitya Non-noble person Prussian State Religious body Community Business ﬁrm Other total

c.1882 Area

Tax value

c.1895 Area

Tax value

c.1907 Area

Tax value

2,124 2,881 3,135 2,188 55 144 48 53

19,374 30,615 34,523 13,140 838 1,480 900 951

2,153 2,787 3,147 2,297 77 153 56 70

19,323 29,686 34,984 14,751 1,039 1,623 961 1,206

2,130 2,614 2,809 2,546 78 204 114 65

19,510 28,668 34,364 14,822 1,026 1,891 1,590 1,043

10,628

102,219

10,739

103,327

10,560

102,913

a High nobility deﬁned as all with rank of count(ess) or above, including royalty; lesser nobility deﬁned as untitled nobles and baron(esse)s.

the data for the Prussian State properties are more problematical, especially early on, but there has nevertheless been some evidence introduced that its holdings probably increased. The increase shown in Table 4.23 probably represents an upper bound to the gains made by the Prussian State. The gains by other owners have already been discussed in the previous section of this chapter. The total number of owners with at least one property of 100 hectares or more stayed quite steady at around 11,000,³⁹ and the average holding per owner remained remarkably static. Because of the problems to identify some individuals, as described in Chapter 2, this is an overstatement of the actual number of owners, at least for the categories of physical persons. Around 600 of the high nobility owned just over a ﬁfth of all the acreage, and just under a ﬁfth of all the tax value of the land found in the holdings of these 11,000 owners. While the high degree of interrelatedness and intermarriage among these aristocrats no doubt led to many family squabbles, especially over inheritance, it also surely ensured a degree of solidarity and cohesiveness in their landholdings as a group. Even when a member of this group died without heirs, we saw above that the properties nevertheless remained entre nous, that is, they passed to other members of the same class of people. One of the most interesting features of the data in Table 4.23 is the opposite trends in the two clearly diminishing groups, the lesser nobility and the bourgeoisie. As the numbers of the lesser nobility decreased, the average size of their holdings grew; they were losing members from the bottom of their ranks. Non-noble persons, on the other hand, increased in number as their average ³⁹ If we accept Alexandra Frank’s claim that in the German Empire at the turn of the 20th century there were about 24,000 large land owners, then the East Elbians accounted for less than half that total. See her Die Entwicklung der ostelbischen Gutswirtschaften im Deutschen Kaiserreich und in den Anfangsjahren der Weimarer Republik (Weiden: Eurotrans-Verlag, 1994), 1.

Land Ownership by Rank or Class of Owner

121

Table 4.23. Number of owners and average area and tax value of their properties, by owner class Owner group

c.1882 High nobility Lesser nobility Non-noble person Prussian State Religious body Community Business ﬁrm Other overall c.1895 High nobility Lesser nobility Non-noble person Prussian State Religious body Community Business ﬁrm Other overall c.1907 High nobility Lesser nobility Non-noble person Prussian State Religious body Community Business ﬁrm Other overall

Number of owners

Average area (ha)

Average tax value (marks)

603 3,017 7,017 1 87 96 57 45

3,522 955 447 2,188,391 630 1,495 847 1,172

32,129 10,147 4,920 13,538,292 9,632 15,415 15,791 21,136

10,923

973

9,358

631 2,870 7,173 1 105 115 82 56

3,411 971 439 2,296,872 730 1,335 679 1,249

30,622 10,344 4,877 14,751,384 9,899 14,113 11,720 21,531

11,033

973

9,365

580 2,666 7,384 1 103 145 112 50

3,672 980 380 2,546,450 761 1,404 1,020 1,294

33,637 10,753 4,654 14,821,694 9,958 13,043 14,198 20,856

11,042

956

9,320

holding fell; they appeared to have been recruiting more members into their lower ranks.⁴⁰ But even as their numbers increased, the total holdings of this group declined, at least in the second half of the period under review. The difference in wealth within the nobility comes clear again in the table, in which the average holding of the high nobility is a multiple of that of the lower nobility: 3.5 or more in area terms and 3 or more in terms of tax value of their ⁴⁰ As pointed out earlier, a signiﬁcant portion of this may simply be better coverage of properties in the 100–199 ha range, especially in the province of East Prussia.

122

Land Ownership by Rank or Class of Owner Table 4.24. Ownership of forest land by major owner groups (area in 000 ha) Owner group

c.1882

c.1895

c.1907

Noble ‘Bourgeois’ Prussian State Other

1,770 563 1,645 131

1,762 548 1,700 148

1,729 423 1,820 198

total

4,109

4,158

4,171

average holding. Comparing the high nobility to non-noble persons, we ﬁnd an average holding just under 8 times the size in terms of area, growing to about 9.6 times, with the tax value discrepancy growing from about 6.5 to 7.2. That there was such a difference between these two sets of multiples is yet another indication that the smaller holdings of non-noble individuals held signiﬁcantly higher proportions of high-value land, compared to the much larger holdings of the high nobility. In closing, it should be noted that the gains of the Prussian State—in forest lands in particular—seemed to come mostly at the expense of non-noble owners, as Table 4.24 shows. The ﬁgures in Table 4.24 are the totals for all forest land in the adjusted data for properties over 100 hectares. Some forest land is no doubt missing from these totals, not only for the reasons enumerated above, but also because for some properties, only the total area was given, with no indication of the makeup of this area (ploughland, meadow, pasture, forest, etc.). While noble owners seemed to be divesting themselves of some forest land over the period, the losses were small, compared to those shown in the totals for bourgeois owners, especially after the mid-1890s. The gains of forest land in the ‘other’ category were primarily in the holdings of cities and business ﬁrms.

5 The Biggest Landowners 5 . 1 . P R EV I EW To err is human, but to inherit, quite divine. That is the lesson of this chapter in a nutshell. It begins with a comparison of my results with Johannes Conrad’s list of the twenty largest landowners in the seven eastern provinces in the early 1880s, with explanations for the few differences found. Only in the case of the Prince of Pless were these differences irreconcilable, although in several other cases it also became apparent that Conrad—or his research assistant¹—had made an error, missed some data, or introduced an inconsistency.² Ranking the top ﬁfty landowners except for the Prussian State,³ ﬁrst by total area owned, then by total tax assessment of their properties, produces two lists that had only about 60 per cent overlap, but both of which were heavily populated by the high nobility (especially the top ﬁfty by total area; the top ﬁfty by land tax assessment was a rather more varied group). Fifty landowners amounted to only about 0.5 per cent of the total number of owners with 100 hectares or more, but they accounted for about 10 per cent of all the land in these larger properties, and over 8 per cent of the total tax assessment. There appeared to be a slight trend toward concentration of landholding, by both area and value, in the hands of the ﬁfty largest landowners. Among physical persons, both the lower nobility and the bourgeoisie were scarcely represented in either top ﬁfty group, and the estates of these landed magnates were heavily concentrated in the province of Silesia, with—in area terms—Brandenburg and Posen contending for a distant second place. When tax value was considered, the surprise entrant into the second-place battle was the province of Pomerania, where both high nobility and several juridical persons ¹ In fact, one can but marvel at the industry and accuracy of Mr Morcinietz, statistician from the Iduna Insurance Company in Halle, who would have had to extract information by hand from seven different volumes of the Handbooks, record it with pencil and paper, and then work with, at best, a crude mechanical adding machine to do his calculations. See Conrad, ‘Die Latifundien’, 122. ² My own data have had to be revised numerous times, and may still be revised or reﬁned further in future, and it is only because I have the advantages offered by a computer that I can claim that my results are in some ways preferable to the Conrad–Morcinietz data. ³ So far and away the biggest landowner that to include it would have completely swamped the rest of the big landowners.

124

The Biggest Landowners

(cities and religious bodies and a university) owned numerous relatively small properties with high tax value per hectare. A look at the top ten landowners shows that (a) to be in this group, being of the high nobility was almost a mandatory prerequisite, even though half the names in the list by value were different from those in the list by area; (b) while individual members might move up or down in the ranking, membership in the top ten was very stable; (c) if a new owner came in, or old owner left, it was almost always the result of death and inheritance; and (d) non-Hohenzollern members of this high elite held their land exclusively, or at least highly concentrated, in a single province, usually Silesia, although inheritance could and did affect the concentration by province.

5 . 2 . I N T RO D U C T I O N : C O N R A D ’ S T W E N T Y L A RG E S T L A N D OW N E R S The most logical place to begin this chapter is with the list of the twenty largest landowners—outside of the Prussian State and the Kaiser himself—in the seven eastern provinces of Prussia compiled and published by Johannes Conrad in 1888.⁴ In the same article Conrad also presented a separate table, arranged by rank, of all owners of 5000 hectares or more in the seven eastern provinces.⁵ For the tables he published in 1888, Conrad used exactly the same sources but perhaps a slightly different selection criterion, compared to what I have used in this book. His wording is ambiguous: Our source does not make a strict boundary, to which size it wishes to consider properties, and thereby goes on down to some very small holdings, although it has given itself only the task of compiling the addresses of ‘landowners’. We have drawn such a line and considered only those that comprise 100 hectares or more area, because we have convinced ourselves that only these play a role in the concentration of land in few hands.⁶

In this text, it is unclear whether ‘these’ (diejenigen) refers to properties or to landowners, so any discrepancies between our ﬁgures, save small ones arising from some of the minor corrections I have made to the original Handbook data (see Ch. 2), would require explanation. Beginning with the twenty largest landowners, I therefore set out his results and mine in Table 5.1; owners’ names in bold type indicate that the discrepancy between our two sets of ﬁgures is large enough that that entry requires detailed examination and explanation. Owners’ names in italics indicate that they were left out of Conrad’s table entirely. In some cases, the ‘explanation’ will necessarily be mere speculation, since I have ⁴ ‘Die Latifundien’, 155–64. ⁵ Ibid. 158–61. ⁶ Ibid. 138. Quotes in original. My translation.

The Biggest Landowners

125

Table 5.1. Conrad’s twenty largest landowners by area c.1882a Rank, Owner

1 Prince Pless 2 Prince Antonb 3 Late Duke of Brunswick 4 Prince HohenloheOehringen 5 Prince Frederik of the Netherlands 6 Duke of Ratibor 7 Imperial Count von Schaffgotsch City of Görlitz 8 Prince Friedrich Karlb 9 Prince von Thurn und Taxis 10 Count Henckel von Donnersmarck 11 Duke zu Sagan 12 Count von Brühl 13 Prince Biron von Curland 14 Duke of AnhaltDessau 15 Friedrich Count zu Solms-Baruth 16 Prince zu Putbus 17 Count zu StolbergWernigerode 18 Count von Redern 19 Prince Hatzfeld

20

11 different owners Archduchess of Sachsen-Weimar sum

Conrad Properties (ha)

Area (ha)

Tax value (marks)

Eddie Properties (ha)

Area (ha)

Tax value (marks)

83 24 65

70,139 59,968 40,186

358,753 199,573 390,078

9 21 3

61,267 43,777 40,191

364,723 129,060 390,078

43

39,365

231,307

12

39,321

232,296

23

35,419

155,662

9

35,418

155,662

53 39

34,026 31,011

179,069 122,603

8 6

33,846 31,455

180,932 125,874

7

29,495

140,547

3

25,690

94,674

3

25,690

94,674

33

24,482

193,422

1

13,789

131,935

33

23,731

131,960

16

23,299

130,351

42 12 29

23,630 22,716 22,691

106,957 123,373 177,633

24 5 7

23,088 22,416 22,895

103,475 123,373 178,433

31

22,650

263,195

34

22,909

261,585

4

22,005

95,987

5

21,935

95,667

62 21

19,752 19,649

354,184 305,517

61 4

19,701 20,619

350,651 319,202

24

18,766

227,919

24

18,766

227,919

30

18,538

168,877

24

13,642

678

588,056

2

18,538

168,877

84

184,389

1,085,897

4

13,646

273,906

4,157,875 349

766,238

5,265,117

277,132

Entries in bold = major discrepancies between Conrad and Eddie; entries in italic = omitted from Conrad’s table. a Excluding the Prussian State and the German Emperor. b Denotes a Hohenzollern prince.

126

The Biggest Landowners

no way of knowing Conrad’s detailed calculations, but most of the discrepancies can be quite simply explained. The difference in numbers of properties for a given owner are simply differences in the working deﬁnition of what Conrad and I regarded as a property, one of the problems in the literature to which I have drawn attention already in Chapter 2. These differences are of no consequence here, and I therefore ignore them in the explanations that follow. The ﬁrst major discrepancy, that for the lands of the Prince of Pless, has already been dealt with in Chapter 2. Sufﬁce it to repeat here that, in my opinion, the original Handbook data, which Conrad seems to have accepted as presented, contain some double-counting, and are therefore an overstatement of the amount of land owned by Prince Pless. In the second entry, that for Prince Anton of Hohenzollern-Sigmaringen, there is not really much of a discrepancy at all, since Conrad has listed him as ‘Prince Anton or his heirs’ (known more popularly as Prince Carl Anton, he died in 1885). The differences between Conrad’s ﬁgures and mine—16,191 hectares and 70,513 marks—match quite closely with my totals for Prince Anton’s son, Leopold: 16,692 hectares and 65,463 marks. Conrad seems to have ignored corporate landowners, hence he had no entry for the City of Görlitz in Silesia, which would have come just before Prince Friedrich Carl of Prussia, no. 8 in Conrad’s table ranked by area owned (29,495 ha vs. 25,690). For Guido Count Henckel von Donnersmarck the discrepancy seems to be the result of two things: 1. Count Guido owned three properties under 100 hectares in size; their tax assessments added to 1609 marks, exactly the difference between Conrad’s ﬁgure and mine. 2. An addition error: the three above-mentioned properties covered 231 hectares total area; allowing for rounding error in the last digit, the area discrepancy of 432 hectares appears to be an addition error on Conrad’s part. With regard to Prince von Thurn und Taxis, Conrad seems to have adopted the same rule as for Prince Carl Anton von Hohenzollern-Sigmaringen, namely that the land of the prince and his heir was conﬂated, whereas I have taken the entries as given in the Handbook: Prince Maximilian Maria died in 1885, and his younger brother, Prince Albert I. Maria, took over the title and the lands. In the Handbooks, Prince Maximilian Maria is listed as owning 13,789 hectares with 131,935 marks of taxable net yield, while Prince Albert Maria is listed as owning 10,693 hectares with 61,487 marks. Added together, these become 24,482 and 193,422—exactly Conrad’s ﬁgures for the Prince of Thurn und Taxis, ﬁrst name unspeciﬁed in his table. Number 17 in Conrad’s list is the Count (later Prince) zu StolbergWernigerode, who ﬁgured so prominently in the previous chapter. Here the

The Biggest Landowners

127

Table 5.2. Reconciliation of discrepancy in holdings of Count Otto zu Stolberg-Wernigerode, c.1882 (a) Data reported in the Handbooks Estate name

District

Province

Area (ha)

Tax value (marks)

Stammgut Schwarza Potarzyce Radenz

Wernigerode Oschersleben Jarotschina Koschminb

Saxony Saxony Posen Posen

14,064 100 1,209 4,277

248,247 1,527 8,953 46,790

Sum

19,650

305,517

Conrad’s ﬁgures My ﬁgures Difference

19,649 20,619 970

305,517 319,202 13,685

(b) Constituent parts of Radenz manor listed separately in Handbook Mycielin Koschmin Posen 666 Bulakow Koschmin Posen 505 Debowiec Koschmin Posen 366 Galazki Koschmin Posen 390 Kaczagorka Koschmin Posen 300 Ludwinow Koschmin Posen 179 Wielowies Koschmin Posen 468 Wronowo Koschmin Posen 408 Wrotkow Koschmin Posen 350 Wziachow Koschmin Posen 1,614

7,833 7,589 2,558 4,745 3,853 2,154 5,569 7,055 3,609 15,510

Sum Conrad’s and Handbook’s totals for the manor of Radenz Difference a b

5,246 4,277

60,475 46,790

969

13,685

Jarotschin district created 1887 from part of Pleschen district. Koschmin district created 1887 from part of Krotoschin district.

discrepancy of 970 hectares and 13,685 marks appears to be the result of Conrad’s failure to check totals against the sums of the parts for manors consisting of several named properties. What Conrad must have done is best seen in Table 5.2. The overall data for the manor of Radenz listed in the Handbook seem to be wrong. They are the exact numbers⁷ listed in an 1872 address book of landowners in Posen,⁸ but the data for the separate properties listed as being constituent ⁷ When converted from morgen and thaler to hectares and marks. ⁸ Wykaz alfabetyczny wszystkich posiadlosci ziemskich w. W. Ksiestwie Poznanskiem/Adressbuch des Grundbesitzes im Grossherzogthum Posen (Berlin: F. Bürde & Co., 1872), 90.

128

The Biggest Landowners

parts of the manor in the 1884 Handbook do not add to those totals, but to the higher totals shown in Table 5.2. Apparently Conrad did not check the reported totals against the sums of the parts. I consider the sum of the individual parts to be the correct total, and the listed area for the manor to be wrong.⁹ Hence the difference between my data and Conrad’s for the overall holdings of Count Otto zu Stolberg-Wernigerode. It is more difﬁcult to explain why Conrad should not have included any one of the owners who held less than the 18,538 hectares of Prince Hatzfeld but more than the 13,642 of Sophie, Grand Duchess of Sachsen-Weimar, in his table of the top 20 landowners by total area owned. These are the ‘11 different owners’ of Table 5.1—of which two are included in Conrad’s totals for Prince (Carl) Anton von Hohenzollern and Prince (Albert) von Thurn und Taxis—so he had nine to choose from. Table 5.3 shows their holdings, according to the Handbook editions of the early to mid-1880s. Since the landholdings of these latifundistas were to be found in several provinces, and they themselves were in some cases royalty, it is very surprising that at least the top one of them, Countess Wilhelmine zu Solms-Roesa, did not appear in Conrad’s table of 20 in place of Archduchess Sophie. It is all the more surprising when we consider that each of the nine apparently omitted landholders appeared in Conrad’s table of all owners with more than 5000 hectares in the same article,¹⁰ in almost every case with exactly the same totals as I have calculated for them. The discrepancy between his higher totals for Princess Marianne of the Netherlands and mine corresponds exactly to the data for her property of Alt Mohrau in Habelschwerdt district of Silesia (92 ha and 623 marks of taxable net yield), which I excluded from the calculation because under 100 hectares. Apparently Conrad’s principle did include under-100 hectare properties in his calculations, at least in this case. Finally, the discrepancy in tax value for Archduchess Sophie of SachsenWeimar no doubt stems from my having had to estimate the tax assessment for one of the subunits (Schimmelei with Wiesenhof, 218 ha) of her manor of Heinrichau (7906 ha in Münsterberg district, Breslau region of Silesia), because it was missing from the Handbook for 1880. My estimate, based on the district averages for the various land types in Münsterberg, was 6774 marks. The actual ﬁgure reported in 1892 for the same subunit was 6781 marks. Conrad must have used some other method to estimate this value, leading to the discrepancy of 3226 marks (his being higher), since our area totals were, for all practical purposes, identical. ⁹ Moreover, the total area for Radenz in the 1896 Handbook is 6779 ha, with 70,551 marks of tax value. All subunits of the manor of Radenz were listed in both years, except that Goreczki and Mittelwalde were new in 1896—total area, 1391 ha, and tax value, 9459 marks. This is close to the difference between my 1884 totals and the Handbook’s 1896 totals: 1553 ha and 10,076 marks. ¹⁰ Conrad, ‘Die Latifundien’, 158–61.

The Biggest Landowners

129

Table 5.3. Large landowners with more land than Grand Duchess of Saxony-Weimar but less than Prince Hatzfeld, c.1882 Name of owner

zu Solms-Roesa, Wilhelmine, Countess HohenzollernSigmaringen, Leopold, Prince Radziwill, Ferdinand, Prince von Arnim, Adolph, Count Princess Marianne of Netherlands zu Carolath-Beuthen, Carl, Prince von Sachsen-Altenburg, Ernst, Duke zu Solms-Baruth, Friedr. Hermann Adolph, Count Henckel von Donnersmarck, Hugo, Count von Mielzynski, Joseph, Count Sk´orzewski, Leo, Count von Thurn und Taxis, Maximilian Maria, Prince sum Omitted by Conrad

Properties owned No. Area (ha)

Notes Tax value (marks)

1

17,413

46,808

Manor of Lasisk, in Silesia Included in Conrad’s data for his father

6

16,692

65463

2

16,398

82,158

14

15,821

112,654

7

15,811

92,828

All in Silesia

14

15,763

114,589

All in Silesia

4

15,651

84,926

6

14,682

74,842

12

14,552

91,732

13

13,932

107,731

4

13,885

80,231

1

13,789

131,935

84 77

184,389 153,908

1,085,897 888,499

Manor +1, all in Posen All in Brandenburg

1 in West Prussia, 3 in Posen Manor of Baruth + others in Brandenburg Manor of Naklo + 11 props., all in Silesia All in Posen Manor in Posen + 3 props. in West Prussia Included in Conrad’s data for his brother

5 . 3 . T H E TO P F I F T Y L A N D OW N E R S B Y TOTA L A R E A A N D TOTA L TA X VA LU E Identifying the largest landowners at one point in time is only one step in the process. To compare how the ranks of these elite landlords changed, we need to make some comparisons over time. Three tables (5.4 through 5.6) that list the top ﬁfty landowners after the Prussian State (hence a ranking that always begins at ‘2’) according to the area owned, are followed by another set of three (5.7 through 5.9) ordered by the assessed tax value of the properties.

130

The Biggest Landowners

Table 5.4. The ﬁfty largest landowners by total area owned, c.1882, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm I Pless, Hans Heinrich XI, Prince Hohenzollern-Sigmaringen, Carl Anton, Prince von Braunschweig, Wilhelm, Duke zu Hohenlohe-Oehringen, Hugo, Prince Prince Frederik of the Netherlands von Ratibor, Viktor Amadeus, Duke Schaffgotsch, Ludwig, Imperial Count City of G¨orlitz Hohenzollern-Sigmaringen, Friedrich Carl, Prince Henckel von Donnersmarck, Guido, Count zu Sagan, Duke Biron von Curland, Calixt, Prince Br¨uhl, Friedrich Stephan, Imperial Count zu Solms-Baruth, Friedrich, Prince von Anhalt-Dessau, Leopold Friedrich Franz, Duke zu Stolberg-Wernigerode, Otto, Count zu Putbus, Prince von Redern, Heinrich, Count von Hatzfeld, Hermann, Prince zu Solms-Roesa, Wilhelmine, Countess Hohenzollern-Sigmaringen, Leopold, Prince Radziwill, Ferdinand, Prince von Arnim, Adolph, Count Princess Marianne of the Netherlands

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

73 9 21

97,348 61,267 43,777

627,007 364,723 129,060

2 3 4

2 4 25

3

40,191

390,078

5

3

12

39,321

232,296

6

9

9

35,418

155,662

7

18

8

33,846

180,932

8

14

6

31,455

125,874

9

26

7 3

29,495 25,690

140,547 94,674

10 11

19 49

16

23,299

130,351

12

24

24 7

23,088 22,895

103,475 178,433

13 14

42 15

5

22,416

123,373

15

27

5

21,935

95,667

16

48

33

21,723

261,585

17

8

4

20,619

319,202

18

6

61 24 2 1

19,701 18,766 18,538 17,413

350,651 227,919 168,877 46,808

19 20 21 22

5 10 16 189

6

16,692

65,463

23

105

2 14 7

16,398 15,821 15,811

82,158 112,654 92,828

24 25 26

71 33 52

The Biggest Landowners

131

Table 5.4. (Continued) Name of owner

zu Carolath-Beuthen, Carl, Prince von Sachsen-Altenburg, Ernst, Duke zu Solms-Baruth, Friedr. Hermann Adolph, Count Henckel von Donnersmarck, Hugo, Count von Mielzynski, Joseph, Count Sk´orzewski, Leo, Count von Thurn und Taxis, Maximilian Maria Lamoral, Prince von Sachsen-Weimar, Sophie, Archduchess von Schleswig-HolsteinSonderburg-Augustenburg, Duke von Eckardstein, Ernst, Baron zu Hohenlohe-Ingelﬁngen, Friedr. Wilhelm, Prince Raczynski, Karl, Count von der Schulenburg, Dietrich, Count von der Schulenburg, Countess, widow von Strachwitz’s heirs, Hyacinth, Count Kennemann, Hermann von Bismarck, Otto, Prince von Thurn und Taxis, Albert Maria Lamoral, Prince von Maltzan, Andreas Joachim Mortimer, Count Reuss j. L., Heinrich XIV, Prince von Waldow-Reitzenstein, Carl Ernst Sigismund von Sydow’s heirs von Frankenberg-Ludwigsdorf, Friedrich, Count zu Dohna-Schlobitten, Richard Friedrich, Count zu Dohna-Finckenstein, Count total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area value

Tax

14

15,763

114,589

27

32

4

15,651

84,926

28

61

6

14,682

74,842

29

84

12

14,552

91,732

30

53

13 4 1

13,932 13,885 13,789

107,731 80,231 131,935

31 32 33

38 76 23

4

13,646

273,906

34

7

2

13,610

76,718

35

80

16 6

13,485 13,196

132,055 41,699

36 37

22 223

6 3

13,022 12,785

82,715 42,425

38 39

69 219

3

12,525

42,437

40

218

1

11,681

49,913

41

169

16 9 2

11,270 11,249 10,693

96,519 56,902 61,487

42 43 44

46 133 118

1

10,663

82,423

45

70

3 3

10,637 10,579

58,673 16,830

46 47

128 1,007

6 2

9997 9872

65,343 69,689

48 49

106 93

12

9587

99,932

50

45

2

9499

56,635

51

137

513

1,043,173

6,892,584

132

The Biggest Landowners

Table 5.5. The ﬁfty largest landowners by total area owned, c.1895, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm II Pless, Hans Heinrich XI., Prince Hohenzollern-Sigmaringen, Leopold, Prince zu Hohenlohe-Oehringen, Hugo, Prince von Ratibor, Viktor Amadeus, Duke von Arnim, Traugott Herrmann, Count zu Stolberg-Wernigerode, Otto, Prince City of G¨orlitz Henckel von Donnersmarck, Edgar, Count Hohenzollern-Sigmaringen, Friedrich Leopold, Prince von Thurn und Taxis, Albert Maria Lamoral, Prince zu Hohenlohe-Ingelﬁngen, Friedr. Wilhelm, Prince zu Sagan, Duke von Anhalt-Dessau, Leopold Friedr. Franz Nicolaus, Duke von Br¨uhl, Friedrich Franz, Imperial Count von Sachsen, Friedrich August, King zu Solms-Baruth, Friedrich, Prince Biron von Curland, Gustav, Prince von Redern, Wilhelm, Count Radziwill, Ferdinand, Prince zu Putbus, Prince von Hatzfeld, Hermann, Prince Neuzelle monastery Schaffgotsch, Countess, and her children von Arnim, Detloff, Count

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

79 10 20

97,180 61,138 44,901

697,353 368,766 162,974

2 3 4

2 3 19

15

41,661

255,164

5

8

10

38,211

177,401

6

14

5

34,878

117,781

7

34

7

33,697

135,944

8

26

8 21

29,722 29,648

144,751 168,477

9 10

23 24

22

25,552

118,384

11

33

2

24,870

192,595

12

12

18

24,114

57,262

13

133

31 33

23,113 23,096

103,475 263,505

14 15

50 6

6

22,805

125,421

16

28

5

21,920

228,207

17

10

4

21,909

95,667

18

55

5

21,502

163,785

19

18

26 6 60 2 5 1

20,358 19,065 19,734 18,796 18,425 18,289

258,654 105,744 338,958 168,499 145,925 67,270

20 21 22 23 24 25

7 4 47 16 22 100

14

17,711

90,494

26

64

The Biggest Landowners

133

Table 5.5. (Continued) Name of owner

Hohenzollern-Sigmaringen, Albrecht, Prince zu Carolath-Beuthen, Carl, Prince von Eckardstein, Ernst, Baron zu Dohna-Waldburg, Rich. Friedr. Count Henckel von Donnersmarck, Hugo, Count von Frankenberg-Ludwigsdorf, Friedrich, Count von Flemming, Count von Sachsen-Weimar, Sophie, Archduchess zu Stolberg-Wernigerode, Christian Ernst, Prince Kennemann, Hermann von Waldow-Reitzenstein, Carl Ernst Sigismund Hohenzollern, Wilhelm, Crown Prince von Schleswig-HolsteinSonderburg-Augustenburg, Duke von der Schulenburg, Dietrich, Count von Sachsen-Altenburg, Ernst, Duke Hohenzollern-Sigmaringen, Carl Anton, Prince von der Schulenburg, Louise, Countess, widow Royal Court Chamber, in Berlin zu Solms-Baruth, Prince Sk´orzewski, Leo, Count Reuss j.L., Heinrich XIV, Prince von Mielzynski, Joseph, Count Schaffgotsch, Imperial Count, a minor von Thiele-Winckler, Hubert von Lehndorff, Carl Meinhard, Count total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

9

16,674

97,742

27

53

14

16,454

134,676

28

27

14 32

16,427 16,275

121,473 171,418

29 30

32 15

12

15,625

98,243

31

51

3

15,306

92,982

32

63

24 4

14,851 14,774

109,877 303,140

33 34

42 5

10

14,064

248,247

35

9

18 3

13,897 13,888

110,958 16,830

36 37

41 1,005

4

13,702

149,805

38

20

2

13,627

77,198

39

85

4

13,496

44,851

40

200

3

13,442

68,775

41

98

7

12,775

41,664

42

226

4

12,671

42,802

43

213

7 1 10 4 9 1

12,661 12,213 12,040 11,373 11,002 10,940

76,768 50,140 73,908 72,367 80,789 48,915

44 45 46 47 48 49

86 165 88 91 80 176

18 3

10,902 10,708

96,552 34,312

50 51

54 322

635

1,092,082

7,216,888

134

The Biggest Landowners

Table 5.6. The ﬁfty largest landowners by total area owned, c.1907, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm II Pless, Hans Heinrich XV, Prince zu Hohenlohe-Oehringen, Christian Kraft, Prince zu Stolberg-Wernigerode, Christian-Ernst, Prince zu Solms-Baruth, Friedrich, Prince von Ratibor, Viktor Amadeus, Duke Hohenzollern-Sigmaringen, Friedrich Leopold, Prince Hohenzollern, Wilhelm, Crown Prince Schaffgotsch, Friedrich, Imperial Count City of G¨orlitz von Anhalt-Dessau, Friedrich, Duke von Arnim-Muskau, Traugott Hermann, Count von Thurn und Taxis, Albert Maria Lamoral, Prince zu Hohenlohe-Ingelﬁngen, Karl Gottfried, Prince Royal Settlement Commissiona Landbank A.-G., Berlin, Behrenstr. 14/16 Duc de Talleyrand, Boson, Duke of Sagan von Sachsen, Friedrich August, King von Br¨uhl, Maria Friedrich Franz, Imperial Count Henckel von Donnersmarck, Guido, Prince Hohenzollern-Sigmaringen, Friedrich Heinrich, Prince Neuzelle monastery von Redern, Wilhelm, Count zu Trachenberg, Hermann, Prince Biron von Curland, Gustav, Prince

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

89 5 11

110,764 50,505 43,732

836,393 319,364 281,099

2 3 4

2 4 8

3

40,581

299,616

5

6

10

38,707

178,191

6

15

5

33,393

173,927

7

16

23

32,246

146,415

8

22

18

32,215

99,653

9

49

3

30,851

117,989

10

39

8 34

29,412 28,472

143,603 322,692

11 12

24 3

2

26,814

79,525

13

78

2

25,642

202,981

14

12

3

25,489

65,485

15

108

47 37

24,073 23,479

212,234 171,944

16 17

11 17

1

23,418

111,148

18

45

25

22,534

231,305

19

10

3

22,183

109,585

20

47

6

21,876

119,955

21

37

4

20,772

131,279

22

32

7 24 2

20,057 19,288 19,265

148,953 248,329 171,169

23 24 25

21 10 18

1

18,305

142,262

26

26

The Biggest Landowners

135

Table 5.6. (Continued) Name of owner

von Tiele-Winckler, Franz Hubert, Count Kennemann, Hermann von Arnim, Detloff, Count zu Putbus, Prince zu Carolath-Beuthen, Karl, Prince Radziwill, Ferdinand, Prince Royal University of Greifswald Henckel von Donnersmarck, Lazy, Arthur, Edgar, Counts von Sachsen-Weimar, Wilhelm Ernst, Grand Duke von der Schulenburg, Louise, Countess, widow zu Schleswig-Holstein, Ernst G¨unther, Duke von der Schulenburg, Dietrich, Count Sk´orzewski, Witold, Count Pomeranian Settlement Society, Inc. Reuss j. L., Heinrich XIV, Prince City of Berlin von Maltzan, Andreas, Count von Waldow und Reitzenstein, Karl zu Stolberg-Wernigerode, Udo, Count von Hochberg, Wilhelm, Imperial Count Hohenzollern, Friedrich Wilhelm Viktor August Ernst, Crown Prince City of Bunzlau von Reichenbach-Gosch¨utz, Heinrich, Count zu Dohna-Schlodien, Hermann, Count Finck von Finckenstein, Count total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

13

17,674

166,647

27

19

25 13 56 2

17,315 16,719 16,666 16,154

137,511 39,622 295,104 122,259

28 29 30 31

28 269 7 35

2 32 4

15,563 15,304 14,916

62,544 266,827 96,220

32 33 34

119 9 52

10

14,733

305,284

35

5

7

14,263

70,226

36

98

2

13,899

83,782

37

73

4

13,796

43,888

38

229

1 15

12,200 11,966

73,225 109,671

39 40

94 46

5 24 3 2

11,752 11,404 11,246 10,366

76,795 143,781 85,481 23,336

41 42 43 44

85 23 70 629

9

10,347

138,538

45

27

3

10,098

39,074

46

274

1

10,094

142,551

47

25

2 4

10,037 9,994

48,187 70,391

48 49

190 96

3

9,603

61,537

50

123

3

9,523

50,685

51

174

618

1,109,705

7,818,262

a Properties of the Royal Settlement Commission still intact and administered as large units. Properties subdivided or in process of subdivision not included.

136

The Biggest Landowners

The top ﬁfty landowners as a group owned more than 1 million hectares— therefore an average of over 20,000 hectares each—in each of the three reference years, and this accounted for almost exactly 10 per cent of the total area of all large properties in the seven provinces in each of the three years, that is, less than 0.5 per cent of all owners owned 5 per cent of all the land.¹¹ To join this club required very close to 10,000 hectares of land (slightly higher c.1895), and—at least at ﬁrst—a rank in the high nobility. In the ﬁrst year there were only four exceptions to the noble rank rule: the city of Görlitz, a baron, heirs of an untitled nobleman, and one lone commoner, Hermann Kennemann.¹² At the mid-1890s the number of exceptions remained at four—the City of Görlitz, the Neuzelle monastery, one untitled nobleman, and the same Mr Kennemann.¹³ Compared to the two earlier years, the 1907 group of exceptions had expanded to ten and was practically ecumenical: two other cities, Berlin and Bunzlau, had joined Görlitz in the ranks, along with the University of Greifswald; the Royal Settlement Commission appeared temporarily as a large landlord,¹⁴ along with the Pomeranian Settlement Society and the Landbank Corporation of Berlin. The Neuzelle monastery remained in the list, Herrmann Kennemann moved up considerably in the ranks, and the one untitled nobleman from c.1895, Karl von Waldow und Reitzenstein, slid down the ladder a bit, but still maintained his position in the top ﬁfty. Tables 5.4 through 5.6 show a slight trend toward concentration of landholding by area, as both the total area and the total tax value of the landholdings of this top-ﬁfty group increased over the period. Tables 5.7 through 5.9 show a similar slight trend, this time presenting the top ﬁfty landowners according to the land tax assessment of their properties. When we switch focus from simple extent of landholdings to their value—even if tax assessment be an imperfect proxy for value—the picture of who the top landlords were changes considerably. No fewer than 22 of the 50 top landowners by tax assessment c.1882 (Table 5.7) were not among the top 50 by total area. Of ¹¹ Were we to factor in one more owner, the Prussian State, then 0.5% of all owners held about one sixth of all the land in the seven eastern provinces. ¹² Kennemann was a rather uncommon commoner. He was one of the founders of the Deutscher Ostmarkenverein (German Eastern Marches Society), an avowedly anti-Polish organization, which the Poles called ‘Hakata’ after the initials of its founders Hansemann, Kennemann, and Tiedemann. (See e.g. Zdislaw Kaczmarczyk, Kolonizacja niemiecka na wschod od Odry [German colonization east of the Oder], (Pozna´n: Wydawnictwo Instytutu Zachodniego, 1945), 217.) All three were substantial landowners: Herr von Hansemann was owner of nearly 8000 ha of land, mostly in Posen, and von Tiedemann, 1378 ha, all in Posen, in the mid-1890s. ¹³ I took properties listed that year as belonging to the Royal Court Chamber (Königlicher Hofkammer) to be the personal property of the king, especially since each of them was listed c.1907 as a royal entail (Fideikommiss). Cf. Ch. 4 n. 8. ¹⁴ One could leave out the Royal Settlement Commission on the grounds that its properties, until subdivided and sold off, were State properties. Doing so would have added Prince Jobst Christian zu Stolberg-Rossla, with 9233 ha having 128,127 marks of taxable value, to the bottom of the list.

The Biggest Landowners

137

Table 5.7. The ﬁfty largest landowners by tax assessment, c.1882, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm I von Braunschweig, Wilhelm, Duke Pless, Hans Heinrich XI, Prince zu Putbus, Prince zu Stolberg-Wernigerode, Otto, Count von Sachsen-Weimar, Sophie, Grand Duchess von Anhalt-Dessau, Leopold Friedrich Franz, Duke zu Hohenlohe-Oehringen, Hugo, Prince von Redern, Heinrich, Count J. G. Boltze Co., in Salzm¨unde Royal University of Greifswald zu Stolberg-Rossla, Botho, Count von Ratibor, Viktor Amadeus, Duke Biron von Curland, Calixt, Prince von Hatzfeld, Hermann, Prince Brunswick Ducal Fiscus Prince Frederik of the Netherlands City of G¨orlitz von Oppersdorff sen., Eduard, Imperial Count zu Lichnowsky, Carl, Prince von Eckardstein, Ernst, Baron von Thurn und Taxis, Maximilian Maria Lamoral, Prince Henckel von Donnersmarck, Guido, Count Hohenzollern-Sigmaringen, Carl Anton, Prince Schaffgotsch, Ludwig, Imperial Count

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

73 3

97,348 40,191

627,007 390,078

2 5

2 3

9 61 4

61,267 19,701 20,619

364,723 350,651 319,202

3 19 18

4 5 6

4

13,646

273,906

34

7

33

21,723

261,585

17

8

12

39,321

232,296

6

9

24 7 20 2

18,766 4,235 9,153 8,874

227,919 218,961 189,840 186,484

20 216 57 60

10 11 12 13

8

33,846

180,932

8

14

7

22,895

178,433

14

15

2 5 9

18,538 3,285 35,418

168,877 160,475 155,662

21 313 7

16 17 18

7 3

29,495 6,618

140,547 133,289

10 94

19 20

15 16 1

8,691 13,485 13,789

132,704 132,055 131,935

65 36 33

21 22 23

16

23,299

130,351

12

24

21

43,777

129,060

4

25

6

31,455

125,874

9

26

138

The Biggest Landowners

Table 5.7. (Continued) Name of owner

von Br¨uhl, Friedrich Stephan, Imperial Count Convent St. Annen und Brigitten, in Stralsund City of Stralsund von Buch, Johann Georg zu Stolberg-Stolberg, Alfred, Prince zu Carolath-Beuthen, Carl, Prince von Arnim, Adolph, Count von Krassow-Pansevitz, Count von Rothenburg, Friedrich Wilhelm Carl, Count von Heyden-Cartlow, Adam Werner, Count City of Greifswald von Mielzynski, Joseph, Count von Hansemann, in Berlin Friedenthal, Carl Rudolph, Dr Ducal Chamber of Brunswick zu Sagan, Duke Zimmermann, Max von Tschirschky-Renard, Mortimer, Count zu Dohna-Schlobitten, Richard Friedrich, Count Kennemann, Hermann von Maltzahn, Helmuth, Baron zu Solms-Baruth, Friedrich, Prince Hohenzollern-Sigmaringen, Friedrich Carl, Prince von Hochberg, Hans Heinrich XIV Bolko, Imperial Count von der Lancken-Wakenitz, Baron total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

5

22,416

123,373

15

27

13

3,558

122,328

275

28

9 8 1

6,018 6,928 8,794

118,477 118,260 115,864

116 88 62

29 30 31

14

15,763

114,589

27

32

14 9 8

15,821 4,612 9,304

112,654 112,652 111,885

25 191 55

33 34 35

9

4,533

108,970

197

36

12 13 5 8 3 24 3 8

6,176 13,932 8,733 8,507 1,812 23,088 1,862 7,015

108,476 107,731 105,558 105,550 104,426 103,475 102,542 100,973

112 31 64 67 775 13 749 86

37 38 39 40 41 42 43 44

12

9,587

99,932

50

45

16 5 5

11,270 4,258 21,935

96,519 95,761 95,667

42 215 16

46 47 48

3

25,690

94,674

11

49

9

8,612

94,395

66

50

8

4,081

93,297

228

51

592

893,740

8,210,874

The Biggest Landowners

139

Table 5.8. The ﬁfty largest landowners by tax assessment, c.1895, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm II Pless, Hans Heinrich XI, Prince zu Putbus, Prince von Sachsen-Weimar, Sophie, Grand Duchess von Anhalt-Dessau, Leopold Friedr. Franz Nicolaus, Duke von Redern, Wilhelm, Count zu Hohenlohe-Oehringen, Hugo, Prince zu Stolberg-Wernigerode, Christian Ernst, Prince von Sachsen, Friedrich August, King Royal University of Greifswald von Thurn und Taxis, Albert Maria Lamoral, Prince zu Stolberg-Rossla, Jobst Christian, Prince von Ratibor, Viktor Amadeus, Duke zu Dohna-Waldburg, Rich. Friedr. Count von Hatzfeld, Hermann, Prince Brunswick Ducal Fiscus Biron von Curland, Gustav, Prince Hohenzollern-Sigmaringen, Leopold, Prince Hohenzollern, Wilhelm, Crown Prince von Buch, Johann Georg Neuzelle Monastery City of G¨orlitz Henckel von Donnersmarck, Guido, Count zu Lichnowsky, Carl, Prince zu Stolberg-Wernigerode, Otto, Prince

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

79 10 60 4

97,180 61,138 19,734 14,774

697,353 368,766 338,958 303,140

2 3 20 34

2 3 4 5

33

23,096

263,505

15

6

26 15

20,358 41,661

258,654 255,164

20 5

7 8

10

14,064

248,247

35

9

5

21,920

228,207

17

10

21 2

9,626 24,870

204,909 192,595

56 12

11 12

2

8,989

192,009

63

13

10

38,211

177,401

6

14

32

16,275

171,418

30

15

2 5 5

18,796 3,624 21,502

168,499 168,469 163,785

22 263 19

16 17 18

20

44,901

162,974

4

19

4

13,702

149,805

38

20

10 5 8 18

8,184 18,425 29,722 26,756

147,438 145,925 144,751 137,194

72 23 9 10

21 22 23 24

18 7

8,844 33,697

136,814 135,944

65 8

25 26

140

The Biggest Landowners

Table 5.8. (Continued) Name of owner

zu Carolath-Beuthen, Carl, Prince von Br¨uhl, Friedrich Franz, Imperial Count von Oppersdorff, Hans Georg, Imperial Count Convent St. Annen und Brigitten, Stralsund J. G. Boltze Co. von Eckardstein, Ernst, Baron Hohenzollern-Sigmaringen, Friedrich Leopold, Prince von Arnim, Traugott Herrmann, Count von Rothenburg, Friedrich Wilhelm Carl, Count zu Stolberg-Stolberg, Alfred, Prince City of Greifswald von Krassow-Pansevitz, Count von Klot-Trautvetter, Count von Zimmermann, Max Kennemann, Hermann von Flemming, Count von Heyden, J¨urgen, minor City of Berlin St. Jürgen monastery von Seydlitz, Ernst Julius, minor Radziwill, Ferdinand, Prince City of Stralsund Ducal Chamber of Brunswick zu Sagan, Duke Henckel von Donnersmarck, Hugo, Count total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

14

16,454

134,676

28

27

6

22,805

125,421

16

28

7

5,326

124,679

136

29

13

3,550

122,254

267

30

2 14 22

2,493 16,427 25,552

121,662 121,473 118,384

467 29 11

31 32 33

5

34,878

117,781

7

34

9

9,572

117,155

57

35

1

8,794

116,137

67

36

13 9 4 4 18 24 9 12 15 2 6 9 3 31 12

6,472 4,614 7,130 2,065 13,897 14,851 4,535 8,320 4,570 5,861 19,065 5,814 1,812 23,113 15,625

113,019 112,634 111,377 111,144 110,958 109,877 108,973 106,932 106,820 105,787 105,744 105,366 104,426 103,475 98,243

93 181 82 642 36 33 188 70 184 116 21 120 764 14 31

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

596

826,464

7,698,968

The Biggest Landowners

141

Table 5.9. The ﬁfty largest landowners by tax assessment, c.1907, excluding the Prussian State Name of owner

Hohenzollern, King Wilhelm II von Anhalt-Dessau, Friedrich, Duke Pless, Hans Heinrich XV, Prince von Sachsen-Weimar, Wilhelm Ernst, Grand Duke zu Stolberg-Wernigerode, Christian-Ernst, Prince zu Putbus, Prince zu Hohenlohe-Oehringen, Christian Kraft, Prince Royal University of Greifswald von Redern, Wilhelm, Count von Sachsen, Friedrich August, King Royal Settlement Commissiona von Thurn und Taxis, Albert Maria Lamoral, Prince zu Stolberg-Rossla, Jobst Christian, Prince zu Solms-Baruth, Friedrich, Prince von Ratibor, Viktor Amadeus, Duke Landbank A.-G., Berlin, Behrenstr. 14/16 zu Trachenberg, Hermann, Prince von Tiele-Winckler, Franz Hubert, Count von Seidlitz-Sandreczki, Ernst Julius, Count Neuzelle monastery Hohenzollern-Sigmaringen, Friedrich Leopold, Prince City of Berlin City of G¨orlitz Hohenzollern, Friedrich Wilhelm, Crown Prince Biron von Curland, Gustav, Prince

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

89 34

110,764 28,472

836,393 322,692

2 12

2 3

5 10

50,505 14,733

319,364 305,284

3 35

4 5

3

40,581

299,616

5

6

56 11

16,666 43,732

295,104 281,099

29 4

7 8

32 24 25

15,304 19,288 22,534

266,827 248,329 231,305

33 24 19

9 10 11

47 2

24,073 25,642

212,234 202,981

16 14

12 13

2

9,489

185,342

52

14

10

38,707

178,191

6

15

5

33,393

173,927

7

16

37

23,479

171,944

17

17

2

19,265

171,169

24

18

13

17,674

166,647

26

19

3

7,524

152,115

74

20

7 23

20,057 32,246

148,953 146,415

23 8

21 22

24 8 1

11,404 29,412 10,094

143,781 143,603 142,551

42 11 47

23 24 25

1

18,305

142,262

26

26

142

The Biggest Landowners

Table 5.9. (Continued) Name of owner

zu Stolberg-Wernigerode, Udo, Count Kennemann, Hermann von Lichnowsky, Karl Max, Prince von Zimmermann, August von Oppersdorff, Hans Georg, Imperial Count Hohenzollern-Sigmaringen, Friedrich Heinrich, Prince von Eickstedt-Peterswaldt, Ernst, Count von Magnis, Anton, Imperial Count zu Carolath-Beuthen, Karl, Prince City of Greifswald Henckel von Donnersmarck, Guido, Prince von Veltheim, Franz Carl Adolph Werner, Baron Schaffgotsch, Friedrich, Imperial Count Convent St. Annen und Brigitten, Stralsund City of Greifswald & Hospital St. Spiritus zu Dohna-Schlobitten, Richard, Prince vom Rath, Schoeller u. Skene, LLC zu Stolberg-Stolberg, Alfred, Prince Duc de Talleyrand, Boson, Duke of Sagan Pomeranian Settlement Society von Br¨uhl, Maria Friedrich Franz, Imperial Count von Rothenburg, Friedrich Wilhelm Karl, Count Hohenzollern, Wilhelm, Crown Prince von Behr-Negendank, Count von Hochberg, Hans Heinrich XIV Bolko, Imperial Count total

Properties owned Number Area (ha)

Tax value (marks)

Rank by Area

Tax value

9

10,347

138,538

45

27

25 1

17,315 8,839

137,511 136,822

28 60

28 29

1 2

2,798 6,070

136,467 134,811

383 115

30 31

4

20,772

131,279

22

32

7

9,233

128,127

56

33

19

9,383

127,828

55

34

2

16,154

122,259

31

35

24 6

8,735 21,876

120,295 119,955

62 21

36 37

6

5,383

119,789

138

38

3

30,851

117,989

10

39

13

3,550

117,688

275

40

12

5,224

116,983

149

41

15

9,182

116,315

57

42

11

3,931

116,058

241

43

1

8,794

115,864

61

44

1

23,418

111,148

18

45

15 3

11,966 22,183

109,671 109,585

40 20

46 47

6

9,402

104,364

53

48

18

32,215

99,653

9

49

9 2

5,382 9,040

99,135 97,745

139 58

50 51

689

995,386

8,874,007

a Properties of the Royal Settlement Commission still intact and administered as large units. Properties subdivided or in process of subdivision not included.

The Biggest Landowners

143

these 22, 9 were from the high nobility and four from the lower nobility, and the rest were scattered: 2 non-noble individuals, 2 cities, 1 convent, 1 university, 1 business ﬁrm, and 2 state organs from the duchy of Brunswick. Two of these made it into the top 50 by tax assessment with less than 2000 hectares of land: the Ducal Chamber of Brunswick owned three properties totalling 1812 hectares in Wanzleben district of the province of Saxony that were nearly all ploughland and meadow, bringing their average tax assessment up to 57.6 marks/hectare. Max Zimmermann (later to become von Zimmermann) owned 1862 hectares in three properties in Merseburg district of Saxony, with 55.1 marks per hectare of land tax net yield. The concentration of tax assessment in the hands of the top ﬁfty landowners was less than that of area: 8.1 to 8.6 per cent, depending on the year. This is, of course, a reﬂection of the already-noted fact that smaller owners tended to have higher proportions of more highly assessed land, principally ploughland and meadow, in their properties, while large owners tended to have higher proportions of less valuable land, principally pasture and forest. The picture from the mid-1890s is similar to that from the early 1880s: twenty of the ﬁfty largest owners by tax value were not in the list of the ﬁfty largest owners by total area of holdings. These twenty were, by and large, the same entities as in the earlier year: the cities of Greifswald and Stralsund were repeats in the group, as were the two arms of the Brunswick ﬁscus, the University of Greifswald, the ﬁrm J. G. Boltze, and the convent of St Annen und Brigitten. New to the list of corporate ‘upstarts’ were the City of Berlin and the St Jürgen monastery on the island of Rügen in Pomerania. Individual persons in the group of twenty included seven members of the high nobility and four members of the lower nobility, including the former commoner, Max von Zimmermann, but no non-noble individuals. By c.1907 the group of owners not on both lists of ﬁfty top owners had shrunk a little further, to eighteen: thirteen of these were members of the high nobility, only one of the lower nobility; the city of Greifswald ﬁgured twice, once on its own and once in partnership with St Spiritus Hospital; a different business ﬁrm, vom Rath, Schoeller u. Skene LLC, and only one convent, St Annen und Brigitten, rounded out the group. This all seems part of the general trend toward holdings that were more valuable per hectare, including particularly on the part of the high nobility. Big landowners were shedding low-value land and acquiring high-value land, so the two rankings have begun to show some convergence. As we have seen in the previous chapter, the chief purchaser of the lower-valued land, mainly forest, appears to have been the biggest landowner of all, the Prussian State. A focus on entities that made their way into one list but not the other should not, however, obscure the vision of who and what these biggest of all owners were overall. Table 5.10 presents a summary of the contents of the preceding six tables, according to the general socio-economic position of the owners.

144

The Biggest Landowners

Table 5.10. The top ﬁfty landowners by area and rank, grouped by socio-economic status Owner class

Number c.1882 By By tax area (ha) value (marks)

Number c.1895 By By tax area (ha) value (marks)

Number c.1907 By By tax area (ha) value (marks)

High nobilitya Lesser nobilitya ‘Bourgeois’ Religious body Non-Prussian State Cityb Bank or business ﬁrm Otherb

46 2 1

35 4 3 1 2

44 3 1 1

33 5 1 3 2

40 1 1 1

36 2 1 2

1

3 1

1

4 1

3 1

3.5 2

1

3

3.5

total

50

50

50

50

50

50

1

a

High nobility deﬁned as rank of count(ess) and above, including royalty; lesser nobility deﬁned as untitled nobles and baron(esse)s. b In ‘City’ and ‘Other’ c.1907 0.5 represents joint ownership by the City of Greifswald and the St. Spiritus Hospital.

In terms of total area owned, the socioeconomic makeup of the top ﬁfty landowners did not change in any signiﬁcant way between the early 1880s and the mid-1890s. The high nobility were overwhelmingly dominant in all three years, but their dominance shrank between c.1895 and c.1907, and the lower nobility had nearly disappeared from the top ﬁfty by c.1907. Even after these losses, the aristocracy as I have deﬁned it still made up between 80 and 90 per cent of the top ﬁfty owners ranked by total area of their estates. The picture is considerably more varied when we consider the taxable value of the estates. The top ﬁfty by this measure were not so heavily dominated by the high nobility, but—on the other hand—the degree of their dominance did not decrease by the end of the period under review. The lower nobility clearly lost out on this measure, especially considering that Max Zimmermann moved into the ranks of the lower nobility by 1895, and was still one of the top ﬁfty landowners c.1907. The Duchy of Brunswick disappeared after 1895, as the ruling house died out, a fact already noted in Ch. 4. Where did these great landlords have their estates? Table 5.11 shows the distribution by province: from the table the position of Silesia as the home of the great landed magnates is very clear, although its relative position declined slightly as time passed. Among the other provinces, an unexpected result of this comparison is the disproportion in the position of Pomerania when its share of the tax value of the top ﬁfty owners by tax value is compared to its share of the total area of the top ﬁfty by area. Of the 675 properties owned by the top ﬁfty

The Biggest Landowners

145

Table 5.11. Share of each province in holdings of top ﬁfty landowners (%) Province

c.1882 Area (ha)

Tax value (marks)

c.1895 Area (ha)

(a) Top ﬁfty landowners by total area owned East Prussia 1.7 2.6 West Prussia 5.1 3.3 Brandenburg 17.8 15.6 Pomerania 5.5 7.1 Posen 17.5 16.0 Silesia 49.9 48.5 Saxony 2.5 7.0 total

100.0

100.0

Tax value (marks)

100.0

100.0

Tax value (marks)

3.3 4.0 19.0 6.8 15.8 48.9 2.3

3.8 2.5 16.2 9.2 15.0 46.7 6.5

1.6 9.0 17.3 6.6 16.7 46.7 2.0

2.2 8.4 14.3 10.2 17.3 42.3 5.3

100.0

100.0

100.0

100.0

2.9 4.7 17.7 11.0 8.5 50.4 4.8

3.5 2.9 14.0 15.8 7.9 44.9 11.0

100.0

100.0

(b) Top ﬁfty landowners by total tax assessment of their properties East Prussia 2.0 2.1 2.7 2.9 West Prussia 3.2 1.3 3.2 1.5 Brandenburg 14.3 11.4 16.9 14.1 Pomerania 10.3 17.4 12.1 19.0 Posen 12.7 9.2 13.0 9.5 Silesia 51.6 42.1 46.5 38.0 Saxony 5.9 16.4 5.6 15.1 total

c.1907 Area (ha)

100.0

100.0

owners by value c.1895, 192 were to be found in the province of Pomerania. These were not particularly large properties, hence their rather high number. They were dominated by the relatively high-tax-per-hectare properties of the St Annen und Brigitten convent, Count von Klot-Trautvetter, Jürgen von Heyden, and the University of Greifswald. The largest property with more than the median value of taxable net yield per hectare for these properties in Pomerania¹⁵ was the 2292-hectare manor of Hohendorf in Franzburg district of the Stralsund administrative region, with 2159 hectares of ‘agricultural area’ as deﬁned in the German agricultural censuses¹⁶ and owned by Count von Klot-Trautvetter. The next largest was an entailed estate of Jürgen von Heyden, 1059 hectares in the district of Demmin, Stettin administrative region. By the time we get to the ﬁfth-largest property with abovemedian tax value per hectare, we are already below the 1000-hectare threshold. ¹⁵ The median value was 19.3 marks per ha. The median value of taxable net yield for all properties of 100 ha or more in Pomerania was only half of that: 9.7 marks/ha. ¹⁶ Deﬁned as ‘ploughland, garden (except for ﬂower gardens), meadow, pasture, and areas devoted to cultivation of vines, fruit, vegetables, tobacco, etc.’ Germany. Kaiserliches Statistisches Amt, Statistik des Deutschen Reichs, ns 5, Landwirthschaftliche Betriebstatistik nach der allgemeinen Berufszählung vom 5. Juni 1882 (Berlin: Puttkammer & Mühlbrecht, 1885), 1*.

146

The Biggest Landowners

Of the 21 properties owned by the University of Greifswald in Pomerania, only 4 fell below the median tax value per hectare for those properties of the 50 biggest owners to be found in that province, and these were generally high-value properties to start with. Similarly, the convent of St Annen und Brigitten and the monastery of St Jürgen together owned 28 properties totalling 8120 hectares, of which only four fell below this median. The provinces of Posen and Brandenburg had similar positions in each of the panels, although Brandenburg’s advantage was somewhat more marked when considering the properties of the 50 owners with the highest tax assessments. Both East and West Prussia were relatively minor players in these ranks, but the row for Saxony—the province with the highest-value land on a per-hectare basis—reveals the effect of this high value, especially among the lands of the top 50 owners by tax value. 5 . 4 . T H E S U PE R E L I T E : T H E TO P T E N L A N D OW N E R S Except for the Hohenzollerns, it was common for a great landlord to have his or her holdings concentrated in a single province. Table 5.12, for the ten largest landlords by total area owned, illustrates this point. In Table 5.12, none of the top ten had any property in the province of East Prussia, and the non-Hohenzollern members of the group were overwhelmingly concentrated in the province of Silesia. In fact, c.1882 only one of the top ten great landowners who was not a Hohenzollern had land in more than one province, namely Prince Pless, whose principal seat was in Silesia. When Prince Hans Heinrich XI died and his lands in Posen went to his other son, Imperial Count Wilhelm von Hochberg, the new Prince Hans Heinrich XV of Pless also became a uni-provincial landowner, as he appears in the third panel of the table. In the mid-1890s two other non-Hohenzollerns held land in two provinces: Count Traugott von Arnim and Prince Otto zu Stolberg-Wernigerode, although both of them were clearly primarily in a single province, Silesia in both cases. The family seat of Prince Otto, however, was in the province of Saxony, but the result in the table follows from my principle of not adjusting the name of the owner of a property in any version of the Handbook, even if this causes apparent anomalies under some circumstances: as earlier came up in reconciling discrepancies between my ﬁgures and Conrad’s with respect to Prince Carl Anton von Hohenzollern-Sigmaringen and Prince Maximilian von Thurn und Taxis for the early 1880s. Prince Otto zu Stolberg-Wernigerode died in 1896, between publication dates of Handbooks used in the c.1895 database. His properties in Saxony therefore appeared as owned by Prince Christian Ernst zu StolbergWernigerode in the c.1895 database, for that is how they were recorded in the 1899 Handbook for that province. Using Conrad’s principle, Prince Otto would have appeared as landowner in three provinces and would have occupied a higher

The Biggest Landowners

147

Table 5.12. Top ten owners by total area excluding the Prussian State, by province (ha) Owner

Province EP WP

c.1882 King Wilhelm I Prince Pless Prince Carl Antona Duke of Braunschweig Prince HohenloheOehringen Prince Frederik of NL Duke of Ratibor Count Schaffgotsch City of G¨orlitz Prince Friedrich Carla total

POM

POS

SL

SA

51,348

13,411

6,381

21,239

4,969

13,105

9,744 29,612

51,523 1,060

61,267 43,777

40,191

40,191

39,321

39,321

35,418

35,418

33,846 31,455

33,846 31,455

29,495

29,495 25,690

25,690 0

c.1895 King Wilhelm II Prince Pless Prince Leopolda Prince HohenloheOehringen Duke of Ratibor Count von Arnim Prince StolbergWernigerode City of G¨orlitz Count Henckel von Donnersmarck Prince Friedrich Carla total

Total BBG

25,690

51,348

26,515

45,737

283,548

4,969

437,807

1,073

39,415

22,842 6,425

21,723 51,039 888 41,661

4,970

20,162

7,158 10,099 17,426

97,180 61,138 44,901 41,661

3,246 7,988

0

97,348

23,829

675

24,902

63,498

38,211 31,632 25,709

38,211 34,878 33,697

29,722 26,756

29,722 26,756

1,048 29,267

43,719

25,552 267,341

4,970

433,696

148

The Biggest Landowners

Table 5.12. (Continued) Owner

Province EP WP

c.1907 King Wilhelm II Prince Pless Prince HohenloheOehringen Prince StolbergWernigerode Prince Solms-Baruth Duke of Ratibor Prince Friedrich Leopolda Crown Prince Wilhelma Count Schaffgotsch City of G¨orlitz total

3,235

Total BBG

POM

POS

SL

SA

53,871

17,494

7,171

24,986

4,007

742

50,505 42,990

26,517 15,429

30,656

542 20,160

0

33,891

90,002

50,505 43,732

14,064

28,342

8,961

40,581

23,278

38,707

33,393

33,393 3 2,246

1,207

32,215

30,851

30,851

29,412

29,412

1,048 10,848

110,764

263,138

18,071

442,406

Notes: EP = East Prussia; WP = West Prussia; BBG = Brandenburg; POM = Pomerania; POS = Posen; SL = Silesia; SA = Saxony. a Indicates a Hohenzollern prince.

position in the list. An additional 14,263 hectares and 258,715 marks of tax value in Saxony would have raised Prince Otto to third position, just behind Prince Pless, in the top ten by area table for c.1895. Prince Christian Ernst zu Stolberg-Wernigerode appears in two provinces in the table c.1907, but now those provinces are Silesia and Saxony, because the manor of Radenz in Posen was divided between Prince Otto’s other two sons, Wilhelm and Hermann, while Christian Ernst inherited the family seat. In the third year Prince Friedrich zu Solms-Baruth replaced Count Traugott von Arnim in the top-ten list, but with holdings in the same two provinces. Since Prince Christian Kraft zu Hohenlohe-Oehringen had already owned the estate of Garthe (742 ha) in Lissa district of Posen in 1896, inheriting his father’s estates in Silesia after Prince Hugo’s death in 1897 made him the second of the non-Hohenzollerns with holdings in two provinces (if only barely) among the top ten in 1907. Entrance into, or exit from, the top ten list seems to have come mostly via death and inheritance. We have seen some examples in the preceding paragraphs. The Duke of Brunswick disappeared upon the death of the last Duke of

The Biggest Landowners

149

Brunswick-Wolfenbüttel; ditto Prince Frederik of the Netherlands, as his properties wound up in the hands of three new owners, only one of whom appeared to be an heir, Princess Marie zu Wied, his daughter. The others were Count Traugott von Arnim, new owner of the 30,000-hectare estate of Muskau in Silesia, and a Major (ret.) Alfred Wietersheim, who acquired Prince Frederik’s smaller properties in Silesia. The case was similar for Imperial Count Ludwig Schaffgotsch (d. 1891), whose properties were split among his wife and children. New to the 1895 top ten besides Count von Arnim was Count Guido Henckel von Donnersmarck. He had been no. 12 c.1882; the expansion of his estates by about 3400 hectares was enough to push him into the top ten. About 1907 these two dropped down again (Arnim to 13th and Henckel to 21st position), making room for Count Friedrich Schaffgotsch and Prince Friedrich zu Solms-Baruth to enter the top ranks. Ranking the top ten owners by the tax value of their properties produces a picture that is very similar overall, but different in its details, from the listing of the top ten by area. This can easily be seen in Table 5.13. In the list of the ten biggest landowners by tax value of their properties half of the names are different from those in the list of the ten biggest by land area, but the dominance of the high nobility is the same: nine of the ten enjoyed the rank of count or above in both cases. Despite the differences in who was a member of the top ten, there was much less change among the top ten in value than among the ten largest in terms of area. When the Duke of Brunswick died, enough of his property went to the King of Saxony (as we saw in the previous chapter) to bring the Saxon King into the top ten. Between c.1882 and c.1895, only the J. G. Boltze Co. dropped out of the list, to be replaced by the University of Greifswald. This, of course, had no effect on the 9 : 1 ratio of high nobility to other owners. Between 1885 and 1899, the ﬁrm of J. G. Boltze Co. apparently transferred some of its properties to family members, notably to August Zimmermann (listed as Inhaber or ‘proprietor’ of the ﬁrm in the early year) and Kurt Zimmermann. Under company ownership remained only the estates of Salzmünde (2186 ha in 1899) and Wurdenburg (307 ha) in the district Mansfelder See, Merseburg administrative region, Saxony. In the 1907 Handbook, the J. G. Boltze Co. no longer existed, and the by now von Zimmermanns, Kurt and August, had only one property each (with August as owner of Salzmünde). The estate of Wurdenburg was missing entirely from the Handbook. Between c.1895 and c.1907, the only changes in the top ten were changes of position of owners already in the list.¹⁷ The upward move of Prince Christian ¹⁷ Although, of course, in some cases an heir inherited the position, e.g. Grand Duke Wilhelm Ernst replaced Grand Duchess Sophie as head of the house of Sachsen-Weimar, Hans Heinrich XV replaced Hans Heinrich XI as Prince of Pless, or Christian Kraft became Prince zu HohenloheOehringen upon the passing of his father, Hugo.

150

The Biggest Landowners

Table 5.13. Top ten owners by tax value excluding the prussian state, by province (000 marks) Owner

c.1882 King Wilhelm I Duke of Brunswick Prince Pless Prince Putbus Count StolbergWernigerode Grand Duchess of SaxonyWeimar Duke of Anhalt-Dessau Prince HohenloheOehringen Count von Redern J. G. Boltze Co. total c.1895 King Wilhelm II Prince Pless Prince Putbus Grand Duchess of SaxonyWeimar Duke of Anhalt-Dessau Count von Redern Prince HohenloheOehringen Prince StolbergWernigerode King of Saxony University of Greifswald total

Province EP

Total WP

BBG

POM

POS

SL

SA

233.7

15.2

63.3

219.0 390.0

95.9

38.7

326.0 22.3

328.3 69.4 48.9

70.2

72.0

249.8 225.0

26.3

13.6

93.1

232.3

59.9

227.9 219.0

219.0

473.8

403.4

246.6

1,414.7

657.7

3,266.3

184.6

95.2

79.0 38.7

229.0 330.0 22.3 254.2

95.9

697.4 368.8 339.0 303.1

93.4

263.5

48.9

73.6 198.8

26.3 59.9

258.7 255.2

255.2

248.2 228.2

13.6

457.0

676.6

193.0

1,319.0

248.2 228.2 204.9

205.0 70.2

261.6

168.1

316.6

70.2

364.7 350.7 319.2 273.9

232.3

70.2

627.0 390.0

437.5

3,166.9

The Biggest Landowners

151

Table 5.13. (Continued) Owner

c.1907 King Wilhelm II Duke of Anhalt-Dessau Prince Pless Grand Duke of SaxonyWeimar Prince StolbergWernigerode Prince Putbus Prince HohenloheOehringen University of Greifswald Count von Redern King of Saxony total

Province EP WP

33.6 91.6

Total BBG

POM

POS

SL

SA

289.9 95.2

108.9

79.3 28.4

249.6

75.2 107.5

48.8

319.4 256.5

51.4

319.4 305.3

248.2

295.1 7.1

33.6

299.6 295.1 281.1

274.0

266.8

266.8

188.5

59.9

248.3

573.5

730.7

231.3 91.6

836.4 322.7

163.6

1,382.1

231.3 430.9

3,406.0

Ernst zu Stolberg-Wernigerode was, as discussed under the rankings by total area, mere statistical artefact. Moreover, the properties he seemed to have gained in Silesia had indeed already been owned by his father in 1892, as noted above, but in 1880 they had been owned by either the Countess zu Solms-Roesa or by the heirs of Count von Tschirsky-Rénard, so it was apparent that the then Count Otto had acquired these properties sometime in the interim. In sum, then, although half the names might have been different, the stability of membership in the top ten owners by tax value of their properties was even greater than that of the top ten by total land owned, but both groups constituted a nearly exclusive club for the high nobility only.

6 Knight’s Estates: A Special Class of Property 6 . 1 . P R EV I EW If 1848 marked a bourgeois revolution in German politics, the bourgeois revolution in rural landownership had already happened, and was followed by a steady bourgeois evolution at least up to the 1880s. This becomes clear from consideration of knight’s estates (Rittergüter, often hereafter RG for short), those properties originally intended to sustain a knight obligated to bear arms for his monarch, but later simply politically privileged properties. The principal questions considered are the following: Who owned these properties? Did this ownership change? Was it specialized? By the mid-1850s non-noble owners possessed about 45 per cent of these special properties, although the share varied considerably by province, from just over one-third to very nearly two-thirds. This level of ownership did not, however, translate into large numbers of bourgeois members of the provincial assemblies (Landtage), where a speciﬁed number of seats was reserved for the Ritterschaft (knightage), deﬁned in the laws regarding membership in the provincial assemblies as all owners of knight’s estates, irrespective of conditions of rank or birth. Gradually increasing numbers of bourgeois in these legislatures, however, did provide continuing incentive for purchase of knight’s estates to realize political or social ends. The special case of East and West Prussia, where non-nobles had nearly two-thirds of all the knight’s estates in the early 1880s, could be examined back to 1834. Here it turned out that most of the inroads by non-noble owners had already been made by 1834, although nobles continued to lose ground in the half-century between 1834 and 1884/5 (the bourgeois evolution). There was considerable aggrandizement of the area of knight’s estates in these two provinces, and properties owned by nobles tended to be about double the size of those owned by non-nobles. I use cluster analysis to consider the question of whether there was any specialization in ownership, that is, did owners from one or another distinct group tend to own knight’s estates of a particular character? This showed that the nobility were over-represented, and the bourgeoisie under-represented, among those knight’s estates that were either strongly or at least partially specialized in forestry. On the other hand, the bourgeoisie were disproportionately highly

Knight’s Estates: A Special Property Class

153

represented among owners of knight’s estates strongly specialized on arable agriculture, although in this group of properties the bourgeoisie lost out somewhat over time, particularly to the miscellaneous class of ‘other owners’.

6 . 2 . I N T RO D U C T I O N In the literature on landownership in Prussia, the term Rittergut (knight’s estate) is often used loosely, frequently to mean any property owned by a member of the nobility, sometimes even to mean any large estate. But this ignores the continuing political function of those properties that actually bore the designation Rittergut. Originally a Rittergut was a property which provided the sustenance for a knight who was obligated to bear arms for his sovereign in case of need, and up until the law was changed in the reforms in 1807, only the nobility were supposed to be owners of a knight’s estate. Although the obligation to bear arms no longer burdened the owner of a knight’s estate in the nineteenth century, such ownership carried other burdens and other privileges. The administration of police power on its territory, at least for minor infractions, was the obligation of the owner of a knight’s estate, and ownership normally entitled him (or, rarely, her) to a vote in the local district assembly (the Kreistag) and potential election by this body to one of the reserved seats in the provincial assembly (Landtag). A series of laws passed in the 1820s regulated the conditions of representation in all the provinces of Prussia and established the estates (Stände) who were to have representation in the district and provincial assemblies.¹ These rights to membership in the First Estate (as the Ritterschaft was ofﬁcially designated) rested on inclusion in a list (Matrikel), sanctioned by the king, of properties possessing the quality of being a Rittergut.² Membership in the First Estate was open to all physical persons who owned a Rittergut, regardless of their social status.³ ¹ In the province of Prussia—but only in this province—there were two parts to the list: section I contained the Rittergüter per se, and section II those free-peasant properties (köllmische Güter) that were independent properties of sufﬁcient size and paying a sufﬁcient land tax to qualify their owners for representation. Other provinces did not give representation to such properties. ‘Gesetz wegen Anordnung der Provinzialstände für das Königreich Preussen vom 1. Juli 1823’ § 8, Gesetzessammlung 1822/23, 138 ff. ² After the Congress of Vienna in 1815 ratiﬁed her annexation of Polish territories in the late 18th century and the gains in the west at the conclusion of the Napoleonic Wars, Prussia set about to establish a uniﬁed system of provinces and regulate their governing. Separate laws for each province promulgated in the early 1820s set out the size, membership, and terms of the several provincial assemblies (Landtage). ³ For example, with reference to the provincial assembly of Brandenburg, the law stated that ‘the right to be elected as a representative to the ﬁrst estate for the knightage is grounded on the ownership of a knight’s estate in the province, without regard to noble birth of the owner.’ ‘Gesetz wegen Anordnung der Provinzialstände für die Provinz Brandenburg vom 1. Juli 1823’, Gesetzessammlung 1822/23; my translation. The same language was used in each of the six provincial laws.

154

Knight’s Estates: A Special Property Class

Thus the owner of a knight’s estate, whether noble or bourgeois, was guaranteed a modicum of political inﬂuence, at least in the area where the property was located. While this was a valuable asset, it was not a designation ﬁxed for all time: under speciﬁed conditions (usually related to the size of pieces after parcellization) a property could automatically lose its designation as a Rittergut, or—upon petition of the owner, and with approval by the king—a property not previously included in the Matrikel could win this designation. That, however, was not a beneﬁt given lightly, not least because the king was aware that it added value to a property. In a letter of 15 March 1837 Interior Minister Gustav von Rochow wrote to provincial governor (Oberpräsident) of the province of Prussia Theodor von Schön that ‘the King does not wish that the acquisition of the Rittergut quality should become an object of speculation, to raise the sale value of properties and so make them easier to sell.’⁴ When the king did approve the granting of designation as a Rittergut, he often tried to prevent such speculation by limiting the life of the designation to the uninterrupted period of ownership of the owner to whom it was granted and his legitimate descendants (eheliche Descendanz). Thus if the property were sold or otherwise transferred out of the direct line of the family, or inherited by an illegitimate offspring, it would no longer be a Rittergut, and the new owner would not have any voting rights in or for any legislative assembly. Sometimes, too, the king would limit the designation to only a right to representation in the district assembly (Kreistag), but not in the provincial legislature (Landtag), but this was only an infrequently imposed condition.⁵ The laws of the early 1820s regulating representation in the provincial assemblies all included a clause that owners of knight’s estates listed in the Matrikel, regardless of noble or common birth, were entitled to be elected as delegates of the First Estate in these legislatures. The district ordinances (Kreisverordnungen) of the years 1872–5, however, removed the privileged position and the reserved seats in the provincial assemblies for the owners of knight’s estates. Thus there was a period of almost exactly half a century during which a bourgeois person, with the purchase of a knight’s estate, could buy the right to vote for, or even to be elected as, a representative of the landed elite in the provincial assemblies of the six (later seven) East Elbian provinces of Prussia that form the object of our focus in this book.⁶ ⁴ Geheimes Staatsarchiv Preussischer Kulturbesitz (hereafter GStA), I Rep. 77 Ministerium des Innern, Tit. 438, Nr. 62, Band 3, unnumbered. My translation. ⁵ Rauer found that of 1798 Rittergüter in the province of Brandenburg in 1855, 118 were limited to representation in the district assembly; similarly, 53 of 1742 in the province of Pomerania and only 5 of 1024 in the province of Saxony. He found none in the other provinces. See K. Fr. Rauer, Hand-Matrikel der in sämmtlichen Kreisen des Preussischen Staates auf Kreis- und Landtagen vertretenen Rittergüter ( Berlin: self-published, 1857), passim. ⁶ Brandenburg, Pomerania, Posen, Silesia, province of Saxony and province of Prussia. The six became seven when the province of Prussia was redivided into East and West Prussia in 1878. In

Knight’s Estates: A Special Property Class

155

Non-noble owners of knight’s estates did not easily win seats in the provincial assemblies, however, but their numbers gradually increased. In the provincial assembly of Brandenburg, for example, all thirty representatives of the First Estate in 1824 were noblemen; in 1831 there were two non-nobles, three in 1861, and ﬁve in 1871. With the seats reserved for the head of the Cathedral Chapter of Berlin and Standesherren (mediatized princes),⁷ the nobility still held 33 of the 67 seats in the provincial assembly of 1824. In 1841, they accounted for 31 of 68 seats, and in 1871 for 30 of 69.⁸ Thus the opportunity to purchase political inﬂuence was there, and the gradual increase in actual representation of non-noble landowners in the provincial assembly provided further incentive for such purchase. To what extent did non-noble persons take advantage of this opportunity? 6 . 3 . N U M B E R S A N D A R E A O F K N I G H T ’ S E S TAT E S IN EAST ELBIA

6.3.1. General Considerations In the absence of data from the original approved lists (Matrikeln) of the 1820s and 1830s,⁹ the ﬁrst quantitative look at this question is provided by Rauer’s data from 1855,¹⁰ summarized in Table 6.1. From the table we see that already by 1855 about 45 per cent of the 11,449¹¹ properties which entitled their owners to membership in the First Estate of their respective provinces were in the hands of non-noble owners.¹² Outside of the province of Prussia, the average ﬁgure was 39 per cent, ranging from 33.5 per cent in Posen to 45 per cent in the province of Saxony. The province of Prussia was an outlier in many respects: it was the only province to include ‘free peasant properties’ (köllmische Güter) in its Matrikel (provided that they met certain size and land tax criteria), and the this chapter, as I have in the rest of the book, I will continue to focus on these ‘core provinces’ of the Kingdom of Prussia. ⁷ Three seats in total in 1824, increased to four in 1827 and to ﬁve in 1843. ⁸ Calculated from lists in Friedrich Beck, ‘Die brandenburgischen Provinzialstände 1823–1872/ 75’, in Kurt Adamy and Kristina Hübener, eds., Geschichte der Brandenburgischen Landtage. Von den Anfängen 1823 bis in die Gegenwart, Brandenburg Historische Studien, 3 (Potsdam: Verlag für Berlin-Brandenburg, 1998), 62–74. ⁹ Except for the 1834 Matrikel for the province of Prussia, which I treat in Sect. 6.4, below. ¹⁰ Rauer, Handmatrikel. ¹¹ Of this total, 54 were ‘privileged estate complexes’ (bevorrechtete Güter-Complexe), i.e. the estates of the mediatized princes who had a personal right to representation in the provincial assembly. The number also included 379 ‘free peasant properties’ from the province of Prussia and 176 knight’s estates that had rights only to representation in the district assemblies of three of the provinces (Brandenburg, Pomerania, and province of Saxony). ¹² Although Rauer calls these owners ‘bourgeois’, the fact that the two categories add to the total number of properties means that the category of ‘bourgeois’ must include also non-physical legal persons such as business ﬁrms, banks, and—probably most importantly—the Prussian State.

156

Knight’s Estates: A Special Property Class Table 6.1. Data from Rauer regarding numbers of knight’s estates Province

Prussiaa Brandenburg Pomerania Silesia Posen Saxony total

Number of properties in Matrikel, Owned by 1855 nobility

Owned by ‘bourgeois’

2,313 1,798 1,742 3,132 1,440 1,024

788 1,116 1,046 1,857 957 563

1,525 682 696 1,275 483 461

11,449

6,327

5,122

a East and West Prussia combined. Sources: See text.

only one to see a large majority (nearly 66 per cent) of its First Estate properties in ‘bourgeois’ hands by 1855. From these data it is clear that, already by the mid-1850s, non-noble persons had taken considerable advantage of the opportunity to acquire landed estates that carried with them political duties and privileges. They do not tell us, however, when these acquisitions occurred. Since the opportunity had been there since 1807, perhaps the major acquisitions occurred in the 1820s, when impressionistic evidence tells us that there was unprecedented turnover in ownership of large estates as a result of a debt crisis caused in large part by a slump in grain prices resulting from the passage of the Corn Laws in Britain.¹³ If this be true, it would then be difﬁcult to determine whether purchases of knight’s estates by non-noble buyers in the 1820s were the result of the landowning nobility’s not being able to service the debts they had taken on during the boom years near the end of the Napoleonic Wars or the result of the creation of the opportunity for the bourgeoisie to buy representation in legislative assemblies. Although Rauer does break down his calculations to the district level, he provides no further information on, for example, the share of the total area ¹³ There are some fragmentary reports from individual districts. For example, Wilhelm Treue, in his Wirtschaftszustände und Wirtschaftspolitik in Preussen, 1815–1825 (Stuttgart/Berlin: Verlag W. Kohlhammer, 1937), 9, reports that in the period 1806 to 1830, ‘of the 84 knight’s estates of Preussisch-Eylau district only 3 allodial and 6 [entailed] knight’s estates survived the ordeals and remained in the hands of their ﬁrst owner.’ In Preussisch Eylau district in 1884 the Handbook listed 106 knight’s estates, of which 3 were under 100 ha. Those 3 and 70 others were in the hands of non-noble persons, two in the hands of juridical persons, and 31 in noble hands. Whether this lends support to Treue’s claim is problematic, since I do not know how long their owners had held these estates, and could only show (based on the Matrikel) whether the same family held the estate in 1884 that had held it ﬁfty years earlier.

Knight’s Estates: A Special Property Class

157

Table 6.2. Comparison of Rauer’s data with Handbook ( HB) data Province

Prussia (E. and W.) Brandenburg Pomerania Silesia Posen Saxony total

Numbers of knight’s estates c.1882: HB 1855: Rauer

Area in 000 ha c.1882: HB 1855: Rauera

1,921

1,929

1,293

1,422

1,405 1,778 3,000 1,577 1,014

1,666 1,688 3,106 1,437 1,014

1,017 1,286 1,258 1,223 377

1,188 1,217 1,588 1,250 294

10,695

10,840

6,454

6,959

a For the provinces of Prussia, Brandenburg, Pomerania, and Silesia Rauer’s area includes substantial numbers of properties which were not knight’s estates, hence is an overstatement and includes more properties than are enumerated in the second column. The area data from the Handbooks refer only to the knight’s estates enumerated in the ﬁrst column. Hence the area data from the Handbooks are not strictly comparable to those from Rauer.

of the properties that was owned by nobility or bourgeois, nor any other information about whether the estates owned by nobles or non-nobles differed in any signiﬁcant way. Comparing Rauer’s ﬁgures with data from the Handbooks is instructive (see Table 6.2).

6.3.2. Coverage and Inclusion Because Rauer’s data for area of properties included all properties that he counted, and not just knight’s estates, his area data for all provinces are an overstatement of the area of knight’s estates. The overstatement is especially great for the province of Prussia, where 379 ‘free peasant properties’ are included, along with ﬁve ‘privileged estate complexes’ (bevorrechtete GüterComplexe).¹⁴ The next strongest overstatement is likely for Silesia, where the ‘privileged estate complexes’ of the great landed magnates were both numerous and huge.¹⁵ The Handbooks included many smaller properties, their 100-hectare size criterion for inclusion notwithstanding. There appeared, however, no systematic process of inclusion of these smaller properties, yet the numbers of knight’s estates counted in the Handbooks for the early 1880s are, overall, remarkably close to Rauer’s ﬁgures, despite the unsystematic coverage of smaller units and the passage of time.¹⁶ Just as East and West Prussia had to be aggregated in the 1880s ¹⁴ These were the great estates of the Standesherren, the mediatized princes who had personal rights to seats in the provincial assembly (cf. n. 11). ¹⁵ See previous chapter, and esp. Tables 5.12 and 5.13. ¹⁶ This might suggest that the compilers of the Handbooks tried to include as many of the knight’s estates as they could, despite the general inclusion criterion of 100 ha and/or 1500 marks

158

Knight’s Estates: A Special Property Class

data because they formed a single province in Rauer’s time, so, too, should Brandenburg and the province of Saxony have been combined, since Rauer included several districts of Saxony in the data for Brandenburg. Were this to have been done, the area data for the two combined provinces would have come out very close, indeed: 1.39 million hectares in the Handbooks to 1.48 million in Rauer. To some more minor extent this shifting of properties occurred for other provinces as well, since Rauer followed older boundaries, set up for a different purpose, in assigning district data to a particular province. That the data from the Handbooks came out so close to Rauer’s data indicates quite similar coverage of properties, and that relatively little change in the ownership of knight’s estates by social category of owner occurred in the roughly three decades that had elapsed since 1855.

6 . 4 . E A S T A N D W E S T P RU S S I A , 1 8 3 4 – 1 8 8 4 / 5 The Matrikel for the combined province of Prussia for 1834¹⁷ offers the chance to extend the consideration of the questions raised above back a further two decades for the provinces of East and West Prussia, and to look at ownership changes in more detail. The 1834 Matrikel listed 2237 properties totalling 1.19 million hectares. Of these, 1873 (1.12 million ha) were noble properties (adlige Güter) and 364 (only 71 thousand ha) were the so-called free peasant properties (köllmische Güter). Of the area of these 2237 properties, a total of 166,000 hectares had ‘gone over into peasant ownership’,¹⁸ leaving a net 1.02 million hectares in 1834. But of these properties, twelve (7764 ha) were listed as having gone over totally to peasant ownership, and one other was listed with no area at all. Another seven showed anomalous results: a negative net area resulted from the amount listed as having gone over to peasant ownership being larger than the original recorded total area of the property. These data are clearly wrong, since all seven estates appear, with areas ranging from 224 to 18,540 hectares, in the Handbooks for 1884 (East Prussia) and 1885 (West Prussia). I have had to leave these properties out of consideration, so that we wind up in the end with 2217 properties, originally with 1.17 million hectares, but with only 1.02 million net total hectares in 1834, of tax assessment, but nowhere do they indicate explicitly that this was or was not the case. In fact, having the quality of a Rittergut does not appear to have had anything to do with inclusion of a property under 100 ha in size into the Handbooks. From the 1882 database, only 10.7% of all principal properties of less than 100 ha included in the Handbooks were knight’s estates, compared to 63% of all principal properties of 100 ha or more. In the size category 100–199 ha, 35.6% of all principal properties were knight’s estates. ¹⁷ Matrikel zum Erscheinen im Stande der Ritterschaft befähigenden Güter der Provinz Preussen (dated 31 December 1834), in GStA: I Rep. 77 Ministerium des Innern tit. 438, Nr. 62, Band 2, pages unnumbered. ¹⁸ ‘In bäuerlichen Besitz übergegangen’, in the language of the Matrikel.

Knight’s Estates: A Special Property Class

159

because 542 of those 2217 properties had been partially transferred to peasant ownership (some 147,446 ha in all), leaving 459,930 hectares remaining in the hands of the current owners. The ﬁnal tally, then, for the 1834 Matrikel is 1853 noble properties comprising 952,000 hectares and 364 free peasant properties totalling only 71,000 hectares, in the combined province of East and West Prussia. These are the data from which we may begin our more detailed analysis. Let us ﬁrst consider turnover in ownership, and whether this turnover resulted in an embourgeoisement¹⁹ of ownership of these estates. This will be done using a relatively large random sample from among all the properties listed in the Matrikel. To produce the random sample used to examine the continuity of ownership of properties listed in the Matrikel, I ﬁrst assigned each property a random number between 1 and 10. To select from among these properties, I generated another random number between 1 and 10, in this case, 2. This double random procedure produced, for further examination, 286 properties that had been assigned the random number 2 in step one. Of these, three were properties that had gone over in their entirety to peasant ownership, so were eliminated, as was a fourth property listed as ‘subdivided’ (parzelliert). The ﬁnal random sample to be checked then consisted of 282 properties, a 12.6 per cent sample of the database, almost exactly 1 of every 8 properties. Of the 282 properties,²⁰ I found 257 in the 1882 land database,²¹ 23 could not be found there, and there were two ‘maybes’. Of the 23 not found in the later database, 6 were owned by nobles, 9 by bourgeois, and 8 by multiple owners—unspeciﬁed, but assumed non-noble. The one speciﬁed multiple owner was still ambiguous: ‘the inhabitants of the village’ (die Dorfs Einsassen) of Ploewken, listed as a ‘noble free village’ (adliges Freidorf ). There is no reason to suspect that these residents of the village belonged to the nobility, despite the designation of the village itself. Of the 257 properties for which a later pair was found, 109 were owned by nobility in 1834, 137 by ‘bourgeois’, 7 by ‘multiple owners’ (mehrere Besitzer), 2 by religious bodies, 1 by a hospital, and 1 by the ‘Academic Senate in Königsberg’. It is these for which we undertake an analysis of what happened to their ownership in the half-century following 1834. ¹⁹ I use the term here to mean simply an increase in the share of non-nobles in the ownership of large estates. Puhle sees another form of embourgeoisement: He writes that ‘signs of the large landowners’ ‘‘embourgeoisement’’ were apparent in their market dependence and capitalist orientation’ in the latter half of the 19th century. Puhle, ‘Lords and peasants’, 84. I prefer to avoid the argument over whether noble owners became more bourgeois, or bourgeois owners more feudal, in consequence of the bourgeoisie’s entrance into the large landowner group, ﬁnding any argument based on purported group attitudes difﬁcult to evaluate empirically. ²⁰ Twenty of these had a net total area under 100 ha, therefore 262 had 100 or more. ²¹ The actual dates were 1884 for East Prussia and 1885 for West Prussia, so that ‘c.1882’ will mean 1884/5 in this section, and ‘1882 land database’ will refer to the entries for East and West Prussia in 1884 and 1885, respectively.

160

Knight’s Estates: A Special Property Class

In one case I had to make an assumption that a large property from the 1834 Matrikel was included as several smaller properties listed separately in the 1882 database. This was Capustigall (2635 ha in 1834) in the district of Königsberg in East Prussia, owned at the time by ‘Truchsess zu Waldburg’ (Steward at Waldburg).²² In the 1884 Handbook for East Prussia, Count zu Dohna-Schlobitten and Eberhard zu Dohna-Waldburg (not identiﬁed as a count, but he no doubt was one) jointly owned nine properties totalling 2604 hectares in Königsberg district. Eight of the nine were listed as belonging to the knight’s estate of Waldburg. Because of the single ownership and the very close correspondence of reported size in the two years, I assumed that these properties were parts of the single estate of Capustigall in 1834. I have further guessed that the Dohna family were the stewards at Waldburg in 1834, and so tentatively assigned this property to the group of those owned by the same family in the two years. Another property of 3712 hectares from the 1834 Matrikel was listed in the 1882 database, but without any data. Neither the area nor the owner was known. These two represent the only signiﬁcant problems in the data of Table 6.3. There are some signiﬁcant overall trends revealed by the data in the table. There was a strong tendency toward either aggrandizement of properties or their Table 6.3. Continuity and change of ownership of sample properties in the 1834 Matrikel also found in the 1882 database Owner or change in owner between 1834 and 1884/5 Bourgeois to Bourgeois Bourgeois to Noble Bourgeois to Unknown Noble to Bourgeois Noble to Fiscus Noble to Noble Other to Bourgeois Other to Noble ‘Same family’: Bourgeois ‘Same family’: Bourgeois? ‘Same family’: Noble ‘Same family’: Noble? ‘Same family’: Othera total

No. of properties

Area in 1834 (ha)

Area c.1882 (ha)

80 28 1 57 1 24 3 3 25 2 25 3 5

29,520 12,003 3,712 25,356 339 21,743 1,210 588 8,501 269 27,383 5,029 2,476

39,057 24,209 0 38,305 510 29,117 1,403 2,041 14,256 392 38,032 6,830 2,324

257

138,128

196,476

a For non-physical legal persons, ‘same family’ indicates that the same body or organization owned the property in both years.

²² The position of Truchsess, usually translated as ‘steward’, was the head of the administration of the court, with special responsibility for the table.

Knight’s Estates: A Special Property Class

161

inclusion into larger property complexes. Of the 257 properties here examined, 53 showed losses of area over the half-century, 4 stayed the same size,²³ but 200 gained territory. Overall the gain amounted to more than 58,000 hectares, 42 per cent of their total area in 1834. In 1834, noble properties were, on average, considerably larger than those of non-noble owners: 109 properties totalling 78,716 hectares, so an average of 722 hectares each, compared to an average of 401 hectares for all other properties (59,412 ha in 148 properties). The 85 noble-owned properties in 1884/5 averaged 1193 hectares in size, while the rest averaged 553 hectares, so that—even as the number of nobleowned estates was declining—the relative size difference grew (from a ratio of 1.8 to 2.2).²⁴ There was a signiﬁcant reduction in the number of properties owned by nobility: from 109 out of 257 (42.4%) to 85 (33.1%). The Prussian ﬁscus gained one of these properties, and one bourgeois property changed to unknown owner; all the rest of the gain was registered entirely by ‘bourgeois’ owners. Of properties that experienced a change in owner status, twice as many went from noble hands into bourgeois hands as from bourgeois to noble. Of the 257 properties, 60 (23.3%) appeared to remain in the same hands; the rest appeared to have changed hands at least once during the half-century that elapsed between the Matrikel and the Handbooks. The turnover might be somewhat exaggerated, since the criterion for ownership remaining in the ‘same family’ was for the later owner to have the same surname, or be the same organization or group, as the owner in 1834. Under this criterion, inheritance by a married female family member, for example, would look like a change in ownership because of the change of surname. A property that was in noble hands in 1834 was more likely to have remained in the same family half a century later, as compared to a property owned by a non-noble person: 28, or just over a quarter of the 109 noble properties in 1834, appeared to be in the hands of the same family in 1884/5, compared to 27 of the 137 (just under one-ﬁfth) owned by non-noble persons in 1834. Therefore, although non-noble persons as a group gained properties, they also exhibited higher rates of turnover in property ownership than did the nobles. The overall apparent turnover rate in ownership of 76.6 per cent (197 of the 257 properties appeared to have changed hands at least once in the interval examined) seems to me to be rather high, but it could merely be a sign of an active and healthy market for land. Ownership by groups declined: all six of the cases of transfer of ownership from ‘other’ to an individual were properties owned in 1834 by a group—very ²³ ‘Same size’ is taken to mean a difference of 2 ha or less in the area in the two years. ²⁴ This could be an overstatement: if we examine the six properties for which the 1834 property was included as part of another estate or manor in 1884/5, only two of these were owned by nobles in 1884/5. Removing the two manors containing these two properties reduces the noble average only a little, to 1169 hectares, so does not alter the conclusion reached.

162

Knight’s Estates: A Special Property Class Table 6.4. Noble ownership of knight’s estates in East and West Prussia, 1834–1907/09 Source and property type

No. of properties

Owned by nobles

% of noble owners

Matrikel, 1834 Noble properties only Of those, 100+ ha Free peasant properties

2,217 1,853 1,716 364

874 822 778 52

39.4 44.4 45.3 14.3

Rauer, 1855 RGa only

2,313 1,929

788 [733]b

33.7 [38.0]b

Handbooks RG only, 1884/5

1,921

656

34.1

RG 100+ ha 1884/5

1,897

656

34.6

RG only, 1894/5

1,761

547

31.1

RG 100+ ha only 1894/5

1,733

541

31.2

RG only, 1907/9 RG 100+ ha only, 1907/9

1,370 1,355

487 485

35.5 35.8

RG = knight’s estate. Estimates: no data in Rauer (see text). Sources: See text. a

b

likely a communal group, but the designation in the Matrikel was only ‘multiple owners’, with no further elaboration. For the provinces of East and West Prussia, therefore, we may conclude that there was a signiﬁcant embourgeoisement of the ownership of knight’s estates, and by implication of large estates in general, in the half-century following 1834. Nevertheless, the principal incursion of non-noble persons into the ranks of large landowners in eastern Prussia must have occurred before 1834, if what happened in East and West Prussia is representative of what happened in the other ﬁve provinces as well. In Table 6.4, the ﬁgures in brackets for ‘Rauer 1855, Rittergüter only’ are estimates based on the assumption that there was no further embourgeoisement of the ownership of free peasant properties in Rauer’s data between 1834 and 1855, that is, noble ownership remained at 14.3 per cent. This assumption produces a residual ﬁgure of 733 noble owners of knight’s estates, or 38 per cent ownership by nobles of the 1929 knight’s estates in Rauer’s list for 1855. That in turn implies that the rate of embourgeoisement of ownership of knight’s estates slowed down after 1855. These considerations simply reinforce the conclusion reached above—the major gains in ownership of properties carrying rights of representation in the provincial assembly occurred before the early 1830s. While the process continued after 1834, it appeared to decelerate after that time, perhaps

Knight’s Estates: A Special Property Class

163

even reversing (but only in a relative sense, since the absolute number of knight’s estates declined—see Sect. 6.5 below) after the mid-1890s.²⁵ This reversal of trend coincided chronologically with a period of improving terms of trade for agriculture, both in Germany and in Europe in general. If we were to conﬁne our attention only to properties of 100 hectares or more, in order to avoid the problems associated with incomplete coverage of properties smaller than that size in the Handbooks, the conclusions would nevertheless remain exactly the same, as the data of the table show.

6 . 5 . OW N E R S H I P C H A N G E S I N A L L S EV E N P ROV I N C E S A F T E R c. 1 8 8 2 Let us now conﬁne our attention to what happened to the ownership of knight’s estates after the early 1880s. Table 6.5 shows a summary of the overall trends in the data of the Handbooks. Table 6.5. Comparison of numbers of knight’s estates in Rauer and in the Handbooks Province

Rauer

Handbooks

1855

c.1882

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

c.1895

c.1907

1,666 1,688 1,437 3,106 1,014

1, 140 781 1,405 1,778 1,577 3,000 1,014

1, 135 622 1,397 1,780 1,414 3,044 997

869 499 1,307 1,621 1,107 3,034 936

10,840

10,695

10,389

9,373

1,929

% change over previous year c.1882 c.1895

c.1907

−15.7 5.3 9.7 −3.4 0.0

−0.4 −20.4 −0.6 0.1 −10.3 1.5 −1.7

−23.4 −19.8 −6.4 −8.9 −21.7 −0.3 −6.1

−1.3

−2.9

−9.8

−0.4

Because of the previous comparison of Rauer’s data with those of the Handbooks c.1882, we can concentrate on the last two columns of the table. There is almost no change in the numbers of knight’s estates between the years 1882 and 1895, except in the provinces of West Prussia and Posen. The declines in these two provinces are almost entirely the result (a) of the activities of the Royal Settlement Commission for West Prussia and Posen, which bought up large properties for ²⁵ As Hartmut Berghoff noted (‘Adel und Industriekapitalismus im Deutschen Kaiserreich—Abstossungskräfte und Annäherungstendenzen zweier Lebenswelten’, in Reif, Adel und Bürgertum, 266), in comparing bourgeois with noble landowners: ‘[Bourgeois owners] were more strongly market-oriented and would have invested more and modernized earlier. By the same token, they would separate themselves more quickly and easily from the properties than the long-seated (alteingesessen) nobility.’ My translation.

164

Knight’s Estates: A Special Property Class Table 6.6. Numbers of knight’s estates of 100 ha or less in the Handbooks Province

c.1882

c.1895

c.1907

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

15 9 27 21 47 211 121

19 5 33 16 2 201 134

9 6 28 16 2 143 71

total

452

410

275

subdivision into small farms on which to settle German farmers, and (b) of the Polish parcellization banks and societies that competed with the Settlement Commission in the same two provinces.²⁶ Inconsistent coverage of properties, either across provinces or over time, could mask or exaggerate changes in the ownership patterns of knight’s estates. Because all Handbooks were supposed to include all properties of 100 hectares or more, I looked at knight’s estates under 100 hectares to see whether such inconsistencies might be apparent. Table 6.6 shows the results of these calculations. The general reduction in numbers of smaller knight’s estates included in the data over time is consistent with the decline in the overall numbers of such properties, but the patterns for the individual provinces are very different. In any given year, the difference between Silesia and Saxony, on the one hand, and the rest of the provinces on the other, is striking, but the patterns for these two provinces differ from one another. Rather than risk an element of bias because of the differences among provinces and years with respect to inclusion of smaller properties, I have chosen to concentrate, in the rest of the chapter, on properties of 100 hectares or more. Table 6.7 summarizes the trends in both numbers and area of knight’s estates in the seven provinces of our concern. After about 1895, the number and area of knight’s estates included in the Handbooks declined in all provinces except Silesia,²⁷ and beside the further reduction in numbers in West Prussia and Posen, an over-proportional decline also occurred in the province of East Prussia, where the Generalkommission (the body originally charged with implementing the peasant emancipation after 1807, but now a promoter of ‘inner colonization’) and private parcellization efforts were both active in subdividing properties for sale to small farmers. The same was true, but to a considerably lesser extent, for the province of Pomerania. ²⁶ See esp. Ludwig Bernhard, Die Polenfrage, 2nd edn. (Leipzig: Duncker & Humblot, 1910), section on ‘the battle for land’, 483–620. ²⁷ I suspect that the small increases shown are the result primarily of better coverage of large properties in this province.

Knight’s Estates: A Special Property Class

165

Table 6.7. Changes in numbers and area of knight’s estates of 100 ha or more (%) Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Change over previous year c.1895 c.1907 Numbers Area Numbers

Area

Change: 1907 values over 1882 Numbers Area

−0.8 −20.1 −1.2 0.4 −7.7 1.9 −3.4

−1.3 −18.9 0.9 −1.6 −5.8 8.4 −3.4

−22.9 −20.1 −6.2 −9.0 −21.7 1.7 0.2

−18.7 −19.3 −6.5 −8.9 −27.0 4.0 10.3

−23.6 −36.1 −7.4 −8.7 −27.8 3.7 −3.1

−19.8 −34.6 −5.6 −10.3 −31.2 12.8 6.6

−2.6

−1.8

−8.8

−9.8

−11.2

−11.4

Overall, the decline in both numbers and area of knight’s estates over 100 hectares in size between c.1882 and c.1907 was on the order of 11 per cent. There is no way to tell if all of the loss of these knight’s estates was through parcellization or subdivision into small farms. Some may have been sold off piecemeal, others divided among heirs, or perhaps simply missed in the enumeration.²⁸ The Handbooks for c.1907 still carried in them 442 properties listed as ‘parcelled’ or ‘subdivided’ (or in the process thereof), of which 121 were knight’s estates—29 of these were in Pomerania, 57 in Posen, 5 in Silesia, and 30 in West Prussia. This is only a small fraction of the more than 1000 knight’s estates that disappeared from the data between 1895 and 1907. I assume that that is the case for two principal reasons: 1. It is likely that the Handbooks included only the most recent parcellizations, including those in the process of subdivision (‘aufgeteilt, oder in der Aufteilung begriffen’, to use the phrase employed in the Handbooks). 2. The compilers of the Handbooks for some provinces chose not to include any properties which had been, or were being, subdivided into parcels. This seems to be most obviously the case for East Prussia, where a government-assisted parcellization programme was under way.

6 . 6 . W H O OW N E D H OW M U C H ? Conﬁning our purview henceforth only to knight’s estates over 100 hectares in size, Tables 6.8 and 6.9 show the numbers and area of such properties by province ²⁸ This would have been very unlikely, in that the compilers of the Handbooks had previous editions to go by, and the probability of any large property’s being missed should have gone down, not up, over the years.

166

Knight’s Estates: A Special Property Class

Table 6.8. Number of knight’s estates of 100 ha or more owned in various years, by owner category Province

Nobility 1882 1895

1907

Bourgeois 1882 1895

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

371 285 819 1,033 936 1,672 487

337 204 764 1,021 804 1,702 471

298 187 730 926 676 1,802 463

741 473 533 705 579 1,013 377

765 395 559 714 521 1,025 358

total

5,603

5,303

5,082

4,421

4,337

Other 1882

1895

1907

537 267 493 637 371 908 372

13 14 29 19 15 104 29

14 18 41 29 87 116 34

25 39 56 42 58 181 30

3,585

223

339

431

1907

Table 6.9. Area (000 ha) of knight’s estates of 100 ha or more owned in various years, by owner category Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Nobility 1882 1895

1907

Bourgeois 1882 1895

1907

Other 1882

1895

1907

324 282 690 830 843 825 246

307 207 678 821 761 931 234

261 188 645 735 581 999 266

355 315 309 442 363 354 112

360 261 322 429 325 349 111

273 176 284 381 220 293 117

5 11 20 12 14 66 12

8 25 29 14 63 68 13

14 33 34 36 38 112 11

4,040

3,939

3,675

2,250

2,158

1,744

139

220

279

and by category of owner for our three reference periods. While the numbers of these larger knight’s estates declined about 8.8 per cent from c.1895 to c.1907 and the area by 9.8 per cent, the comparable ﬁgures for ‘bourgeois’ owners were 17.3 per cent and 19.2 per cent, so nearly double the overall rate. The declines for the nobility were much smaller (4.2% and 6.8%, respectively), while the category of other owners saw an increase in its ownership of knight’s estates. The decline in ownership of knight’s estates by the bourgeoisie was especially marked in East Prussia, West Prussia, and Posen—exactly those provinces in which the government and private programmes of subdividing large estates into parcels for sale to small farmers were most active. It appears that just as the bourgeoisie had taken up the opportunity to acquire knight’s estates when they carried political privileges, so were they quite ready to sell off such properties after these political privileges had been withdrawn. The lag between the change in privileges and the beginning of noticeable sell-offs was probably the result of the decline in grain

Knight’s Estates: A Special Property Class

167

prices to the mid-1890s, after which agricultural protectionism in Germany and the general improvement in agriculture’s terms of trade in the world once again raised the prices of farm outputs and with them, prices of the land as well.

6.6.1. Was there Specialization in Ownership? The previous two tables show that bourgeois owners of knight’s estates clearly tended to own smaller properties than did the nobility or other owners. The nobility owned a larger share of the area of these properties than they did of the number, while for the bourgeois owners exactly the reverse was true. Fig. 6.1 summarizes this quite clearly. There are other important characteristics besides just average size, however. Might there have been a particular ‘proﬁle’ of the typical bourgeois-owned knight’s estate that differed from that of a typical noble-owned estate? In previous work on Hungary I have shown that, for good historical reasons, there was considerable specialization in the types of property owned by different classes of owners. Communities and communal groups tended to own properties made up primarily of pasture with some forest, the nobility and the State tended to own the primarily forest properties, religious organizations (and a large share of the nobility) tended to own properties whose land proﬁle was best suited to mixed farming, while the bourgeois owners tended to specialize in properties made up primarily of arable land.²⁹ In the Hungarian case there was no knight’s

Fig. 6.1. Ownership of knight’s estates by number and area. ²⁹ Eddie, ‘Cluster Analízis’, 17–36.

168

Knight’s Estates: A Special Property Class

estate, and knight’s estates in Prussia may have differed both from other large properties in Prussia and from those in Hungary. To investigate this question for Prussia, I have again employed cluster analysis to group properties according to the proﬁle of their land use.

6.6.2. Grouping Properties by their Land-Use Proﬁle I have divided agricultural use into two subcategories, arable agriculture and animal husbandry, and have assumed that ploughland and garden (vineyards hardly existed in the properties under review) were devoted to arable agriculture,³⁰ while meadow and pasture were devoted to animal husbandry. Moorland, unproductive land, bodies of water, roads, farmyards, etc., because they served either none of these economic functions or more than one of them, fall into the category ‘other land’. Thus any given property may have land devoted to some combination of arable agriculture, animal husbandry, forestry, or other use. To group all large properties according to their land-use proﬁle, the share of each of the four land types in the total area of each property became the relevant set of variables. Thus each property had a four-dimensional land-use proﬁle, in which each of the variables took on a value between 0 and 1. Take for example the estate of Closterfelde in the district of Arnswalde, province of Brandenburg, owned by the Schiel heirs in 1885, with areas of the several land types as follows, in hectares: Ploughland: 318 (84.4%) Wood: 12 (3.2%) Total Area: 377

Meadow: 35 (9.3%) Moor, etc.: 6 (1.6%)

Pasture: 5 (1.3%) Water: 1 (0.3%)

This produces the following land-use proﬁle for the estate of Closterfelde: Arable: 0.844

Animal Husbandry: 0.106

Forestry: 0.032

Other: 0.019

Not all of the knight’s estates used for the foregoing analysis could be included in the analysis of their land-use proﬁle, for three primary reasons: 1. For some properties, only the total area of the property was reported, and no details of land type could be found to ﬁll in the blanks, so no proﬁle could be calculated. 2. Sometimes certain land types (most often ploughland and meadow, but also pasture and forest, or pasture and water, etc.) were reported only as a ³⁰ In fact, all types of arable agricultural land were included under the rubric ‘ploughland’ in the sources used.

Knight’s Estates: A Special Property Class

169

combined total, and had to be allocated arbitrarily to the several land type categories, making any calculated land-use proﬁle also arbitrary. 3. The data contained irresoluble errors: the sum of the different land types did not add to the total reported area of a given property. If the discrepancy was larger than 5 per cent of the total area, I excluded the property from the analysis. After the foregoing exclusions, I was left with totals of 10,190; 9812; and 8837 knight’s estates of 100 hectares or more (for 1882, 1895, and 1907, respectively) usable for grouping according to their land-use proﬁles. Among the various statistical techniques for multidimensional data analysis, cluster analysis affords the most convenient and useful method of assigning properties to various groups according to their land-use proﬁles. The use of this technique is as much art as it is science, for it requires the analyst to specify the number of clusters to be used, after which the statistical software (in this case SPSS), using the k-means cluster algorithm, assigns properties to clusters in an iterative process that seeks to maximize the n-dimensional (in this case n = 4) distance between cluster means, that is, to make the clusters as distinct as possible from each other. For the grouping here presented, cluster analysis was used to divide the properties into 3, then 4, 5, 6, 7, 8, and 9 clusters. Choosing which of these sets of clusters provides the best and clearest summary of the information contained in the dataset—in which proliferation of detail does not obscure the general picture (the classic trees and forest notion)—falls to the judgement of the analyst. In the present case, ﬁve clusters seemed to me to be the optimum, rather than the six used in Chapters 3 and 7. Not only did the data show ﬁve quite clearly different clusters, they also included one very small cluster (under 200 properties in each of the three years, between 1.6 and 1.9% of the total number of properties used). Since cluster analysis is sensitive to outliers, one can at least hope that the outliers tended to lie in a particular direction, and thus the existence of this small cluster would indicate that the other clusters were quite robust. Expanding the clusters to six or more produced additional clusters which I judged to be but minor variations on one of the ﬁve existing clusters. Table 6.10 shows the ﬁve clusters chosen for each of the three years. The ﬁrst thing that is clear from Table 6.10 is that the clusters—chosen independently for each year—hardly differed at all in their average proﬁles over the years. Nor, for that matter, did the relative number of properties to be found in each of the clusters change to any appreciable extent. There was a slight gain for the biggest cluster—properties that were, on average, over 80 per cent arable—at the expense of the ‘mixed farming’ cluster (‘arable with signiﬁcant husbandry’ in the table). Thus although the numbers of knight’s estates fell, the basic groupings into which still-existing knight’s estates fell scarcely changed at all.

170

Knight’s Estates: A Special Property Class

Table 6.10. Mean share of each land type in clusters Land-use proﬁle

1882

1895

Cluster type

Specialized for arable

1907

1882

1895

1907

Arable with signif. forestry

Arable Animal husbandry Forestry Other Number in cluster

0.83 0.10

0.83 0.10

0.83 0.09

0.55 0.12

0.55 0.12

0.53 0.11

0.05 0.03 4588

0.05 0.03 4379

0.05 0.03 4189

0.31 0.03 2573

0.31 0.03 2502

0.33 0.03 2126

Share of total (%)

45.0

44.6

47.4

25.3

25.5

24.1

Cluster type Arable Animal husbandry Forestry Other Number in cluster

Arable with signif. husbandry 0.58 0.57 0.58 0.30 0.31 0.29 0.09 0.03 1670

0.09 0.03 1550

0.10 0.03 1317

0.62 0.03 1200

0.63 0.03 1191

0.66 0.03 1056

Share of total (%)

16.4

15.8

14.9

11.8

12.1

11.9

Cluster type Arable Animal husbandry Forestry Other Number in cluster

Primarily arable and ‘other’ 0.40 0.43 0.38 0.14 0.14 0.14

Share of total (%)

Forestry with signif. arable 0.26 0.25 0.23 0.09 0.09 0.08

Totals for the given year

0.12 0.34 159

0.12 0.31 190

0.13 0.36 149

10,190

9812

8837

1.6

1.9

1.7

100.0

100.0

100.0

6.6.3. Did Different Classes of Owners Specialize in What Types of Properties They Owned and Where? We have already seen earlier that the average size of a noble-owned knight’s estate was considerably larger than the average for those of bourgeois owners. Further examination of the results of the cluster analysis can reveal several more dimensions of the difference between nobles and non-nobles as owners. Table 6.11 shows how properties in the various clusters were distributed according to who owned them c.1882. It is noteworthy here that—compared to the overall average—nobility were particularly concentrated in the cluster where forestry was the principal land use for knight’s estates, while bourgeois owners were correspondingly under-represented there, as well as in the cluster where forestry was the secondary land use. These non-noble owners tended to concentrate in the heavily arable cluster, but not so strongly—neither absolutely nor relatively—as the nobility in the forestry cluster. The very small group of

Knight’s Estates: A Special Property Class

171

Table 6.11. Share of each owner group in number of properties in cluster (%) Owner

Arable

Arableforest

ArableAnHus

ForestArable

ArableOther

Noble Nonnoble Other

46.8 51.1

61.4 37.1

53.8 43.8

71.0 25.8

54.1 42.8

54.6 43.2

2.2

1.5

2.3

3.3

3.1

2.2

100.0

100.0

100.0

100.0

100.0

100.0

total

Total

Table 6.12. Size category share for each owner group in number of properties c.1882 (%) Size category (ha) 100–199 200–499 500–999 1000–1999 2000–4999 5000–9999 10,000+ total

Noble

Non-noble

Other

Total

8.6 38.5 33.8 15.3 3.3 0.4 0.1

16.0 46.1 29.4 7.4 1.0 0.0 0.0

20.3 45.5 25.2 6.3 2.3 0.0 0.5

12.1 41.9 31.7 11.7 2.3 0.2 0.1

100.0

100.0

100.0

100.0

other owners distributed itself approximately as the average for the whole group, although with relatively fewer of its members in the arable–forest cluster, and relatively more in the forest–arable cluster. Since this group of owners was so small, possessing only 2.2 per cent of the properties under analysis, it need not concern us much in any further analysis. We have noted earlier that non-noble owners were most prevalent in the provinces of East and West Prussia, the only provinces in which they owned the majority of knight’s estates. The absolute number of their knight’s estates in the current database, however, was greatest in Silesia (1012), with Pomerania not far behind East Prussia (703 vs. 741, respectively) for second place. Cross-tabulation by owner and size category shows details of the results implied by the earlier observation that estates of the nobility tended to be larger than those of non-noble persons. Table 6.12 shows those details for c.1882. As the size category rises, the share of the nobility in the number of knight’s estates in that category grows, except for the very largest category of properties of 10,000 hectares or more. Only eight properties fell into this category; the city of Görlitz in Silesia owned the largest of these (27,800 ha, of which nearly 25,800 were woodland). For non-noble owners, precisely the opposite pattern was true: no one from this group owned a knight’s estate larger than 10,000 hectares, and only two owned properties larger than 5000 hectares. The biggest single

172

Knight’s Estates: A Special Property Class

property owned by non-noble landowners was the previously mentioned estate of Stolzenburg in Pomerania: 7676 hectares, of which 1193 were arable, 1681 meadow and pasture, 4716 forest, and 287 other land; the owners were August Rehder and Hermann Schwenke. The other was the estate of Hammerstein in West Prussia, 6373 hectares (mostly forest), owned by the ‘Jaffé Brothers’ with no further information about them. It is interesting that no other Rehde is to be found among owners of any properties in the database, and that only one other property (Petershain in Brandenburg, 787 ha) lists Hermann Schwenke as owner. (On average, nonnoble owners of knight’s estates possessed about 1.4 such properties each: for the 10,673 properties in the original database without regard to overall size I found a total of 7721 distinct owners.) As the number of knight’s estates declined over time, there was a slight trend toward further specialization by owners. It is easiest to see this from a comparison of 1882 with 1907, as shown in Fig. 6.2. Here we can see that in the biggest cluster (heavily arable), other owners gained share at the expense of bourgeois owners, as they did in all but the smallest cluster (arable with other), where they gained at the expense of the nobility. The nobility held or increased their share of the properties in each of the other clusters. Overall, the nobility’s share, and especially the share of other owners, increased at the

Fig. 6.2. Share of each owner in total for cluster.

Knight’s Estates: A Special Property Class

173

expense of bourgeois owners. The decline in the number of knight’s estates therefore occurred disproportionately among the estates of non-noble persons. It appears that just as they had earlier bought their way in, now they were selling—to some extent, at least, piecemeal—their way out. Rising land prices, the opportunities created by government and private parcellization schemes, and a reduction in the socio-political value of ownership of a Rittergut were probably behind this trend.

6.6.4. Why did Knight’s Estates Appear to Command a Premium over Other Large Properties in the Market? In a previously published paper, using data on private sales of individual properties for a single district in East Prussia, I found that, ceteris paribus, a knight’s estate commanded a price premium on the order of 20,000 marks of 1913 purchasing power over a property that did not have that quality.³¹ In that paper I attributed this premium to the political privileges attached to a knight’s estate, but, given that many of those privileges had been extinguished by 1875, an alternative hypothesis suggests itself: maybe knight’s estates were systematically under-taxed. If their ‘land tax net yield’ had been under-assessed relative to that of similar properties that were not knight’s estates, this form of economic privilege would also have translated itself into a price premium in the land market. This question, along with other hypotheses regarding possible undertaxation of large properties, will be examined in detail in the ﬁnal chapter of this book.

6 . 7 . S U M M A RY O F E X A M I N AT I O N O F K N I G H T ’ S E S TAT E S What I believe the foregoing discussion has established is the following: 1. With regard to the question of embourgeoisement of the land-owning elite, it had indeed occurred on a signiﬁcant scale, but nearly all of this had happened by the mid-1850s. At that time the bourgeoisie owned about 45 per cent of all knight’s estates, compared to about 43 per cent c.1882. After the mid-1890s, the trend was reversed, so that there was both an absolute and a relative desembourgeoisement up to the years just before the Great War. 2. If the evidence for the provinces of East and West Prussia is representative of the eastern provinces as a group, then the bulk of the embourgeoisement ³¹ Eddie, ‘Price of Land’, 195–216.

174

Knight’s Estates: A Special Property Class

of the ownership of knight’s estates had already occurred by 1834. The representativeness might be questionable, however, given the major differences between these two provinces and the others, especially as noted in the preceding section above, as well as in Chapter 4. 3. There appeared to be relatively little specialization of ownership of knight’s estates of a particular character, except for the strong dominance of the nobility in forestry. 4. Geographically speaking, bourgeois owners of knight’s estates were most important relatively in the provinces of East and West Prussia, the only provinces in which they owned a majority of such properties. 5. The share of noble owners increased as the size of knight’s estates increased, so that necessarily the opposite was true for bourgeois owners.

7 Industry on the Land 7 . 1 . P R EV I EW ‘How ya gonna keep ’em down on the farm, after they’ve seen Berlin?’ One way might have been to provide off-season by-employment on the farm, and in fact many of the larger estates did just that. The stereotypical picture of Germany with an industrial west and an agrarian east ignores the phenomenon of rural-based industry in the east, a subject also largely overlooked in the historical literature. This chapter will begin to redress this imbalance. The Handbooks that are the primary sources for this study contain information on the number and type of industrial establishments on the properties listed in them, and these data are used to provide a picture of rural-based industry in the seven eastern provinces of Prussia. A surprisingly dense network of (mostly small-scale) industrial establishments was found, averaging about one unit for every 2.5 square kilometres. In American terms, this would translate into one industrial plant for every section (640 acres) of rural land. More than one-third of all estates of 100 hectares or more had at least one industrial enterprise located on their land. As German industrialization proceeded apace in the late nineteenth century, some of these small-scale enterprises started to disappear; a decline of 15 per cent or more occurred between c.1882 and c.1907. Comparison with data from tax records for brewing, distilling, and sugar manufacturing shows that the bulk of brewing took place in the cities, while distilling occurred overwhelmingly on the land. In the early 1880s about twothirds of sugar mills were found on landed properties of 100 hectares or more, but by about 1907 the ﬁgure had dropped to little more than two-ﬁfths. Comparison with data from the industrial censuses of 1882, 1895, and 1907 for other typical rural-based industries shows rather remarkably high shares of the total number of establishments located on large landed estates, but it also shows great inter-industry and inter-province variety. Further sections of the chapter deal with location of industrial enterprises by province, by administrative region, by size of landed property, and by type of owner of the land. No one will be surprised to learn that Silesia was the foremost home of these rural-based industries. Perhaps surprisingly, the Baltic provinces (East and West Prussia and Pomerania) had about the same number of industrial

176

Industry on the Land

plants as the east-south-east provinces (Posen and Silesia), while the west central provinces (Saxony and Brandenburg) had only about half as many. The bulk of industrial establishments was found on properties of 200 to 999 hectares in size; add in properties of 1000 to 1999 hectares, and it became the overwhelming bulk. If a property had an industrial establishment on it, it tended to have more than one, with the average for a property owned by a bank or business ﬁrm approaching two. Different types of owners tended to specialize in different types of industries: physical persons in brickworks and grain mills, with the Prussian State and monasteries concentrating on distilleries. The owners of land housing the most diverse group of industries were business ﬁrms and banks. The very largest owners, on whose estates were to be found twelve or more industrial establishments, were a considerably more varied group than the largest owners of land per se, with a quarter of their number being non-noble persons. As with the list of top landowners, the list of top owners of land with industry on it showed marked changes over time, with new entrants as well as movement up and down of existing members. The Prussian State and the King of Prussia were always numbers 1 and 2, and one of the most remarkable features was Hermann Kennemann’s move up from 12th to 7th place as the years passed. The chapter ends with a consideration of factors affecting the location of these rural-based industries: a simple model uses regression analysis to assess what would increase or decrease the probability of ﬁnding an industrial enterprise on a rural estate. Increasing the probability were status as a knight’s estate, having a military ofﬁcer, non-noble person, or business ﬁrm as owner; and increasing size of property. Decreasing the probability was ownership of the land by religious groups, the State, or communities. Location in some administrative regions increased the probability, in others decreased it. Here there were evident both inter-province and intra-province differences.

7 . 2 . I N T RO D U C T I O N That the Industrial Revolution in nineteenth-century Germany was mostly a west German affair seems obvious to even the most casual observer. The great smokestack industries of the Ruhr Basin, the textile centres along the lower Rhine, the shipbuilding in Hamburg and Bremen, potash and fertilizer in Thuringia, all lay well west and/or south of Berlin. The east, the land of the Junkers, was the home of extensive agriculture only recently emerged from the fetters of the so-called second serfdom. Alexander Gerschenkron, in his classic Bread and Democracy in Germany,¹ emphasized the east–west specialization in ¹ Berkeley: University of California Press, 1943.

Industry on the Land

177

production in his discussion of how the import certiﬁcate scheme subsidized the grain-producing, export-oriented landlords in the east and allowed for large grain imports into the food-deﬁcit areas of the industrial west.² But in fact the east was not devoid of industry: a mostly overlooked feature of the nineteenth-century economy of eastern Germany is the signiﬁcant role that industrial production on agricultural estates played in that economy.³ The current chapter will examine the extent of such industry ‘on the land’ in the seven East Elbian provinces of our concern.

7.2.1. The Data The data for this examination come primarily from two sources: one is the Handbooks, the other is the industrial censuses of Germany taken in 1882, 1895, and 1907. From the censuses, I took the data from the volumes presenting data for the smallest administrative divisions of the country.⁴ The Handbooks, besides the other information that they listed, also recorded what if any ‘industrial establishments’ (Industrie-Anlagen) were to be found on the property. In the Handbooks, an industrial establishment was fairly broadly deﬁned: besides plants that processed agricultural products, such as ﬂour mills, breweries, distilleries, starch or sugar factories, and dairy plants, the deﬁnition also included works that won or processed other products of the land, such as brickworks, mines, and quarries, along with a diverse set of other production activities such as paper factories, chemical works, glass factories, blast furnaces, iron and copper smelters, and even one pinecone-seed drying plant. ² He could also have pointed out that in the context of increasing tariff protection for grains, the scheme probably saved shipping costs for western consumers, or helped to prevent a corresponding reduction in the farm-gate price for eastern producers—and even saved investment costs for the government—by substituting grains from nearby (albeit foreign) sources for grains from eastern Germany. Domestic grains would otherwise have had to have been shipped, often by land over areas where railways were still sparse, from Germany’s ‘Far East’ to places like Duisburg and Düsseldorf, Krefeld and Barmen. ³ Most available studies tend to focus on a particular estate or region. One of the best and most recent of these is Klemens Skibicki, Industrie im Oberschlesischen Fürstentum Pless im 18. und 19. Jahrhundert. Zur ökonomischen Logik des Übergangs vom feudalen Magnatenwirtschaftsbetrieb zum modernen Industrieunternehmen, Regionale Industrialisierung, 2 (Stuttgart: Franz Steiner Verlag, 2002). ⁴ These were the Regierungsbezirk (administrative region, a subunit of the province) in 1882, and the Kreis (district, a subunit of the administrative region) in 1895 and 1907. The volumes are all from the series Statistik des Deutschen Reichs: For 1882: ns 7, Gewerbestatistik nach der allgemeinen Berufszählung vom 2. Juni 1882: Gewerbe-Statistik der Staaten und grösseren Verwaltungsbezirke, Erster Abschnitt (Anzahl und Personal der Betriebe) (Berlin: Puttkammer & Mühlbrecht, 1886); for 1895: ns 117, Berufs- und Gewerbezählung vom 14. Juni 1895: Gewerbe-Statistik der Verwaltungsbezirke, Erster Theil (Kleinere Verwaltungsbezirke Preussens) (Berlin: Puttkammer & Mühlbrecht, 1898); for 1907: ns 218, Berufs- und Betriebszählung vom 12. Juni 1907: Gewerbliche Betriebstatistik. Abteilung VI—Kleinere Verwaltungsbezirke; Preussen (Berlin: Puttkammer & Mühlbrecht, 1909).

178

Industry on the Land

7.2.2. Summary of Distribution of Industrial Establishments When we examine properties for which the data are available, it is quite remarkable how many industrial establishments existed on agricultural estates in the East Elbian provinces.⁵ In what follows I have had to exclude all dairy plants (Molkerei) from the data, since the enumeration of such was too inconsistent in the Handbooks to produce reliable data. Sometimes a property was listed as having a Molkerei when it was clear from the context that all that was there was one or more cream separators for the milk taken from the herd on that property; sometimes the Handbooks named a big plant on the property, run by steam (Dampfmolkerei), but other times a large dairy plant—listed by name—was clearly located somewhere else. For these reasons, none of the data in this chapter includes any dairy establishment. The increasing number of properties under 100 hectares shown in Table 7.1 as time passes is an indication of their increasing coverage of properties in general.⁶ Therefore we can be quite conﬁdent that the nearly 15 per cent decline in the number of industrial establishments enumerated between c.1882 and c.1907 is not simply a statistical artefact, but rather a real indication of trend. Indeed, Table 7.1. Industrial establishments on large and small properties enumerated in the Handbooks (excluding dairy plants) Date, size

Principal properties Total No. with number indl. estab.

Industrial establishments Total Per property number

c.1882 Area < 100 ha Area >= 100 ha All

3,906 15,929 19,835

359 5,802 6,161

514 9,190 9,704

0.131 0.577 0.489

1.432 1.584 1.575

c.1895 Area < 100 ha Area >= 100 ha All

4,322 16,145 20,467

349 5,517 5,866

509 8,735 9,244

0.118 0.541 0.452

1.450 1.583 1.576

c.1907 Area < 100 ha Area >= 100 ha All

8,207 15,780 23,987

332 5,056 5,388

457 7,805 8,262

0.056 0.495 0.344

1.377 1.544 1.533

Per property with indl. estab.

⁵ In this chapter, ‘industrial establishments’ always refers only to those establishments enumerated in the Handbooks. ⁶ The increasing coverage is no doubt also the source of the increasing number of properties for which no area data are given. Were these all small properties, they would represent 3.6%, 3.8%, and 2.8% of the under-100 ha group in the three years, respectively.

Industry on the Land

179

given the increasing inclusion of smaller properties in the Handbooks over time, this has to represent a lower bound to the rate of decline. It is likely that the ongoing industrialization of Germany was directly responsible for the decline noted, as growth in the size, scale, and specialization of factory industry gave it an increasing advantage in productivity over smaller-scale enterprises, just as the increasing urbanization of Germany gave ever-stronger locational advantage to urban-based industry, compared to its competitors located in rural areas. Improvements in the transport network, especially expansion of the rail system, also contributed to this trend by eroding any ‘natural protection’ rural-based industry may have had in its local area.⁷ We cannot, alas, follow this point directly, since the Handbook data here used do not indicate anything about the size of the industrial establishments found on the agricultural properties—nothing about capacity,⁸ employment, or output—only that an establishment existed at a particular place in the given year. There are nevertheless important results to be won from these data. For example, let us consider the distribution of industrial establishments on agricultural properties per square kilometre of the seven provinces, as shown in Table 7.2. These data conﬁrm that Silesia led all other provinces, not just in absolute number of industrial establishments enumerated on agricultural properties, but in nearly every area measure as well. Saxony’s large ﬁgure for enumerated industrial establishments per square kilometre of enumerated properties is statistical artefact, more than anything else: Saxony’s very different size distribution of properties—only 30 per cent of her total area in properties of 100 hectares or more vs. 49 per cent in Silesia—skews the area ﬁgure, since the enumerated industry per square kilometre was nearly 30 per cent greater in Silesia than in the province of Saxony. By the measures of the table, West Prussia comes closest to being the ‘average’ province of the group, and will serve as our reference province in the regressions to follow later. From the foregoing tables we can see that the density of industrial establishments in rural areas was quite remarkable throughout the East Elbian provinces, especially when we consider that our data establish only a lower bound to their number: insofar as the Handbooks did not include all agricultural properties, ⁷ For numerous individual examples of land-based industrial establishments forced to close because they could no longer compete with larger, more modern plants, see Berghoff, ‘Adel und Industriekapitalismus’, esp. 254–7. Berghoff also adds another factor: insofar as modernization of their industrial plants would have required behaviour considered inappropriate to their class, the nobility generally would not undertake such behaviour. ⁸ Except in very rare instances. Examples from the province of East Prussia in 1895: on the knight’s estate of Jackunowen, in Angerburg district, the steam distillery had a mash capacity of 5250 litres; the knight’s estate of Kallischken in Johannisburg district had a steam distillery based on 6000 litres of mash; in the district of Lyck, the knight’s estate Baitkowen had a 2000-litre distillery and a brickworks producing 100,000 pieces; the brickworks at Paslöpen in Pillkallen district produced 2 million bricks. In no case was the time period speciﬁed. The royal domain of Göritten in Stallupönen district had a distillery with an annual capacity of 500,000 litres of mash.

180

Industry on the Land

Table 7.2. Enumerated industrial establishments (excluding dairy plants) on agricultural estates of 100 hectares or more, by province and area, c.1882 Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Area of province (sq. km)

100+ ha properties as % of prov. area

Enumerated industry Per sq. Per sq. km km of of 100+ ha province properties

Industry on 100+ ha property Per sq. Per sq. km km of of 100+ ha province properties

3,699 2,552 3,984 3,011 2,896 4,031 2,524

34.9 47.7 44.3 54.1 58.3 49.2 30.3

0.310 0.447 0.344 0.427 0.490 0.556 0.433

0.876 0.922 0.876 0.787 0.836 1.105 1.260

0.294 0.426 0.337 0.425 0.478 0.537 0.334

0.841 0.893 0.762 0.785 0.819 1.093 1.099

22,697

45.6

0.427

0.919

0.405

0.889

the industrial establishments enumerated must be only a proper subset of all the industrial establishments located on agricultural properties. Even the lowest ﬁgure—for East Prussia, at just over 0.3 industrial establishments per square kilometre of the province—is larger than I would have expected. The highest ﬁgure, just under 0.56 enumerated establishments per square kilometre in Silesia, works out, in imperial measurement terms, to 1.46 per square mile. The average for all six provinces, about 0.43 industrial establishments per square kilometre, is equivalent to 1.12 per square mile. Translated to late nineteenth-century American terms, that would be as if every fourth homesteader had had not just his sod hut, but also an industrial establishment, on his quarter of land.⁹ Thus the stereotypical picture of Junkerland as being a vast expanse of extensively cultivated estates with almost no industry is clearly wrong, even if these rural-based industries might have been, in the main, rather small in scale.¹⁰

7.2.3. Numbers of Certain Industrial Establishments from Tax Data Certain industries were specially taxed, so that there is an annual record of the numbers of breweries, distilleries, and sugar factories available from the Statistical Yearbook. From these yearbooks I have taken the numbers of breweries, distilleries, and sugar factories in each province for each year of the Handbook from which the data for that province have been taken.¹¹ We can compare ⁹ One quarter of a square mile, i.e. 160 acres (64.75 ha). ¹⁰ Some of them deﬁnitely were not: in a rare indication of the scale of enterprise, the Handbook for Silesia in 1892 notes that on his estate of Sacrau (812 ha) in the district of Gross-Strehlitz, First Lieutenant (ret.) Viet. Madelung had a ‘powerful lime industry, with twelve lime kilns’. ¹¹ See Table 7.1.

Industry on the Land

181

Table 7.3. Industrial establishments on principal properties of all sizes Province

Breweries East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

Yearbook c.1882

On the Per land cent c.1882

Yearbook c.1895

On the land c.1895

Per cent

Year booka 1906/07

On the Per land cent c.1907

282 103 551 346 157 971 689

45 16 31 2 25 211 63

16.0 15.5 5.6 0.6 15.9 21.7 9.1

190 96 558 322 143 815 510

39 22 26 5 13 228 50

20.5 22.9 4.7 1.6 9.1 28.0 9.8

137 82 512 163 122 559 375

8 7 29 4 7 83 33

5.8 8.5 3.9 2.5 5.7 14.8 8.8

3,099

393

12.7

2,634

383

14.5

1950

171

8.8

Distilleries East Prussia 395 West Prussia 268 Brandenburg 645 Pomerania 363 Posen 439 Silesia 1,076 Saxony 386

309 256 558 340 431 598 260

78.2 95.5 86.5 93.7 98.2 55.6 67.4

291 268 599 373 443 845 323

255 246 454 360 418 671 233

87.6 91.8 75.8 96.5 94.4 79.4 72.1

303 355 651 499 566 902 327

281 264 494 424 434 600 226

92.7 74.4 75.9 85.0 76.7 66.5 69.1

2,752

77.0

3,142

2,637

83.9

3,603

2,723

75.6

0 6 9 2 4 35 94

0.0 33.3 64.3 40.0 30.8 77.8 72.3

3 19 14 10 19 59 118

0 6 10 3 5 43 74

0.0 31.6 71.4 30.0 26.3 72.9 62.7

2 19 13 11 20 52 106

0 3 6 0 7 17 61

0.0 15.8 46.2 0.0 35.0 32.7 57.5

150

65.8

242

141

58.3

223

94

42.2

total

total

3,572

Sugar factories East Prussia 3 West Prussia 18 Brandenburg 14 Pomerania 5 Posen 13 Silesia 45 Saxony 130 total

228

Note: Yearbook data taken from same year as Handbook unless otherwise noted. a No provincial details published after 1908 (for 1906/07 ﬁscal year), so data for that year had to be used here.

the total number for the given province to the total number found on landed properties in that province; Table 7.3 contains the results of those calculations. Variety among the provinces is the principal feature of these calculations, but for distilleries the overwhelming majority of the establishments was located on the land in all three years. Breweries, on the other hand, were not: the share of landed properties in the total number of breweries exceeded one-quarter only in the province of Silesia in 1892. The decline in the number of sugar

182

Industry on the Land

factories over the years took place disproportionately among those located on the land, where—despite the increasing coverage of small properties in the Handbooks —both the numbers and the share of the provincial total declined by a third. The decline in breweries was even more pronounced: there were only about half the number on landed properties c.1907 that there had been twenty-ﬁve years earlier, and this number represented only one in ten of all breweries, where it had been one in eight in the early 1880s. Because of the huge relative weight of distilleries, a weight that grew because of the decline in the other two industries, the share of landed properties in the overall number of the three industries taken together actually rose, from just under 48 per cent to just over 55 per cent, over the two and a half decades covered by these data. Combining these data with the data on industrial establishments per square kilometre of territory in the seven provinces, we see that ‘industry on the land’ was an extremely important part of all industry to be found in these provinces. It is time, therefore, to look in more detail at the geographical distribution of these industrial establishments, at their distribution by size of property on which they were to be found, and at who owned those properties.¹² First, however, the data of the German industrial censuses will help us round out the context in which rural-based industry was to be found.

7.2.4. Industrial Establishments according to the Industrial Censuses Along with its censuses of agriculture, Germany also ran parallel censuses of industry, commerce, and employment in the years 1882, 1895, and 1907.¹³ In this section I have chosen to highlight, from these censuses, the seven most common types of industrial establishments to be found on agricultural estates, to show their trends in the economy as a whole and the relative place of rural-based establishments within the overall totals. These seven industries are brickworks, sawmills, grain mills, breweries, distilleries, starch factories, and sugar mills. Because of an irresoluble anomaly in the criterion of enumeration, the 1895 employment data for sugar mills have had to be estimated: in the data from the 1882 and 1907 censuses, employment was shown as that of the day of the census (5 June and 12 June, respectively). For 1895, it was the yearly or seasonal average. Since sugar beet milling is seasonal, and the season starts in autumn, the average ¹² Some of the data entries for ‘industrial establishments’ make it clear that the landowner sometimes was not the owner of the industrial enterprise; on the other hand, Handbook data for some provinces in some years also show that certain landowners owned shares in industrial enterprises that were not on their property. Both of these facts provide added conﬁrmation of the perhaps surprising conclusion from the data so far, that the landlords of eastern Prussia were not specialists in agriculture alone, but had substantial interests in non-agricultural enterprises. ¹³ These censuses counted the number of establishments and employment not only in industry per se, but also in animal husbandry, market gardening, ﬁsheries, mining, trade and commerce, construction, graphic arts, money and ﬁnance, insurance, transport, and guest services (entertainment, lodging, food and drink).

Industry on the Land

183

seasonal employment recorded was, overall, on the order of three times higher than the actual ﬁgure for the date of the census: 14 June 1895. Fig. 7.2 shows the relative size of the anomaly. For 1882, both the average employment and the employment on census day were available for sugar factories; the ratio of the former to the latter was 2.68. I used this ratio to adjust the average employment for sugar factories in 1895 to an estimate for employment on census day, in order to be comparable with the other two censuses.¹⁴ Figs. 7.1 through 7.3 show the trends in overall numbers of industrial establishments in the seven provinces, in the number of employed personnel in those plants, and in the number of workers per plant. From Figs. 7.1 through 7.3 we can see that the number of plants generally declined, except for sawmills, but that employment in these branches of industry generally increased, except for grain milling and distilling. Even in these two industries, the decline in employment was relatively less than the decline in the number of plants. Therefore a trend toward larger scale of plant in all seven of these branches of industry shows up very clearly in the increase in numbers of workers per plant, although that increase was very small for both grain milling and distilling. By 1907 at least, the scale of rural-based industrial plant, as measured by employment, was generally smaller than that of plants in cities. Between 1898 and 1907, 18 towns were separated from their hinterlands into independent

Fig. 7.1. Number of plants by industry, seven East Elbian provinces. Source: Calculated from data of the German industrial censuses.

¹⁴ This procedure produced values of 65.2, 82.4, and 85.0 employees per plant on the census day in the respective census years. Had I made the correction the other way, using the 1882 average to census day ratio to adjust the 1907 census day ﬁgure up to an estimate of the average, the result would have been as follows: 177.5, 220.9, 227.8. The ratio of the 1907 to 1882 ﬁgure is 1.30 for the census day ﬁgures, 1.28 for the average ﬁgure. Such a small discrepancy makes no difference for our analysis here.

184

Industry on the Land

Fig. 7.2. Number of employed personnel by industry, seven East Elbian provinces. Source: Calculated from data of the German industrial censuses.

Fig. 7.3. Average number of workers per plant, seven East Elbian provinces. NB: Sugar c.1885 estimated from c.1882 average to census day employment ratio. Source: Calculated from data of the German industrial censuses.

‘city districts’ (Stadtkreise), to add to the 25 city districts already existing. Of the 306 districts covered by the 1907 industrial census, therefore, 43 were separate city districts. For the principal industries of our concern here, Table 7.4 shows the ratio of average number of workers per plant in city districts vs. rural districts.¹⁵ Most rural grain mills usually had only between 1 and 3 workers, so mills in the city districts were considerably larger on average, although themselves still ¹⁵ Since the property data do not include any properties in the city of Berlin, neither do the industry data used here. Parts of present-day Berlin (Spandau, Charlottenburg, Wilmersdorf ) were not yet included in Berlin, but were separate, independent city districts in 1907. Data from these districts are, therefore, included in the calculations for Table 7.4.

Industry on the Land

185

Table 7.4. City/rural ratio of employees per plant, selected industries, 1907: overall average for seven eastern provinces Province

Brickworks

Sawmill

Grain mill

Brewery

Distillery

Starch factory

Sugar factory

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

2.4 4.6 0.0 6.7 4.5 2.1 1.9

2.9 2.2 1.6 2.6 4.5 3.2 2.4

6.6 9.1 4.1 3.8 11.3 9.1 4.2

4.1 2.4 2.2 3.4 2.3 2.3 3.3

4.1 2.1 1.5 4.8 2.2 1.8 2.2

0.0 4.5 1.3 3.9 0.1 — 0.4

— 4.0 0.0 1.8 — 0.7 2.1

total

2.3

2.2

7.0

3.8

2.2

1.0

1.2

Note: — indicates no factory of that type found in rural districts of the province; 0. 0 indicates no factory of that type found in city districts of the province.

small by present-day standards. Starch factories were also small (typically 10–15 workers), and they did not differ in size as between factories in the city and factories on the farm. Sugar factories, although considerably larger, also did not differ much in scale as between those located in towns and those situated on rural estates. From this comparison we may conclude that, in general, the share of the total number of industrial establishments found on rural properties overstates the share measured by employment and, therefore, also the share of output of those establishments. This bias should be kept in mind when evaluating the data that follow.

7 . 3 . I N D U S T R I A L E S TA B L I S H M E N TS O N T H E L A RG E R E S TAT E S Our detailed examination of the data requires datasets of uniform and consistent coverage. Therefore, beside having to exclude dairies from the list of industrial establishments because of the inconsistency in their coverage, we shall—from this point onwards—also have to limit our purview to properties of 100 hectares or larger, given the inconsistent coverage of properties below that threshold that we have seen, both across provinces and over time. Since all of the Handbooks strove to include all properties of at least 100 hectares in size, references to ‘properties’ in the rest of this chapter are to be understood as references to principal properties of at least 100 hectares in size. Two tables summarize the outcome of this narrower purview at the beginning of the period under review (c.1882): Table 7.5 shows the distribution by province, and Table 7.6 by size category of estate.

186

Industry on the Land

Table 7.5. Enumerated industrial establishments (excluding dairy plants) on agricultural estates of 100 hectares or more, by province, c.1882 Province

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

Number of props. in Handbooks

No. with indl. estab.

% with indl. estab.

Number of indl. estab.

Industrial estab. per listed per prop. with prop. indl. estab.

2,298 1,687 2,061 2,551 2,367 3,339 1,626

709 676 942 840 879 1,225 531

30.9 40.1 45.7 32.9 37.1 36.7 32.7

1,086 1,086 1,344 1,284 1,381 2,167 842

0.473 0.644 0.652 0.503 0.583 0.649 0.518

1.532 1.607 1.427 1.529 1.571 1.769 1.586

15,929

5,802

36.4

9,190

0.577

1.584

Table 7.6. Distribution of industrial establishments (excluding dairy plants) by size of estate on which they were found, c.1882 Province

100–199 200–499 500–999 1,000–1,999 2,000–4,999 5,000–9,999 10,000+ total

Number of props. in Handbooks

No. with indl. estab.

% with indl. estab.

Number of indl. estab.

Industrial estab. per listed per prop. with prop. indl. estab.

3,365 6,562 3,941 1,375 428 216 42

591 1,916 2,037 938 232 59 29

17.6 29.2 51.7 68.2 54.2 27.3 69.0

817 2,618 3,136 1,700 570 178 171

0.243 0.399 0.796 1.236 1.332 0.824 4.071

1.382 1.366 1.540 1.812 2.457 3.017 5.897

15,929

5,802

36.4

9,190

0.577

1.584

Table 7.6 looks approximately as we would expect, i.e. the share of the number of estates with an industrial establishment and the number of establishments per property tends to rise with the size of property. But, as Table 7.7 shows, the number of industrial establishments per thousand hectares of estates declines with the size of property, the only exception being the rise after we cross the 10,000 hectare threshold. Specialization in ownership explains this relationship: of the 189 properties between 5000 and 9999 hectares in size, 138 were mostly forest properties owned by the Prussian State, and two more were forests owned by cities; of these 140 properties, only 15 had any industry on them. On the other hand, of the 41 latifundia of 10,000 or more hectares, only 11 were owned by the Prussian state, and one by a city. Of the remaining 29, owned by royalty and high nobility

Industry on the Land

187

Table 7.7. Industrial establishments (excluding dairy plants) per thousand hectares of estates of 100+ ha, by province and by size category of estate, c.1882 Province

Indl. estab. per 000 ha

Property size (ha)

Indl. estab. per 000 ha

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

0.841 0.893 0.762 0.785 0.819 1.093 1.099

100–199 200–499 500–999 1,000–1,999 2,000–4,999 5,000–9,999 10,000 +

1.651 1.200 1.149 0.934 0.468 0.138 0.276

total

0.889

total

0.889

exclusively,¹⁶ all but three had at least one industrial establishment on their territory. Among the provinces, we ﬁnd the province of Saxony just edging out Silesia for the highest number of industries per 1000 hectares of large estates. But since only about 30 per cent of the land area of Saxony was to be found in large estates, vs. about 49 per cent for Silesia, that does not mean that the province of Saxony was somehow more industrialized than Silesia. When rural industries are the focus, it is still the fact that Silesia was the most industrialized of the seven provinces under review.

7.3.1. Share of Census-Listed Industrial Establishments on Large Estates Table 7.8 shows what fraction of the total number of industrial establishments in selected industries was to be found on large estates. In the table, an anomaly is immediately apparent: the number of distilleries on large estates in the province of East Prussia was greater than the total number of distilleries enumerated by the census in 1907. In 1907, 277 large estates had a total of 281 distilleries located on their territory, according to the Handbook for East Prussia of that year. The industrial census found only 253 distilleries in East Prussia, although the German statistical yearbook, using tax data, counted 303 in the 1906/7 tax year. According to the Yearbook, there were 395 distilleries in East Prussia in 1882 (369, according to the 1882 census); in 1895 the ﬁgures were 291 and 286, respectively. I have been unable to resolve this anomaly.¹⁷ Since, however, the discrepancies are not great, and the ﬁgures of the census always the lowest, I would assume that the difference ¹⁶ No one in this group was below the rank of count or countess. ¹⁷ The same anomaly occurs, but in more than one province in all three census years, for starch factories, which were therefore omitted from the table entirely. The discrepancies could result as I have supposed they did for distilleries, but also might be the result of a deﬁnitional difference in what constitutes a starch factory in each of the two sources; this problem also remains unresolved.

188

Industry on the Land Table 7.8. Number of establishments on estates of 100 hectares or more as share of total number enumerated in the census(%) Provious

Brick

Saw

Mill

Brew.

Distill.

Sugar

1882 East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

71.9 76.2 27.2 72.9 71.9 41.9 17.4

11.4 31.1 7.3 26.9 6.7 9.3 3.8

18.2 16.5 2.9 11.9 9.3 5.2 3.4

17.4 15.0 7.1 1.3 17.9 24.2 9.2

82.9 72.9 95.7 51.4 54.8 88.3 40.9

0.0 45.5 52.9 50.0 44.4 64.0 48.6

total

44.7

11.2

7.7

14.6

62.9

51.7

1895 East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

53.8 60.3 24.5 82.0 62.1 39.0 20.3

6.0 24.4 5.5 26.3 9.3 11.0 1.0

16.6 16.5 2.7 14.8 9.5 8.7 3.7

22.1 22.1 5.9 3.7 9.7 32.5 8.3

88.1 76.8 69.8 84.9 72.7 60.6 53.0

0.0 28.6 66.7 27.3 31.3 61.7 45.5

total

42.3

10.8

8.9

16.3

65.9

46.9

1907 East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

60.3 45.6 28.5 75.0 45.9 30.9 22.5

17.8 11.8 5.4 6.6 11.7 16.8 2.4

12.4 16.3 2.9 17.3 6.7 8.4 3.7

7.2 7.6 6.9 3.7 6.5 16.1 8.1

111.1 73.2 76.3 96.4 79.6 62.2 55.6

— 15.0 31.3 0.0 35.0 25.9 38.1

total

38.4

10.7

8.3

9.3

72.5

28.7

arises because the census reﬂects the position on a given date in the summer, when some distilleries oriented toward off-season by-employment of farm workers were temporarily closed for the peak agricultural season, whereas the Yearbook data would reﬂect how many distilleries paid their taxes at any time during the year. Table 7.8 conﬁrms the impression from Table 7.3, that of great variety, across both different industries and the several provinces, in the share of industry to be found on rural estates.

7.3.2. Distribution by Province and Size Category over Time We have seen above that the number of industrial establishments on landed properties declined over the period under review. Figs. 7.4 and 7.5 summarize this development: the decline in numbers was very evenly spread out regionally. The

Industry on the Land

189

Fig. 7.4. Industrial establishments on properties of 100 ha or more: number by province or area.

Fig. 7.5. Industrial establishments on larger landed properties: total number by size category.

proportion among the Baltic provinces (Pomerania, East and West Prussia), westcentral provinces (Brandenburg, Saxony) and the east-south-eastern provinces (Posen, Silesia) changed almost imperceptibly, well within the margin of error of the data. The reason for such stable numbers was that within a given region, one province’s gain was offset by another’s loss: Silesia’s share in the number of industrial establishments rose about 2.6 percentage points, while the share of Posen dropped about 2 percentage points. The only other provinces for which

190

Industry on the Land

the change in share between c.1882 and c.1907 was more than 1 percentage point were West Prussia (−1.8) and Pomerania (+1.3). Thus the decline in numbers can be viewed as mostly an across-the-board change, without disproportionately large losses in one or two individual provinces. The case is different, however, when we view the distribution according to the size category of property on which the industrial establishments were found. The decline in the number of industrial enterprises on the larger landed properties was heavily concentrated among the properties under 1000 hectares: 1222 of the 1381 fewer industrial establishments were lost from properties between 100 and 999 hectares, only 159 from those of 1000 hectares or more. Some of the change may be artiﬁcial, in that differing treatment in the Handbooks over time of what was a property might account for some of this (smaller principal properties now being counted as subunits of some larger property complex, for example, would simply move industrial establishments from the smaller size categories to the larger ones). Fig. 7.6 shows, however, that this appears unlikely to have been the case: the total number of properties changed only slightly between c.1882 and c.1907 (15,929 vs. 15,780), and the changes in numbers seem to have fallen mostly in the 200–499 hectare class, with the 100–199 hectare class actually showing gains.

7.3.3. Who Owned the Properties with Industrial Establishments on Them? In the above, we have seen that it apparently made a difference to the presence or absence of industry who the property owner was. Who were their owners? Table 7.9 summarizes the social distribution of ownership of these establishments.

Fig. 7.6. Distribution of properties of 100 ha or more: number by size category of property in hectares. NB: Too few properties in ‘10,000 or more’ group to show on graph of this scale.

Industry on the Land

191

Table 7.9. Distribution of industrial establishments c.1882, by type of owner of the estate on which they were found Type of owner

No. properties

Royalty & nobility ‘Bourgeois’ Religious body Statea Communityb Business, bank Other Unknown total a b

No. with industry

Total no. of indl. estab.

Indl. estab./ property

Indl. estab./ no. with ind.

6,520

2,593

4,304

0.66

1.660

7,770 159 1,093 203 96 76 12

2,827 20 264 32 44 19 3

4,315 33 368 53 81 33 3

0.56 0.21 0.34 0.26 0.84 0.43 0.25

1.526 1.650 1.394 1.656 1.841 1.737 1.000

15,929

5,802

9,190

0.58

1.584

Includes 16 properties with 5 industries owned by other German states and Sweden. Village, city, district, region, province, or communal group.

Table 7.10. Share of each kind of industry in total number of industrial establishments located on the properties of each owner category c.1882 Owner

Mill

Brick

Brew.

Distill.

Starch Sugar

Chem. Saw

Other

Royalty & nobility ‘Bourgeois’ Religious body State Communitya Business, bank Other Unknown

0.187

0.336

0.052

0.273

0.029

0.007

0

0.029

0.087

0.178 0.152

0.307 0.303

0.029 0.152

0.311 0.303

0.062 0

0.011 0

0 0

0.028 0

0.074 0.091

0.130 0.151 0.136

0.182 0.264 0.210

0.038 0.132 0.037

0.397 0.132 0.185

0.068 0.075 0.049

0.065 0.019 0.210

0 0 0

0.038 0.019 0

0.082 0.208 0.173

0.273 0.333

0.121 0

0 0

0.364 0.333

0.091 0.333

0.061 0

0 0

0 0

0.091 0

overall average

0.180

0.314

0.041

0.295

0.047

0.013

0

0.028

0.082

a b

b

Village, city, district, administrative region, province, or communal group. Only one chemical plant in total, share therefore < 0.0005.

Did different types of owners tend to specialize on different types of industries? Table 7.10 shows that, to some extent, they did: physical persons (commoners as well as royalty and nobility) exhibit very similar proﬁles in the tables, with about half of their industrial establishments being either brickworks or grain mills, and another quarter or more, distilleries. These three types made up just short of 80 per cent of the number of industrial establishments on the properties of each of

192

Industry on the Land

those two groups. The Prussian State, along with monasteries and other religious bodies, tended to concentrate on distilleries, but the state paid far less attention to brewing and brick making than did the monks and priests. Over 40 per cent of the industrial establishments owned by communities, on the other hand, were either mills or brickworks, and another quarter were breweries and distilleries. The most diverse proﬁle was that of the category ‘banks and business ﬁrms’, the only category for which the largest single type of industry owned was sugar factories (tied with brickworks for ﬁrst place). The category of ‘other industries’, which formed a signiﬁcantly large component of the proﬁles of communities, along with banks and business ﬁrms, was a diverse set of establishments. Communities owned 11 ‘other’ establishments located on ﬁve properties (one in Pomerania and four in Silesia): a cheese factory, a pinecone drying plant, a basalt quarry, a stone quarry, two mines, and six blast furnaces. Business ﬁrms, on the other hand, had 14 ‘other’ industrial establishments on nine properties in four provinces, and a very diverse group of enterprises this was: an oil mill, a glass factory, and a chicory drying plant; a cement factory, a leather and wood plant, two brown coal pits, another coal mine, an iron mine, a sandstone quarry, a lime pit, an iron smelter, a locksmith shop, and a joinery. Concern with categories of owners other than the two categories of physical persons is, however, important only for ﬁlling in details: calculating from the data in Table 7.11, we ﬁnd that physical persons owned almost 94 per cent of all the estates on which industrial establishments were found c.1882; indeed, they owned over 90 per cent in every individual category except sugar factories, where their share was still nearly 70 per cent. These two dominant groups, the aristocracy on the one hand,¹⁸ and the bourgeoisie on the other, will therefore command the overwhelming bulk of our attention from this point forward. The Prussian State—the biggest single owner, but as a social category a relatively minor player with only 4 per cent of rural-based industrial establishments on its properties—will draw secondary attention only. Over time, the decline in the numbers of industrial establishments on large landed properties was not shared equally among owners. The decline in the number of plants found on the properties of royalty and nobility was almost exactly in proportion to the overall decline, so the share of this group of owners remained nearly constant (46.9% c.1882 vs. 47.7% c.1907). But the share of ‘bourgeois’ owners fell, from 47 per cent to under 42 per cent; nearly all of the decline occurred after c.1895. The State share rose from 4.0 per cent to 6.2 per cent, and that of the remainder of the owner categories from 2.2 to 4.1 per cent (see Fig. 7.7.) The expansion of the ‘all others’ category was an expansion of each of its constituent parts: religious bodies and communal groups gained properties, as ¹⁸ Deﬁned for purposes of this chapter, as before, as being of royal or noble status.

Industry on the Land

193

Fig. 7.7. Ownership of industrial establishments on larger landed properties: number by type of landowner.

did the ‘miscellaneous’ category, but the biggest gain was by the category of banks and business ﬁrms, which doubled the number of properties it held. Part of this doubling was, in a sense, mere statistical artefact—the conversion of several of Countess Johanna Schaffgotsch’s properties in Silesia from personal ownership to ownership by Schaffgotsch Works LLC in 1904.¹⁹ But it also represents a real phenomenon, exempliﬁed by the rise of the Landbank Corporation of Berlin as an owner by buying individual properties from individual sellers, most of them non-noble persons.²⁰ Nevertheless, we should not make too much of these phenomena, since even in the ﬁnal year individual physical persons, nobles and commoners taken together, still owned 90 per cent of all properties of 100 hectares or larger on which industrial establishments were found.

7.3.4. Owners with Twelve or more Industrial Establishments on their Land The tables in this section show the landowners with the greatest number of industrial establishments (excluding dairy plants) on properties of 100 hectares or more. In contrast to the case of landownership per se, fully one-quarter of the top twenty owners with industrial plants on their properties were non-noble individuals.²¹ All of the nineteen physical persons in this list were substantial ¹⁹ The former Johanna Gryzik, daughter of poverty, foster child of Karl Godulla, a major zinc producer, inherited all of Godulla’s wealth and properties and later married ‘the more or less destitute’ Count Hans-Ulrich von Schaffgotsch. Skibicki, Pless, 238–9. ²⁰ Please see further discussion of this point in the following section. ²¹ In Tables 7.11 to 7.14 names of non-noble owners are italicized for easier reference.

194

Industry on the Land

Table 7.11. Owners of properties having twelve or more industrial establishments (excl. dairies) on larger estates in the seven eastern provinces of Prussia, c.1882 Rank

1 2 3 4 5

6 7 8

9 10

11 12 13

14 15 16 17 18 19 20

Owner

Prussian State King Wilhelm I of Prussia Brunswick, Duke of Pless, Hans Heinrich XI, Prince Henckel von Donnersmarck, Guido, Count Biron von Curland, Calixt, Prince Prince Frederick of the Netherlands zu HohenloheOehringen, Hugo, Prince Ratibor, Viktor, Duke of HohenzollernSigmaringen, Carl Anton, Prince Friedenthal, Dr. Carl Kennemann, Hermann Schaffgotsch, Ludwig, Imp. Count Krieger, Ernst Schulz, Friedrich Hugo Anhalt-Dessau, Duke of Redern, Heinrich von, Count Sydow, C. F. W. von Schlieffen, Max von Schulz, H.

Propertiesa Number

Area

Tax net yield

357 38

1,072 73

2,185,124 97,348

13,513,025 627,007

30 29

3 9

40,191 61,267

390,078 364,723

27

16

23,299

130,351

24

7

22,895

178,433

22

9

35,418

155,662

21

12

39,321

232,296

21

8

33,846

180,932

21

21

43,777

129,060

18 17

8 16

8,507 11,270

105,550 96,519

16

6

31,455

125,874

15 14

11 10

7,274 7,936

49,296 87,876

13

33

21,723

261,585

13

24

18,766

227,919

12 12 12

5 6 3

5,476 6,815 5,373

81,307 72,361 34,426

Industrial estabs.

Note: Italicized names are of non-nobles. a

Area of properties in hectares; tax net yield in marks.

Industry on the Land

195

landowners; none possessed less than 5000 hectares in properties of 100 hectares or larger, and thirteen of the nineteen owned more than 10,000 hectares (but among these, only one of the ‘bourgeois’). The top of the list is overwhelmingly populated with owners whose properties lay in the province of Silesia. Of the nine physical persons with more than 20 industrial establishments each on their estates c.1882,²² only three had any properties with industrial establishments outside of Silesia: (1) Prince Calixt Biron von Curland (rank 6), but of his 24 industrial establishments, only 4 were outside of Silesia (a mill, two brickworks, and a distillery in the province of Posen). On the other hand, (2) Prince Carl Anton von HohenzollernSigmaringen (rank 9), with estates also in Pomerania and Posen, had only 3 of his 21 industrial properties in Silesia; and (3) the King of Prussia and Emperor of Germany (rank 2) owned estates with industrial properties in Brandenburg, Posen, and the province of Saxony, as well as in Silesia, in which province 16 of the 38 industrial establishments on his land were located.²³ The ranks of the biggest owners did not remain static as the period progressed. Tables 7.12 and 7.13 reveal this clearly, but it is perhaps most easily seen in Table 7.14. One of the most obvious changes is the ascent of the biggest bourgeois owner with industrial establishments on his land, Hermann Kennemann, who moved up from 12th to 8th to 7th on the three lists, as he added more than 5000 hectares and six industrial establishments to his holdings between 1880 and 1909. As pointed out earlier in the chapter on the largest landowners, this is the same Kennemann who was a co-founder of the radical German Eastern Marches Society (Deutscher Ostmarken-Verein), the organization often referred to as H-K-T, and its members as Hakatisten, from the initials of the three founders.²⁴ All of Kennemann’s properties were to be found in the province of Posen.²⁵ The other big bourgeois owner in 1880, Dr Carl Friedenthal, did not appear in either of the later years: he must have died before 1892, for his estates had been divided among heirs by that time. None of the heirs had twelve ²² All these owners were royal or noble, except, of course, for the Prussian State. ²³ Of the 16, 8 were distilleries and 1 a brewery. ²⁴ To my knowledge, the standard work in English on this society is still Richard Wonser Tims, Germanizing Prussian Poland: The H-K-T Society and the Struggle for the Eastern Marches in the German Empire, 1894 –1919 (New York and London: Columbia University Press and P. S. King & Son, 1941). ²⁵ Kennemann, who liked to style himself Kennemann-Klenka—adding his estate of residence to his name, as the nobility often did, especially to differentiate branches of a larger family—died on 11 April 1910. According to Hans Rosenberg, Kennemann ‘refuse[d] the ‘‘von’’ ’—i.e. refused to be ennobled. Hans Rosenberg, ‘Die ‘‘Demokratisierung’’ der Rittergutsbesitzerklasse’, in Wilhelm Berges and Carl Hinrichs, eds., Zur Geschichte und Problematik der Demokratie. Festgabe für Hans Herzfeld anlässlich seines 65. Geburtstages am 22. Juni 1957 (Berlin: Duncker & Humblot, 1958), 466.

196

Industry on the Land

Table 7.12. Owners of properties having twelve or more industrial establishments (excl. dairies) on larger estates in the seven eastern provinces of Prussia, c.1895 Rank

1 2 3

4

5 6 7

8 9 10 11

12 13

14

15 16 17 18 19 20

Owner

Prussian State King Wilhelm II of Prussia zu HohenloheOehringen, Hugo, Prince Henckel von Donnersmarck, Edgar, Count Biron von Curland, Gustav, Prince Pless, Hans Heinrich XI, Prince HohenzollernSigmaringen, Leopold, Prince Kennemann, Hermann Ratibor, Viktor, Duke of King Friedrich August of Saxony Sachsen-Weimar, Sophie, Grand Duchess of Arnim, Traugott Herrmann, Count Henckel von Donnersmarck, Hugo, Count HohenloheIngelﬁngen, Friedr. Wilhelm, Prince Krieger, Ernst Redern, Wilhelm von, Count Thiele-Winckler, Hubert von von der Osten-Plathe, Count Anhalt-Dessau, Duke of Neuzelle Monastery

Propertiesa Number

Area

Tax net yield

370 38

1,211 79

2,296,057 97,180

14,745,384 697,353

30

15

41,661

255,164

26

18

26,756

137,194

25

5

21,502

163,785

22

10

61,138

368,766

22

20

44,901

162,974

22 21

18 10

13,897 38,211

110,958 177,401

20

5

21,920

228,207

17

4

14,774

303,140

16

5

34,878

117,781

16

12

15,625

98,243

15

18

24,114

57,262

15 14

10 26

7,095 20,358

45,256 258,654

14

18

10,902

96,552

14

10

10,489

79,841

13

33

23,096

263,505

12

5

18,425

145,925

Industrial estabs.

Note: Italicized names are of non-noble owners. a Area of properties in hectares; tax net yield in marks.

Industry on the Land

197

Table 7.13. Owners of properties having twelve or more industrial establishments (excl. dairies) on larger estates in the seven eastern provinces of Prussia, c.1907 Rank

1 2 3

4 5

6 7 8 9

10 11 12 13 14 15

16 17 18 19 20

Owner

Industrial estabs.

Prussian state 455 King Wilhelm II of 48 Prussia zu Hohenlohe28 Oehringen, Christian Kraft, Prince Royal Settlement 25 Commission Henckel von 25 Donnersmarck, Lazy, Arthur, and Edgar Biron von Curland, 23 Gustav, Prince Kennemann, Hermann 23 Pless, Hans Heinrich 21 XV, Prince Henckel von 21 Donnersmarck, Guido, Prince Ratibor, Viktor, Duke of 16 Neuzelle Monastery 16 Schaffgotsch Works LLC 16 Redern, Wilhelm von, 15 Count Hohenzollern, Wilhelm, 15 Crown Prince Hohenzollern14 Sigmaringen, Friedrich Heinrich, Prince King Friedrich August of 13 Saxony Landbank Corporation 13 Anhalt-Dessau, Duke of 12 Tiele-Winckler, Franz 12 Hubert, Count Arnim-Muskau, 12 Herrmann, Count

Note: Italicized names are of non-noble owners. a

Area of properties in hectares; tax net yield in marks.

Propertiesa Number

Area

Tax net yield

1,393 89

2,520,007 110,764

14,585,920 836,393

11

43,732

281,099

47

24,073

212,234

4

14,916

96,220

1

18,305

142,262

25 5

17,315 50,505

137,511 319,364

6

21,876

119,955

5 7 4 24

33,393 20,057 3,651 19,288

173,927 148,953 35,303 248,329

18

32,215

99,653

4

20,772

131,279

25

22,534

231,305

37 34 13

23,479 28,472 17,674

171,944 322,692 166,647

2

26,814

79,525

198

Industry on the Land

Table 7.14. Owners of properties having twelve or more industrial establishments, in rank order (excluding the Prussian State, which was no. 1 in all years) Rank

c.1882

c.1895

c.1907

2 3

King Wilhelm I of Prussia Brunswick, Duke of

4

Pless, Hans Heinrich XI, Prince

King Wilhelm II of Prussia zu Hohenlohe-Oehringen, Christian Kraft, Prince Royal Settlement Commission

5

Henckel von Donnersmarck, Guido, Prince Biron von Curland, Calixt, Prince Prince Frederick of the Netherlands zu Hohenlohe-Oehringen, Hugo, Prince Ratibor, Viktor, Duke of

King Wilhelm II of Prussia zu Hohenlohe-Oehringen, Hugo, Prince Henckel von Donnersmarck, Edgar, Count Biron von Curland, Gustav, Prince

6 7 8 9 10

Pless, Hans Heinrich XI, Prince Hohenzollern-Sigmaringen, Leopold, Prince Kennemann, Hermann Ratibor, Viktor, Duke of

11

Hohenzollern-Sigmaringen, Carl Anton, Prince Friedenthal, Dr. Carl

12

Kennemann, Hermann

13 14

Schaffgotsch, Ludwig, Imperial Count Krieger, Ernst

15

Schulz, Friedrich Hugo

16

Anhalt-Dessau, Duke of

17 18

Redern, Heinrich von, Count Sydow, C. F. W. von

19

Schlieffen, Max von

Redern, Wilhelm von, Count Thiele-Winckler, Hubert von von der Osten-Plathe, Count Anhalt-Dessau, Duke of

20

Schulz, H.

Neuzelle Monastery

21 22 23

King Friedrich August of Saxony Sachsen-Weimar, Sophie, Grand Duchess of Arnim, Traugott Herrmann, Count Henckel von Donnersmarck, Hugo, Count Hohenlohe-Ingelﬁngen, Friedr. Wilhelm, Prince Krieger, Ernst

Hecker & Co. Schaffgotsch, Johanna, Countess

Henckel von Donnersmarck, Lazy, Arthur, and Edgar Biron von Curland, Gustav, Prince Kennemann, Hermann Pless, Hans Heinrich XV, Prince Henckel von Donnersmarck, Guido, Prince Ratibor, Viktor, Duke of Neuzelle Monastery Schaffgotsch Works LLC Redern, Wilhelm von, Count Hohenzollern, Wilhelm, Crown Prince Hohenzollern-Sigmaringen, Friedrich Heinrich, Prince King Friedrich August of Saxony Landbank Corporation Anhalt-Dessau, Duke of Thiele-Winckler, Franz Hubert, Count Arnim-Muskau, Herrmann, Count Martin, Friedrich von Becker, Hermann von Knebel Doeberitz, Martha von, née von der OstenPlathe

Industry on the Land

199

or more industrial establishments on their properties. The Friedenthal story appears to have had its classic elements: in 1883, Elisabeth Friedenthal married Freiherr (Baron) Ernst von Falkenhausen, owner of some 721 hectares in four properties (one under 100 ha), who became von Friedenthal-Falkenhausen in respect of their taking up residence on the manor of Friedenthal.²⁶ She died in 1897, so that in 1909 her son Baron Ernst Carl von Friedenthal-Falkenhausen owned and was resident at Friedenthal, at that time about 1500 hectares. Her other heirs (she had six other children) owned the estates of Glasendorf, Tscheschdorf, Bielau, and Friedrichseck, some 2382 hectares in all. Friedrich Hugo Schulz, the third major bourgeois landowner in the list, also apparently died between 1884/5 and 1896, with his estates in Brandenburg and Posen going to three different heirs: Richard, Rudolf, and Dr W. Schulz. Ernst Krieger disappears from the list after c.1895. Several of his properties in West Prussia appear in the hands of several owners (including the Prussian State) in 1909, while some of them disappear entirely from the Handbooks (perhaps subdivided, but none of them appears in the Handbook of 1909 as being ‘subdivided or in the process of subdivision’). Only two properties appear to have remained in the family—in the hands of Margarete Hoeltzel, née Krieger, most likely his daughter. Several of the other changes observed in the rankings also appear to have been the result of death and inheritance: with 11 industrial establishments on properties of 100 hectares or more c.1882, Grand Duchess Sophie of SachsenWeimar just missed the list of top owners, but she made it in 1895 with 17. By 1909, Grand Duke Wilhelm Ernst had inherited all 14,000+ hectares of his mother’s land in Posen and Silesia, but the number of industrial establishments listed was only 10, so he did not make the list. The Emperor-King, biggest owner among the physical persons in all three years,²⁷ also added substantially to his landholdings (an extra 13,400 ha between c.1882 and c.1907) along with 10 more industrial establishments. Countess Johanna von Schaffgotsch disappeared from the list after 1895, but her properties did not: they are still there, ownership having been converted to a corporation, the Schaffgotsch Works LLC, in 1904, as pointed out in the previous section. An interesting interloper in the list is the Landbank Corporation of Berlin, which appears as a substantial owner of properties with industrial establishments c.1907. From whom did it acquire these properties? The answer is simple: from a great diversity of owners. Of the 37 properties of 100 or more hectares owned ²⁶ Made up of parts of the former manor of Giessmannsdorf (in Grottkau district) and at least four properties in the district of Neisse. Interestingly, almost all nobles who added the name of their family seat to their name put the family seat after their name (e.g. von Arnim-Muskau, ranked 19 in 1907, or von der Osten-Plathe, ranked 17 in 1895, or zu Hohenlohe-Oehringen, ranked 8 in 1882 and 3 in both 1895 and 1907). ²⁷ As the succession went from Wilhelm I to Friedrich III to Wilhelm II.

200

Industry on the Land

by the Landbank in six of the seven provinces (excepting the province of Saxony) c.1907, 10 had been owned c.1895 by eight nobles, and 23 more by 23 different non-noble individuals. Of the remaining four, I could not ﬁnd two in 1895; one of the other two was ‘in bankruptcy’ in 1895 and the last had been owned by a company. It seems quite obvious that the Landbank bought properties individually, or perhaps foreclosed on a mortgage or two, but, to achieve its position as a substantial landowner with over 23,000 hectares, it certainly did not buy in large chunks from one or a few great land owners.

7 . 4 . FAC TO R S I N F LU E N C I N G T H E LO C AT I O N O F RU R A L - B A S E D I N D U S T R I A L E N T E R P R I S E S What effect did certain characteristics of the property have on whether or not an industrial establishment was likely to be located there? To analyse the effect, I ran a series of regressions using a simple Linear Probability Model (hereafter LPM) of the following sort: The dependent variable, HasInd, was = 1 if there was at least one industrial establishment²⁸ enumerated on the property, and 0 otherwise. The principal independent variables were the following: 1. RG (= 1 if the property was a Rittergut [knight’s estate], 0 otherwise²⁹). 2. OF (= 1 if the owner of the property was a military ofﬁcer—active or retired—and 0 otherwise). Note that both ‘aristocrats’ and ‘bourgeois’ became ofﬁcers. 3. Size categories: Area2C5C = 1 if the property was between 200 and 499 hectares in size; Area5C1K if 500 to 999; Area1K2K if 1000 to 1999; Area2K5K if 2000 to 4999; Area5K10K if 5000 to 9999; and Area10Kplus if 10,000 or more. These are the standard property size categories used in virtually all German studies of landownership. Properties in the range of 100 to 199 hectares thus form the reference category for these dummy variables. 4. ArblCat: Took on the values 0 through 9; = percentage decile (truncated value of percentage of arable land in the property divided by 10, so if 43% arable, ArblCat = 4, if 97% arable land, ArblCat = 9, and so forth). 5. IsBOUR (= 1 if the owner was a non-noble person, 0 otherwise). ²⁸ Because of very inconsistent use of the term ‘Molkerei’ (dairy plant), I ignored this designation. Thus the variable HasInd received a value of 1 if any of the following was non-zero for the property: ﬂour mill, brickworks, brewery, distillery, starch factory, sugar factory, chemical works, sawmill, or ‘other’. See the explanation in Sect. 7.3.3 for examples of ‘other’ industrial establishments. ²⁹ A manor with one or more knight’s estates among its subunits was considered to be a knight’s estate in assigning a value for this dummy variable.

Industry on the Land

201

6. IsRLGN (= 1 if the owner was a religious organization or church, 0 otherwise). 7. IsSTATE (= 1 if the owner was the Prussian State or other foreign or German state, 0 otherwise). 8. IsCOMM (= 1 if the owner was a community or communal group,³⁰ 0 otherwise). 9. IsBSNS (= 1 if the owner was a bank or business ﬁrm, 0 otherwise). 10. IsOTHR (= 1 if the owner was neither a non-noble physical person nor a member of any of the other groups in nos. 5–8. Included are such things as non-religious academic institutions, hospitals, secular foundations, and even the ‘Association for the Combating of Vagabondage’,³¹ as well as a tiny number of unknown owners.) These variables comprise the core set of independent variables. Note that the reference group for ownership is the group of royal and noble physical persons, so coefﬁcients for variables 5 through 10 will be relative to this group, chosen as the reference group because its share of total industrial enterprises remained nearly constant over the period here considered. The rationale for including each of the independent variables is as follows: 1. Knight’s estates had been privileged properties, with political powers only recently (c.1878) reduced. One could speculate that it had been easier for the owners of such properties to establish industrial plants, both in terms of access to credit and to any ofﬁcial permissions that may have been necessary. 2. In order to maintain their expected lifestyle, military ofﬁcers often had to supplement their pay from other sources. They might therefore have been more likely to seek and/or seize opportunities to increase the revenue from their estates to provide such a supplement. 3. One might expect that the larger the estate, the more likely it would be that it could support one or more industrial establishments. This is conﬁrmed by the data of Table 7.7. 4. The decile of arable land as a share of the entire property is a crude proxy for value per hectare of the estate as a whole; one could expect that owners with more valuable land could more easily set up an industrial establishment on their estates. 5. Different categories of owner might react differently to, or even perceive less or more readily, proﬁt opportunities through establishing some industrial ³⁰ Village, district, region, province, or such things as local ﬂood control societies or pasturage cooperatives. ³¹ Verein für die Bekämpfung des Vagabondentums.

202

Industry on the Land plant. While I prefer to let the sign of the coefﬁcients for different owner types be determined by the data, intuition might suggest that we should expect priests, politicians, and communal groups to be less likely to set up industries, and ‘bourgeois’ individuals and business ﬁrms perhaps more likely, compared to aristocrats. This may be because some groups may have different attitudes toward proﬁt or because they have different types of properties that may be less well-suited, or better suited, to the setting up or maintaining of industrial establishments. The ‘other’ category is such a grab-bag that it is difﬁcult to predict how this group might go.

This simplest version of the core regression produced the results shown in Table 7.15. While the amount of variation explained is very small indeed, the size and signiﬁcance of the individual coefﬁcients are very interesting. According to these results, one was more likely to ﬁnd an industrial establishment on a knight’s estate, ceteris paribus, and on one owned by a current or former military ofﬁcer or by a bank or business ﬁrm, but less likely to ﬁnd some industry on an estate owned by a religious group, a community, or a member of the ‘other’ category, compared to one owned by a member of the aristocracy. Ownership by the Prussian State or a ‘bourgeois’ person produced mixed results. While many of the coefﬁcients were statistically signiﬁcant at better than the 0.1 per cent level, the extremely small amount of variability explained by the equation suggests that more variables would need to be added for a fuller explanation. The data allow numerous variations on the basic theme. It is useful to ask if the chances of ﬁnding an industrial establishment on a property varied by location or size. For example: did it make a difference which province was home to a particular property, or how big it was? Taking West Prussia as the reference province (see argument above—p. 179—under industrial establishments per Table 7.15. Comparison of simple regression results Indep. vbls.

Constant RG OF ArblCat IsBOUR IsRLGN IsSTATE IsCOMM IsBSNS IsOTHR RSQ, N

c.1882 Coeff.

Signif.

c.1895 Coeff.

Signif.

c.1907 Coeff.

Signif.

0.332 0.184 0.044 −0.017 0.036 −0.155 0.149 −0.165 0.149 −0.050

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.327

0.317 0.170 0.034 −0.013 0.006 −0.161 −0.033 −0.173 0.152 −0.097

0.000 0.000 0.002 0.000 0.509 0.000 0.043 0.000 0.001 0.029

0.252 0.199 0.046 −0.006 −0.016 −0.135 0.001 −0.144 0.057 −0.011

0.000 0.000 0.000 0.000 0.060 0.000 0.925 0.000 0.071 0.810

0.048

15,871

0.045

16,097

0.057

15,799

Industry on the Land

203

Table 7.16. Core regression results with provinces and size categories Indep. vbls.

Constant RG OF ArblCat

1882 Coeff.

Signif.

1895 Coeff.

Signif.

Coeff.

1907 Signif.

0.037 0.070 0.039 0.010

0.060 0.000 0.000 0.000

0.023 0.071 0.017 0.010

0.229 0.000 0.108 0.000

0.012 0.082 0.020 0.011

0.469 0.000 0.063 0.000

IsBOUR IsRLGN IsSTATE IsCOMM IsBSNS IsOTHR

0.086 −0.076 −0.025 −0.133 0.141 −0.040

0.000 0.025 0.494 0.000 0.002 0.404

0.068 −0.081 −0.102 −0.142 0.158 −0.092

0.000 0.017 0.000 0.000 0.000 0.029

0.070 −0.055 −0.069 −0.107 0.094 −0.002

0.000 0.087 0.000 0.000 0.002 0.966

IsBB IsPM IsEP IsPS IsSL IsSA

0.031 −0.107 −0.055 −0.052 0.017 0.009

0.040 0.000 0.000 0.000 0.216 0.571

−0.022 −0.085 −0.061 −0.048 0.065 0.025

0.133 0.000 0.000 0.001 0.000 0.107

−0.014 −0.098 −0.054 −0.027 0.051 −0.033

0.334 0.000 0.000 0.063 0.000 0.026

Area2C5C Area5C1K Area1K2K Area2K5K Area5K10K Area10KPlus

0.129 0.370 0.558 0.517 0.332 0.643

0.000 0.000 0.000 0.000 0.000 0.000

0.130 0.356 0.538 0.486 0.312 0.664

0.000 0.000 0.000 0.000 0.000 0.000

0.119 0.354 0.541 0.462 0.280 0.741

0.000 0.000 0.000 0.000 0.000 0.000

RSQ

0.145

0.135

0.154

square kilometre), I used the following dummy variables to indicate whether a given property was found in a particular province: 1. 2. 3. 4. 5. 6.

IsEP (East Prussia) IsBB (Brandenburg) IsPM (Pomerania) IsSL (Silesia) IsSA (province of Saxony) IsPS (Posen)

The core regression with these dummies added produced the results of Table 7.16. The ﬁrst thing to note is that adding the provincial and size dummies creates another set of highly signiﬁcant variables (t-values for the size category variables typically in the range of 9 to 30+). The probability of ﬁnding an industrial establishment increases with the size of the property, except for the 2000 to 9999 hectare range. This reafﬁrms the result we found in Table 7.8. Adding these dummies has reduced the size, but not the signiﬁcance, of the effect of the property’s being a knight’s estate, and the coefﬁcient of ArblCat

204

Industry on the Land

changes from negative to positive. The coefﬁcient for being a military ofﬁcer appears to rise a bit, but loses some signiﬁcance. Compared to the probability of ﬁnding an industrial establishment on an estate owned by one of the aristocracy, the greater probability associated with ownership by a non-noble person becomes somewhat smaller and less signiﬁcant, just as the lesser probability associated with ownership by a religious body becomes less negative and less signiﬁcant. We ﬁnd that ownership by the State now reduces the probability of there being an industrial establishment on the property, at least in the later two periods, and the lower probability for properties owned by communities stays negative and signiﬁcant, while the higher probability associated with ownership by a bank or business ﬁrm stays high and strongly signiﬁcant. For ‘other’ owners, our miscellaneous group, the probability of ﬁnding an industry on their estates was not signiﬁcantly different from that of ﬁnding one on the estates of our reference group, the aristocracy. In the provinces of East Prussia, Posen, and Pomerania, an industrial enterprise was less likely to appear on an estate there than on an estate in West Prussia, ceteris paribus. The probability for Brandenburg and Saxony was about the same as that for West Prussia, while it was higher if the property were in the province of Silesia, according to the results of this regression. An F-test (available from the author on request) showed the form of the equation to be highly signiﬁcant, even though the total variation explained by this set of variables was still low. Provinces, however, are not uniform territories. Would it make a difference to our results if we divided the seven provinces into their constituent administrative regions (Regierungsbezirke)? In the next equation, new dummy variables to designate the administrative regions were substituted for the provincial dummies. These were the following: 1. IsKBG (Königsberg, East Prussia) 2. IsGUM (Gumbinnen, East Prussia) 3. The region of Danzig (West Prussia) was taken as the reference region, so there is no dummy variable for this region. 4. IsMRW (Marienwerder, West Prussia) 5. IsPOT (Potsdam, Brandenburg) 6. IsFFO (Frankfurt/Oder, Brandenburg) 7. IsSTN (Stettin, Pomerania) 8. IsCOS (Cöslin, Pomerania) 9. IsSTR (Stralsund, Pomerania) 10. IsPOS (Posen, Posen) 11. IsBRM (Bromberg, Posen) 12. IsBRE (Breslau, Silesia) 13. IsLGN (Liegnitz, Silesia) 14. IsOPL (Oppeln, Silesia) 15. IsMGB (Magdeburg, Saxony)

Industry on the Land

205

16. IsMRB (Merseburg, Saxony) 17. IsERF (Erfurt, Saxony) The results are shown in Table 7.17. It is well to remember that we are not comparing apples to apples between Tables 7.16 and 7.17: the reference area for the provincial regressions was the entire province of West Prussia, while the reference area for the regional regressions was only a part of that province, namely the region of Danzig. Even more important, however, is the meaning of Table 7.17. Core regression results for regions and size categories Indep. vbls.

1882 Coeff.

Signif.

1895 Coeff.

Signif.

1907 Coeff.

Signif.

0.021 0.071 0.037 0.012

0.364 0.000 0.000 0.000

0.007 0.076 0.013 0.012

0.759 0.000 0.217 0.000

−0.012 0.074 0.017 0.013

0.575 0.000 0.107 0.000

IsBOUR IsRLGN IsSTATE IsCOMM IsBSNS IsOTHR

0.071 −0.043 −0.054 −0.106 0.123 −0.008

0.000 0.204 0.138 0.001 0.007 0.866

0.054 −0.042 −0.101 −0.112 0.150 −0.060

0.000 0.211 0.000 0.000 0.000 0.155

0.054 −0.019 −0.071 −0.078 0.080 0.050

0.000 0.559 0.000 0.002 0.008 0.231

IsKBG IsGUM IsMRW IsPOT IsFFO IsSTN IsCOS IsSTR IsPOS IsBRM IsBRE IsLGN IsOPL IsMGB IsMRB IsERF

−0.099 0.051 0.012 −0.045 0.122 −0.055 −0.050 −0.251 −0.054 −0.041 −0.020 0.083 0.016 −0.014 0.075 −0.121

0.000 0.030 0.601 0.045 0.000 0.019 0.029 0.000 0.012 0.067 0.361 0.000 0.507 0.572 0.002 0.001

−0.120 0.081 0.015 −0.124 0.099 −0.020 −0.036 −0.232 −0.042 −0.042 0.032 0.124 0.067 0.012 0.074 −0.066

0.000 0.001 0.493 0.000 0.000 0.390 0.109 0.000 0.052 0.061 0.135 0.000 0.004 0.627 0.002 0.069

−0.046 0.017 0.045 −0.089 0.123 −0.043 0.015 −0.240 0.025 −0.036 0.059 0.097 0.101 −0.032 0.040 −0.083

0.019 0.440 0.035 0.000 0.000 0.055 0.501 0.000 0.246 0.111 0.005 0.000 0.000 0.159 0.072 0.018

Area2C5C Area5C1K Area1K2K Area2K5K Area5K10K Area10KPlus

0.130 0.367 0.551 0.521 0.331 0.641

0.000 0.000 0.000 0.000 0.000 0.000

0.131 0.352 0.530 0.486 0.313 0.661

0.000 0.000 0.000 0.000 0.000 0.000

0.119 0.349 0.528 0.450 0.272 0.728

0.000 0.000 0.000 0.000 0.000 0.000

RSQ

0.161

Constant RG OF ArblCat

0.155

0.170

206

Industry on the Land

an insigniﬁcant coefﬁcient: in this version of the LPM, a coefﬁcient for a regional dummy that is not signiﬁcantly different from zero simply means that the data tell us that the probability of ﬁnding an industrial establishment on an estate in that region is not signiﬁcantly different from ﬁnding one on an estate in the region of Danzig, ceteris paribus. They do not tell us the absolute level of that probability. If we set our acceptable signiﬁcance level at the commonly used 5 per cent, then the rest of the regions divide themselves almost equally into the three possible different groups, based on relative probability of ﬁnding an industrial establishment on a property there, compared to the administrative region of Danzig: 1. Higher probability of ﬁnding an industrial establishment on a property of 100 hectares or more: Gumbinnen (East Prussia), except c.1907; Frankfurt/Oder (Brandenburg); Liegnitz and Oppeln (Silesia), except Oppeln in 1882; Merseburg (Saxony). 2. About the same probability: Marienwerder (West Prussia); Posen and Bromberg (Posen), except Posen in 1882; Breslau (Silesia), except 1907; Magdeburg (Saxony). 3. Lesser probability: Königsberg (East Prussia); Potsdam (Brandenburg); Stettin, Cöslin, and Stralsund (Pomerania), except Cöslin in 1907; Erfurt (Saxony), except 1895. Membership in the three groups emphasizes the point that there were intraprovince differences, especially in Saxony, which had one administrative region in each group. Gumbinnen and Königsberg in East Prussia wound up in opposing groups, as did Potsdam and Frankfurt/Oder in Brandenburg. In fact, only Pomerania, Posen, and West Prussia showed any uniformity by the sign of the probability measure. What have we come to in the last variant of the core regression? We have found that size mattered: both the absolute value of the coefﬁcients, and their level of statistical signiﬁcance, were highest for the size category variables. Moreover, the one break in the pattern that the larger the property, the greater the probability that it would have an industrial establishment, could be explained by the unusual composition of ownership and type of property in the 5000 to 9999-hectare groups. Geographic location mattered, too, but not in any systematic fashion; the extent of intra-province variation showed that. With the number of observations at our disposal, we could have introduced dummy variables to examine the locational effect district by district (Kreis), but since there were 259 such districts in the property databases, the tables would have become impossibly unwieldy, and the lack of systematic variation likely even more striking. The coefﬁcient for the variable ArblCat, measuring the share of arable land in the total area of the property, was always highly signiﬁcant, but its coefﬁcient

Industry on the Land

207

was very small (+1.2 to 1.3 per cent)—deﬁnitely lacking in ‘oomph’, to use Professor McCloskey’s characterization.³² The coefﬁcients for various classes of property owners turned out to be about as expected: the State, religious organizations, and community groups were less likely to have industrial establishments on their properties, compared to royal and noble owners, while non-noble individuals and especially business ﬁrms were more likely, and the miscellaneous group (‘other’) about equally likely. Business ﬁrms showed the largest positive coefﬁcient, and communities the largest negative one. Interestingly, the two were not very different in absolute value year by year, just as there was rough equality in absolute value, but opposite sign, for the coefﬁcients of non-noble (‘bourgeois’) persons and religious bodies. But, while many of the coefﬁcients were both highly signiﬁcant and of the expected sign, the equation does not explain very much of the variation in the data. There are certainly many variables that—had we had data for them—would have added to the explanatory power of the equation: density of population in the local area, levels of income, density and quality of the transport network, and so forth. There might still have remained a large random component, however. Nevertheless, the data provided by the Handbooks have given us considerable insight into, and important quantitative dimensions of, the extent of industrial enterprises operating on agricultural estates.

7 . 5 . S U M M A RY A N D C O N C LU S I O N S We have seen above that there was a surprisingly dense network of industrial enterprises on landed estates in the seven eastern provinces of Prussia, and that—despite a decline of at least 15 per cent in the number of such enterprises during the period under review—that density still remained at an average in the seven provinces of about one industrial establishment per square mile at the end of the period. The decline was geographically very evenly shared across the three major areas of East Elbia: the northern, or Baltic, provinces of Pomerania, West and East Prussia; the central-western provinces of Brandenburg and Saxony, and those in the east-south-east, namely Posen and Silesia. Looking at property size or ownership, however, the losses fell disproportionately on properties from 200 to 999 hectares in size and on ‘bourgeois’ owners. Of the branches of industry for which we could calculate the share of the total number of plants to be found on large rural estates, there was great interprovincial variety and great inter-industry variety. Only for distilling was the overwhelming majority of production units rural-based. Considering rural-based ³² Deirdre McCloskey, ‘The Trouble with Mathematics and Statistics in Economics,’ History of Economic Ideas, 13/3 (2005), 85–102.

208

Industry on the Land

industry, Silesia had the most, and the province of Saxony the least, but Saxony led all others in the number of sugar factories. The greatest absolute number of industrial establishments was found on properties of 500–999 hectares in size, the greatest number per property on the very largest properties (10,000 hectares or more), and the greatest number per unit of area on properties of between 100 and 199 hectares. Industrial establishments were overwhelmingly (90 per cent or more) located on properties owned by physical persons, not legal entities, in large part the consequence of the fact that rural land was overwhelmingly owned by physical persons.³³ The Prussian State was by far the largest single owner of land per se, and therefore of land on which industrial establishments were located, even if the density of industrial establishments on State domains was low. After the State, the largest of the owners was the Kaiser and King himself, followed by a group of owners consisting very strongly of persons of royal or noble status. There were a few non-noble persons in this group, but they tended to appear and disappear as the years passed, usually because inheritance divided their properties among several heirs. In the ﬁnal year, c.1907, two business ﬁrms entered the list of the top twenty owners of land on which industry was to be found, one of them because of conversion of personal property to property of a corporation, the other by piecemeal acquisition of individual properties from diverse owners. Using regression analysis to identify the relative importance of certain characteristics of a property in determining the probability of ﬁnding an industrial establishment on that property, we saw that location effects could be signiﬁcant, but were not systematic, while property size made a considerable difference, and in a very predictable pattern. Perhaps more interesting was the ﬁnding that industrial establishments were more likely to be located on knight’s estates than on other types of landed properties, ceteris paribus, although the increased probability was not particularly large.³⁴ Who owned the land also made a difference to the probability of ﬁnding an industrial establishment on the land, more or less as expected: compared to aristocratic owners, non-noble individuals and business ﬁrms were more likely to have industrial establishments on their properties, while the State, religious bodies, and communities were less likely. Most of the variation in the dependent variable, the existence of industrial establishments on agricultural properties, at the end still remained unexplained by the regression analysis. The characteristics of landed properties provided by the Handbooks, when combined with some other readily available data, ³³ Of the total area in properties of 100 ha or more, physical persons owned about 77% c.1895 (calculated from data in the Handbooks). ³⁴ This might also be a partial explanation for the price premium commanded by knight’s estates compared to other properties, as mentioned in Ch. 6 and analysed in Ch. 9.

Industry on the Land

209

nevertheless provided new insights into, and some solid quantitative measures of, the remarkable extent of rural-based industry in what used to be regarded as an exclusively agricultural area, lacking any signiﬁcant industry. The landlords of East Elbia were not simply export-oriented agriculturists; they had substantial industrial interests as well.

8 Classifying Land by its Use 8 . 1 . P R EV I EW Arranging the seven provinces into three bands (north, central, south) and within the bands from west to east, shows a clear east–west gradient in a land yield index by province, the uniformity of which is somewhat disturbed by breaking the bands down into administrative regions. By arranging the average land tax net yield per hectare in the 259 districts whose data are contained in the Handbooks, I show the distribution by district across provinces in overall terms and then by land type, all calculations based on the land tax cadastre of the early 1860s. In these calculations the province of Saxony stands out as having the highest-quality land in general, and also individually for ploughland, meadow, pasture, and forest. East Prussia stood generally at the bottom of the list. From distribution by district, the chapter next moves to classifying large properties by their four-dimensional land-use proﬁles, the share of total area devoted to arable agriculture, animal husbandry, forestry, and other uses. Analysis of these data, when properties are grouped into six clusters, forms the main body of the chapter. The six clusters identiﬁed by the pattern of land-use proﬁles are (1) a mostly arable cluster, (2) a mostly forestry cluster, (3) a cluster with land devoted to arable and forestry in about equal shares, (4) a mixed use cluster: arable with animal husbandry and some forestry, (5) an arable and animal husbandry cluster, and (6) a very small cluster where the land use is arable and other in about equal shares. The ﬁrst three clusters accounted for about two-thirds of all large properties in all three years, and the general structure of the clusters was remarkably stable over time. It was also remarkable that there appeared no cluster of properties mostly devoted to animal husbandry, despite the widespread practice of animal husbandry in these seven provinces. Examining the clusters by total area or total tax value contained within each cluster produces a somewhat more nuanced picture: the mostly arable cluster ranked ﬁrst by total tax value, but only fourth by total area, whereas the mostly forest cluster showed exactly the opposite rankings. Properties specializing in arable agriculture were to be found mostly in the east and south, forestry specialization tended to concentrate in Brandenburg and Silesia, the mixed farming with some forestry cluster in East Prussia, Pomerania, and Silesia.

Classifying Land by its Use

211

The last section of the chapter looks at possible specialization in ownership of properties with particular land-use proﬁles. Bourgeois landowners tended to concentrate their holdings in arable agriculture, the nobility in properties with mixed land use, and the Prussian State in forestry. Other landowners as a group had about half the tax value of their properties in properties specialized on arable agriculture, with the rest spread out over all different combinations of land use.

8 . 2 . I N T RO D U C T I O N : T H E D I S T R I BU T I O N O F L A N D QUA L I T Y A N D VA LU E Much has been made of the ‘East–West gradient’ in German economic history, especially in respect of the location of industry and the level of income per capita. Even within the mostly agrarian east, a similar east–west gradient existed, and it rested in large part on the quality of land and proximity to markets. When carrying out the land tax cadastre in the early 1860s, the Prussian tax authorities took precisely these factors into account when assessing the ‘land tax net yield’ upon which the land tax paid was to be based. Different types of land were graded according to quality—up to a maximum of eight quality grades for any type of land,¹ but typically eight only for ploughland, garden, and meadow; six for pasture and forest, two for water and moorland—and then each quality grade was assigned a potential net yield under normal management, given average local produce prices for the years 1837–60.² Thus the tax yield depended not only on the quality of land, but also on the market and transport conditions that determined farm-gate prices in the local area. Properly carried out, such a procedure should produce estimated ‘net yields’ that are a near-perfect proxy for the value of land in agricultural use.³ The estimated productivity ﬁgures from the land tax assessment could therefore be used as an index of land value, and—with reservations—even as a rough indicator of land quality. In general, ploughland, garden, and meadow had the highest assessments, followed by pasture and forest, with bodies of water and moorland showing the lowest assessments. The assessments of the highest land types were often a substantial multiple of those of the lowest. Of the high-value land types, ploughland and meadow had very similar, often identical, assessments for the respective quality grades. As a summary measure of land productivity, therefore, I have calculated the average assessment for ploughland and meadow ¹ Willy Wygodzinski, Die Besteuerung des landwirtschaftlichen Grundbesitzes in Preussen: Eine kritische Studie ( Jena: Gustav Fischer, 1906), 12. ² Ibid. 7, 11–12. For the actual numbers, see the volumes of results for each administrative district, published as Prussia. Ministerium der Finanzen: Die Grundsteuerveranlagung. Ergebnisse für den Regierungsbezirk (Berlin: Ministerium der Finanzen, 1869– ), 25 vols, one volume for each administrative region in Prussia. ³ See Eddie, ‘Value and Area’.

212

Classifying Land by its Use

Fig. 8.1. ‘Land yield index’ by province, early 1860s.

combined in each province and each administrative region, and then taken the ratio of this average to the overall weighted average for the seven eastern provinces taken together. I have called this ratio the ‘land yield index’.⁴ An index value of less than one indicates below-average net yield for the land tax, and above one an above-average yield.⁵ Fig. 8.1 shows this index for the seven provinces. In the chart the provinces are arranged into three bands (north, central, and south) and within each band from west to east. All three bands show a deﬁnite east–west gradient, and only the southern band shows net yield values above the average for all seven provinces. What we can see is an overall gradient that runs broadly from north-east to south-west, and which represents a rough measure of land quality in the seven provinces. The picture is not so uniform if we examine the data by administrative region, the ﬁrst subunit of the province in the Prussian administrative system. Using the same three horizontal bands, I have ordered the administrative regions from west to east where possible (in the provinces of Pomerania, East Prussia, and Brandenburg), otherwise from north to south (West Prussia, Posen, Saxony, Silesia). The results are shown in Fig. 8.2. ⁴ This is different from the land quality index that I calculate in the ﬁnal chapter, in that it rests on the mean, or average value, rather than on the median. ⁵ Note, however, that this index does not take into account other types of land, nor does it use any weighting for the share of ploughland and meadow in the total land area. It is thus a less comprehensive measure than that used at the beginning of Ch. 3 to discuss the land endowment of the seven eastern provinces.

Classifying Land by its Use

213

Fig. 8.2. ‘Land yield index’ by administrative region, early 1860s

The smallest administrative region (Stralsund, with only four districts) is the only area of the northern band to rise above the seventeen-region average; beyond that, the general east–west (or north–south) gradient in net yield holds across each of the three provinces, although not smoothly and uniformly across all three taken together. The pattern repeats roughly in the central⁶ and southern bands, but since in the provinces of Posen, Saxony, and Silesia the administrative regions run north–south rather than west–east, the existence of an east–west land productivity gradient within these provinces cannot be inferred, but only surmised. The foregoing data set the stage in broad terms. Clearly, a recalculation of the land yield index by district would produce an even more varied picture, but with 236 districts in the land tax cadastre, attempting to present the results graphically would probably be more confusing than enlightening, especially when going into detail about the individual land types. I have therefore adopted a classiﬁcation schema that can use both the data of the land tax cadastre as well as the data for large properties from the Handbooks. Landed properties may be grouped or classiﬁed in a number of different ways, primarily determined by the purposes to which the data are to be put. In other chapters I group them by various measures of size, by location, by who owned them, and by whether or not they were home to some sort of industrial enterprise. In this chapter I return to grouping them by their makeup, i.e. by their land-use proﬁles (see Chs. 3 and 6), based on how the land was used for different income-producing activities. Rural land produces principally crops, animals and ⁶ Had the city of Berlin, which was a separate administrative region by itself with very high land-tax assessments, been included with Potsdam region, as it often was for other statistical purposes, the value of the index for Potsdam region would not have been less than that for Frankfurt/Oder.

214

Classifying Land by its Use

animal products, or wood and wood products;⁷ consequently, this chapter will focus on these three primary economic activities. It is obvious that arable agriculture can only be undertaken where there is ploughland, just as forestry requires woodland. Thus the regional distribution of rural economic activities will depend in large measure on the location and quality of the suitable land types.⁸ It is therefore appropriate to examine ﬁrst the location and quality of the major land types within the seven provinces of our concern here. The type and quality of land were not evenly distributed within the seven provinces. From published volumes of the land tax assessment of the early 1860s used in calculating the land yield index above, I have calculated the average tax assessment per hectare of various types of land in each administrative unit down to the district level.⁹ Since the tax assessment data are not categorized by size of property, the averages could be calculated only for all properties irrespective of size. Presenting these averages by land type is much preferable to the common practice of calculating an overall average tax assessment per hectare for a given area, for it eliminates the composition effect, as we have seen earlier (Ch. 1) from the discussion of Hess’s analysis of the relative quality of land in entailed properties. In the tables for this chapter I present the frequency distribution of districts within the larger administrative divisions, to show how land quality varied both within and across provinces.¹⁰ For the sake of comparison, I ﬁrst present the overall averages, despite their shortcomings referred to above (Table 8.1). More than half the districts fall into the range of 0–9.9 marks per hectare, and it is clear that the province of Saxony stands way above the other provinces ⁷ Mining and industry have already been covered in Ch. 7. Despite its importance to noble large-estate owners in particular, I leave out hunting entirely, since I have no data by which to assess this activity. ⁸ Endowments of land types are not ﬁxed, however. They can be changed in response to demand or other factors, so forests may be uprooted and changed to ploughland, or ploughland changed to pasture, to mention only two possibilities. Large-scale examples of such land-use changes can be found in the Highland Clearances in Scotland in the early 19th century, for example, or in the settlement of the Oderbruch region in Brandenburg in the 18th century. The most massive of all was probably the extension of cultivation into the prairies of the United States in the later 19th century. ⁹ To make the tax data from the 1860s comparable to the data for large properties taken from the Handbooks, I have left out of consideration those city districts that in the three Handbook years had no property of 100 ha or more within their boundaries. If a rural district that existed in the 1860s was later subdivided into other districts (especially in the administrative reform in West Prussia and Posen in 1887), I assigned the average value for that district to each of its new component parts. If a new district was formed out of parts of two districts from the 1860s, I used the average of the two former districts for the new district. From the 236 rural districts listed in the tax assessments of the 1860s there emerged 259 districts—based on their 1895 administrative boundaries—to be evaluated in this chapter. ¹⁰ Since local prices of agricultural produce also ﬁgured in the calculation of the ‘land tax net yield’, this is not strictly a quality measure, since nearness to large markets and the state of the transport system also affected the net yield through their effect on the local price. But it is still a much better measure than some overall average.

Classifying Land by its Use

215

Table 8.1. Number of districts in each province having land tax net yield average for all land in the following ranges (marks/ha), c.1862 Province East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

0–9.9

10–19.9

20–29.9

30–39.9

40–49.9

50–60

Total

24 15 19 17 31 21 4

11 9 11 8 9 23 10

10 7

7 7

9

2

35 25 31 28 40 61 39

131

81

21

15

9

2

259

1 1 3

in terms of the taxable value of its land. East Prussia and Posen have no districts in which the average taxable value of all land is above 20 marks per hectare, Brandenburg and West Prussia only 1, and Pomerania only 3 (the unweighted mean value for all 259 districts was almost exactly 14 marks/ha, the weighted mean almost exactly 12). So, on average, the least valuable land was to be found in the north-east and east, the most valuable in the south-west. The highest overall total average was found in the district of Wanzleben in Magdeburg region of Saxony (57.0 marks ha), the lowest in Schlochau district of Marienwerder region in West Prussia (only 3.1 marks ha). The foregoing ﬁgures are a blend of both land quality and the mix of land between high-value and low-value land types in the district. But since various kinds of land were taxed at very different rates, the variety of distribution of land by taxable value can only be captured by considering different land types separately. Accordingly, in what follows, I present the data for ploughland, meadow, pasture, and woodland.¹¹ As an introduction to this exercise, Table 8.2 shows the great dispersion in average net yield for the land tax in the seven provinces, giving further motive for considering land types separately and locally. Note that in the case of the four major land types, the highest average assessment per hectare is always a large multiple of the lowest average among the 259 districts of the seven provinces, ranging from 11 in the case of meadow to over 32 in the case of pasture. If we could examine units even smaller than the district, then these multiples would grow, as the lowest observed value would become smaller and the highest, larger. Having noted the dispersion of average assessments by district, let us now turn to a more detailed examination of the distribution of the four major land types. Tables 8.3A, B, and C begin this exercise by showing the distribution of ploughland according to average net yield for the land tax. ¹¹ To reduce clutter, I have left out separate tables for moorland and water, and ‘unproductive land’—all minor parts of total area, and very lowly taxed, or not at all.

216

Classifying Land by its Use Table 8.2. Average tax assessments (marks/ha) of six land types in 259 districts of the seven eastern provinces, c.1862 Land type

Mean

Standard deviation

Minimuma

Maximum

Max/Min ratio

Ploughland Meadow Pasture Woodland Water Moorland

16.26 18.67 4.82 4.88 3.37 0.87

10.86 10.52 3.17 3.71 3.44 0.68

4.12 4.77 0.76 0.90 0.39 0.27

58.13 53.33 24.63 23.61 22.91 6.13

14.1 11.2 32.6 26.2 59.0 22.4

all taxable land

14.03

10.36

3.14

57.01

18.2

a.

Excluding zero values for districts with no pasture, water, or moorland.

Table 8.3A. Number of districts in each province having land tax net yield for ploughland in the following ranges (marks/hectare) Province

0–9.9

10–19.9

20–29.9

30–39.9

40–49.9

50–60

Total

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

21 12 7 12 15 9 1

14 11 21 11 25 26 12

1 3 5

1

17 3

7 11

2 8

4

35 25 31 28 40 61 39

total

77

120

29

19

10

4

259

Table 8.3A shows, so far as quality of ploughland in the province was concerned, that East Prussia, West Prussia, and Posen were the least favoured provinces, and that Saxony, followed at some distance by Silesia, was the most favoured. The highest assessment—58.1 marks per hectare—was found in the aforementioned Wanzleben, the lowest—4.1 marks per hectare—in Ortelsburg district, Königsberg region, East Prussia. The ﬁgures for Saxony and Silesia also show a great intra-provincial range of quality of ploughland. Table 8.3B perhaps makes the picture somewhat clearer by presenting the same ﬁgures as percentages of the total number of districts in the province. The percentage ﬁgures in Table 8.3B also reveal that there was no such thing as the average province, so far as quality of ploughland was concerned. The province that comes closest to being ‘average’ is, rather surprisingly, Silesia, but that is primarily simply because it had a range of quality higher than any province save Saxony. Saxony had such high-quality land that it pulled the entire quality distribution upwards. Silesia, simply because it had a range of quality higher than

Classifying Land by its Use

217

Table 8.3B. % of districts in each province having land tax net yield for ploughland in the following ranges (marks/ha) Province

0–9.9

10–19.9

20–29.9

30–39.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

60.0 48.0 22.6 42.9 37.5 14.8 2.6

40.0 44.0 67.7 39.3 62.5 42.6 30.8

4.0 9.7 17.9

4.0

27.9 7.7

11.5 28.2

total

29.7

46.3

11.2

7.3

40–49.9

50–60

Total

3.3 20.5

10.3

100.0 100.0 100.0 100.0 100.0 100.0 100.0

3.9

1.5

100.0

Table 8.3C. % of each province in number of districts having land tax net yield for ploughland in the following ranges (marks/ha) Province East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

0–9.9

10–19.9

27.3 15.6 9.1 15.6 19.5 11.7 1.3

11.8 8.4 17.6 9.2 21.0 21.8 10.1

100.0

100.0

20–29.9

30–39.9

40–49.9

50–60

Total

3.4 10.3 17.2

5.3

58.6 10.3

36.8 57.9

20.0 80.0

100.0

13.6 9.3 12.0 10.9 15.5 23.6 15.1

100.0

100.0

100.0

100.0

100.0

any save Saxony, therefore rather surprisingly appears to come closest to being ‘average’. Table 8.3C contains the column percentages, i.e. the share of each province in the total number of districts falling within the given range of land tax net yield per hectare. By comparing the share in various ranges with the share in the total number of districts in Table 8.3C we can see that the ‘quality bias’ in ploughland was deﬁnitely downward for East Prussia and West Prussia, but upward for Brandenburg and especially for Silesia and Saxony, and ambiguous for Pomerania and Posen. What is striking from Table 8.3C is that of all districts in the seven provinces with averages above 20 marks per hectare, around 70 per cent or more of them were to be found in Silesia and Saxony (indeed, above 40 marks per hectare, fully 100 per cent). So the low-yield ploughland was concentrated in the north-east and east (East Prussia, West Prussia, and Posen), while the high-yield ploughland was concentrated in the south-east and south-west (Silesia and Saxony). Tables 8.4A through 8.4C present the same data for meadow. The picture is broadly the same, except that in each province the meadow land (or hayﬁeld)

218

Classifying Land by its Use

Table 8.4A. Number of districts in each province having land tax net yield for meadow in the following ranges (marks/ha) Province

0–9.9

10–19.9

20–29.9

30–39.9

40–49.9

50–60

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

16 8 4 9 4 1 1

18 11 20 17 33 26 8

1 5 6 2 3 23 4

10 12

1 10

4

35 25 31 28 40 61 39

total

43

133

44

24

11

4

259

1 1

Total

Table 8.4B. % of districts in each province having land tax net yield for meadow in the following ranges (marks/ha) Province

0–9.9

10–19.9

20–29.9

30–39.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

45.7 32.0 12.9 32.1 10.0 1.6 2.6

51.4 44.0 64.5 60.7 82.5 42.6 20.5

2.9 20.0 19.4 7.1 7.5 37.7 10.3

16.4 30.8

total

16.6

51.4

17.0

9.3

40–49.9

50–60

Total

1.6 25.6

10.3

100.0 100.0 100.0 100.0 100.0 100.0 100.0

4.2

1.5

100.0

4.0 3.2

tends to have a slightly higher net yield, so that in all provinces the frequency distribution stands a little bit more to the right in the table. For example, over 75 per cent of all districts in the seven provinces fall into the range of 20 marks or less per hectare for ploughland, whereas the same ﬁgure for meadow is just under 68 per cent. The upper three ranges (30–60 marks/ha) for ploughland contain about 14 per cent of the total number of districts, but for meadow the ﬁgure is somewhat over 15 per cent. The shift in the distribution is therefore detectable, but minor, for the provinces taken together, even though it is more marked for individual provinces, such as West Prussia and Posen, as a comparison of Tables 8.3C and 8.4C shows. Again, there is no such thing as an ‘average’ province, but in the case of meadow land, Silesia stands farther away from the average than it did in the case of ploughland. The highest and lowest values for meadow land, 53.3 and 4.1 marks per hectare, were found in Langensalza district (Erfurt region, Saxony) and Ortelsburg, respectively. Reckoning as before, East Prussia and Pomerania show a clear downward bias in the quality of their meadow land; Brandenburg, Silesia, and Saxony a clear

Classifying Land by its Use

219

Table 8.4C. % of each province in number of districts having land tax net yield for meadow in the following ranges (marks/ha) Province

0–9.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

10–19.9

20–29.9

30–39.9

40–49.9

50–60

Total

37.2 18.6 9.3 20.9 9.3 2.3 2.3

13.5 8.3 15.0 12.8 24.8 19.5 6.0

2.3 11.4 13.6 4.5 6.8 52.3 9.1

41.7 50.0

9.1 90.9

100.0

13.5 9.7 12.0 10.8 15.4 23.6 15.1

100.0

100.0

100.0

100.0

100.0

100.0

100.0

4.2 4.2

upward bias; whereas for West Prussia and Posen the case is ambiguous. Lowyield meadow land tends to be concentrated in the north-east and north, while high-yield meadow can be found mostly in the south. Thus the interprovincial quality distributions for ploughland and meadow are very similar, but not identical. The 8.3 and 8.4 sets of tables represent land devoted to the more intensive forms of land use in crop cultivation and animal husbandry. Pasture and woodland represent less intensive forms of land use, and in consequence their net yield is signiﬁcantly lower than for the former two. For that reason, Tables 8.5A through 8.6C contain ranges with a maximum width of 5 marks, rather than 10 marks. The ﬁgures in Table 8.5A and B show that East Prussia, and to some extent Posen also, fared relatively better in its endowment of pasture land than it did for either plough or meadow, while the same was not true for West Prussia or Pomerania. The province of Saxony again led the pack for pasture, as it had for plough and meadow, but Silesia was relatively poorer in the quality of its pasture, compared to the other two land types examined so far. Brandenburg’s pasture Table 8.5A. Number of districts in each province having land tax net yield for pasture in the following ranges (marks/ha) Province

0–1.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

5 10 1 13

total

36

7

2–4.9

5–9.9

19 12 19 6 21 30 13

10 3 8 8 19 21 19

120

88

10–14.9

Over 15

1

Total

3 5

2

35 25 31 28 40 61 39

13

2

259

3 1

220

Classifying Land by its Use Table 8.5B. % of districts in each province having land tax net yield for pasture in the following ranges (marks/ha) Province

0–1.9

2–4.9

5–9.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

14.3 40.0 3.2 46.4

54.3 48.0 61.3 21.4 52.5 49.2 33.3 46.3

28.6 12.0 25.8 28.6 47.5 34.4 48.7 34.0

11.5 13.9

10–14.9

Over 15

Total

5.1 0.8

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

2.9 9.7 3.6 4.9 12.8 5.0

Table 8.5C. % of each province in number of districts having land tax net yield for pasture in the following ranges (marks/ha) Province East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

0–1.9

2–4.9

5–9.9

10–14.9

13.9 27.8 2.8 36.1

15.8 10.0 15.8 5.0 17.5 25.0 10.8 100.0

11.4 3.4 9.1 9.1 21.6 23.9 21.6 100.0

7.7

19.4 100.0

Over 15

Total

100.0 100.0

13.5 9.7 12.0 10.8 15.4 23.6 15.1 100.0

23.1 7.7 23.1 38.5 100.0

land appeared to be by and large of fair to good quality. The highest-value pasture land appeared in Merseburg district of Saxony (24.6 marks/ha), the lowest value in Schlochau (0.76 marks).¹² While neither Posen nor Saxony included districts where the average quality of pasture fell into the bottom category, East Prussia, Brandenburg, Pomerania, and Silesia exhibited a range of pasture that included all types but the extreme best. Of these four, Silesia’s quality distribution was remarkably uniform, in the following sense: accounting for 23.6 per cent of the total number of districts in the table, Silesia’s share was within 1.5 percentage points of that ﬁgure in three of the four ranges, and only slightly more than 4 percentage points below it in the lowest quality category. In this sense, Silesia was clearly the average province when it came to pasture. Which province had the highest-value forest land? Again, Saxony—this province seemed uniquely blessed in the quality of all four of the major land types. Looking at the value of forest land, once again Ortelsburg district in East Prussia trails the pack, with the lowest average, only 0.90 marks per hectare, ¹² Three districts, all in Breslau region of Silesia, had no pasture land at all.

Classifying Land by its Use

221

Table 8.6A. Number of districts in each province having land tax net yield for forest in the following ranges (marks/ha) Province

0–1.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

13 12

total

41

8 7 1

2–4.9

5–9.9

10–14.9

Over 15

Total

21 10 25 11 31 28 7

1 3 6 9 2 29 11

3 13

8

35 25 31 28 40 61 39

133

61

16

8

259

Table 8.6B. % of districts in each province having land tax net yield for forest in the following ranges (marks/ha) Province

0–1.9

2–4.9

5–9.9

East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony

37.1 48.0

60.0 40.0 80.6 39.3 77.5 45.9 17.9

2.9 12.0 19.4 32.1 5.0 47.5 28.2

51.4

23.6

total

28.6 17.5 1.6 15.8

10–14.9

Over 15

Total

4.9 33.3

20.5

100.0 100.0 100.0 100.0 100.0 100.0 100.0

6.2

3.1

100.0

and once again Wanzleben district in Saxony takes the top prize, with 23.6 marks. To put this in better perspective, the average tax assessment for woodland in Wanzleben was higher than the average assessment for both meadow and ploughland in 197 of the other 258 districts in the seven provinces, and more than double the average for woodland in 243 of the other districts.¹³ Of the 30 districts with the highest average value of forest land, 23 were in Saxony, the other 7 in Silesia. In view of this great diversity in land types and quality in the seven provinces, ﬁrm generalizations elude us, but there did again seem to be a rather general gradient in land quality, irrespective of land type, that ran from south-west to north-east, but without sharply deﬁned borders: pockets of high-quality land existed in West Prussia, as did pockets of low-quality land in Saxony, even ¹³ But there were only 1249 ha of woodland in Wanzleben district at the time of the tax assessment, the third-smallest total for any non-city district in all of Saxony. In 1885, the three largest parcels in the district added to 999 (80%) of those hectares, owned by three separate owners: the Prussian State (411 ha), Baron Alexander von Asseburg in Schloss Neindorf (351 ha), and the Cloister of Our Beloved Lady in Magdeburg (237 ha).

222

Classifying Land by its Use Table 8.6C. % of each province in number of districts having land tax net yield for forest in the following ranges (marks/ha) Province East Prussia West Prussia Brandenburg Pomerania Posen Silesia Saxony total

0–1.9

2–4.9

5–9.9

31.7 29.3

15.8 7.5 18.8 8.3 23.3 21.1 5.3

1.6 4.9 9.8 14.8 3.3 47.5 18.0

18.8 81.3

100.0

13.5 9.7 12.0 10.8 15.4 23.6 15.1

100.0

100.0

100.0

100.0

100.0

19.5 17.1 2.4 100.0

10–14.9

Over 15

Total

though East Prussia and Posen had very little high-quality land, while Silesia had rather considerable areas of low-quality land. With this background regarding location and quality of different land types, we may now turn to classifying large properties according to their use of these land types for income-producing purposes.

8 . 3 . C L A S S I F Y I N G P RO PE RT I E S B Y L A N D - U S E P RO F I L E S As pointed out earlier, the principal land-based economic activities of rural properties are arable agriculture, animal husbandry, and forestry. To group properties according to their mix of economic activities, we may note that ploughland and garden are used in arable agriculture, meadow and pasture in animal husbandry, and woodland in forestry. The other land types (moor, water, farmyard, etc.) are either used in peripheral activities, do not produce at all, or serve more than one use. As in Ch. 3, I have assigned land types to four categories: ploughland and garden to arable, meadow and pasture to animal husbandry, woodland to forestry, and have lumped together the others as, simply, ‘other’.¹⁴ The share of each of these four categories of land in the total area of a given property then makes up its ‘land-use proﬁle’, as deﬁned in Ch. 3. For the cluster analysis I selected all those properties that had details of the area of different land types, leaving out any property for which no details were available, any for which one or more ﬁgures in the area details were arbitrary,¹⁵ ¹⁴ In the Handbooks, the category ‘moor’ was a catch-all: it included not only moorland, but also all unproductive land and all untaxed land such as farmyards, pathways, ditches, farm roads, and land used for ‘public purposes’ such as sites for schools. ¹⁵ As e.g. when a total was reported for ploughland and meadow together, I allocated the ﬁgure 50–50 to the two land types. Eliminating arbitrary assignment of land also eliminated all properties that had been assigned to one year from a later or earlier year as a result of the adjustment of State forest land detailed in Ch. 2.

Classifying Land by its Use

223

Table 8.7. Comparison of databases used for cluster analysis with those containing all principal properties of 100 ha or more Properties

Number Areaa Tax valueb a b

c.1882 All over 100

Used in cluster

c.1895 All over 100

Used in cluster

c.1907 All over 100

Used in cluster

15,929 10,628 102,219

15,190 9,675 96,840

16,145 10,739 103,327

14,808 9,356 94,401

15,780 10,560 102,913

13,473 8,664 88,352

000 ha. 000 marks.

and any for which the sum of the details came to less than 95 per cent, or more than 105 per cent, of the total reported land area of the property. That selection resulted in samples of the sizes shown in Table 8.7. As the table shows, the sample became less inclusive over time. The principal sources of the disparities are the following: 1. The number of properties that had to be eliminated from the cluster analysis because only the total area was reported, and no details could be found elsewhere, rose from 69 c.1882 to 177 c.1895 to 744 c.1907. 2. The number of properties that had to be eliminated from the cluster analysis because only the total area was reported, but details had been found elsewhere, rose from 133 to 234 to 697 in the three respective years. 3. The number of properties that had to be eliminated from the cluster analysis because of arbitrary allocation of some land-type details (e.g. a reported combined area for plough and meadow split 50–50 between the two land types) rose from 142 to 343 to 390. These three reasons alone account for 1831 of the 2307 properties that had to be eliminated from the analysis in 1907 because of data problems, whereas they accounted for only 754 of the 1337 that had to be left out in 1895. In other words, the increase in the number of properties eliminated for these three reasons (1077) was larger than the increase in the total number of properties eliminated between 1895 and 1907 (970). Properties that had to be eliminated for all other reasons therefore actually declined between 1895 and 1907. Because there was so little change in the total area encompassed in properties of 100 hectares or more between 1895 and 1907, numbers 1 and 2 above are consistent with either (a) fairly widespread parcellization of large estates, with poor reporting of the data details for those new parcels over 100 hectares by their new owners, or (b) the inclusion of previously-missed properties into the Handbooks —often with poor reporting of details—combined with the loss of large estates through parcellization by the Settlement Commission and other

224

Classifying Land by its Use

agencies. Without detailed case-by-case examination, to what extent each of these explanations can account for the discrepancies cannot be determined with any degree of precision. The increase in the share of East Prussia, where the land market was far more active than in any other province, in the total number of properties in the dataset in 1907 (see Table 8.11), along with a concentration in the provinces of Silesia and Saxony of the number of properties that had to be eliminated for lack of detail, however, suggest that explanation (b) is the more likely of the two: (i) despite what happened in its land market, East Prussia’s share of eliminated properties did not increase, but (ii) in two provinces where no signiﬁcant parcellization activity, either government or private, was to be found, there was a concentration of properties that had to be eliminated for lack of detail.

8.3.1. Calculating Land-Use Proﬁles At this point it might be appropriate to review the notion and calculation of landuse proﬁles. The land types listed in our data source can be divided according to their principal economic use: ploughland and garden for arable agriculture, meadow and pasture for animal husbandry, and woodland for forestry. Water, unproductive land, and such land types as farmyard, roads, paths, and ditches that can serve all three (or none) of these economic functions have been assigned to a fourth category, ‘other land’. From these land categories I construct a land-use proﬁle, which is simply the share of each type of land in the total area of the property.¹⁶ Consider the following two properties, the large estate of Batewitz in Frantzburg district of Pomerania (3928 ha in 1884, owned by Magnus, Count von Klot-Trautvetter), and the rather smaller property (138 ha in 1892) of Ludwigsdorf in Neisse district, Silesia, owned by Maria Seydel. Their makeup, according to the Handbooks, was as follows: Batewitz: 1799 ha ploughland, 210 ha meadow, 190 ha pasture, 125 ha woodland, 1593 ha ‘moor’, and 11 ha water; total area 3928 hectares. Ludwigsdorf: 129 ha ploughland, 7 ha meadow, no pasture, no woodland, 2 ha ‘moor’, and no water; total area 138 hectares. Batewitz therefore had 1799 hectares devoted to arable agriculture, 400 to animal husbandry, 125 to forestry, and 1604 as ‘other land’. Ludwigsdorf had 129 hectares for arable agriculture, 7 for animal husbandry, none for forestry, ¹⁶ The proﬁle could have been measured as simply the actual area of the four land types, but the data for the very largest properties would have so seriously dominated the outcome of the statistical analysis that they would have obscured, rather than clariﬁed, the classiﬁcation of properties according to their land-use proﬁles. Hence the normalization rule chosen, to use the share of total rather than the actual numbers.

Classifying Land by its Use

225

and 2 ‘other’. Calculating these areas as a share of the total area gives the land use proﬁles: Batewitz: Arable: 0.458 Husbandry: 0.102 Forestry: 0.032 Other: 0.408 Ludwigsdorf: Arable: 0.935 Husbandry: 0.051 Forestry: 0.000 Other: 0.015 Note also that the overall size of the property will have no inﬂuence on which group of properties it will be assigned to, since land-use proﬁles are simply percentages.

8.3.2. Cluster Analysis and the Number of Properties For the grouping here presented, cluster analysis was used to divide the properties into 4, then 5, 6, 7, and 8 clusters. In the present case, six clusters seemed to me to be the optimum. Not only did this number show six quite clearly different clusters, it also included one very small cluster (under 200 properties in each of the three years, between 1.2 and 1.4% of the total number of properties used in the given year). Since cluster analysis is sensitive to outliers, one can at least hope that the outliers tended to lie in a particular direction, and thus the existence of this small cluster would indicate that the other clusters were quite robust. Expanding the clusters to seven or more produced additional clusters which I judged to be but minor variations on one of the six existing clusters. Grouping properties by their land-use proﬁles into six clusters produced two large clusters, one medium cluster, two small clusters, and one tiny cluster. The proﬁles of these clusters did not discernibly change from the early 1880s to c.1907; in other words, no restructuring of land use took place on large properties.¹⁷ Table 8.8 shows the mean values and the standard deviations for the land types that determine each cluster’s designation. The ‘mostly arable’ cluster is the easiest to see, where the average property has arable land and little else. If there is any trend to observe over the time period, it is a very slight intensiﬁcation of land use in the properties in this cluster. Note, too, that the distribution of properties in this cluster is ‘tight’, with a standard deviation of only 0.06 compared to mean values for the share of arable in total area right around 88 per cent in all years. The second-biggest cluster features the most mixed land use of all: arable accounts for nearly 70 per cent of the area of the average property, meadow and pasture for around 17 per cent, and woodland another 10 or 11 per cent. Within the cluster, the land use proﬁles are relatively more varied—the share of arable land varies much less than do the shares of the other two land types, however—than in the ‘mostly arable’ cluster, as a comparison of the centre ¹⁷ This conclusion ﬁts perfectly into Hess’s ﬁndings that there was no evidence for a crisis in agriculture in the late 19th century (see Ch. 1, Sect. 1.1.3 for details).

226

Classifying Land by its Use

Table 8.8. Land-use proﬁles of the six clusters: cluster centres (Ctr.) and their standard deviations (Std.) Clusters from largest to smallest

Year

Land types in the land-use proﬁles Arable Animal Forestry husbandry

Other

Ctr.

Std.

Ctr.

Std.

Ctr.

Std.

Ctr.

Std.

Mostly arable

1882 1895 1907

0.878 0.880 0.888

0.060 0.060 0.060

0.080 0.078 0.073

0.050 0.050 0.047

0.021 0.021 0.020

0.035 0.034 0.033

0.021 0.020 0.019

0.022 0.023 0.022

Arable, husbandry, & some forestry

1882 1895 1907

0.684 0.685 0.701

0.064 0.065 0.064

0.176 0.175 0.168

0.078 0.078 0.076

0.109 0.108 0.098

0.083 0.084 0.080

0.031 0.032 0.032

0.036 0.036 0.038

Arable & forestry

1882 1895 1907

0.454 0.455 0.471

0.098 0.098 0.097

0.122 0.119 0.111

0.071 0.071 0.066

0.392 0.395 0.384

0.097 0.097 0.097

0.031 0.031 0.033

0.038 0.036 0.040

Mostly forestry

1882 1895 1907

0.117 0.111 0.115

0.104 0.101 0.103

0.053 0.052 0.054

0.060 0.060 0.062

0.798 0.804 0.795

0.137 0.136 0.141

0.032 0.033 0.036

0.051 0.055 0.063

Arable & husbandry

1882 1895 1907

0.426 0.437 0.427

0.136 0.133 0.144

0.458 0.448 0.457

0.139 0.133 0.148

0.085 0.083 0.083

0.091 0.091 0.092

0.031 0.032 0.033

0.040 0.041 0.045

Arable & other land

1882 1895 1907

0.386 0.388 0.383

0.148 0.150 0.162

0.133 0.132 0.119

0.078 0.079 0.079

0.082 0.094 0.101

0.097 0.112 0.116

0.398 0.385 0.397

0.160 0.163 0.175

means and standard deviations shows. As we will see later, this mixed-use cluster tends to be regionally concentrated. The third cluster contains signiﬁcantly fewer properties than the ﬁrst two (see Table 8.9), and again shows a nearly unvarying land use, with 85 per cent of the area of the average property devoted to arable agriculture and forestry in not-quite-equal shares that hardly deviate from one another across the three years. This cluster, too, exhibits a ‘tight’ distribution of properties that retains its level of tightness from beginning to end of the period under review. The stable nature of these clusters can be most easily seen with the aid of Figs. 8.3 and 8.4. The total number of properties in the remaining three clusters does not even reach the total in the third-biggest cluster alone; these last clusters together account for only about 15 per cent of all the properties in the sample. Somewhat surprisingly, the cluster containing the properties that, on average, devote about 88 per cent of their total land, in roughly equal shares, to arable agriculture and animal husbandry is one of the small clusters. Properties specializing heavily in forestry contain a few more properties (but naturally a lot more area), than does the arable–animal husbandry cluster. A tiny number of properties falls

Classifying Land by its Use

227

Table 8.9. Number of properties in six clusters in the three sample years Cluster type

Mostly arable Arable, husbandry, & some forestry Arable & forestry Mostly forestry Arable & husbandry Arable & other land total

c.1882 Number

%

c.1895 Number

%

c.1907 Number

%

5,256 4,935

34.6 32.5

5,099 4,806

34.4 32.5

4,901 4,372

36.4 32.6

2,703 1,090 1,016 190

17.8 7.2 6.7 1.2

2,562 1,141 1,006 194

17.3 7.7 6.8 1.3

2,196 1,077 740 187

16.3 8.0 5.5 1.4

15,190

100.0

14,808

100.0

13,473

100.0

Fig. 8.3. Composition of clusters by land type c.1882, c.1895, c.1907: mostly arable, mostly forest, and arable + forest clusters.

into the cluster characterized by about equal shares of arable and ‘other land’ (unproductive land, moor, water, farmyards, roads and pathways, etc.) that together account for just under 80 per cent of the total area of the average property in this cluster in all years. Table 8.9 shows the number of properties in each cluster and the share of that cluster in the total number of properties in the sample. Here we see that the relative sizes of the clusters remained nearly constant, that is, the distribution of properties among the clusters hardly changed between c.1882 and c.1907. Despite the generally falling grain prices and deteriorating terms of trade of agriculture from the 1880s to the mid-1890s, and the increasing grain prices and improving terms of trade thereafter, no discernible change in the overall pattern of land use on larger properties occurred. Just as Hess found from the agricultural censuses that the overall size structure of large farm enterprises was

228

Classifying Land by its Use

Fig. 8.4. Composition of clusters by land type c.1882, c.1895, c.1907: arable + animal husbandry, arable + husbandry and forestry, arable + other clusters.

‘nearly rigid’,¹⁸ here we see the same result from the Handbook data for the overall land-use patterns of large rural properties. There was, however, some slight change in the relative size of the clusters that appears to show a very slight trend toward a bit more specialization. The mostly arable and mostly forestry clusters became relatively larger, at the expense of the arable + forestry and the arable + animal husbandry clusters, while the second-largest cluster (arable with some animal husbandry and forestry) and the very smallest cluster (about equal shares of arable and other land—the estate of Batewitz that we examined earlier surely belonged to this cluster) did not exhibit any change at all in the numbers of properties they contained, relative to the other clusters. All this, however, may be simply a size phenomenon, given that the 100–200 hectare size category had relatively more properties in it in 1907 than in other years. In every year the two largest clusters contained two-thirds or more of the total number of properties, while the very smallest one never reach even 1.5 per cent. It is not surprising that the grouping of properties produced two clusters of properties specializing in arable agriculture and forestry, respectively. It is interesting, to say the least, that no cluster of properties specialized in animal husbandry appeared, even when the number of clusters was expanded to eight. This was because, around 1882 for example, there were only 22 properties in the whole dataset comprising 90 per cent meadow and pasture land, only 34 if the threshold is dropped to 80 per cent, and still only 58 with 70 per cent or more. Compared to 15,190, the number 58 is trivial to the point of unnoticeable. Moreover, these were hardly the biggest properties in the sample: the average size of the 34 properties with 80 per cent or more of their land in meadow and ¹⁸ Hess, Grossgrundbesitzer, 45.

Classifying Land by its Use

229

pasture was only 346 hectares; only one of these was over 1000 hectares. So, while there was considerable animal husbandry practised on large properties in East Elbia, it was almost always practised in connection with other agricultural activities, almost never as an exclusive speciality.

8.3.3. Cluster Analysis and the Area and Tax Value of Properties So far we have looked only at the number of properties in each cluster. If we looked at other measures of size, namely the total area or the total taxable net yield of these properties, a rather different—and somewhat more nuanced—picture could emerge. For example, the mostly arable cluster would no doubt turn out to contain not just the largest number of properties but the largest total tax value as well. The largest cluster in terms of total area, however, could well turn out to be the second cluster, if its properties were on average larger than those in the ﬁrst cluster, which is almost certainly the case. But in fact a different cluster contains the largest area. Let us turn now to these comparisons, beginning with Table 8.10. As expected, the cluster of properties specializing in arable agriculture ranks number one in terms of the total tax value of its properties, even though it always comes in a rather distant fourth in total area. And even though the mostly forestry cluster was in fourth place in terms of number of properties, it is the largest cluster in terms of total area comprised.¹⁹ In terms of land value, as measured by its assessed yield for the land tax, land in the mostly arable cluster was nearly twice as valuable as the average, at around 19 marks per hectare vs. around 10 marks per hectare. The ‘average’ cluster, on the Table 8.10. Total area and total tax valuea of properties in six clusters in the three sample years Cluster type

c.1882 Total area

Total tax

c.1895 Total area

Total tax

c.1907 Total area

Total tax

Mostly arable Arable, husbandry & some forestry Arable & forestry Mostly forestry Arable & husbandry Arable & other land

1,743 2,432

32,728 28,291

1,680 2,341

31,934 27,500

1,604 2,084

31,107 25,554

2,368 2,474 490 169

18,609 11,547 4,730 937

2,197 2,482 495 161

17,468 11,587 5,011 900

1,934 2,541 370 131

15,818 11,115 4,016 743

total

9,675

96,841

9,356

94,401

8,664

88,352

a

Total area in 000 ha, total tax value in 000 marks.

¹⁹ Interestingly enough, the value rank of the clusters is the same as their rank for number of properties.

230

Classifying Land by its Use

other hand, was the second smallest one in all measures, the ‘arable and animal husbandry’ cluster, with an tax assessment per hectare just a few pfennigs above, or a few below, the overall average. The ‘mostly forest’ cluster, as expected, had the lowest average per hectare, in all three years somewhat under half the average for all properties in the sample.

8.3.4. Regional Concentration of Properties by Land-Use Proﬁles Did the various clusters tend to concentrate themselves regionally? Table 8.11 helps to ﬁnd the answer to that question. Properties specializing in arable agriculture exclusively or almost exclusively tended to be concentrated in the east and south: over half of them were to be found in the three provinces of Posen, Silesia, and Saxony, which among them accounted for less than half of the total number of properties in each year. Specialized forest properties were particularly the business of Silesia and Brandenburg, which between them accounted for about half of all such properties in the seven provinces. Saxony was also heavily Table 8.11. Distribution of cluster member properties by province (%) Cluster

Mostly arable Arable, husbandry & forestry Arable & forestry Mostly forestry Arable & husbandry Arable & other land total

Year

Province EP WP

BBG

POM

POS

SL

SA

Total

1882 1895 1907 1882 1895 1907 1882 1895 1907 1882 1895 1907 1882 1895 1907 1882 1895 1907

7.4 6.9 15.5 22.5 21.8 25.0 9.6 9.1 6.7 8.4 7.7 8.0 28.7 30.2 23.9 17.8 15.4 21.3

13.1 13.9 13.7 10.4 10.5 9.9 7.0 6.8 6.6 5.1 6.2 9.5 9.1 9.1 9.1 13.5 9.2 8.0

9.5 8.8 8.5 9.5 9.5 10.2 19.6 18.1 20.3 22.2 19.5 22.0 14.6 13.8 18.3 20.5 19.0 22.9

15.7 16.3 14.1 18.5 19.3 18.9 14.2 15.5 18.4 7.6 9.2 12.7 19.8 20.3 19.7 17.8 20.5 19.1

20.2 19.4 15.4 14.4 14.0 12.6 13.7 13.7 13.2 7.5 7.8 7.0 8.0 8.5 9.3 17.8 20.0 16.5

20.7 21.0 17.9 18.5 19.3 17.1 27.8 29.2 26.6 33.9 35.8 26.7 7.1 7.8 5.5 8.6 11.3 6.9

13.5 13.6 14.9 6.2 5.7 6.4 8.1 7.5 8.2 15.3 13.8 14.1 12.7 10.2 14.2 3.8 4.6 5.3

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

1882 1895 1907

14.3 13.9 17.1

10.3 10.6 10.6

12.7 11.9 12.8

16.1 16.9 16.6

15.4 15.0 13.1

21.1 22.0 18.9

10.1 9.7 10.8

100.0 100.0 100.0

Note: EP = East Prussia; WP = West Prussia; BBG = Brandenburg; POM = Pomerania; SL = Silesia; SA = Saxony.

Classifying Land by its Use

231

Table 8.12. Properties of 100 hectares or more in East Prussia:a numbers and total area in 000 ha Property size 1884 in ha Number 100–199 200–499 500–999 1000+ total

Area

1895 Number

Area

Area

530 956 461 233

78.1 307.7 313.2 608.3

520 911 416 205

77.0 292.4 282.0 550.5

638 1,100 386 201

101.1 349.3 260.5 494.6

2,180

1,307.3

2,052

1,201.8

2,325

1,205.4

Ratio of 1907 value to current-year value 100–199 1.20 1.29 1.23 200–499 1.15 1.14 1.21 500–999 0.84 0.83 0.93 1000+ 0.86 0.81 0.98

1.31 1.19 0.92 0.90

total

1.00

a

1907 Number

1.07

0.92

1.13

For the purposes of this analysis I used the unadjusted data for East Prussia.

into forestry: despite accounting for only about 10 per cent of all properties in the sample, Saxony had around 14 or 15 per cent of all the specialized forest properties. More mixed forms of land use tended to concentrate themselves also, but with a pattern different from the other clusters. The cluster of arable with animal husbandry and some forestry was heavily represented in East Prussia and Pomerania, and also in Silesia, while the cluster with properties having about equal shares of arable and animal husbandry tended to be found in East Prussia, Brandenburg, Pomerania, and also in Posen. The mixed arable–forestry properties concentrated more in Pomerania, Brandenburg, Posen, and Silesia. While these patterns were quite stable across provinces and over time, one apparent anomaly stands out: East Prussia’s share of the number of properties in the biggest cluster of mostly arable makes a big jump between 1895 and 1907, as does its share of the total number of properties in the sample. Closer examination of the data reveals a particular pattern to this jump, which in turn warns us not to put too much reliance on analysis based on the number of properties—much better to use either total area or tax value instead. Table 8.12 explains this point. In 1907, a larger number of properties contained fewer hectares (compared to 1884) or about the same number of hectares (compared to 1895). There are at least two alternative explanations, each of them about equally likely, and each of them consistent with the combination of increase in numbers and area of properties under 500 hectares with the decrease for properties over 500: 1. The Handbook for East Prussia may have changed the deﬁnition of property it used in 1907, so that some large manors from previous years were now

232

Classifying Land by its Use

listed according to their constituent parts. This was true at least to the extent that the 1907 Handbook sought to apportion properties with land in two or more districts into the amounts of land in each district, even if they reported them in the district in which the ‘seat’ of the property was located. Properties lying in only one district, but part of a manor seated in another district, may also have been counted separately in some cases, but I have been unable to determine if this is true. 2. Large properties were subdivided and sold off in smaller pieces. This is consistent with Hess’s ﬁnding that the land market was by far the most active in East Prussia, way ahead of any other province.²⁰

8.3.5. Was there Specialization in Ownership of Large Properties? The speciﬁc provisions of the emancipation of the peasants in Hungary in 1849 led to a particular pattern of specialization in landownership: (a) because the peasants’ share of forest and pasture had been allotted to them in communal ownership, while other types of land were given them in individual ownership, properties that were mostly pasture and forest were mostly owned by communities, whereas (b) the nobles’ share of forest and pasture went to them individually, so their properties tended to be more mixed, including important shares of arable land and meadow. The Prussian peasant emancipation dragged on for half a century, and did not include provision for communal land, so the properties created by it were mixed properties in individual ownership.²¹ But, since almost three decades had passed by the time of the publication of the 1880s editions of the Handbooks, and the Prussian land market was very active, there was perhaps time for a certain specialization in landownership to develop there as well. Let us examine whether such a phenomenon shows through in the various clusters. In terms of number of properties, it is evident from Table 8.13 that non-noble (bourgeois) owners dominated the mostly arable cluster, and, to a lesser extent, the mixed activity cluster as well, by concentrating their ownership in these two clusters (76% of all their properties c.1882 and 79% c.1907). Because properties can be of such different overall sizes, however, we should probably not push this table much further, but rather look at the distribution of land by area (Table 8.14) and by tax value (Table 8.15) in these same clusters. In terms of total area, bourgeois landowners still led the ﬁrst two clusters, but their position was clearly less dominant. We also see them getting out of the mixed land use cluster after the mid-1890s. The high nobility were concentrated in the ²⁰ ‘In the province of East Prussia almost 6000 large landholdings changed hands between 1896 and 1914, twice as many as in any other East Elbian province, and more than in the entire Prussian West taken together’. Hess, Grossgrundbesitzer, 287. My translation. ²¹ See Eddie, ‘Junkers and Magnates’, for a comparison.

Classifying Land by its Use

233

Table 8.13. Number of properties in cluster by owner group Date/cluster

c.1882 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total c.1895 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total c.1907 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total

Higher nobility

Lesser nobility

Bourgeois

Prussian State

Other

Total

454 530

1,228 1,506

3,086 2,580

266 217

222 102

5,256 4,935

549 272 87 29

1,074 269 283 56

1,010 152 515 85

7 297 84 18

63 100 47 2

2,703 1,090 1,016 190

1,921

4,416

7,428

889

536

15,190

448 507

1,104 1,394

3,013 2,535

288 244

246 126

5,099 4,806

502 290 103 35

996 264 252 45

972 154 510 90

23 333 93 20

69 100 48 4

2,562 1,141 1,006 194

1,885

4,055

7,274

1,001

593

14,808

389 402

947 1,189

2,963 2,323

336 312

266 146

4,901 4,372

399 209 65 21

895 252 191 33

774 130 363 103

39 370 70 24

89 116 51 6

2,196 1,077 740 187

1,485

3,507

6,656

1,151

674

13,473

Prussian State

Other

Total

Table 8.14. Total area of properties in cluster by owner group (000 ha) Date/cluster

c.1882 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total

Higher nobility

Lesser nobility

Bourgeois

178 332

482 861

895 1,077

121 120

67 43

1,743 2,432

729 690 64 49

937 289 179 42

661 138 184 52

4 1,234 46 24

37 122 17 1

2,368 2,474 490 169

2,042

2,790

3,007

1,550

286

9,675

234

Classifying Land by its Use

Table 8.14. (Continued) Date/cluster

c.1895 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total c.1907 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total

Higher nobility

Lesser nobility

166 317

437 803

627 665 70 49

Bourgeois

Prussian State

Other

Total

872 1,040

130 128

75 52

1,680 2,341

884 296 164 35

607 143 188 51

32 1,247 53 25

46 130 19 2

2,197 2,482 495 161

1,895

2,620

2,901

1,616

324

9,356

166 270

392 705

813 882

151 170

82 58

1,604 2,084

587 650 49 26

796 296 127 24

470 89 124 52

29 1,368 49 24

53 139 20 5

1,934 2,541 370 131

1,747

2,339

2,429

1,790

359

8,664

Table 8.15. Tax value of properties in cluster by owner group (000 marks) Date/cluster

c.1882 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total c.1895 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total

Higher nobility

Lesser nobility

Bourgeois

Prussian State

Other

Total

3,490 4,576

8,980 10,186

15,390 10,824

3,000 2,034

1,868 670

32,728 28,291

6,405 3,394 700 216

7,323 1,459 1,623 253

4,468 600 1,555 290

20 5,400 658 154

392 694 194 24

18,609 11,547 4,730 937

18,781

29,825

33,127

11,266

3,842

96,841

3,254 4,338

8,475 9,602

15,268 10,627

2,897 2,139

2,039 794

31,934 27,500

5,685 3,240 975 232

6,915 1,427 1,550 194

4,266 644 1,557 303

164 5,489 706 141

437 787 223 29

17,468 11,587 5,011 900

17,725

28,164

32,666

11,537

4,309

94,401

Classifying Land by its Use

235

Table 8.15. (Continued) c.1907 Mostly arable Mostly arable + husbandry/forest Arable & forest Mostly forestry Arable & husbandry Arable & other total

3,271 3,918

7,785 8,765

14,937 9,514

3,043 2,487

2,070 869

31,107 25,554

5,481 3,482 566 157

6,434 1,432 1,262 146

3,231 390 1,305 295

157 5,029 649 104

514 782 234 42

15,818 11,115 4,016 743

16,876

25,825

29,673

11,469

4,510

88,352

mostly forestry and the arable + forest clusters, while the lower nobility concentrated its holdings in the mixed land use and arable + forest clusters. The Prussian State was heavily concentrated in the cluster specialized in forestry. The pattern observed by area largely repeats itself when we look at the land tax assessments of the properties enumerated here, except that the high nobility’s ownership by tax value concentrates in the mixed land use cluster and the arable + forest cluster, while for the lower nobility second place in the value sweepstakes now moves to the mostly arable cluster. The Prussian State, because of its valuable domains and model farms, had about as much land in tax value terms in the ﬁrst two clusters as it did in the forestry cluster.²² ²² Too much should not be made of the differences between years with regard to State properties; since my rule of not including properties with arbitrary values for land types automatically excludes all State forest properties added in the adjustments detailed in Ch. 2.

9 Concluding Remarks 9 . 1 . T H E B O O K S TA RT E D W I T H A S E R M O N Chapter 1 began in a very conventional way: it tried to set the historical stage, put the study in context, and review the relevant literature. This chapter closed, however, with a stern warning and admonition: the last section of Chapter 1 showed, in perhaps excruciating detail, why the data from the German farm censuses of 1882, 1895, and 1907 could not, and cannot, ever be used as property statistics or statistics of ownership, not even in the limited sense of being able to serve as a proxy variable for property or ownership. Contemporary observers, including August Meitzen and Johannes Conrad, and very explicitly the authors of the census volumes themselves, recognized that fact and stated it clearly. Some modern scholars, most notably Klaus Hess and René Schiller, have pointed this out again in no uncertain terms. Yet, because some of the data of the censuses are published in handy, ready-made tables organized by size of farm, temptation and confusion still exist in the scholarly literature. This prompted my demonstration that trying to use these tables as indicators of property size or ownership relations is—for both conceptual and empirical reasons—simply wrong, and seriously misleading into the bargain. Chapter 2 then went on from there to set out, again in perhaps excruciating detail, what available data can be used to describe property and ownership relations, and what steps I have taken to make the data used as complete, accurate, and internally consistent as possible. Here ended the sermon.

9 . 2 . W H AT H A S T H I S S T U DY S H OW N ? With regard to questions raised in the Preface, this study has shown that large properties, conventionally deﬁned as those with 100 or more hectares total area, covered just about exactly half of the area outside of cities in the seven ‘core provinces’ of eastern Prussia, and accounted for about 45 per cent of the total land tax value; indeed, they accounted for a ﬁfth of the land area in all of Germany. Roughly 11,000 owners held these landed estates. Physical persons owned just

Concluding Remarks

237

under four-ﬁfths of these lands, and the Prussian State around a ﬁfth, leaving only 3 or 4 per cent in the hands of other non-physical legal persons, depending on the year being examined. Over the period examined most intensively, 1882–1907, the Prussian State and other legal persons gained land, and natural persons lost some, but those losses were sustained by the lesser nobility and bourgeoisie, not by royalty and the higher nobility, who, as a group, maintained their holdings of large estates. One of the more interesting ﬁndings of these calculations is that, despite the rapid industrialization of Germany in the last half of the nineteenth century, there is no evidence of a net inﬂow of funds from the industrial or commercial nouveaux riches into landholding after the 1880s—indeed, a net outﬂow after the mid-1890s—despite the appeal of potential capital gains from the steadily rising price of land.¹ The largest single owner of land was the Prussian State, with at least 2–2.5 million hectares, depending on the year. It was so far and away the biggest owner that, for most calculations, it had to be considered separately from all other owners in order to avoid obscuring their positions. Physical persons, on the other hand, owned just over 8 million hectares c.1882, but only somewhat over 7.5 million c.1907. Within the category of physical persons, the aristocracy—deﬁned as royalty + nobility—owned about 5 million of those hectares c.1882 and about 4.7 million c.1907. The average area owned by an aristocrat was more than three times that of a non-noble person, rising to nearly four times by 1907. Although this difference might seem large, it turned out that differences within the aristocracy were at least as great: the data showed an average total holding of around 3500 hectares per owner from the high nobility vs. under 1000 for the lesser nobility (and about 410 ± 30, depending on the year, for non-noble owners). After the Prussian State, the largest single landowner, in terms of both area and tax value of his estates, was the Prussian king himself, with roughly 100,000 hectares² and 600,000 to 800,000 marks of land tax net yield, depending on the year. Although the king had extensive landholdings in all provinces except East Prussia, the rest of the really big landowners were heavily concentrated in the province of Silesia: nearly half of the land and tax value of the top ﬁfty owners were found in that province. There the Prince of Pless, with between 50,000 and 60,000 hectares, was second only to the king among physical persons in terms of total area owned and third, behind the Duke of Brunswick, in terms of the total tax value of his estates. Comparing the top ﬁfty landowners after the Prussian State, there turned out to be only about 60 per cent overlap between ¹ One contributor to this price rise was no doubt the series of reductions in the land tax won by landowners. See Mark Hallerberg, ‘The Political Economy of Taxation in Prussia 1871–1914’, Jahrbuch für Wirtschaftsgeschichte, 2 (2002), 11–34. ² While an imposing sum, it begins to look like small potatoes when compared to the half-million hectares owned by Prince Esterházy in Hungary at about the same time.

238

Concluding Remarks

the top ﬁfty by total area and the top ﬁfty by tax value. These top ﬁfty (so less than 0.5% per cent of the owners) owned some 10 per cent of all the land in large properties in the seven provinces, and over 8 per cent of the total tax value. Including the Prussian State would add only one owner, leave the top ﬁfty-one still under 0.5 per cent of all owners, but raise their share of the land to about 30 per cent. Again without the Prussian State, the top ﬁfty by value included fewer of the high nobility and more owners from religious bodies, cities, and business ﬁrms—classes of owners almost completely absent from the top ﬁfty by area. The super-elite, the top ten landowners, formed an almost exclusive club of the high nobility, with only the city of Görlitz included in the ranks of the top ten by area. Nine of the top ten by tax value also belonged to the high nobility, except for one interloper in each year—a business ﬁrm in 1882 and the University of Greifswald in the second two years. The bourgeoisie had clearly made large inroads into the ranks of substantial landowners by the time the ﬁrst Handbooks become available. According to Prussian law, before 1807 only the nobility could own knight’s estates, and only the king himself could grant exception to this rule. But by c.1882 about 7000 non-noble owners possessed over 3 million hectares in large properties, with a tax value of nearly 35 million marks. With the help of some older sources, I showed that most of this incursion of the bourgeoisie into ownership of knight’s estates—and by extension of large properties in general—had to have occurred before the mid-1850s, and at least in the provinces of East and West Prussia, before 1834. The incursion continued, at a slower pace, up to perhaps the mid-1880s, then it became clear that the bourgeoisie, as a group, began divesting themselves of landed estates, a movement that became very clear after the mid-1890s. Examination of the data of the land tax cadastre of the early 1860s shows a quality gradient in land running roughly and broadly, and without well-deﬁned borders, from north-east to south-west across the seven provinces, with the lowest-yielding land found in the eastern part of East Prussia, and the highest in the province of Saxony. The range of average tax assessment per hectare for different types of land was surprisingly great, and there were districts in Saxony where the tax value of forest land was greater than the tax value of ploughland in other provinces. Indeed, the average tax assessment per hectare of forest land in Wanzleben district of Saxony was higher than the average assessment for ploughland and meadow in 197 of the other 258 districts in the seven provinces. Because of differences in both makeup and land quality, the average large property in Wanzleben (433 ha, mostly ploughland) had seven times the tax assessment of the average large property in Johannisburg district of East Prussia, despite the latter’s average size being 1622 hectares (mostly forest). Regarding the land endowment of the seven provinces, the land tax cadastre shows that areas best suited for specialization in arable agriculture tended to

Concluding Remarks

239

be found in the provinces of Posen, Silesia, and Saxony, while those best suited for specialization in forestry were found mostly in Brandenburg and Silesia. The pattern of land endowment in the Baltic provinces (East and West Prussia, Pomerania) predisposed their large properties to mixed land use—arable agriculture, animal husbandry, and forestry together—and not to specialization in any one activity. The concept of a land-use proﬁle, deﬁned as the shares of total land of a property devoted to each of four economic functions (arable agriculture, animal husbandry, forestry, or ‘other’), allows the classiﬁcation of large properties according to the degree of specialization, or lack thereof, in one or more of these activities. That in turn led to classiﬁcation of properties into six different clusters by land use: mostly arable, mostly forestry, arable + animal husbandry, arable + forestry, very mixed land use—arable with some animal husbandry and forestry—and a very small cluster of properties having arable + a large share of ‘other’ land. These clusters remained very stable over the period 1882–1907, both in relative size and in average makeup or land-use proﬁle. Interestingly enough, the analysis did not reveal any cluster, even a small one, of large properties mostly devoted to animal husbandry,³ even when the number of clusters was expanded to seven and then eight. The two largest clusters, those of properties specialized mostly in arable farming and properties with very mixed land use, accounted for two-thirds or more of all the properties analysed in each year, while the smallest of the six—arable + other—did not even account for 1.5 per cent of the properties. When examined by total area of the properties in the cluster, the picture changed: the forestry and very mixed land use clusters accounted for about half of the area in these large properties, and the arable + forestry cluster for another 20–25 per cent, depending on the year. When tax value was considered, the picture changed back again: the mostly arable and the very mixed land use clusters contained about 60 per cent of the total tax value of all properties analysed. Striking a balance, the six clusters divided into four major clusters (mostly arable, mostly forestry, arable + forestry, and mixed land use) and two minor clusters (arable + animal husbandry and arable + other), and this division also stayed remarkably stable over time. Unsurprisingly, the regional concentration of properties in the four major clusters followed the regional concentration of land endowment found earlier: properties specializing in arable agriculture tended to concentrate in the east and south, specialized forest properties in Brandenburg and Silesia. Bourgeois owners tended to concentrate their ownership on properties specialized in arable agriculture or in very mixed land use (about two-thirds of the total area they ³ This was similar to the case in Hungary at about the same time, where animal husbandry was primarily the domain of smallholders. See my ‘Agricultural Production and Output per Worker in Hungary, 1870–1913’, Journal of Economic History, 27/2 ( June 1968), esp. table 8, 218.

240

Concluding Remarks

owned were found in these two clusters), both the higher and the lesser nobility tended to concentrate their holdings in the forestry and arable + forestry clusters (on the order of 70% of their total area owned) and the Prussian State—very unsurprisingly—in the forest sector, where three-quarters or more of its acreage was to be found. State properties in the mostly arable plus the mixed land use clusters had about the same amount of tax value as those in the forestry cluster, however, because of the very high land quality of State domains and model farms. Chapter 7 opened a new perspective onto the stereotypical picture of the ‘agrarian East’. Although most of the enterprises were small in scale, there turned out to be a remarkably dense network of industrial establishments located on large properties in the seven provinces of our concern, overall about 0.43 per square kilometre, and 0.56 in the province of Silesia c.1882. Comparing the data from the Handbooks to those of the industrial censuses of 1882, 1895, and 1907, and also to tax data for brewing, distilling, and sugar milling, showed the relative importance of rural-based enterprises in the sectors as a whole. Starting with over three-quarters of all distilleries, almost two-thirds of all sugar mills, but with only about one-eighth of all breweries, the large rural estates held their position in distilling, but saw their positions in brewing and sugar milling erode over time to levels of 9 per cent and 42 per cent, respectively, as the urban-centred industrialization of Germany continued apace. Using data from six industries from the industrial censuses, I showed that the number of brickworks on large rural properties was around 40 per cent of the total of all brickworks in the seven provinces, of sawmills about 11 per cent, and of ﬂour mills around 8 or 9 per cent. The data from the industrial censuses showed broadly similar results to the tax data for brewing, distilling, and sugar milling. As a caution, it must be pointed out that the measure of number of industrial establishments overstates the share of rural-based establishments in the total, if the ‘true’ share be measured by employment or output, for example. The number of industrial establishments per property tended to rise with size of property, but the number per thousand hectares of land fell as property size rose. A dip below trend in this relationship for the 5000–9999 hectare range of properties was the result of State and city forest lands, with very little industry on them, being concentrated in this size category. As the numbers of industrial establishments on rural estates declined over time, the decline was spread out very evenly regionally, but very unevenly by size category of landed property: the decline was concentrated particularly among properties under 1000 hectares. Among owners of large estates, aristocrats and non-noble persons each owned properties that were home to almost exactly equal numbers of industrial establishments c.1882. Physical persons, whether noble or not, owned 90 per cent or more of all properties on which industrial establishments were found; these persons tended to concentrate ownership on properties that were home to brickworks or grain mills, and to a lesser extent,

Concluding Remarks

241

distilleries. These three types of establishment made up almost 80 per cent of the total number of industrial establishments found on the properties of physical persons. The Prussian State, on the other hand, tended to concentrate on distilleries. The most diverse set of industries was to be found on properties owned by banks and business ﬁrms, where sugar factories and brickworks tied for ﬁrst place, but then all sorts of other plants were also represented on the properties of this group. Just as the decline in numbers of industrial plants was concentrated in a certain size range of properties, so was it concentrated among bourgeois landowners, whose share fell from 47 per cent to 42 per cent. Aristocrats held their own in a relative sense, but both the State and ‘other owners’ gained, even in absolute numbers. About 1882 there were 20 owners, one being the Prussian State, who owned properties on which at least a dozen industrial establishments were to be found; this number expanded to 22 c.1895 and 23 c.1907. Thus there appeared to be some tendency toward concentration, even as the total number of rural-based industrial establishments declined. In the ﬁrst year, among the 19 besides the Prussian State, c.1882, there were two people from the lesser nobility and ﬁve non-noble persons, so twelve from royalty and high nobility. In the 21 c.1895 there were only 2 bourgeois owners, one from the lower nobility, one business ﬁrm, and one monastery. Thus the number from the high nobility increased to 17. Among the 22 c.1907 were to be found only one non-noble person, two from the lesser nobility, one monastery, a government agency (the Royal Settlement Commission), and a bank. This would have left 15 from the high nobility, except that Countess Johanna von Schaffgotsch had converted her properties to corporate ownership, so in her place there was now a business ﬁrm, Schaffgotsch Works LLC. Chapter 7 ﬁnishes with a consideration of what factors might have increased or decreased the probability of ﬁnding an industrial establishment on a landed estate. Regions had something to do with it, some raising, some lowering, and others showing about the same probability as the reference region of Danzig. The most inﬂuential factor turned out to be size of property (no surprise); the share of arable land in the property also raised the probability of ﬁnding an industrial establishment there, but not by much. Compared to those of royal and noble owners, properties of the State, religious bodies, and communities were less likely to be home to an industrial enterprise, while those of non-noble persons and especially business ﬁrms were more likely to have an industrial plant on them. While all these effects are interesting, and most of them predictable, most of the variation in whether or not a property had an industrial establishment remained unexplained. Still, it is signiﬁcant to have found that ‘the landlords of East Elbia were not simply export-oriented agriculturists; they had substantial industrial interests as well,’ to quote the concluding sentence of Chapter 7. This, I believe,

242

Concluding Remarks

casts a very different light on those landlords and brings a new perspective into the economic history of East Elbia.

9 . 3 . WA S T H E L A N D TA X B I A S E D TO FAVO U R B I G HOLDINGS? In Chapter 6 I alluded to the price premium paid for knight’s estates over similar properties without that status, and suggested that systematic undertaxation of knight’s estates might be a cause of that phenomenon. In this section I would like to explore the notion of possible undertaxation in more detail. Could the tax assessors, themselves often drawn from the ranks of substantial landholders, have been tempted to ‘give a break’ to others of their kind when making their land tax assessments? If so, did it happen often enough that it would take on the dimensions of a general trend, that is, were larger properties of one kind or another systematically under-assessed? Such systematic under-assessment might take several forms, some not easily distinguishable from each other. An assessment bias in favour of knight’s estates as a class of property, or of high-ranking nobles as a class of owners, or of larger properties in general—if these be considered singly—might each give similar results, since high-ranking nobles tended to own larger properties and a large fraction of the knight’s estates. In other words, these three characteristics are all positively correlated with each other. They therefore needed to be considered together. To test whether knight’s estates, or properties of the high nobility, or large estates in general, appeared to be under-assessed for their land tax obligations, I use a hedonic regression that, in effect, simulates the procedure of tax assessment: assessors determined the area and quality of each plot of each type of land (ploughland, meadow, pasture, forest, etc.), and multiplied the area times a tax coefﬁcient for the relevant quality grade taken from a table for the local district called the ‘cadastral key’, to establish the plot’s assessment.⁴ Adding up these individual assessments for each plot then produced the tax assessment for the entire property. Since my dataset contains the area of each type of land in a property but only its total overall tax assessment, I had to approximate the tax assessment procedure in the estimating equation in order to deduce the implied average coefﬁcients for each land type. Adding dummy variables for various different characteristics would then permit a test of whether there existed a systematic bias toward under-assessment of the land tax on knight’s estates. I will try the dummies in two forms—a simple form and an interactive ⁴ The key showed the assessment to be applied for each quality grade, in silver groschen per morgen. I have converted all the data taken from the land tax assessment into marks per hectare. One mark was the equivalent of 10 silver groschen, and one hectare, of 3.917 morgen.

Concluding Remarks

243

form—to test two alternative variants of the underassessment hypothesis, which are, respectively: (1) a kind of uniform break related simply to the overall size of the property, or (2) a proportional underassessment related to some or all of the component land types of the property. The regression equation would be as follows: Let AV = assessed value of the entire property in marks. Let PL be the area of ploughland and garden, MD the area of meadow, PA the area of pasture, WD the area of woodland, WA the area of water, MO the area of moorland, TA the total area (all in hectares), and DU a group of dummy variables, each equal to 1 if a given property possessed a certain characteristic and 0 otherwise. The simple variant of the regression equation would then be (1) AV = aPL + bMD + cPA + dWD + eWA + f MO + gDU and the simplest form of the variant with the interactive dummy would be (2) AV = aPL + bMD + cPA + dWD + eWA + f MO + g(DUxTA) The coefﬁcients a, b . . . f would be the average tax assessment in marks per hectare of properties in the given district or region; these coefﬁcients should all be positive. A negative value for g would measure either a lump-sum tax break in the simple variant or a per hectare tax break in the interactive variant. Considering the value of g, it is fair to assume that knight’s estates, for example, would probably contain no-worse-than-average quality land. Absent bias in the tax system, we would therefore expect the value of g to be at least zero. A negative and signiﬁcant g would be strong support for a claim of tax favouritism toward knight’s estates. To separate out the effects of land quality from those of putative assessment bias, I introduce another set of independent variables into the equation. These are land quality indexes, deﬁned as follows: LQI = (Average tax assessment per hectare in a given district)/(median tax assessment for all districts) This index was calcuated for each land type⁵ in each of the 259 districts,⁶ using the data of the land tax cadastre of the early 1860s. I used ⁵ The average was for all land in the taxation district, not just land in larger properties. ⁶ For those districts formed from parts of other districts after the land tax cadastre, I used the average data for the parent district, or an average of the parent districts, if there were more than one.

244

Concluding Remarks

the median as divisor instead of the mean in order to reduce the effect of a skewed distribution of assessments. This produced a series of quality-adjusted variables QualPL QualME QualPA QualWO QualWA QualMO each of which was the product of multiplying the appropriate LQI for each land type and district by the number of hectares of that type of land in each property located in that district and used in the regression equation. Since the LQI is simply the ratio of two numbers having the same dimension (marks per ha), it is a dimensionless number (basically equivalent to a percentage). Multiplying a dimensionless LQI times a land area in hectares means that QualPL, etc. will have a dimension of hectares, and their regression coefﬁcients will be in marks per hectare. Adding these variables to the equation then transformed the regression coefﬁcients of the actual hectares of ploughland, meadow, pasture, etc. in each of the properties into measures of quality difference in these land types for large properties as compared to all properties. This separating out of the effect of quality differences in land as between large properties and the general run of properties thus paves the way for introduction of the dummy variables to measure the effect of the characteristics that we are trying to isolate.⁷ Since farm land was not reassessed during the period under review, it makes no sense to run the regressions on any but the ﬁrst dataset from c.1882. I present here the results of three phases of the regression procedure, in order that the reader may judge more easily whether my conclusions about the results of these regressions are reasonable. The ﬁrst phase is simply to estimate equation (2) without any dummy variables. Table 9.1 presents the results of this procedure. All of the quality-adjusted variables have the correct positive sign, are highly signiﬁcant, and have relative values that are quite in order with the taxation procedure and cadastral keys, except for QualMO, the quality-adjusted variable for moorland. But because ‘moor’ in the Handbooks was a catch-all category that included a sometimes very high proportion of untaxed land, it is unsurprising to ﬁnd that its coefﬁcient was not signiﬁcantly different from zero. The constant term also does not meet the normal cut-off of a 5 per cent ⁷ Please note also that because the quality indexes are calculated district by district, their inclusion already takes location into account.

Concluding Remarks

245

Table 9.1. Regression estimation, tax value as dependent variable, simplest form Variable

Coefﬁcient

Std. error

T-value

Signif.

Plough Meadow Pasture Wood Water Moor

1.550 0.339 −6.961 0.073 −1.570 −0.767

0.133 0.61 0.645 0.055 0.243 1.456

11.644 0.552 −10.787 1.322 −6.471 −0.527

0.000 0.581 0.000 0.186 0.000 0.598

QualPL QualME QualPA QualWO QualWA QualMO

11.522 13.055 5.847 3.361 3.376 −0.059

0.085 0.529 0.589 0.046 0.221 0.835

135.481 24.672 9.923 73.095 15.311 −0.070

0.000 0.000 0.000 0.000 0.000 0.944

Constant

58.225

30.583

1.904

0.057

Note: Adj. RSQ = 0.900; N = 15, 190

signiﬁcance level, but this is also to be expected: if the speciﬁcation of the equation were perfect, and the data 100 per cent correct and accurate, the constant term should be exactly equal to zero. If it were not, a positive constant would imply a penalty in assessment for properties of 100 hectares or more, irrespective of size. (Similarly, a negative constant would imply an assessment bonus, irrespective of size, but that eventuality did not occur in any formulation of the estimation equation.) The coefﬁcients for the actual areas of ploughland, meadow, etc. show an interesting pattern: ploughland in large properties seems to have been of higher than average quality, but meadow does not. The coefﬁcients for pasture and water are negative and highly signiﬁcant. Since there is no reason to suspect that the quality of ponds and pastures was worse than average on large properties, there must have been some other reason for this outcome. Perhaps the reason might lie in characteristics of the property or its owner, so we now turn to the simple dummy version of the regression, in which the dummies for knight’s estate, owned by high nobility,⁸ or larger than 1000 hectares are multiplied by the total area of the property. A negative and signiﬁcant coefﬁcient for any of these variables (rgtot, graftot, and over1000tot) would imply a uniform break per hectare given because of the particular characteristic. These results are shown in Table 9.2. Again, the pattern and values of the quality-adjusted land type variables are as they should be, and the pattern of coefﬁcients for the ‘raw’ amounts ⁸ Dummy = 1 if the rank of owner was count(ess) or above, 0 otherwise.

246

Concluding Remarks Table 9.2. Regression estimation, tax value as dependent variable, simple dummy form Variable

Coefﬁcient

Std. Error

T-value

Signif.

Plough Meadow Pasture Wood Water Moor

2.279 0.635 −6.089 0.132 −1.128 −2.312

0.165 0.634 0.668 0.110 0.269 1.487

13.852 1.002 −9.120 1.209 −4.195 −1.555

0.000 0.317 0.000 0.227 0.000 0.120

QualPL QualME QualPA QualWO QualWA QualMO

11.418 13.535 4.364 3.331 3.215 0.518

0.087 0.542 0.584 0.047 0.229 0.850

130.688 24.989 7.470 71.080 14.068 0.609

0.000 0.000 0.000 0.000 0.000 0.542

rgtot graftot over1000tot

−0.563 −0.191 0.091

0.041 0.037 0.088

−13.669 −5.229 1.027

0.000 0.000 0.304

Constant

52.358

42.987

1.218

0.223

Note: Adj. RSQ = 0.895; N = 15, 190

of each land type show little change from the previous form of the regression. Again, the coefﬁcients of the moor variables are insigniﬁcant, and would have little meaning anyway because of the heterogeneous nature of this variable. The negative and signiﬁcant coefﬁcients for both knight’s estates and the high nobility suggest that there may indeed have been some bias toward under-assessment in the land tax cadastre, and that this bias may have been selective, since the coefﬁcient for property over 1000 hectares did not turn out to be signiﬁcant. These results set the stage for the ﬁnal version of the estimation equation, which includes three sets of interactive dummy variables, each of them the product of the dummy times the amount of one type of land contained in the property. We thus have 18 new variables included: 1. rgpl, rgmd . . . rgmo for ploughland, meadow . . . moor on knight’s estates; 2. grafpl, grafmd . . . grafmo for ploughland, meadow . . . moor on properties owned by members of the high nobility; and 3. over1000pl, over1000md . . . over1000mo for ploughland, meadow . . . moor on properties of 1000 or more hectares. Table 9.3 presents the result of this estimation. This formulation basically repeats the pattern and values of the coefﬁcients for the quality-adjusted land

Concluding Remarks

247

Table 9.3. Regression estimation, tax value as dependent variable, all interactive dummy variables used Variable

Coefﬁcient

Std. Error

T-value

Signif.

QualPL QualME QualPA QualWO QualWA QualMO

11.335 13.326 5.353 3.267 3.484 2.340

0.085 0.529 0.597 0.046 0.239 0.875

132.680 25.186 8.967 70.464 14.561 2.674

0.000 0.000 0.000 0.000 0.000 0.008

2.592 1.307 −5.612 −0.761 −2.000 −15.877

0.233 0.906 1.110 0.213 0.985 3.038

11.125 1.442 −5.055 −3.577 −2.031 −5.227

0.000 0.149 0.000 0.000 0.042 0.000

−1.180 −0.769 1.945 −0.780 22.138 −0.826

0.171 0.648 0.981 0.072 2.430 0.559

−6.898 −1.186 1.983 −10.862 9.115 −1.478

0.000 0.236 0.047 0.000 0.000 0.140

grafpl grafme grafpa grafwo grafwa grafmo

2.192 1.591 −5.648 −0.542 0.201 −20.051

0.191 0.708 1.207 0.061 0.559 2.472

11.449 2.247 −4.680 −8.942 0.359 −8.111

0.000 0.025 0.000 0.000 0.719 0.000

over1000pl over1000me over1000pa over1000wo over1000wa over1000mo

−0.753 −1.939 −0.464 1.355 0.706 9.846

0.176 0.735 0.931 0.203 0.934 2.923

−4.270 −2.636 −0.498 6.685 0.756 3.369

0.000 0.008 0.618 0.000 0.450 0.001

Constant

74.588

42.340

1.762

0.078

Plough Meadow Pasture Wood Water Moor rgpl rgme rgpa rgwo rgmo rgwa

Note: Adj. RSQ = 0.904; N = 15, 190

types and the ‘raw’ land types; this part of the speciﬁcation seems very stable indeed. The results with the interactive dummy variables are interesting, but mixed. Consider ﬁrst the negative, highly signiﬁcant coefﬁcients found on the interactive dummy variables: 1. The coefﬁcient for rgpl implies a small downward bias, on the order of 10 per cent of the average assessment per hectare of ploughland, and for rgwo, on the order of 20 per cent for woodland, if the property was a knight’s estate.

248

Concluding Remarks

2. There appeared to be a similar, but smaller, downward bias in the assessment for ploughland on estates over 1000 hectares, and a somewhat larger one for meadow, on the order of 15 per cent, compared to the coefﬁcient for QualME. 3. High nobility appeared to have received a break on the taxation of pasture in particular, and also of woodland, but the pattern of coefﬁcients for pasture in all categories leaves signiﬁcant doubt about whether any of the pasture coefﬁcients means anything. Offsetting these values are signiﬁcant positive coefﬁcients on several other of the interactive variables, in particular ploughland and meadow for the high nobility, along with pasture for knight’s estates and woodland for large properties. We can also note that the addition of the dummy variables adds very little to the overall explanatory power of the regression equation. The R2 adjusted for degrees of freedom was already 0.900 for the very simplest form of the equation, and the addition of the three sets of interactive dummy variables brought the adjusted R2 value only to 0.904. The coefﬁcients for pasture, other than the ‘correct’ one for quality-adjusted pasture land, are often negative and highly signiﬁcant, sometimes even too negative, in the sense that their absolute value is so great that—when taken together with the effects of the other coefﬁcients for pasture in the equation—it would imply an overall negative assessment for pasture land, which is patently impossible. What could account for such a result? I ﬁrst suspected that there might be an outlier effect from the properties with the very largest amounts of pasture land: in the dataset of principal properties c.1882, there were 503 properties that had 100 hectares or more of pasture land, containing in total nearly 97 thousand hectares of pasture. These properties were distributed among all the provinces, with the largest concentration in Pomerania and the smallest in Silesia. Nearly all of these hectares of pasture were owned by the Prussian State (25,000) or nobility (71,000). Other owners possessed a grand total of only 758 hectares of pasture in properties of this group. Among the nobility, these properties were, interestingly enough, concentrated in the hands of the lower nobility: untitled nobles and barons owned 51,500 of these pasture hectares, counts about 16,000, and princes and royalty the rest. Since the 97,000 hectares of pasture in these 503 properties was well under one-quarter of the pasture land recorded for all properties of 100 hectares or more (more than 418,000 ha), it is unlikely that these 503 could have inﬂuenced the regression results so strongly. I would therefore speculate that there may have been considerable confusion over what was to be reported as pasture, and what as moor, or what as pasture and what as meadow, to the compilers of the Handbooks, so that these data categories may be subject to even larger margins of error than were the other land types. Compounding this could have been

Concluding Remarks

249

conversion of land in the two decades or so between the land tax cadastre and the early 1880s when the Handbooks were compiled. Conversion of moorland to pasture, or the draining of swamps and ponds to make pasture, would have increased the amount of pasture without affecting the overall tax assessment of the property. Since moor and water normally had lower tax assessments than did pasture, this would make it seem as though pasture were dragging down the tax assessment of the property. Without much more detailed examination of the data, which is probably impossible in any case, one cannot tell whence the peculiar coefﬁcients for pasture land orginated. What does seem clear from the regression results, however, is that it would be difﬁcult to make any case for a systematic bias in the land tax assessment system that favoured any of the categories of potential favourites, namely knight’s estates, the high nobility, or the very largest properties. Particularly the dummies that interact with each of the several land types produce such a mixed bag of coefﬁcients as to negate any notion of a pattern in the data. The land tax assessment procedure was in principle systematic, and although it may have been applied quite differently in different districts or regions of the country, and even been chaotic at times,⁹ there is no evidence that it, on average, favoured any particular class of property or group of owners.¹⁰

9 . 4 . AU F W I E D E R S E H E N At the end of all this effort, both on my part and on that of the reader, I hope that this volume has brought a new, deeper, and broader understanding of East Elbia, its landowners, its regional and sectoral variety, and its place in German economic history. It is an area that has long caught the fancy of historians, political scientists, sociologists, economists, and even genealogists, perhaps because it has always been a bit ‘different’, and its history full of drama and pathos. Study of this area is, in my opinion, very rewarding, and will remain so. ⁹ Cf. Wygodzinski, Die Besteuerung, esp. ch. 1. In a 4-page (14–17) catalogue of difﬁculties and complaints, modiﬁcations and recalculations of the land tax assessment, there is no mention of any kind of bias in favour of any class of properties or owners. Such complaints as were mentioned were mostly about regional inequities. ¹⁰ This, incidentally, is the same conclusion I came to when examining the same question with very similar data from Hungary c.1910. See my ‘The Social Distribution of Landed Wealth in Hungary c.1910’, Research in Economic History, suppl. 5: George Grantham and Carole Leonard, eds, Agrarian Organization in the Century of Industrialization (Glenwood, Ill.: Johnson Associates, 1989), 219–49.

APPENDIX

Coding the Handbook Data A . 1 . LOC AT I O N C O D I N G To facilitate retrieval, checking, and combining of different groups of properties or owners, each property received special codes for the following: source from which it was taken, geographic location, and type of owner and condition of ownership. Additional codes identiﬁed whether it was a principal property of 100 hectares or more, whether it was usable for cluster analysis, and special ‘unﬁt’ codes which identiﬁed whether it could be used in various kinds of regressions or in the cluster analysis. Source coding is explained in Section A.3.2, ‘Reference Variables and Fields’, as are other codes in their respective sections. The geographic location code identiﬁed the province, the administrative region (Regierungsbezirk) and the district (Kreis) of that administrative region in which the property was located. Since the Handbooks listed properties by district, further narrowing down of location was not possible (nor necessary). The location code was a ﬁve-digit number: the ﬁrst digit identiﬁed the province, the second and third the administrative region, and the fourth and ﬁfth the district within that region. I chose to use the administrative divisions of Prussia as they existed in 1895, and used the standard numbering used by the Prussian Statistical Bureau in its yearbooks and other statistical publications for the provinces and administrative regions, as shown in Table A.1. As one can see from the numbering of the administrative regions, the provincial numbering is redundant, but I retained it anyway because it makes searching for provincial data in the databases considerably easier. To give an example of geographic location coding, all properties located in the district of Ruppin (number 16 within the Potsdam region) received the code 40616: 4 for Brandenburg, 06 for Potsdam, and 16 for Ruppin. Every property in the district of Frankenstein received a location code of 71317: 7 for Silesia, 13 for Breslau, and 17 for Frankenstein district within Breslau administrative region. Berlin, although included in the coding scheme for completeness, in fact does not ﬁgure in the description and analysis of this study because it contained no properties of at least 100 hectares in size.

A . 2 . OWN E R C O D E S Each property also received an owner code that made it possible to group properties by the social or economic position of their owners. While this code is actually a four-digit code, only three digits are relevant for this study (in some owner categories only two), and the fourth has remained unused.

Coding the Handbook Data

251

Table A.1. Geographic location numbering scheme Province No.

Name

Admin. region within province No. Name

1

East Prussia

01 02

Königsberg Gumbinnen

2

West Prussia

03 04

Danzig Marienwerder

3

Berlina

05

Berlinb

4

Brandenburg

06 07

Potsdam Frankfurt/Oder

5

Pomerania

08 09 10

Stettin Köslin Stralsund

6

Posen

11 12

Posen Bromberg

7

Silesia

13 14 15

Breslau Liegnitz Oppeln

8

Saxony

16 17 18

Merseburg Magdeburg Erfurt

a Berlin data included in Brandenburg. Potsdam.

b

Berlin data included in

A.2.1. Physical Persons The ﬁnest divisions of the owner code are found in the category of physical persons, as shown in Table A.2. According to this scheme, a property owned by a baron and his wife would be coded 1320 (the last digit will always be zero), one owned by commoner siblings would be coded 2030, and one owned by the heirs of a count would receive a code of 1440. If a prince were the single owner of the property, its owner code would be 1510, just as it would be 2010 if owned by one person of non-noble status. In Silesia in 1892 the property of Porschütz (264 ha) was owned by the Count Campani Foundation for Girls, so it received the owner code 1450, whereas the entailed estate of Klein Steegen in East Prussia (1792 ha), owned by Captain of Cavalry (ret.) Oscar von Steegen, received an owner code of 1090. Royalty were coded separately. Any member of any branch of the Hohenzollern family was considered to be Prussian royalty. Thus the estate of Krojanke in West Prussia (2734 ha in 1894), owned by Friedrich Leopold Hohenzollern-Sigmaringen, Prince of Prussia, received an owner code of 1710 (aristocrat, Prussian royalty, single ownership), whereas the estate of Kraussnick (490 ha in Brandenburg in 1896), a royal entail of King Wilhelm II, had an owner code of 1770 (aristocrat, Prussian royalty, royal entail). In contrast to Kraussnick, Norkitten in East Prussia (4804 ha in 1907), received a code of 1670 because it was an entailed estate of the Duke of Anhalt-Dessau, one of the 22 recognized

252

Coding the Handbook Data

Table A.2. Owner codes for properties owned by physical persons First digit

Second digit

Third digit

1 Aristocrat

0 1 2 3 4

0 1 2 3 4

unused single owner with spouse several owners heirs

5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

personal foundation royal domaine royal entail unused entail of other nobility unused single owner with spouse several owners heirs personal foundation unused unused bankrupt estate entail of non-noble person

2 Non-noble person (‘bourgeois’)

5 6 7 8 9 0

Untitled noble unused Knight, Baronet Baron, Freiherr, Edler Count, Reichsgraf, Burggraf, Markgraf Prince, duke, archduke Other royalty Prussian royal family unused unused in all cases

sovereigns in Germany. Czeszewo in Silesia (2687 ha in 1896), which belonged jointly to Princess Marie and Hereditary Prince Bernhard of Saxony-Meiningen, from another of the 22 recognized sovereign houses of Germany, received an owner code of 1630.

A.2.2. Non-Physical Legal Persons Non-physical legal persons presented a different kind of variety. Let us consider ﬁrst property owned by a religious body or a state ( Table A.3). Some examples: Nadziejewo (449 ha in Posen in 1896) was owned by the Catholic Priests’ Seminar, so was coded 3140. The Cathedral in Frauenberg in East Prussia, assumed to be Lutheran, owned the 205-hectare property of Grundhof, which therefore received an owner code of 3220. The local (Lutheran) church in Guztkow, Pomerania owned a property called Strellin (223 ha) in 1893, so Strellin’s owner code became 3230. State properties are denoted by a ﬁrst digit of 4. A second digit of 0, 1, or 2 denotes the Prussian State, another German state, or a foreign state, respectively. Finer divisions into different state entities such as forestry, salt production, railways, stud farms, the military, etc. in the third digit were of no consequence for any of the purposes of this study. Thus most state properties received an owner code of 4000 (4010 if owned by the Settlement Commission for Posen and West Prussia), 4100, or 4200 to denote their ownership by Prussia, another German state, or a foreign state. To illustrate: the estate of Winnigen in Saxony, 1126 hectares in 1885, owned by the Duchy of Brunswick (Herzoglicher Braunschweiger Fiscus), received an owner code of 4100. The Swedish crown property of Grünhufe in Pomerania, 332 hectares in 1884,

Coding the Handbook Data

253

Table A.3. Owner codes for religious bodies and states First digit

Second digit

Third digit

3 Religious bodies

1 2 3 4 5 0 1 2

1 2 3 4

4 States

Roman Catholic Reform (protestant) Greek Orthodox Greek Catholic Other Prussian State Other German state Foreign state

0 1 2 3 4 5 6 7 8

Bishopric, archdiocese Church district, abbey Local church (parish) Religious schools, foundations, monastic orders General ﬁscus Prussian settlement commission unused Joint ownership with another owner Forest ﬁscus Domains and stud farms Railway ﬁscus Military ﬁscus Royal salt works

was coded 4200. Of properties of the Prussian State in West Prussia in 1894, Freudenthal (154 ha) received an owner code of 4040 because it was owned by the forest ﬁscus, but Stellinen (814 ha) was coded 4030 because it was jointly owned by the State and Julius Klaasen. The Royal Settlement Commission had purchased, but not yet subdivided, the estate of Zietlonkowo in Posen (559 ha in 1910), so that property received an owner code of 4010. The Prussian military’s 5500-hectare estate of Neuhammer in Silesia in 1909, coded therefore as 4070, was a military exercise ground (Truppenübungsplatz), whereas the property of Ulrichsfelde in West Prussia (137 ha in 1885), received a code of 4060 because owned by the Prussian state railways. Various levels of communities were combined in most tables and other analyses in the study. The scheme for owner coding of community-owned properties was as shown in Table A.4 (the third digit was zero in all cases). Finally we come to the ‘other’ category of owners, where all those owners who do not ﬁt into any of the earlier categories are assigned. The ﬁrst digit is 9, and the third is 0 in all cases. The second digit denotes the kind of owner, as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9.

Association Non-religious foundation Other legal person or society Hospital School Orphanage University or academy Unknown owner ‘Divided or parcelled’—this designation was for large properties listed in the Handbooks as being, or having been, subdivided into smaller units.¹

¹ Some of these properties were still technically owned by the Royal Settlement Commission for Posen and West Prussia, but nevertheless received an owner code of 9900 because they had already been subdivided, or were in the process of subdivision. Only those properties of the Settlement

254

Coding the Handbook Data

Table A.4. Owner codes for communities or business ﬁrms First digit

Second digit

Third digit

5 Community (Gemeinde) 6 Communal owners

0 all cases

0 all cases

0 unused 1 Former serfs 2 Other communal group (e.g. pasturage society) 0 City 1 District (Kreis) 2 Administrative region (Regierungsbezirk) 3 Province

0 all cases

7 City or larger administrative unit

8 Banks and businesses

0 unused 1 Banks, ﬁnancial institutions 2 Indl. & commercial companies 3 Savings associations

0 Sole ownership 1 unused 2 unused 3 Joint ownership with another body 0 All cases

The 9900 code indicates that such properties should be left out of our working database, and not entered into any of the calculations for large properties performed in this study. Thus a property with the owner code of 9900 plays no part in any of the descriptions or analyses of this book, but has been kept in the underlying database in the interests of retaining potentially useful information. Some examples of the coding of ‘other’ owners are the following: the 222-hectare property Friedrichswille in Brandenburg in 1885, owned by the Provincial Society for Combating Vagabondage, received the owner code of 9100; in Saxony in 1885, the estate Nagelstedt (268 ha), owned by the Griefstädter Foundation Fund in Erfurt, received a code of 9200, while the 237-ha property Schönwiese in East Prussia was coded to 9700 because it was owned by the Academic Senate in Königsberg. All three Silesian properties of the Central Hospital in Görlitz received an owner code of 9400, just as the six Silesian properties of the Royal Land School in Pforta were coded as 9500.

A . 3 . OT H E R C O D E S

A.3.1. Additional Characteristics of Owner or Property Separate variables have been set up in the databases to indicate other characteristics of the owner or the property.

Commission that were still intact, or those unsubdivided parts retained as a Restgut (remainder property) by the Settlement Commission received the owner code of 4010.

Coding the Handbook Data

255

1. For example, the variable OF indicates whether the owner was a military ofﬁcer or wife of an ofﬁcer (including active service, reserve, and retired ofﬁcers, in all branches of the military including the Home Guard (Landwehr)). A value of 1 = ‘Yes’ for this variable, 0 = ‘No’. 2. For the chapter on knight’s estates, any property that was a knight’s estate received a value of 1 for the variable ‘RG’ (for Rittergut), otherwise 0. If a property listed as a subunit of a manor was a knight’s estate, where the manor was recorded as the principal property, then the variable ‘ManorRG’ recorded the number of knight’s estates contained within the manor as subunits. 3. A separate variable, called ‘Unﬁt’ indicates whether and why a particular property should not be included in any regression analysis or cluster analysis, according to the schema shown in Table A.5. Depending on the purpose of the statistical analysis, different categories of ‘unﬁt’ observations had to be eliminated. For example, in the cluster analyses of Chapter 7, using proportions of arable, forest, etc. land as variables, only observations unﬁt for reasons 1 through 5 had to be eliminated. Numbers 6, 7, and 9 touched only on the land tax assessment, not used in the cluster analyses mentioned, and number 8 could be included because land data was taken from elsewhere only in cases where (a) totals for a given property were exactly the same in two different years, but the land details were missing in one, or (b) land details were the same in two different years, but the totals did not match (this last most likely because of typographical or transcription error). In regressions such as those of Chapter 9, where the dependent variable was the land tax net yield (GSRE), and independent variables included the areas of ploughland, meadow, forest, etc., not only did observations unﬁt for reasons 1 through 5 have to be eliminated, but also those unﬁt for reasons 6 and 9. Table A.5. Values for the variable ‘Unﬁt’ Value

What the value indicates

0 or blank 1 2 3 4 5 6 7 8 9

Fit to be included in any kind of regression or cluster analysis Arbitrary allocation of land No details of land, only sum, in source Data taken from elsewhere, but anomalies remain Land data estimated on basis of GSREa , or empty rowb Sum of details differs from total by 5% or more GSREa allocated arbitrarily between or among properties GSRE data taken from elsewhere, OK Land data taken from elsewhere, OK GSRE allocated according to Kreis averages

Note: 1–9 indicate why variable should not be used in statistical analysis. a GSRE = Grundsteuerreinertrag, or ‘land tax net yield’, the basis for calculating the land tax payable for a property. b ‘Empty row’ indicates that only the property name, and perhaps the name of the owner, was given, but there was no other information about the property.

256

Coding the Handbook Data

A.3.2. Reference Variables and Fields To facilitate easy reference to any single observation, and for later checks on the accuracy of data entry, several reference variables were added to the database. The most important of these are the following: 1. Province: The Handbooks had separate editions for each province. This identiﬁes the individual province editions from which the data were taken. 2. Year: This is the year of publication of the particular edition of the Handbook used for the province in question. 3. Source: The ‘Source’ code identiﬁes the page and line number of the volume where the given property is to be found: the ﬁrst three digits are the page number, the last two the line number. Thus 45307 means page 453, line 7, and 215 (really 00215, but most spreadsheet and database software does not display initial zeros) means page 2, line 15. 4. SchlPropNr: The 1909 Silesian address book used gave each property a unique sequence number. For the observations from this province only, and only in the c.1907 database, ‘SchlPropNr’ recorded these unique sequence numbers to facilitate even more accurate reference to where the property could be found in the source book. Additionally, each of these properties also received a ﬁve-digit ‘Source’ number as in number 3 above. 5. Notes: This is perhaps the most important of the reference ﬁelds. It contains information about the state of the property and any changes made to the data in the course of ‘cleaning’ and checking those data. (a) Sometimes there was a note about the property in the source book itself. The single most common one of these was ‘Subdivided or in the process of subdivision’, which led to the assigning of owner code 9900 and the exclusion of the property in question from any of the data used in this book. Other notes appeared, such as ‘see appendix’, in which there was usually some modiﬁcation to the information about the property, or ‘data unavailable’, in the case where a property (and usually its owner as well) was listed, but no further details were shown. All such notes were included. (b) When one or more industrial establishments listed in the source book for a given property were assigned to the industrial category ‘other’, they were named in the Notes ﬁeld. (c) Any changes to the data of the property also appear in the Notes. If, for example, the areas of ploughland and meadow were combined in the source, but assigned separate values later, the assignment and reasons for it appear in the Notes. Perhaps the single most common note would be ‘GSRE allocated according to Kreis averages’ (leading to a value of 9 for the ‘Unﬁt’ variable, see above) for those properties for which no land tax net yield was reported in the source, leading to an attribution based on the average land tax net yield per hectare of each of the land types in the district in which the property was located. All other modiﬁcations and attributions, with their reasons, would also appear in the Notes.

Coding the Handbook Data

257

(d ) Even if no modiﬁcation were made, a Note might still be added to indicate an unsuccessful attempt at supplementing or modifying the data. For example, for the estate Gross Tromp in the district of Braunsberg in East Prussia, 482 hectares and home to a brickworks and a dairy, owned by Gustav Fuchs in 1907, no land details were given, only the total hectares and the land tax net yield. An unsuccessful attempt to ﬁnd details in the East Prussian Handbooks of 1884, 1895, and 1929 led to the following Note for this property: ‘No details. Does not exist in 1929. Smaller in 1884 and 1895.’ Or, if the reported details of ploughland, meadow, pasture, etc., did not add up to the total area reported for the property, and the anomaly could not be resolved, a Note would point this out. If the difference were more than ± 5 per cent, this would also lead to a value of 5 for the ‘Unﬁt’ variable, as noted above.

List of Works Cited 1 . A RC H I VA L S O U RC E S A N D P RU S S I A N L AW S

Geheimes Staatsarchiv Preußischer Kulturbesitz, I Rep. 77 Ministerium des Innern, Tit. 438, Nr. 62, Band 2, Matrikel zum Erscheinen im Stande der Ritterschaft bef¨ahigenden G¨uter der Provinz Preußen (dated 31 December 1834). Geheimes Staatsarchiv Preußischer Kulturbesitz, I Rep. 77 Ministerium des Innern, Tit. 438, Nr. 62, Band 3, letter of 15 March 1837 from Interior Minister von Rochow to provincial governor (Oberpr¨asident) von Sch¨on, page unnumbered. Gesetz wegen Anordnung der Provinzialst¨ande f¨ur das K¨onigreich Preußen vom 1. Juli 1823, Gesetzessammlung 1822/23. Gesetz wegen Anordnung der Provinzialst¨ande f¨ur die Provinz Brandenburg vom 1. Juli 1823, Gesetzessammlung 1822/23.

2 . O F F I C I A L G OV E R N M E N T P U B L I C AT I O N S

Germany. Kaiserliches Statistisches Amt, Statistisches Jahrbuch f¨ur das Deutsche Reich (Berlin: Puttkammer & M¨uhlbrecht, 1880– ). Statistik des Deutschen Reichs (Berlin: Puttkammer & M¨uhlbrecht), various volumes of the agricultural censuses: 1882: N.F. (ns) 5, Landwirthschaftliche Betriebstatistik nach der allgemeinen Berufsz¨ahlung vom 5. Juni 1882 (1885); 1895: ns 112, Die Landwirthschaft im Deutschen Reich nach der landwirthschaftlichen Betriebsz¨ahlung vom 14. Juni 1895 (1898); 1907: ns 212 (in two parts), Berufs- und Betriebsz¨ahlung vom 12. Juni 1907. Landwirtschaftliche Betreibstatistik (1909). Statistik des Deutschen Reichs (Berlin: Puttkammer & M¨uhlbrecht), various volumes of the industrial censuses: 1882: ns 7, Gewerbestatistik nach der allgemeinen Berufsz¨ahlung vom 2. Juni 1882. Gewerbe-Statistik der Staaten und gr¨oßeren Verwaltungsbezirke, Erster Abschnitt (Anzahl und Personal der Betriebe) (1886); 1895: ns 117, Berufs- und Gewerbez¨ahlung vom 14. Juni 1895. GewerbeStatistik der Verwaltungsbezirke, Erster Theil (Kleinere Verwaltungsbezirke Preußens) (1898); 1907: ns 218, Berufs- und Betriebsz¨ahlung vom 12. Juni 1907: Gewerbliche Betriebstatistik. Abteilung VI—Kleinere Verwaltungsbezirke; Preußen (1909).

List of Works Cited

259

Great Britain. Local Government Board, The Return of Owners of Land in England and Wales exclusive of the metropolis, presented to both Houses of Parliament (London: HMSO, 1875). Prussia. Ministerium der Finanzen: Die Grundsteuerveranlagung. Ergebnisse f¨ur den Regierungsbezirk (Berlin: Ministerium der Finanzen, 1866– ), 25 vols, one volume for each administrative region in Prussia. Prussia. Preussisches Statistisches Landesamt, ‘Grundbesitzverteilung in Preussen nach den Ergebnissen der l¨andlichen Verschuldungsstatistik’, Zeitschrift des Preussischen Statistischen Landesamts, suppl. 42 (Berlin: PSL, 1921). 3 . D I R E C TO R I E S O F L A N D OW N E R S A N D L A N D E D P RO PE RT I E S

Adressbuch des Grundbesitzes in : dem Areal nach von 500 Morgen aufw¨arts, volumes for Posen and Silesia (Berlin: F. B¨urde, 1872). ¨ General-Adressbuch der Großgrundbesitzer des Deutschen Reiches u. der Osterr.-Ung. Monarchie. Vollst¨andig in 40 B¨anden. While advertised as encompassing 40 volumes, so far as I have been able to determine, only the ﬁrst three volumes were ever published (Berlin: R. K¨uhn, 1878). General-Adressbuch der Ritterguts- und Gutsbesitzer im Deutschen Reich, I. Das K¨onigreich Preussen, 2. Lieferung, Die Provinz Pommern (Berlin: Verlag des Landwirthschaftlich-Statistischen Bureaus Lodemann und Nicolai Verlag, 1879). Pommersches G¨uter-Adressbuch, Niekammers landwirtschaftliche G¨uteradressb¨ucher (Stettin: Niekammer, 1905). Wykaz alfabetyczny wszystkich posiadlosci ziemskich w. W. Ksiestwie Poznanskiem/Adressbuch des Grundbesitzes im Grossherzogthum Posen (Berlin: F. B¨urde & Co., 1872). Ellerholz, Paul, gen. ed., Handbuch des Grundbesitzes im Deutschen Reiche (Berlin: Nicolaischer Verlag, later Paul Parey Verlag, 1879– ). Frantz, Adolf, General-Register der Herrschaften, Ritter- und anderer G¨uter der preussichen Monarchie. Mit Angaben u¨ ber Areal, Ertrag, Grundsteuer, Besitzer, Kauf- und Taxepreise, u. s. w. (Berlin: Grellius, 1863). Klar, Paul Alois, B¨ohmens gr¨oßer Grundbesitz, wie dieser in der k¨onigl. Landtafel inneliegt (Prague: self-published, 1856). Korn, Wilhelm Gottlob, Schlesisches G¨uteradreßbuch, 9th edn (Breslau: Korn, 1909). Rauer, K. Fr. [Karl Friedrich], Alphabetischer Nachweis (Adressbuch) des in den Preussischen Staaten mit Ritterg¨utern angesessenen Adels (Berlin: selfpublished, 1857). Hand-Matrikel der in s¨ammtlichen Kreisen des Preussischen Staates auf Kreisund Landtagen vertretenen Ritterg¨uter ( Berlin: self-published, 1857). Adress-Buch der Rittergutsbesitzer und Ritterg¨uter in den Preussischen Staaten, 2 vols (Berlin: R. K¨uhn, 1857).

260

List of Works Cited

4 . A RT I C L E S , B O O K C H A P T E R S , L E C T U R E S , M O N O G R A PH S , R E F E R E N C E WO R K S

Bateman, John, The Great Landowners of Great Britain and Ireland (Leicester: Leicester University Press, 1971; reprint of 4th edn, 1883). Beck, Friedrich, ‘Die brandenburgischen Provinzialst¨ande 1823–1872/75,’ in Kurt Adamy and Kristina H¨ubener, eds, Geschichte der Brandenburgischen Landtage. Von den Anf¨angen 1823 bis in die Gegenwart, Brandenburg Historische Studien, 3 (Potsdam: Verlag f¨ur Berlin-Brandenburg, 1998), 62–74. Beck, Lajos, A magyar f¨oldbirtok megoszl´asa: Agr´ar-statisztikai tanulm´any [The distribution of Hungarian landed property: An agrarian-statistical investigation] (Budapest: Pallas, 1918). Berghoff, Hartmut, ‘Adel und Industriekapitalismus im Deutschen Kaiserreich—Abstoßungskr¨afte und Ann¨aherungstendenzen zweier Lebenswelten’, in Heinz Reif, ed, Adel und B¨urgertum in Deutschland, i. Entwicklungslinien und Wendepunkte im 19. Jahrhundert, Elitenwandel in der Moderne, 2 (Berlin: Akademie Verlag, 2001), 233–71. Berlin-Brandenburgische Akademie der Wissenschaften, Das digitale W¨orterbuch der deutschen Sprache des 20. Jahrhunderts, on the Internet at http://www.dwds.de. Bernhard, Ludwig, Die Polenfrage, 2nd edn (Leipzig: Duncker & Humblot, 1910). Blanke, Richard, Prussian Poland in the German Empire, East European Monographs, 86 (Boulder: University of Colorado Press, 1981). Brentano, Lujo, Familienﬁdeikommisse und ihre Wirkungen (Berlin: Simion, 1911). ¨ Buchsteiner, Ilona, Grossgrundbesitz in Pommern 1871–1914. Okonomische, soziale und politische Transformation der Grossgrundbesitzer (Berlin: Akademie Verlag, 1993). Conrad, Johannes, ‘Die Besitzverh¨altnisse an Grund und Boden in Schottland’, Jahrb¨ucher f¨ur National¨okomie und Statistik, ns 26/1 (1876), 50–6. ‘Die Grundbesitzverh¨altnisse im brittischen Reiche’, Jahrb¨ucher f¨ur National¨okomie und Statistik, ns 27/5–6 (1877), 377–96. ‘Die Latifundien im Preußischen Osten’, Jahrb¨ucher f¨ur National¨okomie und Statistik, ns 16 (1888), 121–70. ‘Der Grossgrundbesitz in Ostpreussen’, Jahrb¨ucher f¨ur National¨okonomie und Statistik, 3rd series, 2 (1891), 817–44. ‘Die Dom¨anenvorwerke in der Provinz Preußen’, Jahrb¨ucher f¨ur National¨okonomie und Statistik, 3rd series, 6 (1893), 27–59. ‘Der Grossgrundbesitz in Schlesien’, Jahrb¨ucher f¨ur National¨okonomie und Statistik, 3rd series, 15 (1898), 705–29.

List of Works Cited

261

Dillwitz, Sigrid, ‘Quellen zur sozial-¨okonomischen Struktur der Bauernschaft im Deutschen Reich nach 1871’, Jahrbuch f¨ur Wirtschaftsgeschichte, 2 (1977), 237–70. Eddie, Scott M., ‘Agricultural Production and Output per Worker in Hungary, 1870–1913’, Journal of Economic History, 27/2 (June 1968), 197–222. ‘The Terms and Patterns of Hungarian Foreign Trade, 1882–1913’, Journal of Economic History, 37/2 (June 1977), 329–58. ‘The Social Distribution of Landed Wealth in Hungary ca.1910’, Research in Economic History, suppl. 5: George Grantham and Carole Leonard, eds, Agrarian Organization in the Century of Industrialization (Glenwood, Ill.: Johnson Associates, 1989), 219–49. ‘The Distribution of Landed Properties by Value and Area: A Methodological Essay Based on Prussian Data, 1886–1913’, Journal of Income Distribution, 3/1 (summer 1993), 101–40. ‘Junkers and Magnates: The Social Distribution of Landed Wealth in Prussia and Hungary; A Case Study of Pomerania and Transdanubia, 1893’, ¨ Osterreichische Osthefte, 1 (1994), 109–31; ‘Cluster-analízis a f¨oldbirtokszerkezet vizsg´alat´aban’ [Cluster analysis in the examination of the structure of landed property], AETAS 3 (1994), 17–36. ‘Großgrundbesitz im ostelbischen Preußen: Datenbasis und methodologische Probleme’, in Heinz Reif, ed., Ostelbische Agrargesellschaft im Kaiserreich und in der Weimarer Republik (Berlin: Akademie Verlag, 1994), 141–55. ‘The Price of Land in Eastern Prussia: Data from Capital Gains Tax Records, 1891–1907’, Historical Social Research/Historische Sozialforschung, 22/2 (1997), 195–216. Historisches Verzeichnis der Grundbesitzer des Burgenlandes/Burgenland t¨ort´enelmi gazdacímt´ara 1893–1930, Burgenl¨andische Forschungen, 79 (Eisenstadt, Austria: Verlag des burgenl¨andischen Landesarchivs, 1999). ‘What Size and Kind of Agricultural Units? Reﬂections on the Study of Prussian Land Ownership before the Great War’, Begegnungen (forthcoming). and Christa Kouschil, The Ethnopolitics of Land Ownership in Prussian Poland, 1886–1918: The Land Purchases of the Ansiedlungskommission, Trondheim Studies on East European Cultures & Societies, 9 (Trondheim, Norway: NTNU, 2002). Frank, Alexandra, Die Entwicklung der ostelbischen Gutswirtschaften im Deutschen Kaiserreich und in den Anfangsjahren der Weimarer Republik, Beitr¨age zur Wirtschafts- und Sozialgeschichte, 6 (Weiden: Eurotrans-Verlag, 1994). Gerschenkron, Alexander, Bread and Democracy in Germany (Berkeley: University of California Press, 1943). Goltz, Theodor Freiherr von der, Vorlesungen u¨ ber Agrarwesen und Agrarpolitik (Jena: Gustav Fischer, 1904).

262

List of Works Cited

Grzes, Boleslaw, ‘Teoretyczno-propagandowe aspekty dyskryminacji Polakow w Poznanskiem na przelomie XIX i XX wieku’ [Theoretical-propaganda aspects of discrimination against the Poles in turn of the century Poznania], in Lech Trzeciakowski and Stanislaw Kubiak, eds, Rola Wielkopolski w dziejach narodu Polskiego [The Role of Greater Poland (Poznania) in the history of the Polish nation] (Pozna´n: Wydawnictwo Naukowe Uniwersytetu im. Adama Mickiewicza w Poznaniu, 1979). H¨abich, Theodor, Deutsche Latifundien. Ein Beitrag zur Berichtigung unserer Vorstellung von der bestehenden Verteilung des l¨andlichen Grundeigentums (K¨onigsberg: Gr¨afe & Unzer, 1929). Hallerberg, Mark, ‘The Political Economy of Taxation in Prussia 1871–1914’, Jahrbuch f¨ur Wirtschaftsgeschichte, 2 (2002), 11–34. Harnisch, Hartmut, Adel und Großgrundbesitz im ostelbischen Preussen 1800–1914, inaugural lecture, Faculty of Philosophy and Historical Sciences, ¨ Humboldt University in Berlin, 16 June 1992, Offentliche Vorlesungen, 7 (Berlin: Humboldt-Universit¨at, 1993). Henderson, W. O., The Zollverein (London: Frank Cass, 1984). Herde, Werner, Der Grundbesitz der Stadt Greifswald. Eine Untersuchung u¨ ber die volkwirtschaftliche Bedeutung st¨adtischen Grundbesitzes, dissertation: Staatswissenschaftliche Fakult¨at, Universit¨at Greifswald (Greifswald: Julius Abel, 1921). Hess, Klaus, Junker und b¨urgerliche Großgrundbesitzer im Kaiserreich. Landwirtschaftlicher Großbetrieb, Großgrundbesitz und Familienﬁdeikommiß in Preußen (1867/71–1914), Historische Forschungen, 16 (Stuttgart: Steiner Verlag, 1990). Hirsch, Alfred, Ungarns Grundbesitzverh¨altnisse. Agrarstatistische Untersuchungen (Halle a. S.: Ehrhardt Karras, 1893). Kaczmarczyk, Zdislaw, Kolonizacja niemiecka na wschod od Odry [German colonization east of the Oder] (Pozna´n: Wydawnictwo Instytutu Zachodniego, 1945). L´ang, Ludwig, Hundert Jahre Zollpolitik (Vienna & Leipzig: K.u.k. HofBuchdruckerei und Hof-Verlags-Buchhandlung, 1906). Lebovics, Herman, ‘ ‘‘Agrarians’’ versus ‘‘Industrializers’’: Social Conservative Resistance to Industrialism and Capitalism in Late Nineteenth Century Germany’, International Review of Social History, 12 (1967), 31–65. Lesigang, Wilhelm von, ‘Grundsteuer’, in Johannes Conrad, ed., Handw¨orterbuch f¨ur Staatswissenschaft, 2nd edn, vol. iv (Jena: Gustav Fischer, 1900), 885–917. McCloskey, Deirdre, ‘The Trouble with Mathematics and Statistics in Economics’, History of Economic Ideas, 13/3 (2005), 85–102. Magocsi, Paul Robert, Historical Atlas of Central Europe, rev., expanded edn (Seattle: University of Washington Press, 2002).

List of Works Cited

263

Meitzen, August, Der Boden und die landwirtschaftlichen Verh¨altnisse des Preussischen Staates nach dem Gebietsumfange vor 1866, 9 vols (Berlin: Paul Parey, 1874–1906). and Friedrich Grossmann, Der Boden und die landwirtschaftlichen Verh¨altnisse des Preussischen Staates nach dem Gebietsumfange vor 1866, vi (Berlin: Paul Parey, 1901). Moeller, Robert G., ed., Peasants and Lords in Modern Germany: Recent Studies in Agricultural History (Boston: Allen & Unwin, 1986). M¨uller, Eduard, Der Grossgrundbesitz in der Provinz Sachsen. Eine agrarstatistische Untersuchung, Sammlung national¨okomischer und statistischer Abhandlungen des staatswissenschaftlichen Seminars zu Halle a. d. S., 67 (Jena: Gustav Fischer, 1912). Nabert, Thomas, Der Grossgrundbesitz in der preussischen Provinz Sachsen, 1913–1933. Soziale Struktur, o¨konomische Position und politische Rolle (Cologne/Weimar/Vienna: B¨ohlau Verlag, 1992). ¨ Preradovich, Nikolaus von, Die F¨uhrungsschichten in Osterreich und Preußen (1804–1918), Ver¨offentlichungen des Instituts f¨ur Europ¨aische Geschichte Mainz, 11 (Wiesbaden: Franz Steiner Verlag, 1955). Puhle, Hans-J¨urgen, ‘Lords and Peasants in the Kaiserreich’, in Moeller, ed., Peasants and Lords, 81–109. Richter, Hanns-Joachim, Die Entwicklung des Grossgrundbesitzes in Schlesien seit 1891 (Breslau: Wilhelm Gottlob Korn, 1938). Rosenberg, Hans, ‘Die ‘‘Demokratisierung’’ der Rittergutsbesitzerklasse’, in Wilhelm Berges and Carl Hinrichs, eds, Zur Geschichte und Problematik der Demokratie. Festgabe f¨ur Hans Herzfeld anl¨aßlich seines 65. Geburtstages am 22. Juni 1957 (Berlin: Duncker & Humblot, 1958), 459–86. ¨ Schiller, Ren´e, Vom Rittergut zum Grossgrundbesitz. Okonomische und soziale Transformationsprozesse der l¨andlichen Eliten in Brandenburg im 19. Jahrhundert, Elitenwandel in der Moderne, 3 (Berlin: Akademie-Verlag, 2003). Schissler, Hanna, ‘The Junkers: Notes on the Social and Historical Signiﬁcance of the Agrarian Elite in Prussia’, in Moeller, ed., Peasants and Lords, 25–51. Sering, Max, Die innere Kolonisation im o¨stlichen Deutschland, Schriften des Vereins fur Socialpolitik, 56 (Leipzig: Duncker & Humblot, 1893). Skibicki, Klemens, Industrie im Oberschlesischen F¨urstentum Pless im 18. und 19. ¨ Jahrhundert. Zur o¨konomischen Logik des Ubergangs vom feudalen Magnatenwirtschaftsbetrieb zum modernen Industrieunternehmen, Regionale Industrialisierung, 2 (Stuttgart: Franz Steiner Verlag, 2002). Stern, Fritz, The Politics of Cultural Despair: A Study in the Rise of Germanic Ideology (Berkeley: University of California Press, 1961). Tims, Richard Wonser, Germanizing Prussian Poland: The H-K-T Society and the Struggle for the Eastern Marches in the German Empire, 1894–1919, Studies in

264

List of Works Cited

History, Economics and Public Law, 487 (New York and London: Columbia University Press and P. S. King & Son, 1941). Treue, Wilhelm, Wirtschaftszust¨ande und Wirtschaftspolitik in Preußen, 1815–1825 (Stuttgart and Berlin: Verlag W. Kohlhammer, 1937). Wygodzinski, Willy, Die Besteuerung des landwirtschaftlichen Grundbesitzes in Preussen. Eine kritische Studie (Jena: Gustav Fischer, 1906).

Index Abzac-Hoym, Countess d’, see Saurma-Jeltsch, Marie-Antoinette von, Countess (n´ee Countess d’Abzac-Hoym) address books 6, 10, 16, 22–3 administrative regions (Regierungsbezirke) x–xi, 63 industrial establishments in 204–6 large properties by 63–5 quality of land in 212–13 Adressbuch des Grundbesitzes in — — (F. B¨urde) 5 n. 24 agrarianism 2 agricultural area 14–15, 33–4 agricultural censuses 6, 39 adjustment of 1882 data 30–6 and area of properties 14–15 incomparabilities in 23–7 ownership and 19–20, 27–30, 236 property data 236 agriculture: boom 13 crisis in 12–13 intensiﬁcation of 12, 89 allodial properties 23, 156 n. 13 Alt-Latzig (estate) 105 n. 21 Alt Mohrau (estate) 128 Alvensleben, Albert von 107 Alvensleben, Albrecht (later Count) von 107 Alvensleben, Martha von (n´ee von Sch¨onborn) 107 Alvensleben-Sch¨onborn, Marie Mathilde, Countess von 107 animal husbandry 56 and agricultural crisis 12 extent of 58 on knight’s estates 169–70 land types for 222 in land-use proﬁles 57, 210, 226 as speciality 228–9, 239 arable agriculture: in East Prussia 231–2 on knight’s estates 153, 169–70, 172 land types for 59, 214, 222 in land-use proﬁles 56, 57, 226 location of 210 non-noble persons and 153 in Posen 58, 230 provincial specializations in 238–9 in Saxony 58, 230 in Silesia 58, 230

arable land 56–7 entailed properties and 11 and industrial establishments 201, 241 and land-use proﬁles 225–6, 226 area of property 8 agricultural area vs. 14–15, 33–4 distribution of large properties by 69–73 ﬁfty largest landowners by 129–36, 238 increases in 91 owned by aristocracy 91 size of properties by 69–73, 82–3 ten largest owners by 147–8 twenty largest owners by 125 and type of land 43, 44 aristocracy: disparities in landed wealth among 92 extent of holdings 237 among ﬁfty largest landowners 144 increases in area owned per owner 91 and industrial establishments 192 as landowning class 2–3, 88 and large properties 81, 89 size of property owned 91 tax value of properties owned by 88 see also nobility; royalty Arnim, Heinrich Leopold von 76 n. 21 Arnim, Traugott Hermann, Count von 96 n. 12, 146, 148, 149 Astrawischken (estate) 52 Austro-Hungarian Empire 1, 2 n. 7 Bach, Karl 114 Baden, Friedrich, Grand Duke of 97 Baden, Grand Dukes of 97 Baitkowen (estate) 179 n. 8 Baltic provinces: industrial establishments 175–6, 188–9 land endowment 239 land use 54, 57, 58, 59 large properties in 68–9 mixed land use in 59 see also East Prussia; Pomerania; West Prussia banks and business ﬁrms: and industrial establishments 192, 193, 202, 207, 241 landownership 87, 118–19 among top ten owners 238 barons/baronesses: as landowners 103–6, 115

266

Index

barons/baronesses: (cont.) tax value of land 92 see also lower nobility Baruth (manor) 99 Bateman, John 4 Batewitz (estate) 224–5, 228 Bavaria, Marie, Queen of 96–7 Beck, Lajos 10 Bentheim-Tecklenburg, Anna, Princess zu (n´ee Princess Reuss j.L.) 101 Berghoff, Hartmut 163 n. 25 Berlin 1 Landbank Corporation, see Landbank Corporation of Berlin among largest landowners 136, 143 Biron von Curland, Antoinette, Princess see Lazareff-Hoym, Antoinette von, Countess (n´ee Princess Biron von Curland) Biron von Curland, Calixt, Prince 101, 195 Biron von Curland, Gustav, Prince 100 Bismarck, Herbert von 99 Bismarck, Otto Prince von 99 Borzejewo (estate) 100 bourgeoisie, see non-noble persons Brandenburg: barons/baronesses as landowners 104, 105 ﬁfty largest landowners 123, 146 First Estate membership in 155 forest land 51, 230 forestry 58, 210 in Handbooks 50–1 Hohenzollern properties 96 industrial establishments 176, 189, 195, 204, 206, 207 knight’s estates 172 land use 57, 58 large properties 5, 60, 61, 64, 66, 76, 83 meadow land 218–19 mixed land use 231 pasture land 219–20 ploughland 217 princes/dukes as landholders 99 Prussian State landownership 50–1, 85, 116 religious bodies’ holdings 119 Schiller on 17–19 taxable value of land 215 Brandenburg Cathedral 119 Brentano, Lujo 12 breweries 175, 181, 182 brickworks 176, 179 n. 8, 182, 191–2, 240, 241 Bromberg (Bydgoszcz) (administrative region) 63, 100, 105 Br¨unneck, Roland von 103, 108 n. 24 Brunswick: Ducal Chamber 143

duchy of 65, 144 ﬁscus of 143 Brunswick, Wilhelm, Duke of 95, 97, 148, 237 Brunswick-Wolfenb¨uttel, Duke of 148–9 Buchsteiner, Ilona 5 n. 25, 21 Grossgrundbesitz in Pommern 1871–1914 10, 14–17 Habilitationsschrift 14 Bunzlau, City of 136 business ﬁrms, see banks and business ﬁrms cadastral keys 242, 244 Camenz (estate) 95 Capustigall (estate) 160 censuses of agriculture, see agricultural censuses cities: holdings 117–18 among largest landowners 136, 143 see also names of individual cities Closterfelde (estate) 168 cluster analysis: of knight’s estates 152 for land-use proﬁles 56–9, 168–70 selection of properties for 222–4 see also land-use proﬁles communal groups/communities: holdings 117–19 and industrial establishments 192–3, 202, 207 as owners of knight’s estates 161–2 owning large properties 79 Conrad, Johannes 5 n. 25, 6–9, 146 and agricultural censuses 6, 19 classiﬁcation of high nobility 102 and farm vs. property 21 and Handbooks 6, 8, 39, 40 latifundia 7, 81 and Prince of Pless’s holdings 45 n. 13 and property vs. total area owned 69, 71 and Prussian State ownership 8–9, 73 schema for ownership data 8 students/followers of 9–10 twenty largest landowners 123, 124–9 convents 87, 119, 143 Corn Laws (Britain) 1, 156 corporate owners 126 see also banks and business ﬁrms; cities; communal groups/communities; religious bodies counts/countesses: elevation to princes 86 as landowners 89–90, 102–3 landholdings of 86, 114–15

Index Czartoryska, Hedwig, see Dzieduszycka, Hedwig, Countess (later Princess Czartoryska) Czartoryski, Witold, Prince 100 Czartoryski, Wladislaus, Prince 100 dairy plants (Molkereien) 176 Danzig (administrative region) 64, 241 Demmin (district) 145 Deutsch-Lissa (estate) 101 Deutscher Ostmarken-Verein, see German Eastern Marches Society Diestel, A. 94 Dillwitz, Sigrid 4 n. 17 Dino-Talleyrand, Duke of (also Duke zu Sagan) 99 distilleries: distribution of 191–2 employment in 184 Handbooks on 179 n. 8 monasteries and 176 numbers of 182, 183, 187–8 Prussian State and 176 as rural-based 175, 181, 207–8 district assemblies (Kreistage) 153, 154 districts (Kreise): forest land in 220–1 in Handbooks 38 land-use proﬁles of 56–9 large properties by 64–5 meadow land in 217–19 pasture land in 219–20 ploughland in 215–17 tax assessment in 214–15 Dohna family 160 Dohna-Schlobitten, Count zu 160 Dohna-Schlobitten, Richard Wilhelm Ludwig, Prince (formerly Count) zu 98–9 Dohna-Schlodien, Adolf, Count zu 99 Dohna-Waldburg, Eberhard zu 160 Dohna-Waldburg, Richard Friedrich, Count zu 98–9 Drogelwitz (estate) 101 dukes: holdings by 86, 89–90, 98–102 tax value of land owned 92 Dyhernfurth (manor) 101 Dzialyn (manor) 105 Dzieduszycka, Hedwig, Countess (later Princess Czartoryska) 100 Dzieduszycki, Wladimir, Count 100 East Prussia: arable agriculture 231–2 barons/baronesses as landowners 104

267

Conrad and 7 ﬁfty largest landowners 146 forest land 47–8, 52–3, 220–1 forestry 210 Generalkommission 164 in Handbooks 46–7, 52–3 Hohenzollern properties 93–4 industrial establishments 175–6, 180, 187–8, 204, 206, 207 knight’s estates 152, 164, 166, 171, 174 land use 57 large properties 47, 60, 61, 63, 66 meadow land 218 mixed land use 210, 231 non-noble landownership 85, 109–11 pasture land 219, 220 ploughland 216, 217 princes/dukes as landholders 98–9 Prussian State landownership 9, 47, 52–3, 116 quality of land 210, 222, 238 royal properties 97 taxable value of land 215 economic activities, see land use Eddie brothers (Nicolai and William) 21 Ellerholz, Paul, see Handbuch des Grundbesitzes im Deutschen Reiche (Handbooks) employment 240 in industrial establishments 182–4 entailed properties (Fideikommisse) 7, 11–12, 23, 32–3 Esterh´azy, Prince 237 n. 2 Eulenburg-Hertefels, Philipp Count (later Prince) zu 99 Eyssehnen (estate) 98 n. 15 Falkenhausen, (later Friendenthal-Falkenhausen) Ernst, Baron von 199 farm operations, properties vs. 21–3, 31–6 Fideikommisse, see entailed properties ﬁefs (Lehng¨uter) 23 ﬁfty largest owners 123–4, 129–46, 237–8 by area owned 130–5 by tax assessments of properties 136–42 location of estates 144–5 socioeconomic makeup 144 Finck von Finckenstein, Count 103 First Estate (Ritterschaft) 4, 5, 153, 154 Fischbach (estate) 96 ﬂour mills 177, 240

268

Index

forest land 56–7, 214, 220–1 in arable–forestry cluster 226 in Brandenburg 51, 230 in censuses 24–6, 30–1 in East Prussia 47–8, 52–3, 220–1 entailed properties and 11 and forestry 222 in Handbooks 48–53 Hohenzollern family and 91–2 and industrial establishments 186 land tax net yield for 220–1 on large estates 26 in mixed-use cluster 225–6 nobility ownership 89 non-noble ownership 89 in Pomerania 48–9 Prussian State ownership 47–8, 122 quality of 211, 220–1 royal ownership 88–9 in Saxony 220, 221, 230–1 in Silesia 101, 221, 230 tax assessment on 248 tax value of 238 forestry 56, 57, 59, 214, 222 in Brandenburg 58, 210 in East Prussia 210 high nobility and 232–5 knight’s estates and 152, 170 non-noble persons and 152 provincial specialization in 239 Prussian State and 233–5 in Silesia 58, 210 woodland and 222 Frank, Alexandra 120 n. 39 Frankfurt/Oder (administrative region) 64 Frantz, Adolf: General-Register 5 Franzburg (district) 65, 145 Friedenthal, Carl 195, 199 Friedenthal, Elisabeth 199 Friedenthal-Falkenhausen, Ernst Carl von 199 Friedenthal (manor) 199 Friedrich III, Kaiser 99 Fritzen (forest) 52 Garden (estate) 103 garden land: and arable agriculture 222 quality of 211 Garlepow (estate) 42 Garthe (estate) 148 Gerdauen (district) 52 German Eastern Marches Society (Deutscher Ostmarken-Verein) 136 n. 12, 195 German Historical School 4 Gerschenkron, Alexander 176–7 Godulla, Karl 193 n. 19

Goldap (district) 52, 53 Goltz, Theodor, Baron vonder Vorlesungen u¨ ber Agrarwesen und Agrarpolitik 18 n. 86 Goray (estate) 100 G¨orlitz, City of 81 n. 30, 126, 136, 171, 238 Gothen (estate) 99 Grabowo (estate) 100 grain 177 mills 176, 182, 183, 184–5, 191–2 prices 12, 156, 166–7 Greifswald: City 143 University of 87, 119, 136, 143, 145, 146, 149, 238 Grosskmehlen (estate) 102 Grundsteuerveranlagung, see land tax cadastre (1861–1864) Gryzik, Johanna, see also Schaffgotsch, Johanna, Countess von 193 n. 19 GSRE, see land tax net yield (Grundsteuerreinertrag (GSRE)) Guben, City of 81 n. 30 Gumbinnen (administrative region) 52, 63, 65 Guteborn (manor) 101 Guttentag (manor) 97 H¨abich, Theodor 81 n. 29 Hammerstein (estate) 81 n. 30, 172 Handbook of Landholding in the German Empire (Handbuch des Grundbesitzes im Deutschen Reiche) 5, 6, 37–40 areas of properties in 108 for Brandenburg 51, 99 Buchsteiner and 14, 15, 16–17 censuses vs. 31–6 cleaning of data from 41–3 Conrad and 6, 8 deﬁnition of property in 69 ‘disappeared’ properties and 108 n. 23 for East Prussia 46–7, 52–3, 99, 104, 160, 231–2 forest land in 48–53 Hohenzollern properties in 94 incomplete coverage in 46–8 and industrial establishments 177, 179, 180, 182, 185, 190, 208–9, 240 knight’s estates and 156 n. 13, 157–8, 161, 163–4, 165 manors and 72, 127–8, 231–2 moor in 222 n. 14, 244 M¨uller and 9 names of owners in 41–3, 146 names of properties in 108 Netherlands royal family in 95

Index numbers of owners and 71 for Pomerania 40, 48–9 for Posen 40, 105, 117 principality of Pless in 44–5 problems with individual properties in 43–5 property coverage in 39–40, 110, 117, 119 provincial distribution of large properties in 59–60 Prussian State properties in 46, 47–53 rank of owners in 89 n. 4, 111, 119 rank of persons in 89 n. 4 for Saxony 40 for Silesia 40 smaller properties and 71, 108, 117 ‘snapshots’ from 39 types of land in 248–9 usability of data in 20–1 for West Prussia 106, 117 Handbuch des Grundbesitzes im Deutschen Reiches 5, 6, 37–8, see Handbook of Landholding in the German Empire (Handbuch des Grundbesitzes im Deutschen Reiche) Hansemann, Ferdinand von 136 n. 12 Harnisch, Hartmut 111 Hasen (estate) 99 Hatzfeld, Prince 128 Heinrichau (manor) 128 Heinrichsdorf (estate) 76 n. 21 Henckel von Donnersmarck, Guido, Count (later Prince) 100, 126, 149 Herzog, Mr 94 Hess, Klaus 14 n. 68, 19, 20 n. 94, 73, 214, 227–8, 232 Junker und b¨urgerliche Grossgrundbesitzer im Kaiserreich 10–13 Hesse, Elisabeth, Princess, Widow of Grand Duke of 96–7 Hesse, Ernst Ludwig IV, Grand Duke of 97 Heyden, J¨urgen von 145 Heyme, Lt 101 higher nobility: extent of holdings 237 among ﬁfty largest landowners 123, 136, 143, 144 and forestry 232–5 and industrial establishments 186–7 ownership of large properties 237 among ten largest landholders 124, 149, 238 Hirsch, Alfred 10 HKT (‘Hakatisten’ or ‘Hakata’ ), see German Eastern Marches Society. See also Hanseman, Ferdinand von; Kennemann, Hermann; Tiedemann, Heinrich von’

269

Hochberg, Wilhelm, Imperial Count von 100, 146 Hoeltzel, Margarete (n´ee Krieger) 199 Hohendorf (manor) 145 Hohenlohe-Ingelﬁngen, Friedrich Wilhelm, Prince zu 102 Hohenlohe-Oehringen, Christian Kraft, Prince zu 101, 148 Hohenlohe-Oehringen, Hugo, Prince zu 148 Hohenlohe-Schillingsf¨urst, Prince 100 Hohenzollern family 40, 86, 89, 91–2, 93–4, 96 Hohenzollern-Sigmaringen, (Carl) Anton, Prince von 126, 128, 146, 195 hospitals: holdings by 119 landownership by 87 Hungary: emancipation of peasants in 232 landed property in 10 ownership in 167–8 industrial establishments (Industrie-Anlagen) 38, 240–2 administrative regions and 241 arable land and 201, 241 banks/business ﬁrms and 192, 202, 207, 241 in Brandenburg 176, 189, 195, 204, 206, 207 classes of owners of 240–1 communal groups and 192–3, 202, 207 concentration of ownership 241 in East Prussia 175–6, 180, 187–8, 204, 206, 207 employment in 182–4 knight’s estates and 201 on larger estates 185–200, 201, 207–8 location of 200–7 military ofﬁcers and 201, 202 nobility and 207 non-noble persons and 202, 207 numbers of 193–200, 240–1 owner classes and 241 ownership of 190–3 in Pomerania 175–6, 189, 190, 192, 195, 204, 206, 207 in Posen 176, 189, 195, 204, 206, 207 Prussian State and 176, 192, 202, 208 religious bodies and 192–3, 202, 207 royalty and 207 in Saxony 176, 179, 187, 189, 195, 204, 206, 207, 208 in Silesia 175, 176, 179, 180, 187, 189, 192, 195, 204, 207, 208 size of property and 190, 241

270

Index

industrial establishments (Industrie-Anlagen) (cont.) in West Prussia 175–6, 179, 189, 190, 202–3, 204, 206, 207 industrialization 2, 176–7 and ownership of large properties 237 J. G. Boltze (ﬁrm) 143, 149 Jackunowen (estate) 179 n. 8 Jaff´e Brothers 81 n. 30, 172 Jahrb¨ucher f¨ur National¨okonomie und Statistik 6–7 Johannisburg (district) 65, 238 Junkers 3, 80, 81, 176 juridical/legal persons 23 among ﬁfty largest landowners 123–4, 136, 143 landownership by 87 see also non-physical juridical/legal persons; physical persons Kaiserswalde (estate) 100 Kallischken (estate) 179 n. 8 Karmunkau (forestry operation) 101 Kennemann, Hermann 136, 176, 195 Keyserling, Count von 103 Klar, Paul Alois: B¨ohmens grosser Grundbesitz, wie dieser in der k¨onigl. Landtafel inneliegt 4–5 Klein-Astrawischken (estate) 52 Klein-Droniowitz (estate) 102 Klot-Trautvetter, Magnus, Count von 145, 224 knight’s estates (Ritterg¨uter) 4, 153–5 animal husbandry on 169–70 arable agriculture on 153, 169–70, 172 arable–forest cluster 171 area of 157–8 in Brandenburg 172 coverage of 163 in East Prussia 164, 166, 171, 174 embourgeoisement of ownership 162 forest–arable cluster 171 forestry on 152, 170 grain prices and 156, 166–7 group ownership of 161–2 and industrial establishments 201 land-use proﬁles 168–70 limitation of life of designation 154 mixed farming on 169–70 nobility as owners of 152, 153, 158, 159, 161, 162, 166, 167, 170, 171, 172–3, 174, 238 non-noble owners 152, 161, 166, 167, 170, 171–2, 238 numbers of 163–6

‘other’ owners of 153, 161–2, 171, 172 by owner categories 159, 166 ownership changes 161, 163–74 parcellization of 154, 159, 163, 164, 166 political functions of 152, 153–4, 162–3, 166 in Pomerania 164, 165, 171, 172 in Posen 164, 165, 166 prices of 173, 242 quality of land on 243 in Saxony 163–4 Schiller on 17, 18–19 in Silesia 163–4, 165, 171 size of 160–1, 164–6, 167, 171–2, 174 specialization in ownership of 152–3, 167–8, 174 taxation of 173, 242, 249 in West Prussia 164, 165, 166, 172, 174 K¨ollmische G¨uter, see ‘free peasant properties’ under the heading ‘Prussia (former province of )’ K¨onigliche Ansiedlungskommission, see Royal Settlement Commission K¨onigsberg Academic Senate in 159, 254 K¨onigsberg (district) 52, 110–11, 152, 160 K¨onigsberg (administrative region) 52, 204, 206, 216 K¨onigsmarck, Hans Karl Albert, Count von 103 Korn, Wilhelm Gottlob: Schlesische G¨uteradressb¨ucher 6, 9–10 Krieg, Karl 96, 101 Krieger, Ernst 199 Kreise x–xi see also districts (Kreise) Keristage, see district assemblies Krucz (estate) 100 land endowment, see types of land land tax cadastre (1861–1864) (Grundsteuerveranlagung) 55–6, 59, 62, 210, 238–9, 246 land tax net yield (Grundsteuerreinertrag (GSRE)) 38, 44 forest land 220–1 in Handbooks 5, 39, 44 knight’s estates 173 meadow land 217–19 pasture land 219–20 ploughland 215–17 and quality of land 211 land use: in Brandenburg 57, 58 change in 227–8 in East Prussia 57

Index by nobility 240 by non-noble persons 239–40 in Pomerania 57, 58 in Posen 57, 58 by Prussian State 240 in Saxony 57, 58 in Silesia 57, 58 specialization of 228 tax values and 240 in West Prussia 57 land-use proﬁles 169, 239 arable and other 227 arable–forestry 226 arable–husbandry 226 arable–husbandry–forestry 225–6 by area 210, 229–30 calculation of 224–5 changes in size of 228 classiﬁcation by 213–14, 222–35 clusters 210, 225–7 of districts 56–9 of knight’s estates 168–70 in large properties 239 mostly arable 225 mostly forestry 226 numbers of properties in 227–9 ‘other’ land 59, 227 ownership specialization in 232–6 property groupings by 210 regional distribution of 230–2, 239–40 rural land and 213–14 and specialization in ownership 211 tax assessment and 211–12 tax values and 210, 229–30, 239 land yield index 210, 212 Landbank Corporation of Berlin 104, 119, 136, 193, 199–200 Landkreise, see Kreise Landtafel (Bohemia) 4 L´ang, Ludwig 1 n. 3 Langensalza (district) 218 large estates/properties 3 absolute amounts of land held 83 aristocracy and 81, 89 in Baltic provinces 68–9 in Bohemia 4 bourgeoisie and 89 in Brandenburg 5, 60, 61, 64, 66, 76, 83 classiﬁcation by land use 239 communities/communal groups owning 79 coverage by 236–7 distilleries on 187–8 in East Prussia 47, 60, 61, 63, 66 forest land on 26 growth of smaller at expense of 35–6 industrial establishments on 185–200, 201, 207–8, 240–2

271

industrialization and ownership of 237 knight’s estates as 164–5 land-use clusters and 239 lesser nobility and 237 by logarithmic distributions 76–8 noble ownership 79–80, 237 non-noble ownership 80–1, 81, 86, 162, 237, 238 non-physical legal persons owning 237 ownership of more than one 82–3 percentage of land encompassed by 59 percentage of taxable value of land 59 physical and juridical persons owning 87–8 physical persons owning 78–9, 88, 236–7 in Pomerania 54, 60, 61, 63, 65, 66, 83 in Posen 54, 60, 61, 63, 65, 66, 70 table, 83 provincial distribution 83–5 in Prussia 4–5 Prussian State ownership 71–2, 73, 76, 78–9, 237 publication of data on 4–5 regional distribution of 59–67 royalty ownership of 79–80, 81, 237 in Saxony 54, 60, 61–2, 63, 64, 65, 66, 83 Schiller on 17–19 in Silesia 54, 55, 59–60, 61, 64, 65, 66, 70 table, 76, 83–4, 144, 237 tax assessment on 242–9 tax values 55, 61–6, 73–6, 77–8 by total area of 67–9 by total area owned 69–73 by types of owners 55, 83–4, 232–6 in West Prussia 60, 63, 64, 66 Lasisk (manor) 101–2 latifundia 7, 81 Lazareff-Hoym, Antoinette, Countess von (n´ee Princess Biron von Curland) 101 leasing/renting, of land 21 n. 96, 22, 27–30 legal persons, see juridical/legal persons Lehng¨uter, see ﬁefs lesser nobility, see lower nobility Liebenberg (estate) 99 Liechtenstein, Prince of 96 Lindner, Eduard 113–14 Lippe, Rudolph, Prince zur 101 Louisenhof (estate) 94 lower nobility: extent of holdings 237 among ﬁfty largest landowners 123, 136, 143, 144 landholdings of 86, 120 and large properties 237 mixed land use by 233–5 Ludwigsdorf (estate) 224–5 Lynar, Ernst, Prince zu 102

272

Index

Magdeburg (administrative region) 55 n. 2, 64, 65, 215 manors (Herrschaften) 23, 32–3, 44–5, 72–3 Marienburg (district) 64 Marienfelde (estate) 99 Marienwerder (administrative region) 99, 103, 104, 215 Matrikeln 4–5, 153, 154, 155, 158–61 meadow land 56–7, 217–19 and animal husbandry 222 in East Prussia 218 extent of 228–9 land tax net yield for 217–19 in mixed-use cluster 225–6 in Pomerania 218 in Posen 218, 219 quality of 211, 217–19 in Saxony 218–19 in Silesia 218–19 tax assessment on 211–12, 248 tax value of 238 in West Prussia 218, 219 Mecklenburg 2 n. 9, 5 Meitzen, August 6–7, 52 Merseburg (administrative region) 63, 64, 76, 102 Merseburg (district) 220 Mielczynski, Count von 103 Mielzynska, Constantia, Countess 100 military ofﬁcers: by civil status 114–15 and female landowners 111–12 and industrial establishments 201, 202 as landowners 86–7, 111–16 noble 86–7, 111, 112, 115–16 non-noble 87, 111, 112–14, 115 Mittman, G. 99 mixed land use: in Baltic provinces 59 in Brandenburg 231 in East Prussia 210, 231 high nobility and 233–5 on knight’s estates 169–70 lower nobility and 233–5 by non-noble landowners 232–5 in Pomerania 210, 231 in Posen 231 provincial specialization in 239 in Silesia 210, 231 monasteries: and industrial establishments 176, 192 landholdings 87, 119 moor 244, 246, 248–9 as catch-all category 38, 222 n. 14 see also unproductive land; untaxed land moorland: moor vs. 38

quality of 211 M¨uhlhausen (district) 65 M¨uller, Eduard: Der Grossgrundbesitz in der Provinz Sachsen 4 n. 18, 9, 13–14 M¨unsterberg (district) 128 Muskau (manor) 95–6 Myczielska, Countess von 108 n. 24 Nabert, Thomas: Der Grossgrundbesitz in der preussichen Provinz Sachsen 10, 13–14, 22, 47 n. 18 Naehrich, Paul 114 Nassawen (estate) 53 Netherlands, Frederik, Prince of the 96, 149 Netherlands, Marianne, Princess of the 95, 128 Netherlands, royal family of 95–6 Nette, Bernhard 114 Neuzelle Monastery 119, 136 Niekammers landwirtschaftliche Adressb¨ucher 5–6, 16 nobility: difference in wealth within 121–2 forest land owned by 89 identiﬁcation of 42 and industrial establishments 207 and knight’s estates 152, 153, 158, 159, 162, 166, 167, 170–1, 172–3, 174, 238 landownership by 86, 88, 120 land use 240 large properties owned by 79–80 numbers of properties owned 161 ownership in more than one province 42 in Pomerania 83 in Posen 83 Prussian 1–2 as representatives in provincial assemblies 155 in Silesia 83–4 and size of property 81, 82, 161 tax assessment and 248, 249 titled 89–90 untitled 89–90, 106–8, 115 in West Prussia 83 see also higher nobility; lower nobility; and names of individual ranks non-noble persons 86, 109–11 and arable agriculture 153 and arable cluster 232–5 extent of holdings 237 and First Estate properties 156 forest land 89 and forestry 152

Index identiﬁcation of 43 and industrial establishments 192, 193, 195, 199, 202, 207 and knight’s estates 152–3, 166, 167, 170, 171–2, 238 landownership by 120–1 land-use specializations 239–40 large properties owned 80–1, 86, 89, 123, 136, 143, 162, 237, 238 mixed land use 232–5 ownership in East Prussia 85 as representatives in provincial assemblies 154–5 size of property owned 91, 161 tax values of properties 88, 238 in West Prussia 85 non-physical juridical/legal persons: as landowners 117–19 ownership of large properties 237 nouveaux riches 237 Oels (manor) 95, 97 Oginska, Princess 98 n. 15 Oldenburg, Nicolaus Peter, Grand Duke 97 Oldenburg family 97 Ortelsburg (district) 216, 218, 220–1 owner categories 55, 81–2, 83–4, 119–22 of knight’s estates 159, 166 and size of properties 161 and types of properties 167–8 owners: as Besitzer 22–3 ﬁfty top, see ﬁfty largest owners identiﬁcation of 42–3, 120 incomes 20 increases in area owned per 91 of more than one property 82–3 number of 120–1 numbers of properties per 126 recording of names of 41–3 ten largest, see ten largest owners twenty largest, see twenty largest owners women as 3, 42, 111–12 see also juridical/legal persons; non-physical juridical/legal persons; physical persons ownership 3 agricultural censuses and 23–4, 27–30 bourgeois revolution in 152 concentration of 36 concepts of, vs. production-unit concepts 20 directories 4–5 embourgeoisement of 152, 159, 162, 173–4

273 in more than one province 42

Parey series, see Handbuch des Grundbesitzes im Deutschen Reiche (Handbooks) parish churches 119 pasture land 56–7, 219–20 and animal husbandry 222 in Brandenburg 219–20 conversion of moorland to 249 in East Prussia 219, 220 extent of 228–9 land tax net yield for 219–20 in mixed-use cluster 225–6 in Pomerania 219, 220 in Posen 219, 220 quality of 211, 219–20 in Saxony 219, 220 in Silesia 219, 220 tax assessment on 248 in West Prussia 219 peasants 2, 3 emancipation of 164, 232 free, properties of (k¨ollmische G¨uter) 153 n. 1, 155 as landowners 18, 22, 27, 158 Petershain (estate) 172 Pfeilings (estate) 99 physical persons: extent of holdings 237 and industrial establishments 192, 194 table, 240–1 landownership by 86 large properties owned by 78–9, 88, 236–7 Pless, Hans Heinrich XI, Prince of 100, 146 Pless, Hans Heinrich XV, Prince of 100, 146 Pless, Prince of 10, 44, 123, 126, 148, 237 Pless (manor) 44–5 Pless (principality) 44–5 Ploewken (village) 159 Plottnitz (estate) 95 ploughland 214, 215–17 and arable agriculture 222 in Brandenburg 217 in East Prussia 216, 217 land tax net yield for 215–17 in Pomerania 217 in Posen 216, 217 quality of 211, 215–17 in Saxony 216, 217 in Silesia 216–17 tax assessment on 211–12, 248 tax value of 238 in West Prussia 216, 217 Pomerania: barons/baronesses as landowners 105 Buchsteiner on 14, 16–17

274

Index

Pomerania: (cont.) communally owned properties 118 Conrad and 7 counts/countesses as landowners 102 ﬁfty largest landowners 123–4 forest land 48–9 in Handbooks 40, 48–9 Hohenzollern properties 94 industrial establishments 175–6, 189, 190, 192, 195, 204, 206, 207 knight’s estates 164, 165, 171, 172 land use 57, 58 large properties in 54, 60, 61, 63, 65, 66, 83, 144–6 meadow land 218 mixed land use 210, 231 nobility as landowners 83 ownership relations in 28–30 pasture land 219, 220 ploughland 217 princes/dukes as landholders 99–100 Prussian State landownership 9, 48–9, 116 religious bodies’ holdings 119 taxable value of land 215 Pomeranian Settlement Society 136 Posen: administrative division changes in 41 arable agriculture 58, 230 barons/baronesses as landowners 104, 105–6 communally owned properties in 118 Conrad and 7 ﬁfty largest landowners 146 First Estate membership in 155 in Handbooks 40 Hohenzollern properties 96 industrial establishments 176, 189, 195, 204, 206, 207 knight’s estates 164, 165, 166 land use 57, 58 large properties 54, 60, 61, 63, 65, 66, 70 table, 83 meadow land 218, 219 mixed land use 231 nobility as landowners 83 non-noble landownership 109 pasture land 219, 220 ploughland 216, 217 princes/dukes as landholders 99–100 Prussian State landholdings 116, 117 quality of land 222 religious bodies’ holdings 119 royal properties 97 taxable value of land 215 Potocki, Oswald, Count von 108 n. 24 Potsdam (administrative region) 64

Preussen, Heinrich, Prince von, see Prussia, Heinrich, Prince of Preussisch-Eylau (district) 52, 156 n. 13 princes: elevation of counts to 86 as owners 86, 89–90, 98–102 mediatized (Standesherren) 155 tax value of land owned 92 Prinz, Adalbert 75–6 produce prices 12 and quality of land 211 and value of land 62 properties: deﬁnitions of 32–3, 81–2 farming enterprises vs. 21–3, 24, 27, 31–6 renting of 27–30 subunits vs. principal properties 41 provincial assemblies (Landtage) 4, 152, 153, 154 Prussia, Albrecht, Prince of 94–5 Prussia, Friedrich Carl, Prince of 126 Prussia, Heinrich, Prince of 40, 97 Prussia, King of 1, 101, 105 n. 20, 176, 237 Prussia, Wilhelm, Crown Prince of 95, 97 Prussia, Wilhelm I, King of 93–4 Prussia (kingdom): Brandenburg holdings 50–1, 85 East Prussia holdings 9, 47, 52–3 ﬁscus 44, 108, 161 forest land owned by 47–8, 122 forestry and 233–5 Handbooks and 46, 47–53 history of 1–3 and industrial establishments 176, 186, 192, 202, 208, 241 landownership 8–9, 86, 87–8, 116–17, 120, 122 land use 240 large estates in 4–5 large properties owned by 71–2, 73, 76, 78–9, 237 as largest single owner 237 and lower-valued land 143 Pomerania holdings 9, 48–9 Posen holdings 117 Saxony holdings 65, 85 Silesia holdings 9 size of 55 size of properties owned by 81 and top ﬁfty landowners 238 West Prussia holdings 9, 99, 116–17 Prussia (former province of) 157–8 First Estate membership in 155 ‘free peasant properties’ (k¨ollmische G¨uter) 155, 157 , 158, 162 Matrikel for 158–61

Index ‘privileged estate complexes’ (bevorrechtete G¨uter-Complexe) 157 Puhle, Hans J¨urgen 1 n. 1 Putbus, Prince zu 101 quality, of land: in administrative regions 212–13 distribution of 211–22, 214 in East Prussia 210, 222, 238 forest 211, 220–1 garden 211 gradients in 210, 212, 221–2, 238–9 on knight’s estates 243 land yield index vs. 212 n. 4 meadow 211, 217–19 moorland 211 owned by Prussian State 240 pasture 211, 219–20 ploughland 211, 215–17 in Posen 222 produce prices and 211 in Saxony 221, 238 in Silesia 222 tax yield and 211 types of land and 244–6 water 211 in West Prussia 221 Radenz (manor) 127–8, 148 Radolinski, Hugo von, Count (later Prince Hugo von Radolin) 100 rank, social: changes in 89 n. 4, 99, 102, 106, 108–9 and size of landholding 86 and tax value of property 91–2 Ratibor, Duke of 101 Rauer, Karl Friedrich 155–8, 162, 163–4 Adress-Buch der Rittergutsbesitzer und Ritterg¨uter in den Preussischen Staaten 4–5 Hand-Matrikel der in s¨ammtlichen Kreisen des Preussischen Staates auf Kreis- und Landtagen vertretenen Ritterg¨uter 154 n. 5 Recz (estate) 105 Regierungsbezirke, see administrative regions Rehder, August 81 n. 30, 94, 172 Reinfeld B (estate) 99 religious bodies: and industrial establishments 192–3, 202, 207 landownership by 87 landholdings 119 see also convents; monasteries; parish churches Renard, Johannes, Count 102

275

renting, of land, see leasing/renting, of land Return of Owners of Land (Great Britain) 4, 7 Richter, Hanns-Joachim 9–10, 19 n. 90 Riepenhausen, Asta von (n´ee Imperial Countess von Wyttich und Lochum) 101 Ritterg¨uter, see knight’s estates Ritterschaft, see First Estate Rochow, Gustav von 154 Rohrdeck (estate) 76 n. 21 Rothkirch, Major von 102 Royal Settlement Commission (K¨onigliche Ansiedlungskommission): among largest landowners 136, 241 and numbers of large properties 46, 54, 60, 63, 106, 109, 163 in Posen 54, 63, 106 in West Prussia 104, 105 royalty: Bavarian 96–7 forest land owned by 88–9 and industrial establishments 186–7, 207 landownership by 88, 89–90, 93–8 large properties owned by 79–80, 81, 237 Netherlands 95–6 and size of properties 82 among twenty largest landowners 128 see also names of individual rulers Ruckert, Oscar 101 ruling houses, landholdings by 86 rural-based industry, see industrial establishments Sachsen-Weimar, Sophie, Grand Duchess of 128, 199 Sachsen-Weimar, Wilhelm Ernst, Grand Duke of 199 Sagan, Duke zu, see Dino-Talleyrand, Duke of Salzm¨unde (estate) 149 Saurma-Jeltsch, Marie-Antoinette, Countess von (n´ee Countess d’Abzac-Hoym) 101 sawmills 182, 240 Saxony: arable agriculture in 58, 230 banks/business ﬁrm holdings 118, 119 barons/baronesses as landowners 104, 106 counts/countesses as landowners 102–3 ﬁfty largest landowners 146 First Estate membership in 155 forest land 220, 221, 230–1 in Handbooks 40 industrial establishments 176, 179, 187, 189, 195, 204, 206, 207, 208 knight’s estates 163–4 landownership in 13–14

276

Index

Saxony: (cont.) land use 57, 58 large properties 54, 60, 61–2, 63, 64, 65, 66, 83 meadow land 218–19 non-noble landownership 109 pasture land 219, 220 ploughland 216, 217 princes/dukes as landholders 98, 102 Prussian State landownership 65, 85 quality of land 210, 221, 238 richness of soil 55 n. 2 royal property 96 sugar factories 118 taxable value of land 214–15 untitled nobility as landholders 106, 108 value of land 84 Saxony, Friedrich August, King of 97 Saxony, Friedrich August III, King of 97 n. 14, 101 Saxony, Friedrich Georg Albrecht I, King of 97 n. 14 Saxony, Georg I, King of 97 n. 14 Saxony-Meiningen family 97–8 Schaffgotsch, Friedrich, Count 149 Schaffgotsch, Hans-Ulrich, Count von 193 n. 19 Schaffgotsch, Johanna, Countess von 193, 199, 241 Schaffgotsch, Ludwig, Imperial Count 149 Schaffgotsch Works LLC 193, 199, 241 Schaumburg-Lippe, Adolph George, Prince of 97 Schaumburg-Lippe family 97–8 Schiel family 168 Schildau (estate) 96, 100–1 Schildberg (district) 97 Schiller, Ren´e: Vom Rittergut zum Grossgrundbesitz 6 n. 28, 10, 14 n. 68, 17–19, 38 n. 2 Schlochau (district) 215, 220 Schmidt, Victor 103 Schnallenstein (estate) 95 Sch¨onborn, Martha von, see Alvensleben, Martha von (n´ee von Sch¨onborn) Sch¨onborn, Mrs. von 107 Sch¨onburg-Waldenburg, Heinrich, Prince of 99 Sch¨onburg-Waldenburg, Ulrich, Prince of 101 schools: holdings by 119 landownership by 87 see also Greifswald: University of Schulz, Freidrich Hugo 199 Schwenke, Herman 81 n. 30, 94, 172 Seherr-Thoss, Baron von 105

Seherr-Thoss, Roger, Count von 102 Seitenberg (estate) 95 Senfft von Pilsach, Carl, Baron 94, 105 n. 20 Sensburg (district) 110–11 Sering, Max 3 Seydel, Maria 224 Silesia: arable agriculture 58, 230 banks/business ﬁrm holdings 118–19 barons/baronesses as landholders 106 communally owned properties 118 Conrad and 7 ﬁfty largest landowners 123, 144 forest land 101, 221, 230 forestry in 58, 210 in Handbooks 40 industrial establishments 175, 176, 179, 180, 187, 189, 192, 195, 204, 207, 208 knight’s estates 163–4, 164, 165, 171 land use 57, 58 large properties 54, 55, 59–60, 61, 64, 65, 66, 70 table, 76, 83–4, 144, 237 manors 32 meadow land 218–19 mixed land use 210, 231 nobility as landowners 83–4 pasture land 219, 220 ploughland 216–17 princes/dukes as landholders 98, 100–2 ‘privileged estate complexes’ (bevorrechtete G¨uter-Complexe) 157 Prussian State landownership 9 quality of land 222 royal properties 94–5, 95–6, 97 ten largest landowners 146, 151 untitled nobility as landholders 106 value of land 84 size, of properties 39–40 and agricultural area 33–4 in agricultural censuses 25–6 distribution by 67–9, 81–3 and industrial establishments 240–1 knight’s estates 164–5, 167, 171–2, 174 logarithmic distributions 76–8 nobility and 82, 161 of noble vs. non-noble owners 122 non-noble owned 161 owner categories and 81–2, 119–20, 161 ownership relations and 28–30 royals and 82 and size of agricultural enterprises 34–5 and size of farms 31–2 by tax value 73–6 by total area owned 69–73, 82–3 Sk´orzewski, Leo, Count 103

Index Solms-Baruth, Friedrich (I) Hermann Adolf, Prince (formerly Count) zu 99 Solms-Baruth, Friedrich (II), Prince zu 148, 149 Solms-Roesa, Wilhelmine, Countess zu 101–2, 128, 151 Sprenger, Heinrich von 105 Sprottau: City 81 n. 30 district 101 St Annen und Brigitten, convent of 143, 145, 146 St J¨urgen monastery 143, 146 St Spiritus Hospital 143 Stallup¨onen (district) 52, 53, 94 Standesherren, see princes, mediatized starch factories 182, 185 Stern, Fritz 2 Sternbach (estate) 99 Stolberg-Rossla, Count zu 103 Stolberg-Rossla, Jobst Christian, Prince zu 136 n. 14 Stolberg-Wernigerode, Christian Ernst, Prince zu 146, 148, 149, 151 Stolberg-Wernigerode, Hermann, Prince zu 100, 148 Stolberg-Wernigerode, Otto, Prince (formerly Count) zu 89 n. 4, 100, 102, 103, 126–7, 128, 146, 148, 151 Stolberg-Wernigerode, Wilhelm, Prince zu 100, 148 Stolzenburg (estate) 81 n. 30, 94, 172 St¨open (estate) 99 Stralsund: administrative region 65, 145, 213 City 143 sugar factories 118, 175, 181–3, 185, 192, 208, 240, 241 Sumpf (estate) 99 tax assessment 44 bias in 249 conversion of lower-taxed land into higher-taxed land and 62 in districts 214–15 and economic strength 11 on knight’s estates 242 on large properties 242–9 and nobility 248, 249 of properties owned by ﬁfty largest landowners 136–42 types of land and 214, 238, 242, 247–9 and value of land 62, 211–12 see also untaxed land tax values: and entailed properties 11

277

ﬁfty largest landowners by 238 of forest land 238 of land owned by nobility 120 and land use 239, 240 of large properties 61–6, 77–8 of meadow land 238 non-noble owners by 238 by owner category 119–20 of ploughland 238 and size of properties 73–6 ten largest owners by 149–51, 238 and twenty largest landowners 128 ten largest owners 124, 146–51, 238 by tax values of properties 149–51 by total area owned 147–8 Thurn und Taxis, Albert, Prince von 126, 128 Thurn und Taxis, Albert I. Maria, Prince von 126 Thurn und Taxis, Maximilian Maria, Prince von 126, 146 Thurzy (estate) 101 Tiedemann, Heinrich von 136 n. 12 Treue, Wilhelm 156 n. 13 Tschirsky-R´enard, Count von 151 twenty largest owners 123, 124–9 types of land 55–9 and area of properties 43, 44 conversion among 249 distribution of 55–9, 214 in districts 55–9 in Handbooks 38 provincial specialization in 238–9 quality of land and 244–6 regional distribution of 239–40 tax assessment of 214, 238, 242, 247–9 and uses 222, 224 unproductive land 38, 222 n. 14, 224, 227 untaxed land 38, 222 n. 14 value, of land: distribution of 211–22 high-value vs. low-value 143 knight’s estates 154 political privileges and 154 produce prices and 62 productivity and 211 tax assessment and 62, 211–12 vom Rath, Schoeller und Skene LLC 143 Waldau (estate) 104 Waldburg (estate) 160 Waldow und Reitzenstein, Karl von 136 Wanzleben (district) 65, 215, 221, 238

278

Index

Warnen (estate) 52 water land 211, 249 Wendenburg, Erich 114 n. 33 Wendenburg, Gustav 113–14 West Prussia: administrative division changes in 41 barons/baronesses as landowners 104–5 Conrad and 7 counts/countesses as landowners 102, 103 ﬁfty largest landowners 146 industrial establishments 175–6, 179, 189, 190, 202–3, 204, 206, 207 knight’s estates 152, 164, 165, 166, 172, 174 land use 57 large properties 60, 63, 64, 66 meadow land 218, 219 nobility as landowners 83 non-noble landownership 85, 109 pasture land 219 ploughland 216, 217 princes/dukes as landholders 99 Prussian State landownership 9, 99, 116–17

quality of land 221 taxable value of land 215 untitled nobility as landholders 106–8 Wied, Friedrich, Prince zu 96, 100 Wied, Marie, Princess (formerly Princess of the Netherlands) zu 96, 100–1, 149 Wietersheim, Alfred 149 women, as landowners 3, 42, 111–12, see also names of individual women woodland, see forest land Wurdenburg (estate) 149 Wyslawice (estate) 100 Wyttich und Lochum, Asta von, Imperial Countess, see Riepenhausen, Asta von yeoman farmers 3, 27 Zembowitz (manor) 101 Zimmerman, August 149 Zimmerman, Kurt 149 Zimmerman, Max von 143, 144 Zollverein (German Customs Union) 1