The Transformation of Australia's Population, 1970-2030

T H E T R A N S F O R M AT I O N O F A U S T R A L I A’ S P O P U L AT I O N : 1970–2030

THE TRANSFORMATION OF AUSTRALIA’S POPULATION: 1970–2030

edited by Siew-Ean Khoo and Peter McDonald

UNSW PRESS

A UNSW Press book Published by University of New South Wales Press Ltd University of New South Wales UNSW Sydney NSW 2052 AUSTRALIA www.unswpress.com.au © UNSW Press 2003 First published 2003 This book is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced by any process without written permission. While copyright of the work as a whole is vested in UNSW Press, copyright of individual chapters is retained by the chapter authors. Inquiries should be addressed to the publisher. National Library of Australia Cataloguing-in-Publication entry: The transformation of Australia’s population, 1970 to 2030. Bibliography. Includes index. ISBN 0 86840 502 7. 1. Australia – Population. 2. Australia – Economic conditions – 1965- . 3. Australia – Social conditions – 1965– . I. McDonald, Peter F. (Peter Francis), 1946– . II. Khoo, Siew Ean. III. Title. 304.60994 Printer BPA

CONTENTS

Preface Contributors 1

Introduction: Australia’s population history and prospect Geoffrey McNicoll 2 Indigenous Australians: The first transformation John Taylor 3 Fertility trends and differentials Gordon A. Carmichael and Peter McDonald 4 Transformations in the Australian family Peter McDonald 5 The changing dimensions of mortality Heather Booth 6 The management of immigration: Patterns of reform Bob Birrell 7 A greater diversity of origins Siew-Ean Khoo 8 Changing patterns of population distribution Graeme Hugo 9 Transformations in the labour force Bruce Chapman and Cezary A. Kapuscinski 10 An ageing population: Emergence of a new stage of life? Don Rowland 11 Australia’s future population: Population policy in a low-fertility society Peter McDonald Notes Bibliography Index

vii ix 1 17 40 77 104 129 158 185 219 238 266

281 283 297

P R E FA C E

The idea of a book to examine the profound demographic changes that have occurred in Australia since 1970 and to look at their implications for the 21st century was conceived as part of the Reshaping Australia’s Institutions Project, an initiative of the Australian National University’s Research School of Social Sciences to reflect on the social, economic and political changes of the last few decades of the 20th century. A select group of social scientists from the disciplines of demography, economics, geography and sociology were invited to prepare papers in their own areas of expertise in relation to Australia’s population, to examine the pattern of changes, the underlying causes, future trends and implications for policy. The aim was to stimulate further thinking about the size, structure and distribution of Australia’s population in the 21st century. Initial drafts of these papers were presented at a seminar at the Australian National University towards the end of 1999. The seminar drew much interest and participation from both the academic community and the Federal Government bureaucracy. We thank the seminar participants for their thoughtful comments, which have helped all the contributors in revising their papers into chapters for this book. The contributors have extended and updated their earlier papers to include the most recent demographic data and projections, including the first release of data from the 2001 Census, in their chapters. In some cases this has involved substantial rewriting to keep up with the latest demographic trends. We thank all the contributors for their efforts in meeting the publication deadline. We are also grateful to a number of people for their assistance in

VIII

•

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

preparing the manuscript for publication. We thank Wendy Cosford, Jacob Lasen and Beth Thomson for their time and cheerful efforts in reading and checking the manuscript, and preparing the tables, figures, charts and bibliography. We would also like to thank the two anonymous reviewers for their helpful and encouraging comments on earlier drafts of the chapters. There is currently a great deal of interest in Australia’s population, judging from the many articles in the press on Australia’s fertility decline, ageing population, level of immigration and population loss from regional areas. A Population Summit, convened with support from both the private and public sectors and held in Melbourne in February 2002 to debate Australia’s population future and whether the government should have a population policy, drew a large and diverse audience. The Intergenerational Report, released in connection with the Federal Budget for 2002–03, highlights the importance of demographic change in the government’s policy development and planning for the future. We hope that this book will contribute to a better understanding of the demographic shifts that have transformed Australia’s population during the last 30 years of the past century and which will continue to do so for at least the first 30 years of this century. Siew-Ean Khoo and Peter McDonald

CONTRIBUTORS

Bob Birrell is the Director of the Centre for Population and Urban Research and Reader in Sociology at Monash University. He is the joint editor of the demographic journal People and Place published by the Centre. His most recent book is Federation: The Secret Story, Duffy and Snellgrove, Sydney, 2001. Heather Booth is a Senior Lecturer in the Demography and Sociology Program at the Australian National University. Her doctoral research at London University was on modelling fertility and her current research interests are mortality modelling and forecasting and probabilistic population forecasting. She has worked in the USA, UK and Pacific Islands in the areas of demographic estimation and technical assistance and in the demography of ethnicity in the UK and Europe. She is Editor of the Journal of Population Research. Gordon A. Carmichael is a demographer with the Australian National University’s National Centre for Epidemiology and Population Health. He has written extensively on family formation and dissolution in Australia and New Zealand, and on trans-Tasman migration. His research has focused on marriage and fertility patterns, consensual partnering, divorce, and trends in non-marital pregnancy and induced abortion. He is currently engaged in a project using qualitative research methods to examine family formation decision-making in Australia. Bruce Chapman is a Professor in the Economics Program in the Research School of Social Sciences at the Australian National University. He has a PhD from Yale University and is a labour and

X

•

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

education economist with extensive direct policy experience, including the design of the Higher Education Contribution Scheme and as a senior adviser to Prime Minister Paul Keating. A Fellow of the Academy of Social Sciences, he was awarded the Order of Australia (General Division) in 2001 for contributions to the development of economic, labour market and social policy. Graeme Hugo is Professor at the Department of Geographical and Environmental Studies and Director of the National Centre for the Social Applications of Geographical Information Systems at the University of Adelaide. He is the author of Australia’s Changing Population: Trends and Implications, Oxford University Press, 1987. He is a Fellow of the Academy of Social Sciences and is a recipient of the Australian Research Council’s prestigious Federation Fellowship in 2002. Cezary A. Kapuscinski obtained his PhD in econometrics from the Australian National University. An applied econometrician, he has held positions in the public sector, the University of Canberra and the Australian National University. His main research interests include labour economics, applied econometrics and time series modelling. His publications have included work on long-term unemployment, the dynamics of the apprenticeship and traineeship system in Australia, and the determinants of illegal and criminal activities. Siew-Ean Khoo is Executive Director of the Australian Centre for Population Research in the Research School of Social Sciences at the Australian National University. A graduate of Harvard University, she has worked with the East-West Population Institute, Honolulu, the Australian Bureau of Statistics and the Bureau of Immigration, Multicultural and Population Research. Her current research interests include immigration and immigrant settlement issues in Australia. Peter McDonald is Professor of Demography and Head of the Demography and Sociology Program at the Australian National University. A Fellow of the Academy of Social Sciences, he is also CoDirector of the Australian Centre for Population Research. In recent years his research has focused on explanations of low fertility rates in advanced countries and the implications of population dynamics for ageing and the labour force.

C O N T R I B U TO R S

•

XI

Geoffrey McNicoll is a Senior Associate in the Policy Research Division of the Population Council, New York. From 1988 to 2000 he was professor in the Demography Program, Research School of Social Sciences, Australian National University. He has also worked with the Central Bureau of Statistics, Jakarta, and the East-West Population Institute, Honolulu. He has a BSc from the University of Melbourne and a PhD in demography from the University of California, Berkeley. He has written widely on population and development issues. Don Rowland is a Reader in Population Studies at the Australian National University, where he teaches courses in demography and sociology. His research has been concerned with Australian and New Zealand demography, especially migration, the family and ageing. He has published widely in these fields. His most recent book, Demographic Methods and Concepts will be published by Oxford University Press in 2003. John Taylor is a Senior Fellow at the Centre for Aboriginal Economic Policy Research at the Australian National University. His academic roots are in human geography and he has specialised in research on the demography of Indigenous Australians since the mid1980s. He has published widely in this field both within Australia and overseas, and frequently advises government and Indigenous communities on the policy implications of demographic change.

1

INTRODUCTION: A U S T R A L I A’ S P O P U L AT I O N HISTORY AND PROSPECT GEOFFREY McNICOLL

Population issues have long figured prominently in Australia’s perception of itself and its future. Traditionally, these have concerned settlement and immigration. A distinctive, unforgiving ecology, mastered over the millennia by the country’s first inhabitants, has presented formidable challenges to later settlers, confining most of the newcomers to the temperate fringes of the continent (even so, with dire consequences for the Aboriginal population). In common with a relatively few other countries—principally, Canada and the United States—immigration has been accepted as a routine component of population increase, rather than tailing off as aggregate population has grown. Public policy directed at population change has virtually coincided with immigration policy. But there are other population issues, some of long standing, others emerging more recently, that are of comparable interest and importance and that arguably should also be the concern of public policy. Patterns of migration within Australia are one such set of issues, particularly the demographic hollowing-out of major inland areas. Another, larger, set has to do with changes in the family and the labour market and the associated trends towards low fertility and population ageing. In these cases it seems there should be a policy agenda, but consensus on objectives is often lacking and even where it exists policy instruments of proven efficacy may be hard to find. A demographic perspective on modern Australia thus brings together questions of land and distribution, migration and national identity, health and longevity, and family and work. It forces

2 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

consideration of which features of Australia’s situation are sui generis and which are ones where Australia is in the company of virtually all mature industrial societies. And it illuminates areas where public policy has or plausibly should have a role. This, broadly, is the subject of the studies brought together in the present volume. The time scale is 30 years into both the past and future: roughly a generation in each direction. That is too long for economists, too short for environmentalists; but it is a comfortable period for demographers. Looking to the past, it gives time for finding significant behavioural change, while ensuring that data are likely to retain sufficient consistency to measure it. For the future, the period is about the extent of reach over which we can have moderate confidence in population projections, with the age-structured dynamics of population change going some way to limit the range of demographic possibilities. Population change is at first sight a straightforward subject— people are born, they migrate, they form families (and often dissolve them), they have children, they work and retire, and they die—processes that seem to be both concrete and readily measured. But complications abound. Chief among them is that the processes are played out over many decades in the life-cycles of individual men and women and are intricately tied to age. As features of a person’s life-cycle, fertility and mortality, for example, are properly recorded at the end of the reproductive years and at death. But for most of the population that time is likely to be many years in the future, and anyway would shed little light on what counts in assessing the current pace of population change: the number of births and deaths this year. The registration data that record those numbers, however, are behaviourally empty: they need somehow to be linked to life-cycle risks of those events. The conventional way of doing this is to construct summary measures of fertility and mortality that do not refer to the experience of any actual group of people but to the hypothetical experience of a person whose risks at each age resemble the current average for the population at that age. A further complication in describing demographic conditions and change is that the complex behaviour involved is poorly characterised by a single index. The number of children the average woman would bear over her life—the standard measure of fertility, known as the total fertility rate—tells us almost nothing about the extent of childlessness or of large families, both of which are features that are

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 3

critical in giving meaning to the level of fertility. The commonly used summary measure of mortality, life expectancy at birth, has to perform the statistically dubious task of describing in one number a bimodal distribution: mortality risks have peaks in infancy and old age. Mortality in the two age ranges would be better described separately. Discussions of population change thus cannot avoid a certain amount of technical demography. In this volume, however, it is kept to a minimum, and where it intrudes a bemused reader should find it easy to make a brief detour. The organisation of the volume mainly reflects the specialisations within demography: fertility and family, mortality, immigration, internal migration and distribution, labour force, and aging. Aboriginal demography warrants separate attention because of its distinctiveness in comparison to that of the majority population— especially the persisting high levels of adult mortality. And an effort to peer into the population future offers an appropriate coda. In the remainder of this introductory chapter I will touch on some of the volume’s topics and themes to give a sketch of what is happening in Australia, demographically speaking: a view of how population fits into the larger scheme of things. It is not meant as an executive summary and hence does not proceed chapter by chapter.

ENVIRONMENT AND SETTLEMENT For social scientists, a category that includes most demographers, the natural environment is usually just that: a background against which human behaviour is played out. Some are totally dismissive of its influence, seeing human adaptation as endlessly inventive; others acknowledge that it does impose some limits on human activity. Relatively few are concerned to explore the extent to which those limits themselves can be narrowed as well as enlarged by human action. What has happened in the environmental arena in Australia over the last few decades? Two things stand out. First, we have learned a lot about Australia’s ecological history, both before and after human settlement—the latter at least, especially after 1788, a fairly sorry tale of land degradation and loss of biodiversity—and about the dynamics of climate, notably the El Niño cycle and the likely or possible effects of atmospheric warming. And second, perhaps more arguably, our perception of the Australian environment has changed: away from ‘the land’, ‘the outback’, ‘the Mulga’, ‘the dead heart’, and similar clichés, and towards the coast. From the bush to the beach.

4 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

In literary terms, from The Drover’s Wife, or Voss, to Robert Drewe’s The Bodysurfers or Tim Winton’s Shallows. The two things are connected, and both bear on one of the subjects of this book, changing settlement patterns. The effect on the rural economy of better understanding of environmental conditions seeps into everyday thinking and public policy, although always mixed, when it comes to difficult action, with shortrun economic interest and political calculation. But it is now far less radical than it used to be to hold that European farming may be ill suited to much of Australia, that cotton and rice are less than optimal crops for the driest continent, that the emptying out of inland and remote areas reflects something more fundamental going on than a response to bank branch closures and inadequate diesel subsidies. The Murray–Darling Basin Commission’s salinity audit, revealing the extent of the degradation of this vast region (some 100 million hectares) and the time scale involved in amelioration, was a notable event in this transformation of outlook. Ecological awareness and stimulus to policy action of this sort is very far from asserting particular limits on overall population size. That contentious topic is not treated in this volume. Carrying capacity is a popular concept in studies by environmentalists of Australia’s demographic future. Transferring the concept from its animal husbandry origins to the case of humans raises many problems, not least, the dismal level of ambition implied by such a parallel; but for what it is worth, the so-called ‘ecological footprint’ of Australia, that is, a measure of the biologically productive space currently being used, is calculated to be 9 hectares per person while Australia’s total biological capacity available to be 14 hectares per person. By such indices we could theoretically expand another 50 per cent or so in population numbers or consume 50 per cent more profligately, but not both; and not if we wished to retain much non-desert wilderness or keep substantial agricultural exports. The world as a whole, by the same methodology, is estimated to be ecologically overdrawn by a third, signalling rapid depletion of its natural capital. The natural environment enters this study as a major determinant of population distribution. The extreme settlement density ranges in Australia are familiar but still worth recounting, as Graeme Hugo in this volume does: 70 per cent of the land empty, another 20 per cent virtually empty, and still another 8 per cent settled at the very modest level of 1–10 persons per square kilometre; and, at the

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 5

other end of the range, 76 per cent of the population living at densities of over 100 persons per square kilometre, occupying 0.3 per cent of the land. In every country, of course, development is associated with concentration, not dispersal, of population. Even very populous countries, when looked at closely, have huge swathes of near-empty landscape: in China, the entire expanse of the western and north-western regions, 50 per cent of the country’s total land area, contains just 3.6 per cent of the population. Moreover, modern agriculture is a capital-intensive industry that for the most part uses little labour: the United States, with its huge agricultural production, employs just 2.5 per cent of its labour force in the sector: about the same proportion as the occupational category of secretaries, stenographers and typists. Sparseness of inland settlement has always been a cause of anxiety in Australia. We retain the strange idea that empty spaces must be filled, like nature abhorring a vacuum. The sentiment is often heard from immigration boosters, but not only from them. There is a common feeling that settlement should be spread out, that the bush deserves support just for being there, as if to make up for being so far from the beach. This was once manifested in organised rural resettlement schemes, with their ideology of yeoman farmers (typically to be created out of demobilised soldiers) as the salt of the earth. The schemes failed fairly abysmally, from western Victoria to the Ord River. The United States comparison is relevant here. There is plenty of inland settlement in the US, though it tends to be along the navigable waterways. So could Dubbo or Wagga become Australia’s St Louis or Kansas City? The chances are vanishingly small: the Murray–Darling system, even downstream with most of the diverted Snowy River added to it, is a muddy trickle in American terms. In the American West, settlement has also been favoured by access to the salubrious mountain environments of the Rockies and Sierras. So could Wangaratta or even Canberra become an Australian Denver? Again the prospects are slim. Australia’s settlement pattern, like its overall population density, has more in common with Canada’s. Hugo reports that 84 per cent of Australia’s population live within 50 kilometres of the coast, an easy day-trip; Canadians, lacking beach weather, tend to settle along the US border. What of the distributional future? Both environmental and economic forces apparently lie behind Australia’s changing rural

6 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

demography, notably the depopulation of dry farming and pasture areas, and the decline in inland towns as their function as service centres for agricultural enterprise erodes. Such decline can be slowed by a flow of economic subsidies: not blatant like Europe’s or Japan’s, but through a more modest sustenance of economic and social infrastructure, rendered politically acceptable by lingering romantic illusions about the bush and to a degree also by gerrymandered electorates. But there remains a strong political appeal for subsidies to the bush, eliciting support even from city-dwellers, whose taxes or telephone charges must fund them. Demographic decline, however, is not easily halted. The US also heavily subsidises its vast but little-populated interior regions, and in its case they contain entire states and hence represent serious political weight in the federal system; but the Great Plains still experience steady population loss. In both countries, young people in remote rural areas see that opportunity lies elsewhere and move. It is sad in many respects, but probably not remediable, perhaps properly not. As a thought experiment, say an empty Australia was being peopled today: what pattern of settlement would be planned? Almost certainly, Flying Doctors and their internet analogues would be the model, rather than branch offices of Australia Post and the Commonwealth Bank. The rural Aboriginal population shows a different pattern of change. John Taylor presents an interesting sketch of the settlement consequences of the new land rights regimes for Aborigines in remote areas. As his maps indicate, Aboriginal groups are spreading out; but in many cases recreating pre-European patterns of settlement, where that is still feasible. There is little likelihood, it seems, of an Australian Nunavut, the vast region—1.9 million square kilometres, 26 000 inhabitants—that Canada has carved out of its Northwest Territories and assigned to Inuit administration. Turn now to Australia’s cities. A recurrent theme of geographers is the stability of Australia’s urban settlement patterns, despite the high rates of mobility of individuals and families. There may of course be changes at a finer-grained level, but the big picture is one of stasis. There are the large cities—in A.D. Hope’s strikingly inelegant metaphor, the ‘five teeming sores’—and there are the country towns. As Hugo remarks, the geographer Griffith Taylor, in his classic work on Australian settlement, thought the pattern was already fixed when he was writing half a century ago. But lack of change also calls for an explanation. Why, for instance, have such fine

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 7

harbours as Jervis Bay or Twofold Bay not been developed? From the viewpoint of recreation and environmental amenity, it is all to the good, but that does not explain the case. No doubt part of the explanation is the drawing power of Sydney and Melbourne, along with the mutually destructive state rivalries (epitomised by the different rail gauges) that have reinforced hub-and-spokes development within the states. But there is an important theoretical issue here, one with which geographers are familiar. That is the role of self-reinforcing processes in settlement patterns, the phenomenon of lock-in, where an outcome, once established, may perpetuate itself simply by being there. This is part of the ‘new’ economic geography, emphasising historical contingency and increasing returns to scale and possibilities for multiple equilibria. Urban centres or locational systems in general, in this perspective, can be sustained long after their initial geographic rationales have vanished.

I M M I G R AT I O N A N D I D E N T I T Y For the decades covered by this study Australia’s population has been at least one-fifth foreign-born. The fraction is currently even higher, near one-quarter, a remarkable figure by the standards of most other industrialised societies. Comparative UN data for 2000 on the stock of immigrants (not uniformly defined, but usually meaning foreignborn) show Australia at 24.6 per cent, Canada 18.9 per cent, the United States 12.4 per cent, and the United Kingdom 6.8 per cent. A number of Western European countries, such as France, Germany, the Netherlands and Sweden, have seen a rapid, often startling, increase in the proportion over the last decade or two and now lie in the 9–11 per cent range. In Australia the political debate on levels of immigration is mainly concerned with economic effects, and to a lesser extent with environmental effects. The first is exemplified by the enthusiastic populationism of the 1988 FitzGerald Report, in which a half-million migrants a year would have been endorsed had the number survived the economic simulations; the second, a few years later, by the ecologically informed reports of the now-defunct National Population Council and the Parliament’s Jones Committee (the House of Representatives Standing Committee on Long-term Strategies). The economics of immigration is sometimes presented as a settled matter, usually by those who see it as giving support for

8 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

larger target numbers. That is not the case. Much of the recent research comes from the US and offers decidedly equivocal results, with considerable sensitivity of findings to modelling assumptions and parameter values such as the discount rate. That the human capital of migrants is an important factor in assessing net benefits, however, is widely accepted. It is implicit in the points system used for migrant selection by both Australia and Canada. Bob Birrell, in this volume, describes the detailed evolution of Australia’s recent immigration policy, showing the tension between skills and family ties as selection criteria. A focus on skills can amount to a fairly forlorn effort to micro-manage the labour market, seeking to plug its apparent holes. Thus Australia once sought waiters to staff its tourist hotels, and Canada today is seeking farmers from traditional old-world communities to take the place of those who have left the land for the city lights. Elevating family reunion as the main basis for selection, which has been the practice in the US, yields a migrant stream with lower levels of education and English proficiency. In the US these characteristics also apply to the large numbers of unauthorised entrants. The combined effect is to perpetuate a substantial low-wage sector of the economy, especially in services and agriculture, that co-exists with the high-productivity mainstream: an option that Australia has deliberately sought to avoid. The traditional countries of immigration, and increasingly most other industrialised economies as well, are generally eager to admit highly-skilled migrants, although in some cases objections are raised by domestic professional groups threatened by the harsher competitive environment this can produce. The latter difficulty is neatly circumvented if the prospective recipient country also offers the tertiary training of such persons, a practice in which the US, with the attraction of its superb graduate schools, is pre-eminent. Australia competes at some level in this league, but its ability to extract a significant hightechnology payoff from the process, Silicon Valley-style, is uncertain. So migrant numbers matter. But how much? The case of New Zealand suggests a certain arbitrariness in how numbers are handled. In most respects an integrated trans-Tasman economy exists, with free trade and easy mobility of labour. From an Australian standpoint, the Closer Economic Relations treaty with New Zealand is equivalent to the sudden and painless receipt of 30 or 40 years worth of migrants, trained, English-speaking, complete with capital equipment and infrastructure (a bit more obsolete than Australia’s),

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 9

and, by being located across the Tasman, admirably decentralised away from Sydney and Melbourne; thereby accomplishing a goal that has defeated all Australian governments thus far. GDP is effectively raised by one-sixth; GDP per capita is lowered only slightly. We do not look at it in that way merely because of a few national symbols that channel our thinking. The reason immigration remains a sensitive issue in Australia probably has little to do with its putative economic consequences. What matters more in the public mind is the perceived effect on national identity and cultural traditions, a concern manifested in political debate over multiculturalism and assimilation. In this volume, Siew-Ean Khoo explores a salient aspect of national identity in her discussion of ethnic diversity, in particular the increase in Australia’s Asia-born proportion. The 2001 figure is 6 per cent, although considerably higher in the major cities. Charles Price, Australia’s noted arithmetician of ethnicity, calculates much larger future proportions by allowing ancestral lines to continue over the generations, appropriately diluted by intermarriage. Thus he projects that by 2025 the population (or perhaps more strictly, the national gene pool) will be 16 per cent of Asian origin, as compared to 62 per cent ‘Anglo-Celtic’ and another 15 per cent other European. Yet it is that very process of intermarriage that attenuates the change in identity that would result. The forecast for the Asia-born is for a doubling by 2030, to some 9 per cent, a figure comparable to that projected for Canada, which would represent a gradual and most likely unproblematic shift. It is well within the range that could be accepted as a welcome addition to cosmopolitanism—SBS television, fusion cuisine, and so on—rather than a source of ethnic tension. For Australia’s immigrants in general, Khoo’s picture in most cases is one of rapid cultural integration, seen in intermarriage and language loss, at least in the second generation. The view that exposure to diversity erodes conventional ideas of ethnicity seems defensible on some readings of the evidence, but we should at the same time be impressed by experience elsewhere where the outcomes have been very different. Birrell’s comments on the economic and social problems of ethnic communities around Sydney, the dominant migrant destination, suggests that a satisfied appraisal that Australia has got it right would be premature. On the relationship between migration and identity, comparison with the US may be inappropriate. The US situation is dominated by

10 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

the numerical scale of the migrant intake from Mexico and other Latin American and Caribbean countries, forming the so-called Hispanic minority concentrated in a few major states, principally California, Florida, New York, and Texas. By 2030, the child population (under 18 years) in the US, according to the 2000 Census Bureau projections, will be over one-quarter Hispanic and only 50 per cent ‘non-Hispanic white’. Even if accompanied by a substantial pace of assimilation, this trend seems to presage a strengthening pattern of biculturalism, and bilingualism. Nevertheless, even if less culturally specific, migration on the scale Australia has experienced and that will certainly continue over the time scale of this study is inevitably altering and will go on altering the country’s self-image. Plausibly, it produces a shorter-term (or more readily revised) historical memory; in particular, a weakening of the sense of the country’s identity as a society forged over the generations by the exigencies of life. To some, this is to be welcomed for eroding narrowly parochial and racially exclusive views, including views on who should be admitted; to others, for whom those roots are seen as also containing much of value, it is a cause of regret. Where the balance is struck has consequences for the vision of what sort of society, quantitative and qualitative, should be sought. Refugees and asylum-seekers make up another category of potential migrants, but numerically a small component of the total. Unlike European Union member states, Australia tries to insist on offshore determination of eligibility for refugee status and imposes a firm quota on annual admissions. As Birrell notes, that policy attracts strong public support, even as its recent exercise has drawn much media criticism of Australia abroad. The fear of a large demonstration effect should ‘boat-people’ be accorded a more welcoming reception is quite probably justified. Italy, where such landings are common, can adopt a stance of relative equanimity on the realistic assumption that most of those arriving will move on to other EU country destinations; but it is roundly belaboured by those countries for not doing more to halt the flow. There will always be a contrast between the economic and social welfare discourse appropriate for setting and administering government refugee policy and a human rights and individual welfare perspective that nearly every particular case can evoke when the details are examined. Yet a more than sporadic concession to the latter can defeat the very idea of a policy.

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 11

H E A LT H A N D L O N G E V I T Y Unlike the subject of health, a near obsessive concern of many people in the world’s healthiest countries, mortality in societies like Australia is relegated to the margins of attention: it is no longer, so to speak, a fact of life, or at least, of everyday life; it is a property of a particular age group, the elderly. And death itself has retreated into the private domain. That is not to say that mortality rates have ceased to change. Quite the opposite: declines in mortality at old ages have continued apace, to a much greater extent than was anticipated. That has been a significant factor in creating the financing crisis that faces many countries’ public pension schemes. There is of course one major exception in Australia, a group whose mortality is persistently higher than the population average: the Aborigines. Here child and adult mortality must be distinguished. As Taylor recounts, there has been a remarkably rapid drop in Aboriginal infant mortality since the 1970s, when the government belatedly began to pay serious attention to the matter. But infant mortality is still relatively high by developed-country standards: one Aboriginal child in 40 still dies in the first year, several times the risk for Australian infants as a whole. Moreover, this figure, associated with the frequency of low birthweights, has not been responding to further health service interventions. High Aboriginal adult mortality, especially among men, is even more disturbing. Here the rates are sticking at levels that the majority population was experiencing in 1920. The immediate causes are fairly well known: some of them common to indigenous populations in various other countries, others related to the specifics of Aboriginal disadvantage in Australia. But finding a solution is proving highly intractable. For delivery of health services, the remoteness of many Aboriginal settlements is a particular difficulty, indeed a worsening one, given the dispersal of groups to smaller ‘homeland centres’. Government outreach agencies have to pursue their clients to these distant locations. For the Australian population as a whole, Heather Booth sketches what is mostly a cause-of-death picture that could be found in many other developed countries, where infectious disease is largely under control, leaving the field to chronic and degenerative disease. And much of the latter is linked to lifestyle factors, such as consumption of tobacco and alcohol, and obesity. The estimates that smoking accounts for 17 per cent of all male deaths in Australia and

12 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

9 per cent of female deaths are not high by many international comparisons but nonetheless show a major route for further mortality reduction. In this case, the male fraction is trending down and the female fraction rising: Australia’s proportion of women smokers is quite high, nearing one-quarter. Low mortality need not imply low rates of disability, but usually delays the onset of disability. The expansion of the disability-free period of life in the ages after retirement is a major new structural feature of modern societies. Don Rowland, in this volume, uses the late Peter Laslett’s concept of a ‘Third Age’ of life to organise Australian data on independent living. Shifts in the circumstances of older people are potentially as important, if not as consequential for the future of society, as shifts in the circumstances of people at childbearing ages. The demographic realities that Rowland describes, together with writings such as Laslett’s, are serving to institutionalise the concept of an active post-retirement period that is not just fishing and pottering, or golf and gambling. It is a sort of invention, just as earlier the concepts of adolescence and even of childhood were inventions. The Third Age, by definition, requires good health: not too much disability and no decrepitude. Rowland cites estimates that over 80 per cent of Australia’s population of 65 years and older are Third Agers. Provided the health system’s finances can be maintained in good order (admittedly a considerable challenge), Australia’s experience of healthy ageing will no doubt parallel that of other affluent societies. Australia is relatively advanced in converting pay-as-you-go pension arrangements, notoriously unsustainable as old-age dependency rates increase, into funded schemes. But the economic well-being of the elderly still requires a healthy economy since, whatever their formal ownership, most of the resources that are in fact consumed in retirement are being produced by the economically active. While 65 is a conventional retirement age, a substantial and rising proportion in fact retires earlier; or is retrenched and effectively leaves the workforce. Bruce Chapman and Cezary Kapuscinski predict further falls in Australian labour-force participation rates, a trend already further advanced in much of Western Europe. As Rowland reports, in the age group 55–65, only 63 per cent of men and 28 per cent of women were in the labour force. Those who argue that migrants are needed to beef up an otherwise declining workforce do not have in mind active 55-year-old retirees.

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 13

F A M I LY A N D W O R K All affluent Western societies have current levels of total fertility at or below a two-child average, some of them much below two. The United States is at the high end: 2.0; Canada and the United Kingdom, both about 1.6, are mid-range; Italy and Spain are at the bottom: 1.1–1.2. Australia (1.75) is towards the high end. We do not know for certain that low current rates of childbearing will apply to the completed fertility of present-day young women, since they have many years of potential childbearing left in which delays could be made up. Asked about their intentions, young people typically say they want two children. However, the realities of economic and social circumstances—and of biology—intrude to make it unlikely that those intentions will be fulfilled. The average number of children ever borne for women who have all but passed through their childbearing ages provides a behaviourally well founded estimate of the fertility level, albeit describing the pattern prevailing 10 or 20 years ago. Thus far there are few examples of countries where fertility measured in this way is appreciably below two. Countries with equal averages may also exhibit very different distributions. To illustrate this, consider the contrasting patterns of childbearing for women aged around 40 in 2000 (i.e., born around 1960) from the United States and the Czech Republic, using data from the 2000 US Current Population Survey and Observatoire Démographique Européen. In both cases, the average number of children borne per woman was very close to two. The percentage distribution of families by number of children, however, was strikingly different:

United States Czech Republic

Distribution by number of children ever borne (%) 0 1 2 3 4+ Total 19 16 35 19 11 100 7 15 56 17 5 100

The Czech pattern is what one might first assume would characterise low fertility: a bunching of family size at two, with few in the outlier categories, childlessness or four children and above. It is a pattern shared by other eastern European countries. In contrast, the US has a spread-out family size distribution: a third with one child or none, a third with two, and a third with three or above. The proportion childless, around one-fifth, is particularly notable, though not

14 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

unprecedented. A fifth of women childless was the proportion for women born in the 19th century and up to about 1910, but of course reflecting a different kind of economic pressure: the lack of wherewithal for marriage. Childlessness dropped rapidly over subsequent birth cohorts, those reaching peak ages of childbearing in the 1940s and in the baby-boom years that followed. The family size distribution in Australia, as discussed by Gordon Carmichael and Peter McDonald in this volume, is very similar to the US pattern. The social and economic circumstances that generate different present-day fertility outcomes seem mainly to be connected to the family, the education system and the labour market. As women achieve routine access to higher education and to the full spectrum of employment opportunities and especially as they approach equality of treatment, their roles need no longer be defined by the family. A labour market commitment competes with the specialised childrearing role the family had formerly assigned to the wife, providing her with new choices but also, for many, an intricate problem in balancing two conflicting roles, both of them desired. Variations in family patterns and in the institutional design of labour markets across countries make for a diversity of contexts for work and family choices and consequent demographic outcomes. And the family itself is in play in this new institutional and cultural landscape: many young people choose less formalised arrangements—cohabitation or even looser relationships, same-sex partnerships—while ready divorce and adjustments in the legal status of long-term partners alters the meaning of marriage itself. McDonald in this volume discusses these issues in terms of a stylised typology of families: conservative, liberal, and radical. The political terminology, of course, is not entirely transparent, since placement in the political spectrum differs depending on whether the topic is economic or social. Take the familiar ambivalence on the part of governments as to whether the clients of policy should be seen to be families or individuals. A thorough-going economic conservative would pick individuals; a social conservative would be more likely to pick families. On the left, there would be a similar split between oldguard unionists, who would probably choose families, and radicals, who would choose individuals. Under either right- or left-wing government, there is ample scope for policy incoherence in terms of net influences on fertility. Notwithstanding such complications, many countries with very

I N T R O D U C T I O N : A U S T R A L I A ’ S P O P U L AT I O N H I S TO R Y A N D P R O S P E C T

• 15

low fertility will no doubt be seeking policy measures to raise the number of births, not necessarily fully back to replacement level but to a level that slows the pace of ageing and minimises any absolute decline of population. Australia is not as yet in that situation; although should fertility drop further, to 1.6 or 1.5, the demand for such intervention would no doubt quickly increase.

INVENTING SOCIAL POLICY The studies brought together here are in the main descriptive demography. They portray what has happened demographically in Australia over the past generation and what seems likely to happen over the next. That monitoring and projection exercise is of interest in itself and provides a necessary input to the analysis of economic and social change. But the demographic story, even where it is made up of the myriad decisions of individual men and women, does not play out merely in response to historical circumstances and fortuitous events: it is in part the outcome of government action, intended and unintended. Hence, there is always a policy dimension to the analysis. Many of the policy issues are perennial components of Australian political debate. That is the case especially for immigration and rural development, although these are not topics where closure is ever likely to be possible. Other issues have appeared more recently on the public agenda but still have been well-aired: Aboriginal health might figure here, and perhaps ‘healthy ageing’. And one issue in particular is as yet incipient: the reconciliation of family and work so as to sustain fertility levels somewhere near a two-child average. There was an earlier episode in Australian history when low fertility appeared troubling, to the extent that a formal investigation was set up, the New South Wales Royal Commission on the Decline of the Birth Rate, which reported on the matter in 1904; but the parallels with today’s situation are slight. One policy area that Australia is not likely to have to explore, unlike some European countries and Japan, is adaptation to depopulation. Low fertility gives every sign of being a permanent state both in Australia and in the rest of the affluent, post-industrial world: there are compelling forces making for a 1–2 child choice, with 0 acceptable and 3 verging on irresponsible excess. In the new search for acceptable means to sustain fertility, there are many suggestions of what to do but a scant international empirical record of success. It is not of course immaterial who bears the children, in terms of that sig-

16 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

nificant part of upbringing that consists of parental nurture and transmission of values: a consideration that appears to appreciably narrow the feasible policy set. But with many countries engaged in the search, the knowledge base of what works and in what circumstances is likely to expand steadily. And just possibly we will find that values change, in the unpredictable, bandwagon way such things happen, and women (enough of them, or men) will again seek the mixed pleasures of childraising.

2

INDIGENOUS AUSTRALIANS: T H E F I R S T T R A N S F O R M AT I O N J O H N TAY L O R

The primacy of Indigenous peoples in considering past and future transformations of the Australian population is unquestionable. As a proposition this reflects the historic order of demographic events, but it also stems from the likelihood that no other group in the Australian population has been so transformed, nor is likely to be in the years ahead. At various times and to varying degrees over the last two centuries of their current millennium, the initial inhabitants of Australia and their descendants have been decimated, relocated, assimilated, emancipated, and, most importantly in the context of assessing transformations, enumerated. Given this context, it is surprising to note the lack of attention afforded to Indigenous peoples in the overview of progress in Australian demography presented in the tenth anniversary edition of the Journal of the Australian Population Association (volume 11 number 1 of 1994). In this, reference to the content of Indigenous demographic analysis was provided in only one out of five review articles, and then only briefly to note distinct fertility and mortality levels (Hugo 1994a). As pointed out in another of the review articles, analysis of Indigenous Australians was part of a ‘yawning gap’ in Australian demography until the 1970s (Borrie 1994). On the basis of this most recent synopsis, so it would seem to have remained. The fact is, though, as the challenge to demography has increasingly been to demonstrate its potential contribution to new areas of major public interest (Caldwell 1994), a flurry of research activity has focused on the Indigenous population. As a consequence, a sizable literature

18 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

has emerged over the past 30 years detailing the demographic characteristics of Indigenous Australians (Taylor 1997a). A common product of these works is the depiction of what may be described as an ‘enclave demography’ clearly distinct from the mainstream demographic context. This is circumscribed by politico-cultural constructions of identity (Langton 1981; Jordan 1985; Dodson 1994; Anderson 1997) and derives from the historic experience of population decline as a consequence of colonisation by non-indigenes, followed by a period of recuperation and, more recently, rapid growth coinciding with a shift from exclusion to inclusion of Indigenous people in the provisions of the modern state (Altman and Sanders 1994). Structurally, the socio-economic position of Indigenous Australians has been described as resembling that of the ‘Third World in the First’ (Young 1995). However, the demography of Indigenous Australians that has emerged is significantly different from that of Third World populations being as much a manifestation of inequitable power relations and marginalisation in the midst of plenty as it is to do with any lack of development per se (Gray 1985, p. 143). The onset of this research effort was far from coincidental as it commenced at a time when Indigenous peoples of the continent began to emerge, in a statistical sense, from the shadows of Australian society: a process that can be uniquely pegged to the overwhelming ‘yes’ to the 1967 referendum question as to whether section 127 of the Constitution should be deleted. This, of course, had the effect that Indigenous people were to be counted (or rather included) for the first time in ‘reckoning the number of the people of the Commonwealth’. In demographic terms, the significant consequence was the inclusion in the 1971 Census of a question on selfreported Aboriginal and Torres Strait Islander origins. With minor subsequent variation, this form of question in the census, and increasingly in administrative data collections as well, has provided a basis for charting transformation in the population to the present and beyond, although, as we shall see, not without some difficulty and imprecision (Taylor 2002a).

INDIGENOUS DEMOGRAPHIC TRANSITION— ON COURSE? It was clear by the 1970s that the Indigenous population had undergone a series of systematic fluctuations in fertility and mortality levels, uneven over space and time, but ultimately comprehensive and

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 19

uniform in effect. These have been conceived of as four separate, but overlapping, transitions from the pre-European period of stable growth with high mortality and fertility; through a phase of progressive population decline to a stationary state; followed by a period of high growth; and finally into a regime of lower growth based on reduced mortality and fertility (Smith 1980a). Although information on vital rates is sketchy, from the time of first European settlement the Indigenous population suffered a drastic decline in population with reduced fertility and rising mortality accompanying the frontier of European occupation. For the continent as a whole, there was rapid population decline until about 1890. Between 1890 and 1930 the rate of decline dropped to zero with the population levelling off around 1933 at about 20 per cent of its original estimated size to fully complete its first transition. The second transition represented a move towards a stationary state. Aggregate evidence from those groups who were included in statistical collections at the time indicates that stationary populations existed by about 1880 with birth and death rates at around 35 per 1000. The first sign of change to a third transition appeared with a rise in the birth rate to over 40 per 1000 between 1940 and 1950 falling back again to around 35 per 1000 by the 1970s. In the postwar years, this was accompanied by a sudden and substantial drop in the mortality rate which levelled off at around 16 per 1000 by 1960 and heralded a period of rapid population increase with a growth rate between 2 and 2.5 per cent per annum by 1971. The current focus is on the course of a fourth transition first postulated in the report of the National Population Inquiry in 1975 and confirmed in the supplementary findings of 1978 (Commonwealth of Australia 1978). This involves a regime of lower natural growth based on reductions in both fertility and mortality. While recent trends suggest movement towards convergence between Indigenous and total fertility, the same cannot be said of Indigenous mortality which appears stalled in its decline (Gray 1997; Kinfu and Taylor 2002; Ross and Taylor 2002).

T R E N D S I N M O RTA L I T Y Since 1973, when the Australian Government gave itself ten years to raise the standard of health of Indigenous people to the level of that of the rest of the population, mortality rates have been closely monitored, at least to the extent that available statistics have allowed. It is some measure of the overall lack of progress in consolidating a shift

20 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

to low mortality that by the end of this initial policy period the target for health equality was revised to the year 2000, while targets are now viewed in the context of generational change (Anderson and Brady 1995, p. 18). As with other aspects of Indigenous affairs policy, the question has been asked whether this is evidence of optimism fading away or simply realism setting in (Sanders 1991). Initial studies of contemporary trends in mortality relied heavily on a composite of data sources, although it was possible to establish a continuing decline in crude death rates for the period 1965–78 from about 19 per 1000 to about 13 per 1000. What was more certain over this period was a steady and precipitous decline in the infant mortality rate from around 100 per 1000 in the mid-1960s to 26 per 1000 by 1981, with much of this due to improvements in infant mortality (Thomson 1990). This decline is clearly illustrated in Figure 2.1 which also reveals that further improvement in infant survival during the 1980s and 1990s has been less impressive, with Indigenous infant mortality rates remaining consistently around two and a half times the Australian average.

Figure 2.1 Infant mortality rates for Indigenous and total populations, 1880–1998

ABS 2000a; Smith 1980b; Thomson 1983, 1990; Gray 1997; unpublished data provided by Colin Mathers, Australian Institute of Health and Welfare, Canberra.

SOURCES

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 21

For Indigenous Australians, decline in infant deaths coincided with improvements in community infrastructure and the development, in the 1970s, of intensive Indigenous health programs and services. However, there appear to be limitations to the impact of medical intervention as suggested by the persistence of relatively high infant mortality. From Figure 2.1, it can be seen that since 1990 the level of Indigenous infant mortality has been stuck at around the level recorded for all Australians in the 1940s. While access and equity issues remain important in the delivery of health care services to Indigenous Australians (McDermott, Plant and Mooney 1996; Deeble et al. 1998), further significant improvements in infant survival seem more reliant on a decrease in the proportion of low birthweights which, in turn, is heavily correlated with nutritional issues and the socio-economic status of mothers (Dugdale, Musgrave and Streatfield 1990; Streatfield et al. 1990). As for overall mortality, the first reasonable national estimates were obtained from 1981 and 1986 Census data and revealed life expectancies to be around 56 years for males and 64 years for females. Also clear was a relative lack of mortality variation between the states and territories, although life expectancies were lowest in regions with the most remote and rural communities (Gray 1990a), a situation that has persisted (Mathers 1995; Gray 1997). However, the pattern of relatively high death rates at all ages, but especially in middle adulthood between 30 and 50 years, was found to be universal. Once again, this feature has shown little sign of subsequent abatement (Gray 1997; Australian Bureau of Statistics and Australian Institute of Health and Welfare 1999). Data from the 2001 Census indicate little change in overall survival prospects, with Indigenous male life expectancy estimated at 57 years and female life expectancy at 65 years (Kinfu and Taylor 2002). This overall lack of progress in raising Indigenous life expectancies is all the more striking given that the situation for all other Australians has displayed marked improvement since 1981, which covers the period for which reliable Indigenous estimates have been available (Figure 2.2). Also poignant is the fact that the level of mortality observed for Indigenous males at the end of the 20th century is equivalent to that recorded for all Australian males in 1920. Among females, the comparison is no more encouraging, with life expectancy for Indigenous females currently hovering around a level last recorded for females generally in 1930.

22 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 2.2 Life expectancies for Indigenous and total populations, 1881–2001

SOURCE

Kinfu and Taylor 2002.

To summarise, then, mortality levels followed a clear downward trend through the 1970s, largely due to improved infant survival, but have experienced almost no change since the 1980s because of the slowing down of survival gains for infants and the persistence of high adult mortality. This lack of steady improvement in life expectancy with lowered infant mortality is a unique demographic phenomenon, even by comparison with comparable Fourth World populations in New Zealand and North America (Kenen 1987; Kunitz 1990, 1994; Hogg 1992; Ross and Taylor 2002). With mortality rates presently immutable and seemingly capable of remaining so for some time, further progress in demographic transition appears reliant on continued decline in fertility.

T R E N D S I N F E RT I L I T Y Total fertility rates for Indigenous women reached their highest levels in the decade from 1956–66, remained relatively high until 1971 and then tumbled throughout the 1970s to reach levels not recorded since the turn of the century (Figure 2.3). This effectively halved the total fertility rate (TFR) from around 5.9 in 1966–71 to around 3.3 in 1976–81 and lower again to around 3.1 or 3.0 in 1981–86. The 1996 Census pointed to a further lowering of the TFR to 2.7 representing a substantial drop of around 50 per cent since 1970, while the 2001 Census showed a further decline to 2.3, close to replacement level.

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 23

Figure 2.3 Total fertility rates for Indigenous women and all Australian women, 1958–2001

SOURCE

Kinfu and Taylor 2002.

In explaining this decline, the focus has been on the effects of increased participation by Indigenous people, particularly women, in institutional structures that have altered the costs and benefits of children. In this process, three factors—age at leaving school, labour force status and income—have been recorded as particularly instrumental (Gray 1990b; Dugbaza 1994). This begs a broader question regarding the future course of fertility decline in different parts of the country given that socio-economic status is generally higher in the south and east and in major urban areas (Taylor 1993)—certainly, Indigenous fertility rates remain high in remote areas with TFRs often exceeding 3.0 (Taylor and Bell 2002). While the overall expectation is for steady progress towards replacement fertility (Gray 1997), it can be argued that this has already been surpassed given the context of high mortality. One issue that has emerged parallel with the timing of fertility decline among Indigenous women is the growing importance of Indigenous births to non-Indigenous women as a factor in the overall natural increase of the Indigenous population. While indications of this added component of Indigenous population growth were available from 1986 and 1991 Census data on intermarriage (Jain 1989; Gray, Trompf and Houston 1991; Dugbaza 1994), closer measurement of the effect has been consequent on the improvement of paternity records in birth registration data. Against this source,

24 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Gray (1998) has successfully fitted a logistic model of increase in the proportion of Indigenous births to non-Indigenous mothers in each state and territory with clear indication that different stages of a uniform process are under way in each jurisdiction. The form of this model is such that Tasmania, the Australian Capital Territory and Victoria are well advanced in terms of additional fertility due to intermarriage, while the Northern Territory and Western Australia lag many years behind. Over the 1996–2001 intercensal period, 27 per cent of Indigenous births were to non-Indigenous women (Kinfu and Taylor 2002).

P O P U L AT I O N S I Z E – H OW C L O S E TO C L O S U R E ? While the amount needed to make intercensal increase in a population balance after accounting for births, deaths and migration is usually small, research on the demography of socially constructed populations reveals that this ‘error of closure’, as the American demographer Passel has described it, is often large (Passel 1996). The idea of closure here stems from the demographic equation: P1 = P0 + B – D +/– NM + e Where P1 = the population at time 1 (e.g., 2001) P0 = the population at time 0 (e.g., 1996) B = births during the time interval (e.g., 1996–2001) D = deaths during the time interval NM = the net balance of international migration e = error of closure The size of this error of closure for the Indigenous population is often large (50 per cent of population change between 1991 and 1996, and 31 per cent between 1996 and 2001). As a consequence, there is no sense in which the Indigenous population can be described as clearly defined for statistical purposes. Rather, political and cultural processes, including the highly variable way in which states, territories and the Commonwealth have attempted to enumerate and categorise Indigenous people and the choices made by respondents to these efforts, produce the statistical construct referred to as ‘the Indigenous population’ (Smith 1980a; Langton 1981; Jordan 1985; Dodson 1994; Anderson 1997; Ross 1999, pp. 2–10). In the more distant past, these sociological and political

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 25

processes have effectively excluded or devalued Indigenous representation in official statistics. In the more recent politics of data collection, every effort is made to encourage identification (Barnes 1996).

P O P U L AT I O N G R OW T H The consequence, since 1971, has been a 254 per cent increase in the census count of Indigenous Australians, from 115 953 to 410 004 at the 2001 Census. This averages out at an annual increase of around 8 per cent. For the 1996–2001 intercensal period, only two-thirds of the 16 per cent increase in population count could be accounted for by demographic components of change. While the Australian Bureau of Statistics (ABS) attributes this error of closure to an increase in the propensity of individuals to declare Indigenous status on the census form (ABS 2002d, p. 2), striking correspondence in age profiles between ‘old’ and ‘new’ populations has been consistently observed (Gray 1997; Kinfu and Taylor 2002), suggesting the influence of more structural factors possibly related to enumeration procedures. Ultimately, of course, the reasons for high error of closure in selfidentified populations remain unknown. In terms of establishing transformations in the population, this open-ended equation presents a number of dilemmas. First, meaningful intercensal comparison of social indicators requires that base year indicators be adjusted to the level of the newly revealed population using reverse survival techniques (Taylor and Bell 1998). Second, in situations where there is a large error of closure, as is the case in Tasmania for example, it is not clear whether aggregate change observed in population characteristics over time involves an alteration in the circumstances of the original population or whether it merely reflects the particular features of individuals appearing in the population for the first time. The problem for the analysis of transformation is that change in the condition of the original population is undetectable. All that can be noted is different aggregate status in respect of ‘different’ populations. Finally, it undermines the robustness of population estimates and projections. Partly for this reason ABS refers to Indigenous estimates and projections as ‘experimental’ and, unlike for the total population, projects only over a 10-year period (ABS 1998a). The main issue concerns the modelling of change in the propensity of individuals to identify as Indigenous on census forms, as different assumptions regarding this propensity can produce widely differing population estimates as illustrated in Table 2.1.

26 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 2.1 Projections of the Indigenous population under varying assumptions about propensity to identify, 1996–2006 As at 30 June 2006 State and As at Nil Constant 1986–96 1991–96 annual change rate of rate of Territory 30 June 1996 changea of 1% change changeb ’000 ’000 ’000 ’000 ’000 NSW Vic Qld SA WA Tas NT ACT Australia NOTES a SOURCE

109.9 22.6 104.8 22.1 56.2 15.3 51.9 3.1 386.0

132.7 26.5 133.3 26.6 67.0 18.0 60.6 4.1 469.1

148.3 29.7 148.5 29.7 74.8 20.2 67.7 4.6 523.6

181.8 35.1 158.3 29.9 72.3 28.7 62.5 7.3 576.1

216.3 33.1 179.3 32.2 79.6 36.4 64.0 7.8 649.0

Published low series; b Published high series. ABS 1998a, p. 13.

The published low series ABS projections, which yield an Indigenous population of 469 000 by 2006, assume no further growth due to increased identification as Indigenous. On the evidence of past census counts this seems an unreasonable expectation and so higher projections are provided based on varying assumptions about increased propensities to identify. Of these, the published high series assumes a continuation of the rate of new identification observed over the 1991–96 intercensal period. This yields a population of 650 000 by 2006. The effect of these different assumptions on the estimated size of the population is best illustrated at the state and territory level. For example, in Tasmania where negative growth is projected for the population overall, the high series Indigenous projection points to a rise in the Indigenous share of the population from 3 per cent of the state total in 1996 to as much as 8 per cent only ten years later. In the Northern Territory, on the other hand, little effect is predicted. While this modelling suggests that the manner in which individuals respond to the census question on Indigenous status has substantial consequence for future population size and distribution, there are indications from analysis of post-enumeration survey data that point to greater stability of Indigenous identification than prima facie census evidence implies (Hunter 1998).

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 27

A further reason for the experimental nature of the projections derives from increasing levels of marriage between Indigenous men and non-Indigenous women and the difficulty involved in accurately modelling the fertility implications of this process. As the progeny of such liaisons are invariably identified as Indigenous, this results in far higher numbers of Indigenous births than would be the case if the only factor was the fertility of Indigenous women. Overall between 1996 and 2001, births to non-Indigenous women accounted for more than one-quarter of all Indigenous births, and the evidence which is starting to emerge from improved vital statistics is that the proportion of Indigenous births to non-Indigenous mothers has been increasing across the country (Gray 1998). The effect is that even as Indigenous women move towards replacement fertility levels, the additional contribution of Indigenous births to non-Indigenous mothers will continue to boost Indigenous population growth well above the level achieved by the rest of the population.

P O P U L AT I O N D I S T R I B U T I O N — AN URBAN MYTH? One of the more obvious transformations of the Indigenous population in the second half of the 20th century has been a shift in overall geographic distribution away from remote and rural areas in favour of urban and metropolitan centres and consequently towards the south and east of the country. Over the longer term, this process may be viewed as an effect of the European settlement of Australia: the original dispersed distribution of Indigenous peoples broke down as individuals and families moved, or were moved, into government and mission settlements, reserves, towns and cities. Over the shorter term, there are issues unresolved as to whether demographic or sociological processes are more responsible for this redistribution; indeed, just how much redistribution actually has occurred. The proportion of the Indigenous population resident in urban areas rose from just over two-thirds in 1991 (67 per cent) to almost three-quarters in 2001 (74 per cent). Consequently, almost one-third of Indigenous Australians are now resident in major urban areas and while this is still substantially less than the total population (66 per cent), it nonetheless represents a marked increase from the 15 per cent of the Indigenous population counted in 1971 (Table 2.2). As this process of ever-greater population counts in urban areas has unfolded, ipso facto the rural share of the population has continued to decline: down from 56 per cent in 1971 to almost one-quarter in 2001.

28 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 2.2 Census counts of the Indigenous population by section-of-state, 1971 and 2001 1971 2001 No. % No. % Major urban 17 332 14.9 127 505 31.1 Other urban 34 076 29.4 175 322 42.8 Rural 64 545 55.7 107 177 26.1 Total 115 953 100.0 410 004 100.0 SOURCE

ABS Census of Population and Housing, unpublished data.

If anything, these figures understate both the extent and rise of urban living, especially in terms of proximity to metropolitan centres and large cities. ABS criteria for classifying Collection Districts (CDs) as urban or rural are based on measures of population density, landuse and spatial contiguity. This means that many people who may reasonably be regarded as forming part of a city region are not classified as urban dwellers. One way of incorporating such populations is to examine distribution according to the Statistical Divisions (SDs) that are coincident with each major urban area, thereby incorporating populations regardless of land-use and density measures. According to this measure, in 1991 a total of 70 872 Indigenous Australians (27 per cent of the population) lived in major urban SDs. By 1996, this figure had risen to 128 452 (36 per cent of the Indigenous population). Initial research on the causes of apparent urbanisation focused on the role played by migration, especially to metropolitan areas (Gale 1972; Burnley and Routh 1985). Subsequent analysis, however, has pointed to the likelihood that migration to major cities contributed less to Indigenous urban population growth than previously assumed (Smith 1980c; Gray 1989), and that much of the apparent shift in population distribution from the 1960s onwards could have been due simply to increased enumeration of city-based residents. Thus, while data on population distribution point to a process of very rapid urbanisation, this is, in part, an urban myth. Two observations support this. The first stems from analysis of inter-regional migration flows. Since at least 1976, the overall effectiveness of migration flows in redistributing the Indigenous population between metropolitan and non-metropolitan areas has been very low with net migration gains mostly to Brisbane and Perth offset by persistent net losses mostly to Sydney and Melbourne (Gray 1989; Taylor and Bell 1996,

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 29

pp. 400–02, 1999). Part of the perception that rapid urbanisation was underway derived from the focus of early migration studies on movement into cities with almost no attention paid to the net effect of movement out which had also been significant (Gale and Wundersitz 1982, p. 39; Gray 1989). The second observation derives from the general lack of association observed between net migration and regional population growth, especially in regions dominated by urban-based populations. The suggestion here is that part of the process of ‘urbanisation’ has been a realisation that most Indigenous Australians are resident in the cities and regional centres (Langton 1981).

N E T M I G R AT I O N A N D R E G I O N A L P O P U L AT I O N G R OW T H As might be expected, a strong positive relationship exists among the non-Indigenous population of Australia between regional net migration gain and regional population growth. This is clearly illustrated in Figure 2.4 which charts population growth against net migration for the non-Indigenous population aged 5 years and over. Put simply, regions (Statistical Divisions) that experience growth in nonIndigenous population do so largely because of net gains from migration. Conversely, those experiencing decline do so mostly because of net movement out. While the form of this relationship also holds for the Indigenous population, the association is much weaker (Figure 2.5), with many regions experiencing population growth (substantial at times) far above expectation given their prevailing net migration rate. This is underlined by the fact that some regions display high population growth despite experiencing negative net migration. Overall, this lack of association can be traced to nondemographic factors in population growth, mostly an increased propensity for individuals to identify as Indigenous in the census. Accordingly, the data points in Figure 2.5 can be grouped into regions where net migration is a very poor indicator of population growth, and those where population growth is close to expectation on the basis of net migration. Prominent among the former are regions which have high population growth rates despite experiencing net migration loss. These include the highly urbanised regions such as Sydney, Central New South Wales, and Melbourne. Also included are regions where population growth rates far exceed net migration gain such as Moreton, Brisbane, Darling Downs,

30 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 2.4 Relationship between Statistical Division population growth and net migration: non-Indigenous population, 1991–96

SOURCE

Taylor and Bell 1999.

Figure 2.5 Relationship between Statistical Division population growth and net migration: Indigenous population, 1991–96

SOURCE

Taylor and Bell 1999.

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 31

Mid-North Coast New South Wales, Hunter, Illawarra, Canberra and Hobart. Regions which more or less conform with expectation are found mostly in remoter parts of the country, especially in the north. These include the Pilbara and Kimberley regions of Western Australia, Darwin, Northern Territory balance, North West and Far North Queensland, the Far North and Eyre regions of South Australia, the Wimmera region in Victoria and Northern New South Wales.

R E M OT E A U S T R A L I A — A N I N D I G E N O U S D O M A I N ? Reference to ‘remote’ Australia is long-standing in regional analysis. Essentially, the term draws attention to a distinction in social and economic geography between closely settled areas and sparsely settled areas, with economic development and service provision severely impeded in the latter by force of relative locational disadvantage, low accessibility, and a specialisation of economic activity (Holmes 1988). This conception is now incorporated as part of the Australian Standard Geographic Classification (ASGC). In 1996, 28 per cent of the Indigenous population was estimated to be resident in remote and very remote areas as defined by the ASGC compared to only 2 per cent of the non-Indigenous population. In the same year, Indigenous people constituted as much as 42 per cent of the resident population in very remote areas and 11 per cent in remote areas. It is worth noting the parallels here with historic distinctions drawn between ‘colonial’ and ‘settled’ Australia in recognition of the much higher proportions of Indigenous people in remote areas, and the somewhat different manner of their incorporation into wider social and economic structures (Rowley 1971). Indeed, away from the larger mining towns and service centres of the outback, it is possible to talk of Indigenous ‘domains’ in the sense that Indigenous people and their institutions predominate. In recent years, moreover, there has been a substantial transfer of land back to Aboriginal ownership and stakeholder interest across remote regions, with the prospect of more to come via land purchase and native title claims (Pollack 2001). This land transfer is an important element if the post-productivist transition in Australia’s rangelands (Holmes 2002), and newly recognised land values often lie outside the market economy, being more culturally based. These values are manifest in the emergence of a distinct settlement structure on Aboriginal lands involving the formation of numerous, dispersed,

32 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

small, and discrete Indigenous communities, especially in the Northern Territory, Western Australia and the far north of South Australia and Queensland. In ABS parlance, discrete communities are defined as geographic locations that are separated by physical or cadastral boundaries, and inhabited or intended to be inhabited predominantly by Indigenous people (more than 50 per cent), with housing and infrastructure that is either owned or managed on a community basis. Such communities represent Indigenous living areas many of which are formerly constituted as government and mission settlements, or reserves. They also include special purpose lease areas within towns, as well as excision communities on pastoral stations. Many of these communities were established for the purpose of administering Aboriginal welfare policies, or simply as camping areas removed from white society. As such, they required no modern economic base, nor have they subsequently acquired one, at least not in a manner that is presently sustainable beyond the provisions of the welfare state. In the latter decades of the 20th century, demographic trends in remote areas have been volatile. With many Indigenous people residing close to their ancestral homes, their attachment to such places is reflected in a relative lack of net out-migration (Gray 1989; Taylor 1992; Taylor and Bell 1996, 1999). This contrasts with the historically more recent and ephemeral non-Indigenous settlement of the outback with, in recent decades, generalised out-migration leading to population decline in many non-metropolitan districts (Bell 1992b, 1995; McKenzie 1994; ABS 2002d). Between 1981 and 1996, the Indigenous population in areas approximating the remote and very remote categories of the ASGC grew by 23 per cent. By contrast, since 1986, aggregate non-Indigenous population growth in this same area was negative (Taylor 2000). As a consequence, the Indigenous share of the ‘outback’ population is estimated to have increased from 12 per cent to 17 per cent over this same period (Taylor 2000). Thus, in contrast to the image of downward spiral evoked for much of remote Australia, the progressive granting of Aboriginal land rights and financial means to self-determination has led to increased economic activity and a major transformation of Indigenous settlement structure since the 1970s. This is most clearly manifest in the purposeful shift of population towards smaller scale, dispersed clan-based settlements referred to as ‘outstations’ or

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 33

‘homeland centres’ (Coombs, Dexter and Hyatt 1982; Commonwealth of Australia 1987; Taylor 1992). Collectively, in 1999 there were almost 1000 such localities of less than 200 persons across northern and central Australia with a population, semi-permanently at least, estimated at around 29 000 (Taylor 2002). This redistribution and dispersal of rural population is most advanced in the Northern Territory as a consequence of the growing influence of Aboriginal self-determination based on legal access to traditional lands, government spending and income from royalties. Visual representation of this is provided in Figure 2.6 which shows the locations of Aboriginal settlements in the Northern Territory in 1970 compared to the situation some 20 years later. From the bureaucratic perspective of those seeking to provide services and achieve social and economic equity and efficiency goals, such a focus by Aboriginal people on using and residing on Aboriginal lands may be construed as a retrograde step on the grounds that it serves to reinforce the locational disadvantage of an already severely disadvantaged group. From an Indigenous perspective, however, this spatial transformation may be viewed as locationally advantageous given the importance attached to living on one’s own country and the associated capacity to fulfil cultural obligations and assume a degree of autonomous existence in smaller and more socially viable residential units within an Aboriginal domain. The fact is, one of the lasting impacts of European settlement has been a redistribution of the Indigenous population into a wide variety of locational settings, though with an emphasis still on non-metropolitan residence. This provides for quite different structural circumstances in regard to the manner and degree of Indigenous articulation with wider economic and social systems. This locational diversity together with rapid population growth presents a fundamental challenge for policies aimed at improving the economic wellbeing of Indigenous peoples. What then are the implications of demographic transformation for Indigenous economic status?

CHANGING NUMBERS—CHANGING NEEDS? There is now a substantial literature detailing the relatively low economic status of Indigenous Australians and examining underlying causes over the past 30 years (Altman and Nieuwenhuysen 1979; Fisk 1985; Altman 1991; Taylor 1993; Taylor and Hunter 1998). Common threads in determinants of poor economic outcomes have

34 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

Figure 2.6 Distribution of outstation localities in the Northern Territory, c1970 and 1989

SOURCE

ABS 1990, p. 12.

1970–2030

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 35

remained focused around the themes of low human capital endowments and the historic legacy of exclusion from the mainstream provisions of the Australian state. Also found to be of relevance is locational disadvantage with a large share of the Indigenous population far removed from mainstream labour markets. In this context, the prospect that greater Indigenous presence in urban, and particularly metropolitan, areas might alter the overall profile of disadvantage appears unlikely given that locational disadvantage is also characteristic of Indigenous populations within the cities. This is well illustrated in Figure 2.7, which indicates change in population distribution across major urban CDs grouped into deciles of socioeconomic status. Figure 2.7 Distribution of Indigenous and non-Indigenous populations in major urban Collection Districts ranked by SES index, 1991 and 1996

SOURCE

Taylor 1997b.

It is clear that Indigenous people are over-represented in the poorest city neighbourhoods and this pattern appears stable over time despite substantial growth in the major urban population. Moreover, within these poorest neighbourhoods Indigenous people continue to display the worst economic outcomes. While unemployment rates, for example, are typically very high in such areas, Indigenous rates are highest. To take just one example from the 1996 Census, the Elizabeth area of North Adelaide had one of the nation’s highest urban unemployment rates for non-Indigenous people at

36 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

23 per cent. Among Indigenous residents of Elizabeth, however, the unemployment rate was 52 per cent. Analysis of social and economic conditions for the general population in Australian cities has highlighted a spatial dimension to the emergence of an underclass (Gregory and Hunter 1995). Clearly, the continuing gap in life chances between Indigenous people and the rest of the urban community suggests that race forms an added dimension. Overall, despite a 44 per cent increase in the level of Indigenous employment between 1991 and 1996, the underlying rate of employment was little altered, with only 26 per cent of the working-age population engaged in mainstream work (Taylor and Bell 1998). There are two reasons for this anomaly. First, much of the recorded employment growth was due to increased participation in the Community Development Employment Projects (CDEP) scheme which is an Indigenous-specific work-for-the-dole scheme. In 1996, this accounted for as much of one-fifth of Indigenous employment. Also contributory was an increase in wage-subsidised employment and training under the Federal Labor Government’s Working Nation initiatives. Against these government-sponsored labour market interventions, growth in mainstream work was negligible. The other reason is demographic: quite simply, growth in jobs, especially mainstream jobs, has failed to keep up with growth in the workingage population. Consequently, unlike the general population for whom the consequences of old age will be increasingly apposite, the ascendant issues for Indigenous social policy are more to do with needs in the school-to-work transition years and in the years of prime working age and family formation. It is conservatively estimated that the number of Indigenous adults aged 15–55 years will be greater by 70 000, or 33 per cent, by 2006. This is substantially above the projected increase for the rest of the population in this age range which is only 6 per cent. This growth in working-age population has major implications for the future economic status of Indigenous people as it suggests much greater effort just to maintain the status quo. Just how much effort, in terms of job needs, has been estimated by Taylor and Hunter (1998). Just to maintain the status quo (an employment rate of 39 per cent and an unemployment rate of 26 per cent) would require 25 000 extra jobs by the year 2006, whereas on current trends only 21 000 are expected to be created (Table 2.3). To achieve employment equality with the rest of the Australian population, an additional

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 37

77 000 Indigenous people would have to be employed resulting in an overall deficit of some 55 000 jobs. If the focus is on mainstream job requirements, excluding opportunities provided by the CDEP scheme, then the backlog in the number of jobs needed is projected to be much larger, and greater than the number presently employed. Early results from the 2001 Census point to slightly reduced Indigenous employment and labour force participation rates and suggest that even these estimates might be conservative. Table 2.3 Required Indigenous employment growth to maintain the status quo or to achieve employment equality, 1996–2006 Employment/ population ratio

Base employment 1996a

Total jobs required by 2006

New jobs required

New jobs projected

Projected job deficit by 2006

38.9b

90 212

115 307

25 095

21 444

3 651

56.4c

90 212

167 181

76 969

21 444

55 525

NOTES a The estimated number of Indigenous Australians in employment in 1996; b The estimated employment/population ratio for Indigenous Australians based on 1996 population estimates; c The employment/population ratio for non-Indigenous Australians from the 1996 Census. SOURCE Taylor and Hunter 1998.

Thus, in seeking economic transformation of the Indigenous population, it is clear that the time available for decisive action is decreasing rapidly. In respect of employment status, for example, the vital issue for Indigenous policy this century is the distinct prospect that the overall situation will deteriorate. This is primarily because of population growth, but also because of the enormous difficulties of economic catch-up in a rapidly changing and skills-based labour market.

CONCLUSION At the commencement of the 21st century, the Indigenous population of Australia is well into an expansionary phase which one commentator has referred to as an ‘explosion of Aboriginality’ (Gray 1997). Though inconsistency has been a hallmark of census counts of Indigenous Australians, the trend in overall numbers since 1971 has been invariably upwards with growth rates often far above the level accounted for by natural increase. Reasons for this anomaly have been the subject of much speculation with the debate centred around

38 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

the relative effects of an increased willingness of individuals to declare their Indigenous identity in official statistical collections and the greater efforts made by the ABS to achieve better enumeration (Gray 1997; Ross 1999). Given this context of statistical uncertainty regarding Indigenous population dynamics, and in the interpretation of census data purporting to establish this, it is worth recalling the Commonwealth’s three-part definition of an Indigenous Australian: that an individual has Aboriginal or Torres Strait Islander descent; identifies as an Aboriginal or Torres Strait Islander; and is accepted as an Aboriginal or a Torres Strait Islander by the community in which he or she lives. The fact is, of course, that the Indigenous population revealed by the census could only conform with the first and/or second of these criteria (to the extent that these are invoked by the census), and even then only to the extent that a collection of individuals anonymously tick the appropriate box on a census form which asks if they are of Aboriginal or Torres Strait Islander origin. While the third of these criteria may not always be applied when recording Indigenous status in administrative statistical collections, its lack of application in the census methodology means that the census-derived Indigenous population would almost certainly be of a different size to any population based on the full Commonwealth definition. This effectively raises the prospect of different Indigenous ‘populations’ eventuating in different statistical contexts, with that derived from the census being just one of these, though probably the most inclusive, net of any census error. Conceptually, it should be noted that as long as the census question on Aboriginal and Torres Strait Islander origins remains the primary means of delimiting the Indigenous population, then it is likely that the numbers identified in this way will continue to rise steadily owing to improved enumeration, increased self-identification and a growing pool of potential identifiers due to the expansionary effects of intermarriage (Gray 1998; ABS 1998). Given growing pressures for targeted service delivery that is cost-effective and based on demonstrated need, this prospect of an ever-expanding population is likely to draw increasing scrutiny over time. As research on self-identified Indigenous populations in the United States has indicated, this poses a paradox for public policy in that Indigenous populations are considered discrete and homogeneous when in reality they are likely to become less discrete, less homogeneous and more difficult to define unambiguously (Snipp 1997, p. 675).

I N D I G E N O U S A U S T R A L I A N S : T H E F I R S T T R A N S F O R M AT I O N

• 39

Nonetheless, within the time frame set by the present analysis, and for some years beyond, there remain several compelling arguments for an enhanced focus on the demography of Indigenous peoples within the overall analysis of Australian population transformations. Locational arguments: despite national minority status and growing urbanisation, there is considerable regional diversity of Indigenous representation and the demography of large tracts of northern and central Australia is effectively the demography of Indigenous inhabitants. In particular, it should be recognised that the Indigenous share of the resident outback population rose steadily between 1981 and 1996 and currently approaches one-fifth of the total. Conceptual arguments: it is clear that the demographic structure and behaviour of the Indigenous population is not simply a subset of the pattern observed overall and that there are unique historical, cultural and structural factors which produce distinct outcomes. Social science arguments: while Borrie’s (1994) ‘yawning gap’ is slowly closing, fundamental areas of research remain undeveloped with too few researchers dedicated to this endeavour. There is need for work in particular on the socio-economic and cultural precedents of demographic transition, on the complex processes that shape ethnic identity and on the demographic consequences of intermarriage. Contextual arguments: a history of widespread dispossession of land and subsequent existence on the margins of Australian society has created a unique social and economic context for transformation of the Indigenous population. Viewed internationally, however, much of the demographic experience described above has been shared with similarly encapsulated populations in North America and New Zealand, though often with important differences (Kunitz 1990). Consequently, much can be learnt about the potential future transformation of the Australian Indigenous population by crossnational comparison. Social justice and policy arguments: Australian governments are committed to policies aimed at achieving social justice for Indigenous people and improving their socio-economic well-being. Since the broad parameters for this charter are determined by the size, growth, composition and changing location of the Indigenous population, then only by accurate monitoring of change in these factors can needs be adequately assessed and resources fairly and equitably distributed.

3

FERTILITY TRENDS AND DIFFERENTIALS GORDON A. CARMICHAEL and PETER McDONALD

In theory it is possible to study the fertility of women or men, but in practice, for several reasons, it is the numbers of children women have, the ages at which they have them, and the intervals between those births that are important in demographic studies. Not least among these reasons is women’s greater certainty as to, and hence reliability as informants on, births in which they are biologically implicated. Reliable data are thus more readily available for them, from both official sources and special surveys. Women’s lives are also more profoundly affected by fertility than men’s. Their personal health is placed at risk, and the expectation in most societies that they will be the primary carers for young children constrains their life options in ways that men’s are much less likely to be constrained. This chapter outlines recent trends and differentials in childbearing among Australian women. It draws primarily on vital registration and census data to highlight change that has occurred, likely future change, and differences in fertility behaviour across subgroups of women defined by various characteristics. It also looks at the development since the 1970s of an increasing detachment of childbearing from formal marriage.

T R E N D S I N F E R T I L I T Y A N D F A M I LY S I Z E Fertility is conventionally measured, in summary form, by the total fertility rate (TFR). This index measures the average number of children a group of women would have during their lifetimes if they experienced at each age the fertility rates recorded at those ages in a given

FERTILITY TRENDS AND DIFFERENTIALS

• 41

year. For example, Australia’s TFR in 1961 was 3.6 births per woman. This means that if fertility rates at each age in 1961 had applied across the lifetimes of a group of women, those women, on average, would have had 3.6 births each on reaching the end of their childbearing lives (approximately their 50th birthdays). At the low levels of mortality prevailing in Australia in the 1990s, one generation would exactly replace itself if its average fertility rate was 2.07 births per woman: 2.05 births to replace the woman, her male partner and males unpartnered because more male than female children are born (about 105 males for every 100 females); 0.02 of a birth to help replace those who died before having the opportunity to replace themselves. The trend in the TFR in Australia from 1945 to 2000 is shown in Figure 3.1. The TFR rose during the Baby Boom years from 1945 to 1961, when it reached a peak of 3.6 births per woman. It then fell sharply in the first half of the 1960s as the introduction of the oral contraceptive pill enabled women at older reproductive ages to better avoid unwanted higher parity births, and younger women to delay having children after marriage. After levelling off at 2.8–2.9 births per woman during 1967–71, the TFR again fell precipitously, to 1.9 births per woman in 1978. It remained at around this level until the early 1990s, but since 1992 there has been a further slow but sustained decline. From 1.89 in 1992, the TFR had reached 1.75 births per woman by 1999. It remained at 1.75 in 2000, the lowest level ever recorded in Australia. If this level were to be sustained over an extended period, the next generation would be 16 per cent smaller than the current generation of childbearing age. Figure 3.1 Total fertility rate, Australia 1945–2000 4.0

Total fertility rate

3.5 3.0 2.5 2.0 1.5 1.0 1945 SOURCE

1950

1955

1960

1965

1970

ABS, Births Australia 2000, Table 8.1.

1975

1980

1985

1990

1995

2000

42 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

As a measure based on the annual number of births, the TFR is affected by short-term influences such as economic recessions and booms, but more significantly it is affected by shifts in the timing of births. If women shift towards having their children at younger ages than has been the case in the recent past, births will tend to accumulate in the present and the TFR will rise. Such a shift explains much of the rise in the rate from 1945 until 1961. During this period there was a substantial fall in the ages at which women had their first children, related, in turn, to a shift to much earlier ages at marriage (McDonald 1975; Carmichael 1988). The average age of women when first giving birth fell from 26.4 years for women born in 1908–12 to 23.3 years for women born in 1933–37. The latter group also recorded the highest average level of fertility reached by a cohort of women born during the 20th century, 3.0 children, a level well below the 1961 peak of the TFR (Jain and McDonald 1997). The fact that the peak in cohort fertility was achieved by the same group who were youngest when they began having children is not without significance. It supports an argument that age at the commencement of childbearing is a major determinant of ultimate family size. The average age at first birth stopped falling in the early 1960s, contributing to the downturn in the TFR, but it remained at a low level into the early 1970s. This might have been due partly to the introduction in late 1964 of conscription at age 20 for military service in Vietnam, and the provision that a man could defer any liability to serve if he had dependants upon whom ‘exceptional hardship’ would be imposed. In effect, being married before age 20 was a way to avoid conscription, and the male first marriage ratio at age 19 did rise from 37 to 47 per 1000 between 1964 and 1965, remaining at about that level for several years thereafter. But with the chance of being conscripted being less than 10 per cent (the system was based on balloting birth dates), early marriages to avoid it can be overemphasized. A rising incidence of unplanned youthful childbearing stemming from increased non-marital sexual activity (Carmichael 1996) was probably a more potent factor in offsetting the deferment of first births after marriage to which the new contraceptive technology gave rise (Ruzicka and Choi 1981; Carmichael 1988). Since the mid-1970s, the trend to early childbearing has been reversed, and women have been having their first children at increasingly older ages. If the commencement of childbearing is postponed,

FERTILITY TRENDS AND DIFFERENTIALS

• 43

a fall in the TFR is expected as births are deferred. The combination of a delay in childbearing and reduced childbearing across the lifetime produced the steep fall in the TFR during the 1970s. An important facilitator of this trend was the emergence of sterilisation as a popular terminal method of contraception (Young and Ware 1979; Santow 1991). The trend in the TFR over the past 20 years is more difficult to interpret. An examination of more detailed data is required. Delay in the commencement of childbearing has continued up to the most recent year for which data are available. In other words, for more than 25 years, women have been commencing their childbearing at increasingly older ages. This period is so long that the older generations that began the trend have been making up their delayed births at the same time as younger ones have been perpetuating it. Also to be factored in is the reality that a birth delayed is one that may never occur, and the longer it is delayed the more likely is its non-occurrence. Thus, Australia’s near constant TFR from 1978 to 1992 does not indicate a long period of stability in the nation’s fertility. Rather, it is the net result of very complex underlying processes. The delay in the commencement of childbearing is an international phenomenon that has occurred in all advanced countries over the past quarter century. Observation of historically low fertility rates in the mid-1980s led demographers in many countries to predict that the TFR would rise again as delayed births were made up. As Table 3.1 shows, some countries, most notably the Nordic and Benelux countries, and the United States, Canada and New Zealand, did experience rises in the TFR during the late 1980s. Australia, however, was among those countries where the rate remained almost flat during this period, before falling in the 1990s. Other countries with this pattern were the United Kingdom, France, Austria, Germany and Hungary. In a third set of countries, including the Southern European countries, most of the former Eastern bloc countries, Ireland and Japan, fertility fell throughout the decade from 1985 to 1995. This lack of international consistency also makes it difficult to predict future trends in the TFR. However, it is evident from Table 3.1 that between 1990 and 1995 the trend was downward in almost all countries. The argument that fertility will rise in the future because births postponed in the 1980s will be made up now looks very thin. Nevertheless, as noted above, a more detailed examination of the underlying trends is required.

44 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 3.1 Total fertility rates for various countries, 1985, 1990 and 1995 Country

1985

1990

1995

Australia

1.89

1.91

1.82

New Zealand

1.93

2.18

2.04

United States

1.84

2.08

2.01

Canada

1.61

1.71

1.64

United Kingdom

1.79

1.83

1.71

Ireland

2.48

2.15

1.87

Sweden

1.74

2.13

1.73

Denmark

1.45

1.67

1.80

Norway

1.68

1.93

1.87

Finland

1.64

1.78

1.81

France

1.81

1.78

1.70

Belgium

1.51

1.62

1.57

Netherlands

1.51

1.62

1.53

Switzerland

1.52

1.58

1.48

Austria

1.47

1.45

1.40

Germany

1.37

1.45

1.25

Italy

1.42

1.33

1.17

Spain

1.64

1.36

1.18

Portugal

1.72

1.51

1.38

Greece

1.67

1.39

1.32

Poland

2.33

2.04

1.61

Czech Republic

1.96

1.89

1.28

Hungary

1.85

1.87

1.57

Russia

2.05

1.9

1.34

Japan

1.76

1.54

1.42

SOURCE

Monnier 1998.

FERTILITY TRENDS AND DIFFERENTIALS

• 45

T R E N D S I N F E RT I L I T Y R AT E S AT E AC H AG E This examination is facilitated by Figures 3.2 and 3.3, which show trends in fertility rates at individual ages. Note that the TFR for any year is the sum of the fertility rates at individual ages for that year. Figure 3.2 shows that, in 1980, only the rates at the youngest ages were falling (ages 20–22 years). However, moving through the 1980s, one after the other the rates at each successively older age began to fall (age 23 after 1981, age 24 after 1982, age 25 after 1983, age 26 after 1984, age 27 after 1985). All of these declines were led by women born around 1958. These were women whose entire adult lives had been lived in the new era of social behaviour that bloomed in Australia in the 1970s. Figure 3.2 Age-specific fertility rates, aged 20–29, Australia 1980–2000 180 160 140

Fertility rate

120 100 80 60 40 20 0 1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Age 20

Age 22

Age 24

Age 26

Age 28

Age 21

Age 23

Age 25

Age 27

Age 29

SOURCES Calculated from ABS Births Australia (annual volumes), Catalogue No. 3301.0, and ABS, Estimated Resident Population by Sex and Age: State and Territories, of Australia (various issues), Catalogue No. 3201.0.

At ages 28 and 29 there were clear rises in the fertility rates in the early years of the 1980s. The rate at age 28 reached a peak in 1985, flattened, and then began to fall steadily from about 1990. That at age 29 began to fall only from 1993. Thus, the commencement of

46 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 3.3 Age-specific fertility rates, aged 30–39, Australia 1980–2000 180 160 140

Fertility rate

120 100 80 60 40 20 0 1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Age 30

Age 32

Age 34

Age 36

Age 38

Age 31

Age 33

Age 35

Age 37

Age 39

SOURCES Calculated from ABS Births Australia (annual volumes), Catalogue No. 3301.0, and ABS, Estimated Resident Population by Sex and Age: State and Territories, of Australia (various issues), Catalogue No. 3201.0.

the sustained fall in the TFR after 1992 (Table 3.1) approximately coincided with the point in time from which fertility rates at all ages in the 20s were falling. Figure 3.3 shows that, throughout the 1980s, age-specific fertility rates rose at all ages in the 30s. These rises apparently balanced the falls that were occurring at the same time at younger ages, leading to little or no change in the TFR. This period of equilibrium followed the 1970s, during which fertility had declined substantially among both women in their 20s, who were delaying marriage and parenthood, and those in their 30s, who having begun childbearing young, were using sterilisation and other reliable methods of fertility control to avoid unwanted third and subsequent births. Around 1980, the downward momentum of the latter trend was overtaken by the upward momentum the former generated above age 30 as deferred births began to be made up. More recently the fertility rate at age 30 has begun to fall again,

FERTILITY TRENDS AND DIFFERENTIALS

• 47

and rates at ages 31 and 32 have shown signs of levelling off. These developments very much dampen prospects for a future rise in fertility. While sharp rising trends are still evident for ages 34 and 35, these are of insufficient magnitude to counter the falls at younger ages. There is thus little support for the notion that fertility will rise in future because postponed births will be made up. It is extremely unlikely that more than a fraction of these births will be made up after age 35.

T H E AG E AT C O M M E N C E M E N T O F C H I L D B E A R I N G A useful summary measure of change in the age at commencement of childbearing is the age by which each birth cohort reaches an average of one child born (Figure 3.4). At the lowest point, women born in the late 1930s had had one child on average before their 24th birthdays. This represented a substantial fall compared to those born between 1910 and 1915 (the early childbearers’ mothers). Their age by which an average of one child had been born approached 28 years, this late commencement of childbearing being related to the delay of marriage that ensued from the Great Depression of the 1930s (McDonald 1975; Carmichael 1988). Figure 3.4 Age at which women born during 1906–71 had an average of one child 32

Exact age (years)

30

28

26

24

22 1906 1911 1916 1921 1926 1931 1936 1941 1946 1951 1956 1961 1966 1971 Year of birth SOURCE

Calculated from birth registration data.

48 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The age at commencement of childbearing has been increasing steadily for women born from 1940 onwards. For those born in 1967, the age by which an average of one child had been born was above 29 years, higher than the level for those whose childbearing was delayed by the Depression. There are limits to which the curve in Figure 3.4 can continue to rise, and it should be acknowledged that it is deriving some of its upward momentum from a trend for an increasing minority of women to avoid childbearing altogether (Merlo and Rowland 2000). But there is little sign of a slow-down in the extent of the rise. As postulated earlier, while the age at commencement of childbearing continues to rise, the TFR can be expected to continue to fall.

CHANGES IN THE NUMBERS OF CHILDREN T H AT WO M E N H AV E The numbers of children that Australian women aged 15–49 had at the time of the 1986 and 1996 censuses are summarised in Figure 3.5. Separate graphs show, by age, percentages for zero, one, two, three and four or more children. Between 1986 and 1996 the proportions with no children increased substantially. For example, at age 25 in 1986, 56 per cent of women had no children, this figure rising to 68 per cent by 1996. In 1986, 26 per cent of women at age 30 had no children, but by 1996 this had risen to 38 per cent. The increases were above 10 percentage points at all ages between 24 and 31, with a peak of 14 percentage points at age 27. The proportion of women having one child declined at younger ages, consistent with the deferment of childbearing below age 30 during the late 1980s and early 1990s, but rose beyond age 30. This latter trend also reflects the deferment of childbearing, women being more likely to be caught ‘midstream’ in forming families when in their 30s, and possibly also displaying a greater tendency to have only one child. The emergence of the two-child family as the Australian norm is evident in the big increases in the proportion of women aged in their 40s with that number of children, and the large decline in the proportion with four or more children. Below age 40, however, the proportion with two children declined, again highlighting deferred childbearing but possibly also increased voluntary childlessness. The proportion with three children fell at all ages, with the quite minor decline among women aged in their 40s of note. No doubt some who in earlier times might have had four or more children have scaled back to three.

FERTILITY TRENDS AND DIFFERENTIALS

• 49

Figure 3.5 Distribution of women by parity and age, 1986 and 1996 censuses Parity 0

100

80

80

%

60

%

60

40

40

20

20 0

0 15

25

35

15

45

1986

25

35

1986

1996

Parity 2

100

45 1996

Parity 3

100 80

80

p

y

Parity 1

100

60

60

g

%

40

40

p

%

20

20 0

0 15

25

35

15

45

1986

25 1986

1996

Parity 4+

80

p

y

100

g

60 40

p

%

20 0 15

25 1986

SOURCE

Unpublished 1996 Census data.

35

45 1996

35

45 1996

50 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

In summary, there is evidence of large increases in the proportions of women having no children and two children, a small increase in the proportion settling for one child and a small decrease in the proportion settling for three children, and a sharp fall in the proportion bearing four or more children. Obviously, when the average age at commencement of childbearing rises to close to 30 years, the chance of women having four or more children is restricted. McDonald (1998b) has predicted for women aged around 33 in the late 1990s that 22 per cent will ultimately have no children and 16 per cent only one child. Thirty-five per cent will have two children, 20 per cent three, 5 per cent four and 2 per cent five or more. Childlessness of around 20 per cent is not unusual in Australian history, having been the norm for women born from 1860 to 1910 (McDonald 1995). However, the percentage childless fell as low as 9 per cent for women born in the early 1930s, and this tends to be the point of reference for current trends. The average of the distribution by number of children just quoted is 1.77 children per woman, which is close to the current level of the TFR. As fertility rates are still falling for women younger than 33 years, continued decline in the TFR into the future seems almost certain.

F E RT I L I T Y P R E F E R E N C E S While 38 per cent of women aged around 33 in the late 1990s are projected to have no children or one child, survey data on fertility preferences suggest that these numbers of children are by no means the ideal. Evans and Kelley (1999, p. 13), describing the responses of a national sample of men and women of all ages, found that ‘Australians overwhelmingly reject childlessness’, and that ‘one child families are only a little more popular’. In the Australian National University’s Negotiating the Life Course Survey (NLCS), which interviewed a national random sample of 18–54-year-olds in late 1996 and early 1997, 10 per cent of women aged 20–24 expected to have no children and 5 per cent only one child. These figures are well below those projected above for women aged 33 in the late 1990s. Moreover, the expressed fertility expectations of women aged 20–24 equated with an average family size of 2.3 children, well above the level that would be predicted given a TFR headed towards 1.7 and below. Van de Kaa (1998) shows that women in most European countries have expectations when they are young about the number of

FERTILITY TRENDS AND DIFFERENTIALS

• 51

children they will have that, when they get older, they fall well short of realising. The extent of the reduction tends to be higher for those who have more to lose by reducing their attachment to the paid labour force. That is, it is higher for those for whom the lost earnings due to withdrawal from paid employment are highest. The NLCS also shows that the fertility expectations of Australian women who have completed secondary school and have a postschool qualification drop sharply from age group 20–24 (2.55 children on average) to age group 30–34 (1.81 children). By age group 30–34 this more highly educated group had actually had only 1.17 children on average, placing in question its chance of meeting its expected target of 1.81 children (McDonald 1998b). At the opposite end of the education spectrum, women aged 20–24 and 30–34 who had not completed secondary school and had no post-school qualification did not differ in the average numbers of children they expected (2.40 children). Furthermore, the achieved fertility of the older group (1.96 children on average) was closer to its expected fertility than was the case with the more highly educated women. Some commentators expect fertility in low-fertility countries to rise automatically in the future, because desired family sizes of young women are above current actual fertility levels. An alternative view is that achieved fertility will remain below early fertility expectations, as these are modified in the face of institutional constraints.

DIFFERENTIAL FERTILITY C O N C E P T A N D DATA As the term implies, the study of ‘differential fertility’ seeks to highlight differences in the level and pattern of fertility between discrete subgroups of a population. The analysis presented here uses data from the 1996 Census, the most recent census to include a question on number of children ever born alive (CEB) to women. The usual summary measure by which differential fertility comparisons are made is the mean, or average, number of children ever born (mean CEB) to a subgroup of women. It is often calculated specific for age, since the number of CEB is obviously highly variable by age, and apt to increase with it through the reproductive ages. It is frequently also standardised for age, to eliminate the effect of differences in age structure from an all-ages comparison. This procedure entails replacing the observed age structure for each subgroup being compared with that of a single,

52 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

standard population. The resulting age-standardised mean CEB values have no useful meaning individually, but in a comparative context their relativity to one another gives a more reliable indication of how much higher or lower the fertility is in one subgroup than in another.1 Standardisation may also be undertaken by marital status, since just as being older makes a woman likely to have had more children, so, too, does being (or having been) married. Two crucial shortcomings of the data used here, and assumptions made in addressing them, also need to be appreciated. First, there are invariably numbers of women for whom number of CEB, or parity, is not stated. This can occur for a variety of reasons. Women who have not yet had children (or others, often parents, responding on their behalf) may ignore the CEB question as irrelevant. Older women may consider their childbearing days so distant as to also consider the question no longer relevant. And it may additionally be bypassed either because it is deemed too intrusive or because an informant responding on behalf of someone else (e.g., an infirm older woman) simply does not know the answer. Second, the 1996 Census provides single parity detail for parities zero to five, but a final response option of ‘six or more’. Reflecting the reasons cited, the percentage of women failing to state CEB in 1996 fell from 16.6 at age 15 to 6.5 at age 20, was consistently below 4.0 at ages 32–49, then rose again to 6.3 at age 60, 9.8 at age 70, 14.6 at age 80 and 20.8 at age 89. The assumption made in dealing with these women was that their proportionate distribution by parity was the same as that of women the same age who stated their CEB. An alternative possibility was to assume all were childless (parity 0), but this is only realistic at the youngest ages where, because the vast majority of women with CEB stated are childless, the first assumption is a close equivalent anyway. In respect of women coded as having ‘six or more’ children, the assumption made was that on average they had seven. This seemed a reasonable figure based on a 1986 Census average of 7.16 children (more detailed data are available for 1986) and the likelihood that the average had fallen over the intervening decade.

M A R I TA L S TAT U S A N D C O N S E N S UA L PA RT N E R I N G Australian women aged 15 or older at the 1996 Census had an average of 1.81 children each. Predictably, those who had never been married recorded much lower fertility, about a fifth of the average for

FERTILITY TRENDS AND DIFFERENTIALS

• 53

all adult females after standardisation to that population’s age structure. All other marital status groups recorded above average agestandardised mean CEB. Divorced women exceeded the average by 9 per cent, married women by 13 per cent, separated women by 26 per cent, and widowed women by 33 per cent. Suspicion attaches to widows ranking highest here, since age-specific mean CEB data link that ranking to implausibly high figures at younger ages, where widowhood is distinctly uncommon (Figure 3.6). There appear to be aberrations in the data at these ages caused by census authorities’ practice of imputing unstated age and marital status. This seems to have spuriously created a very small, but influential, number of very young widows of improbably high parity. Figure 3.6 shows that at ages where they were more numerous, widows in fact ranked behind separated women, and at ages 35–44, behind married women as well. Figure 3.6 Average number of children by women’s age and marital status, 1996

Average number of children

4

3

2

1

0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75+

SOURCE

Never married

Separated

Married

Divorced

Widowed

Unpublished 1996 Census data.

Beyond ages 25–29, and especially at ages 30–49 where much marital disruption occurs, being divorced is associated with lower average fertility than is recorded for married women. This hardly surprises, as marital discord can be expected to often interrupt

54 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

childbearing, and even discourage its onset. Fertility is much higher among separated women, however, than among divorcees. Separation is an earlier phase in the marriage dissolution process, so that especially at childbearing ages the differential might reflect more recent marital breakdown and hence longer exposure to the risk of childbearing within marriage. It is also possible that women of higher parity have more limited repartnering options when marriages founder, and that this reduces the urgency to divorce. The persistence of the differential at older ages, and the clearly higher fertility of separated than married women at those ages, further suggest that separation is apt to be a less than transient state for higher-parity women of perhaps lower socio-economic status than for other women. Some of these women may even be ‘separated’ from de facto spouses, with divorce not an option. There is nothing strange about fertility levels falling at the oldest ages in Figure 3.6. Cohorts aged in their 60s in 1996, born in the late 1920s and early 1930s, were identified earlier as having had the highest completed fertility of any born in the 20th century. As to the pattern of age-specific mean CEB for never married women, three things underpin it: first, a contemporary morality widely accepting of childbearing within consensual unions; second, a bygone morality that held that spinsters should be childless, and should marry if pregnancy occurred; and third, a tendency, even today, for marriage not infrequently to eventually follow non-marital initiation of childbearing. Standardising for age, mean CEB for never married women who were cohabiting in 1996 was three times the figure for their noncohabiting counterparts, but only half that for all adult women. Never married cohabitors are thus a diverse group, some having embarked on childbearing while others are pursuing a childless phase that may in some cases be perceived as lifelong. Cohabitors of other marital statuses recorded age-standardised mean CEB figures marginally below (divorced, separated and widowed women) or above (married women) those of their non-cohabiting counterparts. The notion of a group of women who were both married and cohabiting perhaps seems somewhat incongruous. It probably comprised women who were not legally married, but perceived their consensual unions to be marriages. At younger ages (20–24, 25–29 and 30–34), the presence of children in consensual unions seems likely to have been associated with such a perception. Women claiming to be both married and consensually partnered at

FERTILITY TRENDS AND DIFFERENTIALS

• 55

those ages recorded appreciably higher mean CEB values than those who were just married (1.21 cf 0.67, 1.76 cf 1.10 and 2.07 cf 1.82, respectively).

E D U C AT I O N Differential fertility by level of education is anticipated, there being much previous research for Australia and other countries to suggest that pursuit of education, and the opportunities its acquisition opens up, have a marked negative effect on the urgency and intensity with which women engage in family formation (Carmichael 1988). The data do not disappoint. Table 3.2 is divided into three panels which separate women with university qualifications from those with other post-school qualifications (including a sizable group who claimed one, but failed to specify what it was), and three subgroups of women with no qualifications. Fertility is clearly lowest among women with higher degrees: about two-thirds of the national average after standardisation for age, rising to three-quarters after further standardisation for marital status. Standardising for marital status quite significantly boosts mean CEB for the first three university qualification categories and holders of associate diplomas (graduates of Technical and Further Education courses). This highlights the disincentives to marriage created by the serious pursuit of education and by post-graduation efforts to justify in the labour market the investment in education. The fertility of women with undergraduate diplomas more closely resembles that of women with vocational qualifications than that of other university-qualified women: below average, but by less than 10 per cent. Not dissimilar are unqualified women who remained at secondary school until at least age 17. Those who left school earlier than this, however, had experienced decidedly above-average fertility, even after standardising to control for their comparatively old age structure. Early school leaving generally means early entry into often rather mundane employment, which provides little disincentive to commencing childbearing comparatively young. And as noted previously, the earlier childbearing begins the larger completed family size is likely to be, modern methods of fertility control notwithstanding. There may also be an element of selection of women bent on motherhood as a career into the early school leaver group, which doubtless includes as well many who became group members as a direct result of unintended pregnancy. The high-fertility group in Table 3.2

56 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Level of post-school education attainment

Average number of children

Age-standardised number of children

Standardised index

Age/marital status standardised average number

Standardised index

Table 3.2 Average number of children of women aged 15 or older by level of post-school educational attainment, 1996

Higher degree Postgraduate diploma Bachelor degree Undergraduate diploma

1.24 1.47 1.17 1.86

1.22 1.42 1.41 1.64

68 79 78 91

1.37 1.54 1.50 1.67

76 85 83 92

Associate diploma Skilled vocational Basic vocational Qualification not stated

1.24 1.59 1.55 1.90

1.54 1.68 1.69 1.78

85 93 93 99

1.62 1.64 1.69 1.77

90 91 94 98

None Left school aged 17+ Left school aged <17 Never attended school

1.21 2.34 3.24

1.71 2.03 2.42

95 113 134

1.72 2.00 2.47

95 111 137

Total

1.81

1.81

100

1.81

100

NOTE Standardised to age and age/marital status structures of total Australian female population aged 15 or older. SOURCE Unpublished 1996 Census data.

who never attended school are chiefly older migrant women, many of whom had most of their children before arriving in Australia. Age-specific mean CEB profiles for the larger education groups (Figure 3.7) confirm differences just discussed. Of special note, however, is how much lower the fertility of higher degree holders is than that of first degree holders above age 30. Chiefly responsible for this are high propensities to have avoided marriage or, if marrying, to have remained childless. At some ages the proportion of higher degree holders never married in 1996 was five times the total population proportion. These more extreme ratios prevailed at older ages; cohorts for which, when women were younger, a professional career and family were apt to be portrayed and perceived as strict

FERTILITY TRENDS AND DIFFERENTIALS

• 57

alternatives. But even higher degree holders in their 40s were 2–3 times as likely to have never married as women in general. Moreover, among married women in 1996, higher degree holders were at ages from 35–39 to 60–64 consistently two-and-a-half times as likely to be childless as married women in general. Ratios were a little lower outside this age range. Figure 3.7 Average number of children by women’s age and post-school educational attainment, 1996

Average number of children

4

3

2

1

0 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74

SOURCE

75+

Higher degree

Skilled vocational

None – left school 17+

Bachelor degree

Basic vocational

None – left school <17

Unpublished 1996 Census data.

L A B O U R F O R C E AT TAC H M E N T Predictably, after standardising for age and marital status, extent of involvement in the labour force is inversely associated with fertility. Women who were not working recorded the highest standardised mean CEB of 2.03 children, followed by those working only 1–15 hours per week at 1.88 and those working 16–24 (1.78), 25–34 (1.70) and 35+ hours (1.57). Age-specific mean CEB profiles (Figure 3.8) strongly reinforce the negative relationship between fertility and commitment to employment. In particular, not working is associated with especially high fertility at ages 20–24, and being

58 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

employed full-time (35+ hours) with the lowest fertility among cohorts aged from 25–29 through to approaching retirement age. Figure 3.8 Average number of children by women’s age and labour force commitment, 1996

Average number of children

3

2

1

0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75+

SOURCE

Not working

16–24 hours

1–15 hours

25–34 hours

35+ hours

Unpublished 1996 Census data.

Age-specific data for married women and for never-married cohabiting women equivalent to those graphed for all women in Figure 3.8 also reveal clear negative associations between the extent of labour force involvement and mean CEB. Fertility levels for the latter group do, however, fall after age 35. This is symptomatic of transition from a situation where never-married cohabitors are having children whilst delaying marriage, to one in which survivors in the category beyond age 35 increasingly are women who intentionally have remained unmarried and childless.

URBAN-RURAL RESIDENCE Age-standardised fertility was clearly lowest (7 per cent below the national average) in major urban areas: cities of at least 100 000 population (Table 3.3). It was 11 per cent above average in other urban areas (towns and cities of 1000–99 999 people), and 17 per cent

FERTILITY TRENDS AND DIFFERENTIALS

• 59

above average in bounded localities (villages and towns of 200–999 people), with the rural balance of the population lying in between at 14 per cent above average. Further standardisation by marital status compresses the differentials slightly but does not change the overall pattern.

Urban-rural residence

Mean CEB

Age-standardised mean CEB

Standardised index

Age/marital status standardised mean CEB

Standardised index

Table 3.3 Average number of children of women aged 15+ by residence, 1996

Major urban Other urban Bounded locality Rural balance Total

1.66 2.03 2.22 2.11 1.81

1.69 2.00 2.11 2.06 1.81

93 111 117 114 100

1.70 1.98 2.08 2.00 1.81

94 110 115 111 100

Standardised to age and age/marital status structures of total Australian female population aged 15 or older. SOURCE Unpublished 1996 Census data. NOTE

The higher fertility of women living in ‘bounded localities’ was due mainly to the much higher fertility of young women in these areas. The fertility of women aged 15–19, 20–24 and 25–29 living in bounded localities was 2–3 times higher than that of their counterparts in major urban areas. To examine this phenomenon further, Table 3.4 shows the mean CEB for women aged 15–29 according to their marital status and locality of residence. Although women who were never married and not living with a partner had the lowest fertility on average, those living in bounded localities and other urban areas also tended to have much higher fertility compared with their counterparts in the major urban areas. Further research is needed to explain this much higher fertility among young unmarried women living outside major urban areas. Possible factors include (a) limited employment opportunity and, perhaps, the attraction of welfare income encouraging an early start to family formation; (b) selective

60 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 3.4 Average number of children born to women aged 15–29 by urban-rural residence and marital/de facto status, 1996 Major urban Age and marital/de facto status Mean CEB Index 15–19 Never married, de facto 0.30 100 Never married, not de facto 0.02 100 Married, not de facto 0.62 100 20–24 Never married, de facto Never married, not de facto Married, not de facto

0.34 0.10 0.59

100 100 100

25–29 Never married, de facto Never married, not de facto Married, not de facto

0.47 0.22 0.96

100 100 100

SOURCE

Unpublished 1996 Census data.

out-migration of women with higher educational and career aspirations to the large cities leaving behind the less well-educated; (c) inmigration of young unemployed parents to smaller, cheaper urban places; (d) concentration of the Aboriginal and Torres Strait Islander population in these smaller towns; and (e) remoteness from, and possibly less favourable attitudes towards, abortion services. It is notable, though, that in the genuinely rural localities, teenaged fertility is not much higher than in major urban areas (only 13 per cent higher; compared to 143 per cent higher in bounded localities). This might reflect isolation from potential partners, possibly coupled with rural conservatism, limiting social interaction, although classification as ‘rural’ of broadacre areas on the outskirts of major cities could also be a factor.

RELIGION After standardisation for age and marital status the fertility of Catholic women was 6 per cent above the national average, and that of Anglican women 2 per cent below it (Table 3.5). The Christian group with the highest fertility was the Pentecostals, followed by the

FERTILITY TRENDS AND DIFFERENTIALS

Other urban Mean CEB Index

Bounded localities Mean CEB Index

• 61

Rural balance Mean CEB Index

0.39 0.04 0.77

128 175 124

0.40 0.05 0.84

132 217 137

0.38 0.02 0.84

125 96 137

0.58 0.24 0.79

171 246 133

0.72 0.30 0.97

212 310 164

0.57 0.14 0.80

166 144 134

0.97 0.55 1.31

204 247 137

1.22 0.65 1.50

259 293 156

0.92 0.36 1.35

194 163 141

‘Other Christian’ category, which includes sects such as Jehovah’s Witnesses, Mormons, Seventh Day Adventists and the Salvation Army. Their fertility was 16 and 12 per cent respectively above the average. Greek Orthodox and Other Orthodox women had the lowest fertility, which was 9 and 13 per cent respectively below average. The two main non-Christian religious groups had well above average fertility. Buddhists were on a par with ‘Other Christians’, and followers of Islam had the highest fertility of all, almost 40 per cent above the average. High pre-migration fertility among older Muslim women together with early marriage among Muslim women generally and the strict gender roles which Islam advocates all contribute to the higher fertility of Muslim women. Claiming to have no religion has become increasingly common over recent censuses, and in 1996, 13.5 per cent of females aged 15 or older placed themselves in that category. Their mean number of CEB was 7 per cent below the national average after standardisation for age and marital status. Still lower was the standardised mean for the small ‘Other religion’ group, 70 per cent of whom were Jews or Hindus.

62 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Religion

Mean CEB

Age-standardised mean CEB

Standardised index

Age/marital statusstandardised mean CEB

Standardised index

Table 3.5 Average number of children of women aged 15 or older by religion, 1996

Anglican Baptist Catholicism Greek Orthodox Lutheran Other Orthodox Pentecostal Presbyterian Uniting Other Christian ‘Christian’

1.91 1.92 1.87 1.64 1.87 1.61 2.00 1.98 1.97 2.10 1.82

1.79 1.89 1.91 1.70 1.77 1.66 2.15 1.73 1.81 2.06 1.92

99 104 106 94 98 91 119 96 100 114 106

1.78 1.87 1.91 1.64 1.75 1.56 2.09 1.73 1.78 2.03 1.91

98 103 106 91 97 87 116 96 98 112 106

Buddhism Islam Other religion

1.61 2.03 1.59

2.03 2.68 1.63

112 148 90

2.02 2.47 1.60

112 137 88

No religion Religion not stated

1.32 1.74

1.64 1.74

91 97

1.68 1.76

93 97

Total

1.81

1.81

100

1.81

100

NOTE Standardised to age and age/marital status structures of total Australian female population aged 15 or older. SOURCE Unpublished 1996 Census data.

The above-average fertility of Catholics and slightly belowaverage fertility of Anglicans were the product of differences at ages 35 and older. At younger ages Catholic fertility was actually slightly lower than Anglican fertility. The below-average fertility of Greek Orthodox women reflected lower than average fertility at ages 20–34 and 50 and older, while that of Other Orthodox women was

FERTILITY TRENDS AND DIFFERENTIALS

• 63

concentrated at ages 40 and older. Cautious childbearing by younger Greek Orthodox women is consistent with contemporary behaviour in Greece itself, and symptomatic of a tension between cultural expectations once women commit to motherhood, and the educational and career opportunities available to them in modern Australia. The above-average fertility of Pentecostals and members of ‘Other’ Christian sects was observed at all ages. Members of these smaller sects probably are more likely than members of larger Christian denominations to actually practise their faith, and thus may adhere more strongly to biblically buttressed family values and gender roles. High Islamic fertility was also an all-ages phenomenon, while above-average Buddhist fertility was characteristic only of women already past their reproductive years. Those who were still of childbearing age in 1996 displayed distinctly below-average fertility. The former element of this peculiar pattern is attributable to the dominance of older age groups by migrants from countries like Vietnam and Laos, much of whose childbearing took place overseas. The latter element probably reflects significant numbers of temporarily resident students, together with fertility restraint among women who had arrived in refugee intakes as children or young adults, or been born in Australia of refugee parents. A combination of inauspicious economic circumstances and the according of priority to acquiring education and establishing careers may have contributed to their lower fertility. Finally, the below-average fertility of women claiming to have no religion was evident only beyond age 25, and more especially beyond age 30. At the youngest ages this group recorded marginally above-average fertility.

PA RT N E R E D M OT H E R S V E R S U S L O N E PA R E N T S Data on relationships within the household allow women with a husband or partner present (partnered women) to be distinguished from lone parents. In comparing the fertility of these groups it is appropriate to eliminate from the first of them childless women, since lone parents, with rare exceptions (e.g., adoptive parents), must have given birth to at least one child. After standardisation for age, the mean number of CEB of partnered mothers was 32 per cent above that of all women in Australia, while that of lone parents was 38 per cent above the national average. Further standardisation for marital status barely alters this picture. Among lone parents, those who had never married

64 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

had distinctly lower age-standardised mean CEB than those who had been married, the former group recording a level 11 per cent above the national average, and divorced, separated and widowed women levels ranging between 39 and 54 per cent above it. Partnered mothers fell into three main subgroups: those who were married, those who had been married but were now cohabiting, and those cohabiting without ever having been married. After standardisation for age the former two subgroups recorded mean CEB respectively 34 and 38 per cent above the national average. The figure for the third subgroup was lower, at 22 per cent above the national average. Age-specific fertility profiles for partnered mothers and lone parents are shown in Figure 3.9. Clearly the higher overall fertility of lone parents noted above is a product of higher fertility at older ages. This is an artefact of the reality that, to be classified as a lone parent, one needs to have at least one co-resident (though not necessarily dependent) child. This means that at older ages women tend to be selected into lone parenthood by virtue either of having continued childbearing until relatively late in their reproductive lives or of having proceeded to higher parities. The more children a woman has, the more likely it is that at least one of them will live with her when she is elderly. At younger ages (15–19 to 40–44), lone-parent fertility was in 1996 actually lower than partnered mother fertility, suggesting some, though not a huge, depressing effect of lone-parent status. The difference was most pronounced, but still modest, at ages 15–19 and 35–39. Teenagers often become lone parents following unintended pregnancy, and this circumstance is less conducive to quickly having a second child than is having a first child when married or cohabiting. By ages 35–39 the typical path to lone parenthood is marital or relationship breakdown, and its occurrence at or just before these ages is especially likely to interrupt and thereby reduce childbearing.

B I RT H P L AC E Table 3.6 presents, in rank order after standardisation for age and marital status, mean CEB and standardised mean CEB figures for all birthplace groups for which 2000 or more females aged 15 or older were enumerated in 1996. Whereas in previous calculations the total Australian population provided the standard, in this instance the standard population was the Australia-born population. The range of standardised mean CEB levels captured is from 58 per cent above to 39 per cent below the Australia-born level.

FERTILITY TRENDS AND DIFFERENTIALS

• 65

Figure 3.9 Average number of children by women’s age and partnering status, 1996

Average number of children

4

3

2

1

0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75+ Partnered mother SOURCE

Lone parent

Unpublished 1996 Census data.

Small Pacific Island groups and source countries of refugee migration feature prominently among the highest-fertility birthplace groups. Significant proportions of older members of many of these groups were already of relatively high parity upon arrival in Australia. At the opposite end of the spectrum a concentration of generally ageing Eastern European birthplace groups stands out as having especially low standardised fertility. Several larger birthplace groups—UK and Ireland, New Zealand, Italy and Greece—also rank below Australiaborn women, but not substantially below. Individual countries making up the first of these groups (not itemised in Table 3.6) all had standardised mean CEB 6–8 per cent below the Australia-born. Age-specific mean CEB profiles are shown for the 15 birthplace groups with the largest numbers of females aged 15 or older in Figure 3.10. In each graph the solid line represents Australia-born women. The first graph covers the three groups that deviate most markedly from this comparator. The profile for the Lebanese simply takes off on a much steeper trajectory than the Australian one, and remains well above it at all ages. Early female marriage and a culture that sees young women as wives and mothers, not tertiary students and employees, generates much higher fertility at younger ages than Australia-born women display, while at older ages high levels of childbearing before arriving in Australia are clearly a major factor.

66 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Country of birth

Mean CEB

Age-standardised mean CEB

Standardised index

Age/marital statusstandardised mean CEB

Standardised index

Table 3.6 Average number of children of women aged 15+ by country of birth, 1996

Cook Islands Lebanon Western Samoa Tonga El Salvador Cambodia Afghanistan Syria Laos Vietnam Iraq Malta Papua New Guinea Fiji Turkey Mauritius Netherlands Cyprus Chile Philippines

2.43 3.11 2.62 2.65 1.96 2.12 2.35 2.73 1.94 1.79 2.18 2.94 1.41 1.75 2.11 2.13 2.69 2.45 1.81 1.60

2.85 3.02 2.75 2.67 2.39 2.51 2.93 2.62 2.43 2.25 2.34 2.19 2.01 2.03 2.11 1.90 1.88 1.95 1.85 1.87

163 173 157 153 137 143 168 150 139 129 134 126 115 116 121 108 108 112 106 107

2.76 2.69 2.57 2.50 2.38 2.36 2.34 2.29 2.28 2.22 2.16 2.05 2.02 1.91 1.85 1.84 1.82 1.81 1.79 1.76

158 154 147 143 136 135 134 131 130 127 123 117 116 109 106 105 104 103 102 100

Australia

1.75

1.75

100

1.75

100

Indonesia Pakistan New Zealand Italy Iran Myanmar Portugal Peru Greece

1.50 1.89 1.63 2.63 1.79 2.06 1.88 1.70 2.33

1.80 1.93 1.70 1.87 1.85 1.68 1.80 1.67 1.78

103 111 97 107 106 96 103 96 102

1.75 1.73 1.72 1.72 1.71 1.68 1.67 1.65 1.62

100 99 98 98 98 96 95 94 93

FERTILITY TRENDS AND DIFFERENTIALS

• 67

UK/Ireland Spain Finland Taiwan India Belgium South Africa Israel Malaysia Former Yugoslavia Egypt South Korea Kenya Sri Lanka Denmark Zimbabwe Singapore Uruguay China Argentina France Hong Kong US Germany Sweden Austria Canada Romania Thailand Poland Switzerland Hungary Former USSR Former Czechoslovakia Japan

2.07 2.07 2.10 1.05 1.92 1.94 1.66 1.79 1.29 2.02 2.19 1.18 1.64 1.73 1.78 1.44 1.10 1.69 1.73 1.58 1.62 1.09 1.40 1.96 1.38 1.89 1.29 1.60 0.96 1.78 1.50 1.77 1.82 1.59 0.82

1.68 1.67 1.64 1.56 1.68 1.60 1.62 1.66 1.560 1.69 1.69 1.56 1.61 1.60 1.52 1.61 1.45 1.52 1.53 1.48 1.42 1.42 1.43 1.43 1.38 1.39 1.35 1.46 1.31 1.36 1.31 1.33 1.34 1.29 1.07

96 96 94 89 96 91 93 95 89 97 97 89 92 92 87 92 83 87 87 85 81 81 82 82 79 80 77 83 75 78 75 76 76 74 61

1.61 1.60 1.60 1.58 1.57 1.57 1.57 1.56 1.560 1.55 1.54 1.52 1.51 1.51 1.48 1.48 1.46 1.44 1.44 1.43 1.42 1.42 1.42 1.40 1.37 1.37 1.35 1.34 1.32 1.28 1.28 1.27 1.25 1.23 1.07

92 92 91 90 90 90 90 89 89 89 88 87 86 86 85 85 83 82 82 82 81 81 81 80 79 78 77 77 76 73 73 73 72 71 61

Total Australia

1.81

1.73

99

1.71

98

NOTES Standardised to age and age/marital status structures of Australia-born females aged 15 or older. Birthplace groups represented are those with 2000 or more female members aged 15 or older. SOURCE Unpublished 1996 Census data.

68 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The Vietnamese profile is not dissimilar to that described earlier for Buddhists, the overlap between the two groups being substantial. Fertility is marginally below that of Australia-born women to age group 35–39, but then diverges sharply from it as it becomes less Australian fertility and increasingly fertility that was achieved overseas. The fertility of Philippines-born women tracks well below that of Australia-born women for cohorts in their 30s, 40s and 50s, but then converges, crosses and moves to very high levels at advanced ages. The former feature is a product of the ‘mail order bride’ phenomenon of the 1980s, and the dampening effect on fertility of marriages to often considerably older Australian males. Comparatively old bridal ages and subsequent marital instability may also be factors. The latter feature once again probably reflects imported fertility, this time associated with Filipina brides sponsoring family reunion migration. The next graph in Figure 3.10 goes to the opposite extreme, showing the similarity to the Australia-born comparator of fertility profiles for women born in the UK and Ireland, New Zealand and (except at the oldest ages) the Netherlands. British fertility levels in particular are generally a little lower than those for the Australia-born, but not substantially so. On the next graph the fertility profiles of other significant European birthplace groups are plotted. Germans and Poles exhibit relatively low fertility from ages 25–29 onward. Greeks, and especially Yugoslavs, aged 20–34 have higher fertility than the Australia-born, but then decidedly lower fertility after about age 50. Greece-born fertility should not be confused with Greek Orthodox fertility. The latter was in 1996 distinctly lower at younger reproductive ages because of the large numbers of Australia-born women of Greek heritage included. Italians have a similar profile to the Australia-born, although they deviate above it at ages 40–54 and below it for cohorts aged in their 60s. Finally, the last graph in Figure 3.10 covers other large Asian birthplace groups. All have fertility profiles that lie below, and often well below, the Australia-born comparator. Almost zero fertility among Malaysia– and Hong Kong-born migrants aged under 25 attests large proportions of students, and high levels of education that are conducive to low levels of fertility are characteristic of all groups at most ages. The notably low fertility of China-born cohorts aged 35–39 and 40–44 may be attributable to their having come under the influence of China’s One Child Family policy before migrating to Australia. Older China-born cohorts record fertility more in keeping with that of the other Asia-born groups represented.

FERTILITY TRENDS AND DIFFERENTIALS

Figure 3.10 Average number of children by women’s age and birthplace, 1996 Average number of children

6 5 4 3 2 1 0 15–19

20–24

25–29

30–34

35–39

Average number of children

Vietnam

Average number of children

50–54

55–59

60–64

65–69

70–74

Lebanon

75+

Australia

5 4 3 2 1 0 20–24

25–29

30–34

35–39

UK/Ireland

40–44

45–49

50–54

New Zealand

55–59

60–64

65–69

70–74

Netherlands

75+

Australia

6 5 4 3 2 1 0 15–19

Average number of children

45–49

6

15–19

20–24

25–29

30–34

35–39

40–44

45–49

50–54

55–59

60–64

65–69

70–74

Italy

Greece

Poland

Former Yugoslavia

Germany

Australia

75+

6 5 4 3 2 1 0 15–19

20–24

China

SOURCE

40–44

Philippines

25–29

30–34

35–39

Malaysia

Unpublished 1996 Census data.

40–44

45–49

India

50–54

55–59

60–64

Hong Kong

65–69

70–74

75+

Australia

• 69

70 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

S E C O N D G E N E R AT I O N Fertility behaviour can change dramatically between the migrant or first generation and their children born in the destination country. The second generation of a birthplace group is defined here as Australia-born persons with at least one parent born in the country in question. A tighter definition would require both parents to have been born in that country, but adopting it substantially reduces the size of the second generation of some groups. Among the birthplaces represented in Figure 3.11 as having reasonably large second generations of females aged 15 or older by the looser definition, for example, percentages that also met the ‘both parents’ definition ranged from 78 for Lebanon, 75 for Greece and 66 for Italy to 25 for UK and Ireland, 19 for Germany and 9 for New Zealand. This variability reflects the different lengths of time over which migration to Australia has taken place, different degrees of cultural dissimilarity from the host population, different gender balances in migrant flows and variable levels of resistance to out-marriage. As it happens, levels of second-generation fertility standardised for age and marital status for the birthplaces shown in Figure 3.11 vary little by definition of ‘second generation’ adopted—never by more than three percentage points. For all but two birthplaces (Lebanon and the Netherlands), standardised second-generation fertility was at the 1996 Census below the fertility of all women born in Australia, the lowest level (15–16 per cent below) being recorded for the Chinese and Greek second generations. The fertility of secondgeneration Lebanese and Dutch women was only marginally (less than 5 per cent) above the Australia-born level. In comparing first and second-generation fertility by age, Figure 3.11 focuses only on the reproductive ages. The largest difference is that observed for the Lebanese, whose second generation behaves much more like other Australia-born women than Lebanon-born women. The Greek, Italian and Former Yugoslav second generations also exhibit clearly lower fertility at all ages than their respective first generations. The pattern for the New Zealand group is similar, if more muted, featuring lower fertility below age 35 among the second than the first generation. The Dutch and German second generations’ fertility, however, tends to exceed that of migrants, while generational differences in UK and Ireland fertility are minimal. The comparison between the Chinese first and second generations is an interesting one, with lower fertility for the second generation up to

FERTILITY TRENDS AND DIFFERENTIALS

• 71

Figure 3.11 Average number of children by women’s age and origin, 1996 4

4

Lebanon

3

Number of children

Number of children

China

2 1

3 2 1 0

0

15–19 20–24 25–29 30–34 35–39 40–44 45–49

15–19 20–24 25–29 30–34 35–39 40–44 45–49

4

4 Italy

3

Number of children

Number of children

Greece

2 1 0

3 2 1 0

15–19 20–24 25–29 30–34 35–39 40–44 45–49

15–19 20–24 25–29 30–34 35–39 40–44 45–49

4

4

3

New Zealand Number of children

Number of children

Netherlands

2 1 0

3 2 1 0

15–19 20–24 25–29 30–34 35–39 40–44 45–49

15–19 20–24 25–29 30–34 35–39 40–44 45–49

4

4

UK/Ireland Number of children

Number of children

Former Yugoslavia 3 2 1

3 2 1 0

0

15–19 20–24 25–29 30–34 35–39 40–44 45–49

15–19 20–24 25–29 30–34 35–39 40–44 45–49

Migrants

4

Second generation Number of children

Germany 3 2 1 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49

SOURCE Unpublished 1996 Census data.

72 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

age group 25–29, and then lower fertility for the first generation at ages 30–44. The latter finding seems likely to reflect the suppressant effect on migrant fertility of China’s One Child Family policy.

I N C R E A S I N G D E TAC H M E N T O F C H I L D B E A R I N G F R O M M A R R I AG E At the same time as the fertility trends outlined earlier have been unfolding, childbearing in Australia has become increasingly detached from formal marriage. Historically this is not a new phenomenon among the non-Aboriginal population of Australia. Childbearing within consensual unions was widespread during the convict era of the late 18th and early 19th centuries (Carmichael 1996). After 1830, however, as free settlement replaced transportation as the major vehicle of European colonisation, far more middleclass family values took root, and by the early 1950s only 4 per cent of births occurred to unmarried women. This figure has subsequently climbed relentlessly. It passed 10 per cent in 1975, 20 per cent in 1989, and by 2000 had reached 29 per cent. Until the early 1970s the increase was mainly generated by rising non-marital fertility among teenagers. This was a period over which, as young people’s access to motor vehicles increased, parental monitoring of adolescent sexual behaviour relied ever less on physical oversight and ever more on much less effective moral entreaty. Thus, between 1950 and 1971 the rate of conception leading to nonmarital live birth among unmarried women aged 15–44 rose from 12 to 27 per 1000, and the percentage of those conceptions occurring to teenagers from 27 to 46. The other component of the upsurge in childbearing brought about by increased adolescent sexual activity was a marked rise in premaritally conceived marital births. This, of course, had no bearing on the proportion of births occurring outside marriage, other than that the ‘shotgun’ marriages implied prevented it from rising more rapidly. However, over the same period (1950–71) the rate of conception leading to marital live births among unmarried women aged 15–44 rose from 16 to 24 per 1000, and the teenage percentage of these conceptions from 49 to 65. The early 1970s were a social watershed in Australia. Among other developments, abortion effectively on request became available following a landmark legal ruling in New South Wales in late 1971, and a trend to widespread consensual partnering that has swept most developed countries to varying degrees (Carmichael 1995) began to

FERTILITY TRENDS AND DIFFERENTIALS

• 73

gather momentum. The latter was a logical development following the introduction of the oral contraceptive pill in 1961. This initially was used by married women aged in their 30s and 40s to avoid unwanted births of higher parity, and by younger couples to defer the commencement of childbearing after marriage, thereby allowing them to marry earlier and place their marriages on a sounder economic footing. The latter development and the rise in childbearing associated with unintended non-marital pregnancy discussed above combined to produce the overall stability, at a historically low level, in the age at commencement of childbearing through the 1960s that was noted earlier. But with the early years of marriage becoming childless (Ruzicka and Choi 1981), the divorce rate rising (Carmichael, Webster and McDonald 1997), and respect for religiously-founded values on the wane (Carmichael 1998), it was inevitable that people would question the necessity for such a lifecycle phase to be sanctioned by marriage. Hence, the trend to consensual partnering also was inevitable, as was its becoming, over time, less a childless phase before marriage, and increasingly one in which family formation would be initiated. Data from the 1996–97 NLCS show that of respondents first married in the 1960s, 9 per cent had lived together before marriage, these figures rising to 29, 52 and 64 per cent for those first married in the 1970s, 1980s and 1990s. The primary vehicle for the continuing, more rapid, increase in the proportion of births taking place outside marriage since the early 1970s has been increased consensual partnering, and a growing preparedness of couples to have children in such unions. This is reflected in a radically altered age distribution of unmarried women conceiving children who were subsequently also born outside marriage (Table 3.7). Whereas in the early 1970s almost half were teenagers, by 1995 only a fifth were, and proportions who were of more normative reproductive ages had risen appreciably. It is also notable from Table 3.7 that the teenage share of non-marital conceptions resulting in live marital births fell even more spectacularly. Rates of conception leading to marital live birth for unmarried women aged 15–19 and 20–24 plummeted from 28 to 10 per 1000 and from 30 to 13 per 1000 between 1971 and 1976, and have continued to fall since. These initial declines clearly owed much to improved access to abortion, although more frequent cohabitation was undoubtedly conducive to more effective contraception in nonmarital relationships as well. But whereas in 1971 less than 7 per cent

74 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

of conceptions followed by ‘shotgun’ marriages involved women aged 25–34, by 1995 the figure was 39 per cent. Marriage when pregnant has become far less common, and when it does occur it tends not to involve teenagers, but older women who, with their partners, decide to formalise their consensual unions before the child is born. Table 3.7 Age distribution of women experiencing non-marital conceptions that resulted in live births by nuptiality at birth, 1971 and 1995 15–19 %

Age group (years) 20–24 25–29 30–34 % % %

35–39 40–44 Total % % %

Non-marital 1971 1995

45.8 20.9

29.1 33.2

13.9 23.2

6.7 15.0

3.5 6.6

1.0 1.1

100.0 100.0

Marital

1971 1995

64.8 11.4

28.1 33.4

4.9 27.7

1.6 18.2

0.5 7.8

0.1 1.5

100.0 100.0

Total

1971 1995

54.7 20.0

28.6 33.2

9.7 23.6

4.3 15.3

2.1 6.7

0.6 1.2

100.0 100.0

Birth type

SOURCE

Year

Calculated from Australian birth statistics for 1971–72 and 1995–96.

CONCLUSION At a superficial level relatively little appears to have changed in the fertility behaviour of Australian women over the past two decades. The TFR, after falling sharply through the early and middle 1960s, and again from the early to the late 1970s, remained stable through the 1980s before declining slowly after 1992. It has been shown, however, that this apparent stability is illusory. It masks a major transition in the timing of fertility that has seen age-specific fertility rates for women in their 20s fall, while those for women in their 30s rose as births deferred during and after the 1970s were made up. More recently the latter trend has eased while the former has continued, the net effect being the slow decline in the TFR currently in progress. As a result of these developments, the age pattern of fertility has shifted decisively. While in 1980 women aged 30 or older contributed 27 per cent of total fertility, by 2000 they accounted for

FERTILITY TRENDS AND DIFFERENTIALS

• 75

48 per cent. Fertility increments at these older reproductive ages have now slowed. How much longer declines at younger ages continue will thus determine how low the TFR falls, and assuming no marked subsequent resurgence in fertility, what the consequences for the growth and ageing of Australia’s population will be (Kippen and McDonald 1998). There is evidence from the NLCS that the contemporary reluctance of Australians to have children represents, to a not insignificant degree, an inability by young women to convert early reproductive expectations into actual reproductive performance. A clear need exists for research that will assess more fully the firmness of these ‘expectations’, and enhance understanding of the forces responsible for difficulties being experienced in realising them. Fertility differentials across a range of variables have been explored. Some, such as those by education level and degree of labour force attachment, were entirely predictable. Greater commitment to education and the labour force are both associated with lower fertility. Further research is needed to disentangle several potential explanations for excess fertility at younger reproductive ages outside the major urban areas. And fertility differentials across religious and birthplace groups, and between migrant and secondgeneration groups, reflect a variety of idiosyncrasies of individual groups that add texture to Australia’s family formation landscape. The early years of European settlement in Australia were quite unique, and in consequence childbearing was distinctively detached from formal marriage. In the late 20th century the wheel has turned full circle, and having children within consensual unions has become commonplace again. Close to one-third of births nowadays occur outside marriage, but the stereotype of the unmarried mother as a single teenager that had some validity through the 1960s and early 1970s has become a gross distortion. People talk openly of having a ‘partner’ rather than a ‘wife’ or a ‘husband’. Pregnancy within a consensual union may trigger marriage, but only if convenient, and it probably more often does not. That does not rule marriage out as a future option, when an acceptable mix of convenience and commitment is judged to exist, although if one or both partners have been previously married there is probably a particular reluctance to exercise that option. Concerning the future, it is hard to see Australia’s TFR rising again without major social policy changes to, in particular, render easier the combining of maternal and labour roles for women. Even

76 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

then, the effect could be more to stem further decline than to generate increase. The debate at present really is one about ‘how much lower?’, and the answer to that question has significant implications for the severity of the ageing population problem Australia will face as the 21st century unfolds. Will a new equilibrium be established at TFR = 1.65? 1.60? Lower? Australia is not currently as afflicted as those European nations that already are grappling with the hyperageing implications of TFRs closer to 1.0 than to 2.0 (Table 3.1), but there is no room for complacency. Individuals and couples will make decisions on whether to have children, when to have them and how many to have in their perceived own best interests, not out of a sense of duty to contribute to an aggregate outcome promoted as being in the interests of society at large. The achievement of a desirable aggregate outcome therefore requires policies that are sensitive to forces that matter at the individual level. The challenge for contemporary research and policy development is to identify clearly those forces and then devise effective ways of addressing them.

4

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N FA M I LY 1 PETER McDONALD

ON DEFINITIONS For official statistical purposes, a family in Australia is defined as two or more persons, one of whom is at least 15 years of age, who are related by blood, marriage (registered or de facto), adoption, step or fostering, and who are usually resident in the same household (McEwin 1998, p. 16). In everyday language, the word ‘family’ is used in ways that are not consistent with this statistical definition. The most obvious way that our everyday concept of the family differs from the statistical definition is that almost all of us frequently refer to people living in other households as members of our family. If we are adults, our siblings, our parents, our grandparents and our children will often be described as ‘family’ even though they may live elsewhere. If we are children, we are likely to see our grandparents, our parents or siblings who might live elsewhere and even our cousins and aunts and uncles as family. In some uses of the word, we refer to our in-laws as ‘family’ but at other times they will be ‘your family’. At more ceremonial occasions, such as weddings and funerals, family might be used in a much broader sense to include cousins, uncles and aunts. The widespread interest in genealogy has extended the concept of family in other directions, most notably, to an ancestor and his or her descendants. The people we consider as family in everyday terms vary according to the purpose and to our life cycle stage. Co-residence is just one of several criteria in defining a ‘family’. Other criteria may include personal circumstances, cultural norms, the nature

78 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

of the relationship (closeness, sense of obligation, etc.) and the context. On the other hand, the word ‘family’ is sometimes not used to describe situations that are considered as family in the official statistical definition. Two sisters sharing a household may not describe themselves as ‘a family’. Even a couple with no children may not in conversation refer to their household unit as ‘my family’. In popular parlance, there is a sense in which ‘family’ is something more than two people of the one generation in the one household. The statistical definition is limited because it requires co-residence. It is also limited because it is static whereas ‘family’ is dynamic. The people we consider as members of our family change as our circumstances change. We are continually adding and subtracting people to our concept of family that we use for different purposes as we move through life. Family type in the statistical definition is a structural type. People live in one-parent families, twoparent families, couple families, or ‘other’ families. More precisely, these are forms of living arrangement rather than forms of ‘family’. For example, most children who live in a one-parent family have another parent living elsewhere whom they would describe as part of their family. Thus, it seems more appropriate to describe ‘family’ in terms of the changing nature of relationships between people that can be considered to be ‘family’ relationships. This implies a more functional approach to the consideration of families, in contrast to the structural approach used in the statistical definition.

ON THEORIES The approach to the family used in this chapter has been classified by the sociologist, Michael Gilding, as neo-functionalist and liberal (Gilding 1997, pp. 37, 254). Gilding portrays this approach as follows: The liberal position is comfortable with the family as a social institution which changes over time. It does not hark back to a glorious past, when life was simpler and people kinder. Nor does it insist upon a narrow definition of family. On the contrary, liberals acknowledge the enduring importance of long-term relationships and children, and the changing meanings attached to these relationships. The account is consistent with the fact that most Australians in the 1990s nominate the family as the most important aspect of

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 79

their lives. In close connection, Australians also define families in more diverse ways than was once the case (Gilding 1997, p. 254). Gilding describes two other theoretical perspectives, the conservative and the radical. The conservative view is reflected in the following quotation: It is simply false to argue that there is no relatively fixed definition of ‘family’. The human record, honestly confronted, shows that the family is a natural, universal, and irreplaceable community rooted in human nature. The ‘family’ in all ages and in all corners of the globe can be defined as a man and a woman bonded together through a socially-approved covenant of marriage to regulate sexuality, to bear, raise, and protect children, to provide mutual care and protection, to create a small home economy, and to maintain loyalty and continuity between the generations, those going before and those coming after (Carlson 1996, p. 8). Although it is not stated in this definition, the conservative approach also normally specifies rigid role segregation of husbands and wives, with husbands being responsible for income earning and wives for the care of children and for household maintenance (Bogle 1996). Conservatives see the family as being in a state of decline because, based on the statistical definition, a much smaller percentage of people today live in the conservative form of family than was the case at the beginning of the 1970s. In 1974, approximately 40 per cent of Australians lived in married-couple families with dependent children where the husband was in the labour force and the wife was not in the labour force. In 1998, the equivalent figure was about 13 per cent (derived from ABS 1976 and ABS 1998b). In the radical perspective, the family is not about intimacy and caring but about power, oppression, abuse and conflict. The family is seen as changing but the changes are very slow in regard to the establishment of equal and co-operative forms of family relationships. The radical perspective sees the family as rooted in patriarchy. This perspective does not see change as necessarily heading in one direction, but as chaotic, fragmentary and uncertain. In the extreme, the radical approach does not mourn the passing of ‘the family’ (Stacey 1993).

80 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

In contrast to the United States and Britain, according to Gilding (1997, p. 254), the liberal position achieved widespread influence in Australia in the 1980s, ‘marginalising the conservative position in terms of religious fanaticism and “backward provincialism”’. The liberal position also generally held sway in the official pronouncements from the United Nations during the 1994 International Year of the Family: Families assume diverse forms and functions from one country to another, and within each national society. These express the diversity of individual preferences and societal conditions (United Nations Proclamation on the International Year of the Family 1994, paragraph I.3.b). Gilding (1997, p. 256) considers the radical perspective to be the most marginal in public debate in Australia to the extent that ‘in the 1990s, feminist and gay activists are more likely to frame their critique in liberal terms, upholding the diversity of families’. Another theoretical theme that crosscuts the conservative-liberal-radical paradigm is the public-private dichotomy. The liberal agenda is one in which private agendas are public in the sense that they should be supported in the public sphere, philosophically, legally and financially, and individual well-being is seen as the end product of a partnership between government, employers and families. In the conservative agenda, the family is played out in the private sphere but governed by norms and rules determined in the public sphere. The public sphere addresses the family through its agent, the father and husband. The principal roles of government in the conservative agenda are to protect the privacy and the stability of the family, and to ensure the economic well-being of the breadwinner. The position of the radical agenda in relation to the publicprivate dichotomy is less clear. Sometimes, for example, in regard to sexual relationships, the radical stance will be that the state has no role in the private affairs of individuals. At other times, it will call for state recognition of homosexual marriages or for the removal of abusive fathers from their wives and children. Gilding (1997, p. 256) points out that in some areas such as new reproductive technologies, the conservative and radical agendas have coalesced around the need for state restriction. This chapter looks at the main changes that have taken place in

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 81

the nature of family relationships since 1970 using the conservativeliberal-radical paradigm. The discussion will also consider the publicprivate dichotomy and, consequently, the role of public policy. Two main family relationships will be addressed: intimate couple relationships and parent-child relationships.

I N T I M AT E C O U P L E R E L AT I O N S H I P S Intimate couple relationships refer to relationships through which people obtain both sexual and emotional intimacy. Marriage has been the common form of such relationships, but other forms exist.2 Survey evidence suggests that most people value such relationships and would prefer to be in one. The conservative perspective specifies marriage as the one and only acceptable form of intimate relationship. Other intimate relationship types are considered inadequate or inappropriate in some way. The liberal perspective emphasises the individual’s need for such intimacy, but is not necessarily prescriptive about the legal status of the relationship. Intimate relationships imply trust and commitment and these may be more or less likely in one form of relationship than in another. The conservative claims that trust and commitment are inherent in marriage. On the contrary, some with a radical perspective argue that, because marriage is a well-established form of patriarchy, freely given intimacy and commitment are more likely in forms of relationship other than marriage. The liberal view is that it is the trust and commitment in the relationship that is important, not the form of the relationship. Consistent with the views of most people, there is a tendency in all three perspectives to favour living in some form of intimate relationship. However, a choice not to be in a relationship, considered to be deviant in the 1950s, is now less an issue of concern for all three perspectives. The conservative may see singleness as an unfortunate but tolerable outcome. The liberal may see singleness as a legitimate choice but not one that would suit most people. The radical might see singleness as the inevitable outcome of a society that does not offer people equality within relationships. The statistical evidence (Table 4.1) shows that, both in 1971 and 1998, marriage was the dominant way in which most Australians lived through the central ages of adult life (30–59 years); however, there have been substantial changes in the proportions married at both the younger and the older ages.

82 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 4.1 Percentages of men and women aged 15 and over who were legally married and not permanently separated, 1971 and 1998 Age group Men Women 1971 1998 1971 1998 15–19

1.4

0.2

8.7

0.7

20–24

35.1

5.8

62.0

13.4

25–29

71.5

31.2

84.3

44.3

30–34

82.6

54.8

88.6

63.0

35–39

85.0

65.5

88.8

69.1

40–44

84.8

70.2

87.0

70.9

45–49

84.6

73.1

84.1

72.1

50–54

84.1

75.4

79.2

72.8

55–59

83.1

77.0

72.3

72.2

60–64

80.7

77.8

61.8

68.8

65–69

76.5

77.1

49.7

61.2

70–74

70.4

75.6

36.1

50.8

75–79

62.5

71.8

24.1

37.8

80–84

51.3

64.2

14.0

23.2

85+

35.5

49.0

6.0

10.5

NOTE The published 1998 percentages for currently married have been adjusted to currently married and living together using proportions married but permanently separated, by age and sex, obtained from the 1996 Census of Population and Housing. It is assumed that the proportion of married persons who were permanently separated was the same in 1998 for each age and sex group as it was in 1996. SOURCES 1971 Census of Population and Housing, Demographic Characteristics Australia, Table 1, Bureau of Census and Statistics Australia; Marriages and Divorces Australia 1998, Table 4.3, Australian Bureau of Statistics, Catalogue No. 3310.0.

I N T I M AT E R E L AT I O N S H I P S : P E R S O N S AG E D 2 0 – 2 9 Y E A R S At the youngest ages (less than 30 years), the fall in the proportion married is massive. In 1971, 62 per cent of women aged 20–24 years were married, but in 1998 only 13 per cent were married. This is the result of a powerful movement away from the early marriage pattern that characterised the 1940–70 period. This is not an area of controversy in the family debate. The conservative perspective does not call for a return to early marriage and would tend to support the notion that there is a ‘proper time to marry’ which involves establishment of economic security and emotional maturity. The liberal perspective

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 83

would strongly support the shift to later marriage on the grounds that years of experience of adulthood before marriage provide young people with a better appreciation of the options open to them. In the liberal way of thinking, later marriage provides women with the opportunity to establish themselves in career terms and hence to be in a position to enter marriage on more equal terms. From the radical perspective, later marriage gives young people time to realise for themselves that marriage is patriarchy and that other options may be preferable. Has the movement away from marriage among young people in their 20s been associated with a burgeoning of other forms of relationships? We can address this question in a static or dynamic way. In the static approach, the living arrangements of 20–29 year-olds at a point in time are examined. Four living arrangement types are shown in Table 4.2. At a point in time, the proportion living together without being married is the smallest of the four possible states. The most prominent state for young people in their 20s is not being in a relationship, followed by marriage and ‘living apart together’ relationships. Thus, from the static viewpoint, the alternative of living together without being married does not appear to be widespread. The dynamic approach considers people’s experience during their lifetime, rather than their circumstances at a point in time. Table 4.3 shows a selection of measures that indicate young Australians’ experience of various relationship situations. There is little difference between the proportions who are currently married and living with their spouse (Table 4.2) and the proportions who have ever been married (Table 4.3). Also, the proportions who have married more than once are very small. That is, breakdown of their own marriage is not an important feature of the lives of 20–29 yearolds. However, the dynamic approach provides a very different perspective from the static approach in relation to the frequency of living-together relationships. Almost 50 per cent of all men and women aged 20–29 years have lived together without being married, including those who lived together before their marriage. Among those who have married, 71 per cent of men and 55 per cent of women had lived together with their spouse before the marriage. Thus, a dynamic approach shows that living-together relationships occur much more commonly than we would conclude from the static approach. The implication of this is that living-together relationships are not lasting relationships.

84 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 4.2 Living arrangements of men and women aged 20–29 years, 1996–97 Living arrangement Not presently in a relationship In a relationship, but not living together Living together, not married Married and living together Total SOURCE

Men % 44 15 13 28 100

Women % 31 20 16 33 100

Negotiating the Life Course Survey, 1996–97.

Table 4.3 Relationship experiences of men and women aged 20–29 years, 1996–97 Men Women All persons: % ever married 29 36 All persons: % previously married but not now married 2 3 All persons: % married now but married more than once 0 1 All persons: % who have ever lived together without being married 46 50 Ever married: % lived together with spouse before first marriage Ever remarried: % lived together with spouse before second marriage Ever married: % has lived together with someone other than a spouse NOTE a SOURCE

71

55

a

–a 14

– 16

Less than 30 respondents. Negotiating the Life Course Survey, 1996–97.

The Negotiating the Life Course Survey showed that about 90 per cent of living-together relationships contracted in the late 1980s had been ended by separation or marriage within about eight years of formation and only 24 per cent were intact after about four years (McDonald 1998a). While living together may be a short-term alternative to marriage, it appears not to be an alternative form of permanent relationship. There may be a small number of people whose preference is for a sequence of short-term, non-marital relationships and a small number who have lasting living-together relationships, but, in Australia in general, living together is not an alternative to marriage. Instead, for a high proportion of people, it is an integral part of the process of getting married. This is contrary to the standpoint of both conservatives and radicals. Both portray living together

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 85

as if it were an alternative form of permanent relationship, one seeing it as an inadequate alternative, the other seeing it as a desirable alternative. For the majority who live together before they marry and for whom living together is part of the marriage process, an anti-living together agenda is effectively an anti-marriage agenda. Conservatives, therefore, tend to marginalise their message about the importance of marriage by not recognising that a sizable majority of people enter marriage through a living-together relationship. Those who do not live together before marriage are identified by their ethnicity and their religiosity. People of Mediterranean or Asian origin are much less likely to live together before marriage than other groups. For example, only 16 per cent of people whose mother was born in a Mediterranean country lived together before marriage. Also, the percentage who live together before marriage rises as the level of religiosity falls. For people for whom religion was not at all important in their lives, 80 per cent lived together before marriage. This fell to 33 per cent for those for whom religion was very important in their lives (McDonald 1998a). In 1996–97, religion was important or very important in the lives of less than 30 per cent of Australians aged 20–29 years. Over the past 30 years, there has been an important change in relationship patterns at this age. The Negotiating the Life Course Survey has shown that, over time, first living-together relationships have become more likely to break up than to end in marriage. Facilitated by the availability of reliable methods of birth control, a pattern of young people living together before marriage began to be prominent in the 1970s. At that time, most living-together relationships led on to marriage. More recently, however, these first living-together relationships have been more likely to break up than to lead to marriage. The shift to a later age at marriage is both a cause and a consequence of this trend. The longer marriage is delayed, however, the less likely it is to occur at all, and certainly the proportion of Australians who ever marry has fallen sharply since the early 1970s. This recent trend presents something of a dilemma for the liberal. If the most desirable end result for most people is marriage, is there a question about a trend in behaviour that clearly makes marriage less likely to occur? The trend may sustain the case of the radical that the gender inequality inherent in relationships or the unfavourable nature of institutional structures relevant to young people (employment, housing) leads to fragmentation and uncertainty. The conservative

86 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

case that people should simply marry without living together is not well supported by this trend. First, as already indicated, most people would not marry at all unless they had lived together beforehand. Second, we could not expect that couples whose living-together relationship broke up would have stayed together if they had married without first living together, and the trauma of ending a marriage would have been greater.

I N T I M AT E R E L AT I O N S H I P S : P E R S O N S AG E D 3 0 – 3 9 Y E A R S The married state is much less common in 1998 than in 1971 for people in their 30s (Table 4.1). This is where the political debate about marriage heats up. The conservatives would prefer to see most people in their 30s in the married state. They would be particularly concerned if people in their 30s were living in some other form of relationship. The radicals would be happier with even lower percentages married and would approve if the shift away from marriage was related to a shift into other, more egalitarian forms of relationship. Liberals are ambivalent and uncertain about the situation. They have a strong sense that being in a relationship is a good thing and, as marriage is by far the most common form of intimate relationship for people in their 30s, perhaps the percentage married, ideally, should be higher. On the other hand, liberals would be loathe to make pronouncements in individual cases and would defend the reasons why people in this age group are not married. The main alternative to being married in this age group is not being in a relationship at all (Table 4.4). Only 8 per cent of people were living together without being married and between 3 and 7 per cent were in relationships but not living together. Table 4.4 Living arrangements of men and women aged 30–39 years, 1996–97 Living arrangement Not presently in a relationship In a relationship, but not living together Living together, not married Married and living together Total SOURCE

Negotiating the Life Course Survey, 1996–97.

Men % 17 7 8 68 100

Women % 14 3 8 75 100

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 87

The dynamic view of the relationships of this age group is shown in Table 4.5. Marriage breakdown is now a feature of the history of relationships. For example, 23 per cent of all women in this age range (26 per cent of all ever-married women) have been previously married or married more than once. This is well on the way to the estimated 40 per cent of first marriages that end in divorce. The rise in the rate of divorce took place over a short period of years, mainly in the 1970s. In the past 20 years, there has been little further increase in the rate of divorce. The high incidence of marriage breakdown is abhorred in the conservative perspective. Indeed, reduction of the divorce rate is a rallying cry of the conservative perspective on the family. The liberal would probably like to see a lower divorce rate, but achieved by means other than making divorce harder to obtain. The liberal would also say that, while it is a good thing to provide supports to marriage so that relationships continue, a high divorce rate is inevitable and we must adjust to this situation. The radical looks upon a high divorce rate as a consequence of all that the radical perspective says is wrong with the institution of marriage, that is, as a justification of the radical position. The debate about marriage breakdown is at the centre of the differences between the perspectives in the conservative-liberal-radical paradigm. The dynamic view provided in Table 4.5 also shows that 60 per cent of all 30–39 year-olds have lived together with a partner to whom they were not married at some time in their lives. This again is a very different picture from that provided in the static analysis of Table 4.4. In most cases, the respondent later married the person with whom they had lived, but 28 per cent of men and 21 per cent of women in this age group had lived with someone that they did not marry. Table 4.5 Relationship experience of men and women aged 30–39 years, 1996–97 Men Women All persons: % ever married 76 88 All persons: % previously married but not now married 8 13 All persons: % married now but married more than once 6 10 All persons: % who have ever lived together without being married 60 62 Ever married: % lived together with spouse before first marriage Ever remarried: % lived together with spouse before second marriage Ever married: % lived together with someone other than a spouse NOTE a SOURCE

Less than 30 respondents. Negotiating the Life Course Survey, 1996–97.

52

51

a

86 21

– 28

88 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

I N T I M AT E R E L AT I O N S H I P S : P E R S O N S AG E D 4 0 – 5 4 Y E A R S A N D 5 5 Y E A R S A N D OV E R The fall during the period from 1971 to 1998 in the proportion of people who are married and living with their spouse is still quite evident for ages 40–54 years but is somewhat less prominent than it was at younger ages (Table 4.1). Again, the main alternative living arrangement was not being in a relationship. Other alternatives accounted for only small proportions of people (Table 4.6). Table 4.6 Living arrangements of men and women aged 40–54 years, 1996–97 Living arrangement Not presently in a relationship In a relationship, but not living together Living together, not married Married and living together Total SOURCE

Men % 14 2 4 80 100

Women % 17 2 7 74 100

Negotiating the Life Course Survey, 1996–97.

The dynamic view (Table 4.7) shows that very high proportions of this age group have married at some time.3 These percentages are much higher than the levels reached by older cohorts. On the other hand, 33 per cent of all women in this age range are either previously married or married more than once. Hence, while those with a conservative perspective would approve of the extent to which this cohort has married, divorce is again the worrying feature for the conservative. The dynamic view given by Table 4.7 also shows that almost half the men in this age group and 40 per cent of the women have lived with a partner outside marriage. Almost 70 per cent of men and 78 per cent of women in this age group who had remarried, had lived with their spouse before the marriage. Again, if remarriage is desirable, then the conservative must expect that people will live together before they marry again. In contrast to the younger age groups, people aged 60 years and over, particularly women, were much more likely to be married in 1998 than they were in 1971 (Table 4.1). This change is due both to an increase in the prominence of marriage after about 1940 and to improved rates of old age survival in the past 25 years. These are not

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 89

Table 4.7 Relationship experience of men and women aged 40–54 years, 1996–97 All All All All

persons: % persons: % persons: % persons: %

ever married previously married but not now married married now but married more than once who have ever lived together without being married

Ever married: % lived together with spouse before first marriage Ever remarried: % lived together with spouse before second marriage Ever married: % lived together with someone other than a spouse SOURCE

Men Women 93 95 13 21 10 13 48 40 33 68 22

23 78 13

Negotiating the Life Course Survey, 1996–97.

changes that would figure highly in the debate about families, but given that spouses are the first carers in the society, this trend has important and positive implications for aged care.

T H E PA R E N T- C H I L D R E L AT I O N S H I P The relationship between parents and children has changed in many ways since the 1970s. The three main changes have been that children are now much more likely to be born outside marriage, the relationship between the child’s parents is much more likely to have ended, and the mother of the child is much more likely to be in the labour force. All of these changes run strongly counter to the conservative perspective on families. This perspective of the family places very heavy emphasis on the situation in which children are raised. The conservative ideal is that children are raised in a marriage of both the natural parents of the child where the mother of the child is not in paid employment. The liberal also would express a preference that children grow up with both their natural parents, but realises that this ideal often will not be possible. For the benefit of children whose parents live apart, the liberal will call for support for sole-parent families and social tolerance for those who are in this situation. The conservative sees such support and tolerance as incentives for parents to end their relationships. The conservative would prefer to see divorce made much more difficult for those who have the care of children. Likewise, where a child is born outside marriage, the conservative sees social support for mothers in this situation as stimulating such behaviour. The liberal takes the view that the child should be

90 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

supported irrespective of the way in which the child was born. The radical perspective is that any person who has a child has a right to social support for that child and for the relationship between the parent and the child. It is the parent-child relationship that is paramount for the radical, not the relationship between the child’s parents. The liberal perspective strongly supports increased levels of gender equity in families. Recognising that young women are now provided with educational and employment opportunities that are equivalent to those of young men, the liberal case argues that it is incumbent upon the state to support and facilitate the employment of mothers in the paid labour market. This involves initiatives such as provision of support for child care and for working conditions that more readily enable parents to combine work and child-rearing. The conservative sees the employment of mothers as the root cause of social problems related to children and young people. The mother at home is able to provide much greater time and much greater care to her children than a mother who is working. A child with a full-time mother feels that he or she is a valued child. Full-time care by the mother is valued at all ages of the child, but especially when the child has not yet commenced school. Governments in Australia, in recognition of the chasm between the views of liberals and conservatives, have attempted to please both through the provision of ‘choice’. Maximum government benefits now flow on one hand to mothers whose work attachment is highest and on the other hand to mothers whose work attachment is lowest, that is, to the extremes. When maximum benefits are provided at the extremes, disincentives are created as we move away from the extremes and towards the centre. In relation to work choices for parents, 6 per cent of dependent children in couple families and 58 per cent of dependent children in sole parent families had no parent employed in 1998 (ABS 1999b).

T H E PA R E N TAG E O F C H I L D R E N In June 1996, the proportions of Australian children who were living with both their natural parents was 85 per cent for 0–4 year-olds, 79 per cent for 5–9 year olds and 73 per cent for 10–14 year-olds. Among those not living with both natural parents, by far the majority lived with only their mother in a one-parent family. Hence, very few Australian children do not live with their mother and the sizable majority live with both their natural parents. The family type of children aged 0–14 years is shown in Table 4.8.

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 91

One-parent families come into being in four main ways: divorce or separation from a marriage, widowhood, separation from a livingtogether relationship, and birth to a woman who is not in a relationship. Except for widowhood which has become increasingly less common, all of these paths to lone parenthood have become more common since the 1970s. As a consequence, the incidence of oneparent families has been increasing. Of all families with dependent children in 1998, 22 per cent were one-parent families. This compares with 17 per cent in 1991, 15 per cent in 1986, 13 per cent in 1981 and 9 per cent in 1974. In 1998, in 89 per cent of all one-parent families with dependent children, the parent was a woman, almost always the mother of the children. The percentage of one-parent families that are headed by a woman has not changed very much during the past 20 years and, if anything, appears to be increasing (McDonald 1995; ABS 1999b). Table 4.8 Family type of children, 1997 Age of child 0–2 years 3–4 years 5–11 years 12–14 years SOURCE

Couple family % 85.3 82.7 80.7 80.8

One-parent family % 14.7 17.3 19.3 19.2

Total % 100.0 100.0 100.0 100.0

ABS 1999b, p. 22.

The registered marital status of lone parents in 1997 shows that 62 per cent were separated or divorced from a marriage, 31 per cent had never married and 7 per cent were widowed (ABS 1999b, p. 23). A high proportion of those who were never married would have had their children as part of a living-together relationship, and the family became a one-parent family because the living-together relationship broke down. Also, some of the births to lone parents who were separated or divorced from a marriage or widowed may have occurred outside the marriage. The proportion of all births that were births to women who were not married (ex-nuptial births) increased from 9 per cent in 1971 to 28 per cent in 1997. However, the paternity of the child was acknowledged on the child’s birth registration for 85 per cent of ex-nuptial births in 1997. Overall, paternity is not

92 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

acknowledged for only 4 per cent of all births. Evidence from the Negotiating the Life Course Survey shows that, for 74 per cent of exnuptial births in the 1990s, the parents of the child were living together at the time of the birth and, in a further 8 per cent of cases, the parents had lived together before or after the birth of the child. Thus, the parents of the child have neither married nor lived together for 18 per cent of ex-nuptial births constituting only 5 per cent of all births, virtually the same as the percentage of births where the father is not acknowledged. Ex-nuptial births are often associated with teenagers in the public mind. In fact, less than one in six of all ex-nuptial births in 1997 were to mothers in their teens. In summary, almost all Australian children live with their mother and the sizable majority live with both their natural parents. Oneparent families have been increasing mainly because of marriage breakdown and the breakdown of living-together relationships, but the father is acknowledged on the birth record for all but 4 per cent of all children. Among those children who are not living with their natural father, again a sizable majority has contact with their father. For children under the age of five who have a parent living elsewhere, 73 per cent have contact with that parent. For those aged 5–11 years, this percentage falls to 63 per cent (ABS 1999b, p. 28). These results suggest that conservative and radical pronouncements of the demise of the family, based on the situation of children, are premature, at least in Australia. There is support here for the liberal case that assistance should be provided to help couples stay together but that the breakdown of some families is inevitable and so, in most cases, parenting relationships should be fostered for the minority of children whose parents do not live together.

L A B O U R F O R C E PA RT I C I PAT I O N O F PA R E N T S Between the 1970s and the 1990s, the proportion of mothers in the labour force has risen from about 25 per cent to 50 per cent for mothers whose youngest child is aged 0–4 years and from about 30 per cent to 70 per cent where the youngest child is aged 5–9 years. Much of the increased participation of mothers has been in the form of part-time work and this form of involvement expanded rapidly in the 1980s. Besides having an effect upon the number of children that women have, postponement of the birth of the first child also fundamentally changes the labour force experience of young women. The longer that the first birth is postponed, the higher is the level of work force experience gained before the birth of the first child. Obtaining

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 93

qualifications and work force experience are the two most important means of human capital formation. In the 1950s and 1960s, when the age at first birth was low, women had very little experience of the labour market before they had their first child. Most were essentially beginning their careers when they returned to the labour force after having children. Without experience and with mainly part-time return to the labour force, the chances of establishment in their careers were lessened. The only advantage they had was that almost all other women were in the same situation. Few women had no children and few had only one child. Today the situation is very different. Most women, before having their first child, have worked full-time for several years and most have established themselves in some form of career. In 1998, 92 per cent of women aged less than 35 who had a partner but no children were in the labour force. Of those who were employed, 82 per cent were employed full-time (ABS 1998b). This changed experience of young women has three important effects. First, it means that women have more to lose if they quit their attachment to the labour force through having children. Second, because they have a greater incentive to return to the labour force, they are likely to do so more quickly than was the case in previous generations. Third, because they have a much higher level of human capital than previous generations, they are more readily employable than mothers of earlier generations. In addition, those who return after childbearing are now in competition with a higher proportion of women who have no children. From this observation of women’s labour market participation in relation to the number of children that they have, it is possible to calculate the loss of lifetime earnings that a woman experiences because of the birth of a child. This was done by Beggs and Chapman (1988) using the 1986 ANU Family Survey and by Chapman et al. (1999) using the Negotiating the Life Course Survey, 1996–97. This research shows that the largest part of lost earnings is due to the birth of the first child. Losses related to subsequent births are considerably lower. However, by comparing the results of the two studies, Chapman et al. found that the loss of earnings related to the birth of the first child dropped substantially between the two surveys. In 1997 dollars, the lifetime earnings lost through having the first child for women who had completed high school fell from $435 000 in 1986 to $200 000 in 1997. This indicates that the greater level of attachment to the labour force that today’s young women gain

94 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

before they have their first child enhances their lifetime earnings through greater attachment and higher earnings after they return to the labour force. From the perspective of gender equity, it is worth noting that the lifetime loss of earnings for men who have children is less than zero. That is, men who have children actually have higher lifetime earnings than those who do not have children. The fact that this is a very different story from that of women who have children confirms the emphasis taken in this chapter upon the changing employment patterns of mothers rather than of fathers. At the same time, with rising expectations and increased insecurity of male employment, the employment of mothers provides both additional family income in the good times and a hedge against unemployment of fathers in the bad times. Most young couples today see two incomes as being necessary to maintain the lifestyle to which they aspire. That two incomes are required in today’s environment is also indicated by the recent expansion of tax rebates to one-income families with children aged less than 5 years.

E M P L OY M E N T C I R C U M S TA N C E S O F M OT H E R S AC C O R D I N G TO T H E N U M B E R A N D AG E S O F T H E I R C H I L D R E N Decisions about the timing and number of children are much more likely to be influenced by labour market considerations when women have developed a higher level of attachment to the labour force. It is important therefore to consider changes in the employment circumstances of mothers with varying numbers and ages of children. The following discussion is based upon analysis of the one per cent sample files from the 1986, 1991 and 1996 Australian censuses. Among couple families with one child, the percentage of mothers employed was around 25 per cent for those whose child was aged less than one year, and there was little change in the percentage employed over the decade (Table 4.9). When the women had two children, one of whom was aged less than one year, the picture was very much the same. Thus, the data suggest that it is the age of the youngest child rather than the number of children that determines the employment of the mother when one of the children is a baby. However, among those with a baby, where there was only one child, women were a little more likely to work 25 hours per week or more: about 50 per cent of those employed with one child and 40 per cent

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 95

of those employed with two children. While the numbers were small and hence less reliable, the percentages of lone parents with a baby who were employed were much lower than for mothers in couple families with a baby (Table 4.10). Among couples with one child, the percentage of mothers who were employed rises as the age of the child increases. In 1991 and 1996, over 50 per cent of women with one child aged 1–2 years were employed, with about half of those employed working for 25 hours or more per week. Again, when there were two children, the younger of whom was aged 1–2 years, the employment rate of mothers in 1996 was similar to that of mothers with one child of this age. That is, again, it was the age of the youngest child that mattered rather than the number of children. However, this was not the case in 1986 and 1991. The employment rate of women with two children where the younger was aged 1–2 years had risen sharply over the decade. Furthermore, the proportion of these women employed for 25 hours or more per week had also tended to rise during the decade. In regard to child care, the most important age group of children is those aged 3–5 years who are not yet at school. In one-child families where the child was aged 3–5 years, the proportion of mothers employed rose substantially over the decade, from 48 per cent in 1986 to 55 per cent in 1991 and to 62 per cent in 1996. Similar rises during the decade are evident for those with two children where the younger was aged 3–5 years. For one-parent families with a child aged 3–5 years, an increase in employment over the decade was also evident but it was not as great as for couple families. Thus, the targeting of child care assistance to this group over the decade from 1986 to 1996 seems to have had a substantial effect upon the employment of mothers where the youngest child was aged 3–5 years. Indeed, in 1996, where there was only one child, the employment rate of mothers in couple families was almost the same for those whose child was aged 3–5 years as for those with an older child. This conclusion provides a different impression from that given by Gregory (1999). Gregory concluded that ‘the rapid expansion of child care places that occurred over the 1991–96 period was associated with a marked decline in the rate of growth of employment of women with dependent children’ (Gregory 1999, p. 14). Two comments can be made. First, it is preferable to examine the employment rate of the mothers with children of the age to which the new child care places were mainly directed rather than the employment rate of

96 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 4.9 Employment of mothers in couple families according to the number and ages of their children, 1986, 1991 and 1996 Ages of children

1986 1991 1996 Percentage with mother employed at least one hour per weeka

Families with one child <1 year old 1–2 years old

24 42

27 53

28 50

3–5 years oldb Primary school Secondary school+

48 56 60

55 60 66

62 63 69

14

28

22

23

28

24

–c

31

27

33 38

41 45

48 53

Both 3–5b

37

45

45

3–5b

45 55 59 63

53 66 69 68

56 65 67 75

Families with three children All preschool Two preschool and one primary+ One preschool and two primary+ All primary Two primary and one secondary+ One primary and two secondary+ All secondary+

22 27 37 52 52 60 63

27 34 45 54 65 62 71

26 33 49 61 64 74 70

Families with four children At least one preschool child All primary+

29 47

32 50

30 52

Families with two children <1 and 1–2 <1 and 3–5

b

<1 and primary+ Both 1–2 or 1–2 and 1–2 and primary+

3–5b

and primary+ Both primary Primary and secondary+ Both secondary+

a Mothers aged less than 55 years; b 3–5 year-olds who are not at primary school; Denominator is less than 100 mothers. SOURCES 1986, 1991 and 1996 censuses (Household Sample File).

NOTES c

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 97

Table 4.10 Employment of mothers in one-parent families according to the number and ages of their children, 1986, 1991 and 1996 Ages of children

1986 1991 1996 Percentage with mother employed at least one hour per weeka

Families with one child <1 year old 1–2 years old

11 20

15c 23

10 30

3–5 years oldb Primary school Secondary school+

33 36 52

36 45 61

44 52 63

Younger is aged <1c Younger is aged 1–2

4 14

17 22

9 21

Younger is aged 3–5b Younger is at primary school Younger is at secondary school+

28 39 52

36 51 67

31 53 71

Families with two children

NOTES a Mothers aged less than 55 years; b c Denominator is less than 100 mothers. SOURCES

3–5 year olds who are not at primary school;

1986, 1991 and 1996 censuses (Household Sample File).

mothers with dependent children of any age. Second, the rate of growth of any percentage will tend to slow down as the percentage gets larger. In these circumstances, it is preferable to look at the absolute shift in the percentage employed rather than the percentage shift in the percentage employed. Of mothers in couple families with one child aged 3–5 years, over 60 per cent worked 25 hours or more per week at all three censuses (63 per cent in 1996). This figure was a little lower where there were two children and the younger was aged 3–5 years (54 per cent in 1996), but, again, there was little change in this proportion across the censuses. For lone mothers with one child aged 3–5 years, however, there was a substantial decrease in the proportion who were working 25 hours or more per week. The percentage fell from 80 per cent in 1986 to 56 per cent in 1996. This was not so much because there was a shift among individual women from full-time to part-time work, but rather that the addition to the employment rate during the decade was made up almost entirely of lone parents working part-time.

98 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Employment rates of mothers in all categories increased across the decade when all children were of school age. The rises were more substantial, however, for those with somewhat greater difficulties in relation to child care, that is, lone mothers and mothers in couple families with three children. Again, the increase during the decade in out-of-school-hours care seems to have been effective in promoting the employment of mothers, especially those with heavier demands. As the employment rate of women is relatively low when they have a baby, irrespective of the number and ages of other children, the timing of births determines whether women have one longer interval out of the labour force (short intervals between births) or a succession of periods in and out of the labour force as they have each successive birth (longer intervals between births). An investigation of women’s preferences and behaviour in this regard would be useful. I have examined the ages of the younger siblings of children aged 3–5 years (not yet at school with no older siblings) at the 1996 Census. This suggests that women are much more likely to fall in the category of one absence from the labour force with the length of the absence being dependent upon the number of children.

F U T U R E E M P L OY M E N T R AT E S O F M OT H E R S AND POLICY RESPONSES Couples and especially women will be making decisions about family formation and about employment within the social and institutional context of a future Australia. Several aspects of that context will influence both decisions about employment and decisions about the number of children that people have. The first set of social and institutional considerations is changes in attitudes and values. As discussed above, young women today, before they have commenced having children, have a much stronger attachment to the labour force than was the case in previous generations. They have more to lose by not working. They are more highly educated and have been socialised through education, family and the workplace to expect to be employed in the paid labour market. Housing mortgages are usually based on an assumption that both parents will be working. There is an indication that young men welcome the additional income that a working partner obtains. In the Negotiating the Life Course Survey, 71 per cent of men aged less than 30 years were opposed to the notion that it was better for the husband to be the principal breadwinner while the wife had primary

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 99

responsibility for the children. Also, young people may have become accustomed to a higher spending lifestyle before their first child was born. Finally, employment rates of mothers of young children are higher in countries such as Canada (Peron et al. 1999, pp. 264–69), the United States and the Nordic countries than they are in Australia. All of these considerations point to an increase in the employment rate of mothers of young children in Australia in the future. On the other hand, there has been little change in the extent to which young men are prepared to share household work and this will be an obstacle to mothers taking up paid employment (Baxter 1996). The second set of considerations relates to the labour market. If the economy remains strong, there should be a rising demand for the employment of young mothers because of the skills that they hold as a result of their education and workforce experience. There are also signs of a tightening of the skilled end of the labour market as the cohorts entering the labour force become smaller. The advance of technology and the growth of the service and information sectors also favour the employment of women. On the other side of the equation, mothers of young children require job security and predictability, and family-friendly workplaces. They also need to be able to work for the number of hours that best fits their family responsibilities. If they are unable to return to work part-time or if their employer puts excessive demands upon them, then the employment rate of mothers will fall. Alternatively, women will have fewer children. Flexibility of work hours benefits women with young children, but this is not necessarily the case if the flexibility is all in the control of the employer. Twelve-hour shifts or early morning starts, beneficial to many workers without young children, are less likely to be in the interests of those with young children. An argument is often made that employers will benefit from family-friendly workplace policies because they will be able to retain or attract the best workers. If this were so we would expect employers to support new approaches to family-friendly workplaces. The evidence is counter to this. Employer organisations almost always oppose suggested new initiatives to make workplaces more familyfriendly. The third set of factors relates to the economic returns from employment. This includes the wage rate that employed mothers can obtain, the costs of child care, the costs of working including transport costs and the changes in entitlements through the tax-transfer

100 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

system as mothers take on employment. Costs of child care have been increasing in recent years and the loss of benefits from the tax-transfer system resulting from employment has increased through the introduction of single-income family tax rebates. The costs of child care and the single-income family tax rebate provide an incentive for the father to work more hours than he does at present and for the mother not to work. For example, where the husband works 40 hours per week and the wife none, the couple will be better off if he works an additional 10 hours per week than if she works 10 hours per week. The effect upon employment decisions of mothers of these changes is yet to be assessed. Finally, employment decisions of mothers of young children will be affected by the availability of quality, affordable and accessible child care. It has been argued that many parents are unable to obtain cheap, informal care through the extended family. This implies that good, affordable and accessible child care must be provided through the formal sector. Australia has taken great strides to make provision for this type of care over the past 15 years. Progressively, we have been building a system of which we can be justly proud. The argument here would be that we should continue to build the system. While a careful modelling exercise is required, the present indications are that the demand for child care places, which rose sharply during the 1990s, is likely to level off in the near future. In summary, Australian women tend to increase their labour force participation as their youngest child ages. Participation rates step up when the child reaches the first and third birthdays. Only a minority are working when the child is aged less than one year, but a clear majority are working when the youngest is aged 3–4 years. Part-time work is common. This middle pathway of participation is not supported by current policy. Under the banner of the provision of choice, current family policy allocates the highest financial rewards to those who take the two most extreme choices, that is, mothers who spend no time in paid employment while they have young children and mothers who work full-time from soon after the birth of their child. Mothers who spend no time in paid employment have benefited from new tax policies that provide tax rebates for single-income families. At the other end of the spectrum, the child care benefit arrangements provide the highest benefits to mothers who work the longest hours. The former Child Care Rebate, available to all working parents, was based on the cost

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 101

of child care irrespective of the hours worked. Its replacement bases benefits on hours worked with the maximum benefit flowing to those who work the longest hours. Thus, present policy provides incentives for women to take the extreme choices and, consequently, disincentives to make choices that lie between the extremes. Yet all the evidence of preference and behaviour is that most families prefer a middle course. Policy favours the extreme choices because governments have attempted to accommodate views about the employment of mothers that extend from the conservative to the radical instead of recognising that the reality of most people’s lives lies between the extremes. More specifically, the problem is that policy only distinguishes between those who have children under and over the age of five years. All families with a child under the age of five (or not at primary school) are treated in the same way irrespective of whether their youngest child is aged six months or four years. Behaviour and preference data, however, show that families themselves make very stark distinctions between these circumstances. Families prefer to increase their labour force participation and the time that a child is in child care as the child ages from birth to school age. Policy would be more effective if it were more attuned to these preferences. Essentially this means greater emphasis on leave and income support when children are very young (under one year); a mixture of income support and child care much like the present child care system, but more generous, when children are aged 1–2 years; and universal, free early childhood education with child care at the beginning and the end of the school day for children who are aged 3–4 years.

DISCUSSION In the past 30 years in Australia, rules and norms governing family relationships have been liberalised. Much of the change occurred in the 1970s but it has been sustained since that time with less assault from the conservative agenda than has been the case in the United States or the United Kingdom. In regard to the public-private dichotomy, public support of the liberalised agenda is much more prominent in Australia than in the United States. In Australia, state support takes the following main forms: a very liberal divorce law, equal treatment of marriage and living-together relationships in most state systems (immigration, social security, taxation, health, child custody), provision of a livable sole-parent pension that is not work-

102 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

tested, abortion free to all through the national health system, statesubsidised housing, state-subsidised child care, and various provisions that support the employment of women. At the same time, aspects of the liberalisation of family relationships have not extended quite so far in Australia as they have in the United States. The divorce rate, the labour force participation rate of mothers, and the proportion of children born outside marriage all stand at significantly lower levels in Australia than in the United States. People’s lives can be characterised in three broad spheres: the self, the intimate and the social (McDonald 1996). In the conservative agenda, rules and norms of behaviour are set in the public or social sphere that largely determine how people behave in the intimate sphere of family relationships. The third sphere of life, the self, is expected to conform to these rules and norms. In the radical agenda, autonomy is provided to individuals (the self) to determine their own intimate outcomes free of social rules and norms. The liberal agenda floats between these two. Social changes in the past 30 years have provided more autonomy to individuals to determine their own individual and intimate lives. In other words, conservative social rules and norms have been greatly relaxed and, indeed, often the state has paid the financial costs of liberalisation. However, autonomy provides an increased range of choices such that most people experience conflict between the three broad spheres of life: their social sphere (particularly work), their sphere of self (personal autonomy) and their intimate sphere (couple relationships and parent-child relationships). The need to resolve the conflict between individuation (autonomy) and fusion (intimacy) was the subject of the presidential address to the American Psychological Association by Janet Spence in 1985 (Stevens-Long and Commons 1992). These conflicts are acute in the transition periods of the life course such as forming and ending couple relationships and having children. The changes in women’s and men’s lives over the past 30 years have accentuated the degree of conflict. The liberal approach to this increased conflict has been to allow people greater flexibility in the ways in which they arrange their relationships and to provide new supports in combining work and family responsibilities. The growth of cohabiting unions, divorce, later marriage, later childbearing and having children outside of marriage can all be seen as social experiments that the society has been prepared to tolerate as people seek their own solutions to these conflicts.

T R A N S F O R M AT I O N S I N T H E A U S T R A L I A N F A M I LY

• 103

Rather than being driven by conformity to rules and norms, behaviour becomes ethical, driven by personal principles (Stevens-Long and Commons 1992, p. 85). This reliance on ethical behaviour, however, is at once the strength and the weakness of this social approach. In Australia, the liberal social experimentation is holding up relatively well and, in general, has community support. However, as in the United States and the United Kingdom, the argument is mounted that the financial cost of these social experiments should be borne by those who wish to engage in them rather than by the state. Through this means, changes in family relationships become caught up in social debate about welfare dependency, and the conservative-liberalradical debate about the family continues.

5

THE CHANGING DIMENSIONS O F M O R TA L I T Y H E AT H E R B O OT H

Mortality is one of the three underlying factors determining the demographic characteristics of any population, and is thus instrumental in determining the nature and rapidity of the transformation of Australia’s population. Along with other industrialised populations, the Australian population underwent a major mortality transition over the course of the 20th century (Young and Ruzicka 1982). The decline in mortality was such that life expectancy increased from 57 years in 1901–10 to 80 years in 1999. The causes of the decline follow the classic epidemiologic transition model (Omran 1971, 1977). The transition from the second to the third stage of the epidemiologic model was dominated by two main trends: a dramatic reduction in infectious diseases affecting mainly younger age groups in the first half of the century; and from about 1970 a decrease, after an initial increase, in chronic diseases affecting older age groups (Australian Institute of Health and Welfare (AIHW) 2000, ch 8; Taylor et al. 1998a, 1998b). The balance of these two trends resulted in a period of almost static mortality in the 1960s, but from 1968 the decline has been relatively rapid. By the late 20th century, Australia had moved into the fourth stage of the epidemiologic transition characterised by high life expectancy and delayed degenerative diseases (Olshansky and Ault 1986). The reduction in mortality has had two fundamental effects. First, decreased death rates have resulted in more rapid population growth than would otherwise have been the case. Second, the increased longevity of individuals has contributed to the changing

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 105

structure of the population towards greater proportions at older ages. The exact manner in which mortality affects population change depends on the age-sex patterns of mortality and of mortality change. These in turn are dependent on disease patterns and fatality rates which determine the levels and patterns of mortality by cause of death. This chapter examines the age, sex and cause-of-death dimensions of mortality in Australia since 1970. It shows the contributions that each age and cause of death have made to increases in life expectancy between 1971 and 1995 for females and males, and to sex differences in life expectancy at different points in time. The Australian mortality experience is also compared internationally. Finally, the prospects for future mortality are considered.

D ATA A N D M E T H O D S This analysis is based on three-year central death rates (three-year average annual deaths per mid-central-year population). These are referred to by the central year. The numerators are death registration data (ABS, annual-a) and the denominators are mid-year estimated resident population data (ABS, annual-b). Rates for 1971–91 were obtained from Jain (1994). The latest accessible detailed death registration data were for the years 1994–96, and the requisite tabulations were provided by the National Centre for Epidemiology and Population Health at the Australian National University. Death registration data are classified by underlying cause of death according to the World Health Organization International Classification of Diseases (ICD), which is periodically revised. Deaths before 1979, which were originally classified according to the eighth revision of ICD, were reclassified according to the ninth revision for Jain’s (1994) study. Hence the time-series data in this analysis are consistent with respect to the classification of cause of death. The grouping of detailed causes of death into the 11 broad underlying cause of death categories used in this analysis is shown in Table 5.1. Standardisation of death rates is achieved by direct standardisation and is based on the 1986 total estimated resident population using ages 0, 1–4, 5–9, … , 90–94, 95+. Standardised rates are comparable over time and between sexes. Life expectancies are from abridged life tables in order to maintain consistency with the causeof-death analysis. The methodology for the decomposition by age of differences, over time or between sexes, in life expectancies follows Pollard (1982; see also Jain 1994).

106 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 5.1 Underlying cause-of-death categories by detailed ICD code and distribution by sex, 1995 Underlying cause-ofdeath category

ICD codes (ninth revision)

Heart disease 393–398, 402, 404, 410–416, 420–429 Malignant neoplasms (cancer) 140–208 Cerebrovascular disease (stroke) 430–438 Chronic obstructive pulmonary disease and allied conditions (incl. asthma, emphysema and bronchitis) 490–496 Motor vehicle traffic accidents 810–819 Other accidents 800–809, 820–949 Pneumonia and influenza 480–487 Diabetes mellitus 250 Suicide 950–959 Nephritis, nephrotic syndrome and nephrosis 580–589 Chronic liver disease and cirrhosis 571 Other causes Remainder Total

Distribution in 1995 Male Female 29.0 28.7 7.7

30.3 24.7 12.7

6.0 2.1 2.5 1.2 2.1 2.8 1.1 1.1 15.7 100.0

4.4 1.0 1.6 1.6 2.4 0.8 1.5 0.5 18.5 100.0

Jain 1994; tabulated data from death registration files held by the National Centre for Epidemiology and Population Health, Australian National University.

SOURCES

T R E N D S I N O V E R A L L M O R TA L I T Y Over the period 1971–95, the trend in overall mortality was one of decline. This decline was a continuation of a longer-term trend, although there was a period of little change in the 1960s. Even with the ageing of the population, the crude death rate declined from 8.5 per 1000 in 1971 to 6.9 in 1995. When the population age structure is taken into account, the standardised death rate shows a decline of approximately 40 per cent over the same period: for females from 8.0 to 4.7 per 1000 and for males from 13.1 to 7.9, as seen in Figure 5.1. Corresponding to this decline in mortality, female life expectancy increased from 74.9 years in 1971 to 81.2 in 1995, and male life expectancy increased from 68.2 to 75.4 years. Figure 5.1 shows the trend in life expectancy since 1921.

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 107

Figure 5.1 Life expectancy and standardised death rate by sex, 1921–95 90

18

85

16

Life expectancy (years)

12

75

10

70

8

65

6

60

Deaths per 1,000

14

80

4

55

2 0

50 1921 1931 1947 1954 1961 1966 1971 1976 1981 1986 1991 1995

SOURCE

Female e(0)

Male e(0)

Female SDR

Male SDR

Jain 1994; and author’s calculations.

C A U S E S O F D E AT H In 1995, the main causes of death were heart disease and cancer, and to a lesser extent stroke: these three diseases accounted for about two-thirds of all deaths. Table 5.1 shows the distribution of cause of death by sex in 1995. Heart disease accounted for roughly 30 per cent of deaths for both females and males. Other causes of death affect the sexes less equally: cancer is more common among males (29 per cent compared with 25 per cent among females), while stroke is more common among females (13 per cent compared with 8 per cent among males). Of the remaining causes of death, chronic obstructive pulmonary disease (COPD), motor vehicle traffic accidents (MVA), other accidents, suicide and liver disease are all more frequent among males than among females, while pneumonia and influenza, diabetes mellitus and kidney disease (nephritis, nephrotic syndrome and nephrosis) are more frequent among females. To a large extent, this reflects lifestyle differences between the sexes and the fact that females survive on average to older ages.

108 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The distribution of cause of death is partly dependent on the age structure of the population. Standardised death rates, which take into account age-structure differences over time and between the sexes, are shown for the main causes of death in Figure 5.2 for the period 1971 to 1995. For both sexes, cause of death patterns have changed considerably over time, the main changes being significant reductions in deaths due to heart disease, especially among males, and stroke. In contrast, standardised death rates due to cancer have remained almost constant over the 24-year period, and by 1995 cancer almost equalled heart disease as a cause of death for both females and males. For most other causes of death, rates have declined albeit from relatively low levels. The exceptions are COPD among females, due to an increase in smoking in previous decades; suicide among males; and kidney disease among both females and males.

A G E - S P E C I F I C M O R TA L I T Y The age pattern of mortality is seen in Figure 5.3 which shows agespecific death rates (deaths per 1000 persons of the appropriate age) in 1971, 1981 and 1995 for males and females. The 1995 pattern is very similar to the 1991 pattern (not shown). The pattern follows the typical J-shaped curve found in developed societies. In 1995, the lowest rate occurred at age 5–9. This represents a change since the 1970s when the minimum occurred at age 10–14. The decline in mortality since 1971 has occurred fairly consistently at all ages and over all time periods. This is clearly the case among females, as Figure 5.3 shows. For males, however, the decline is less consistent in youth and young adulthood and small increases have occurred in recent periods at ages 25 to 34. This is discussed in more detail below. The reader is reminded that by adopting the standard practice of using a logarithmic scale in Figure 5.3, the visual comparison is one of relative change: a constant difference between the mortality curves indicates a constant percentage change. In making comparisons, changes over time can be expressed in either absolute or relative terms. The pictures portrayed by these two approaches will differ considerably. Figure 5.4 shows both absolute and relative reductions in age-specific death rates by sex over the period 1971–95. It is immediately seen that the absolute reductions are greatest at older ages, rising steeply from about age 50 or 60, whereas relative to 1971

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 109

Figure 5.2 Standardised death rates for selected causes of death, 1971–95, males and females.

Males 6

Deaths per 1,000

5 4 3 2 1 0 1971

1976

1981

1986

1991

1995

1986

1991

1995

Females 6

Deaths per 1,000

5 4 3 2 1 0 1971

1976 Heart

SOURCE

1981 Cancer

Jain 1994; and author’s calculations.

Stroke

COPD

110 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 5.3 Age-specific death rates, 1971, 1981 and 1995, males and females

Males

Deaths per 1,000 (log scale)

1,000

100

10

1

0.1 0

5

15

25

35

45

55

65

75

85

95

55

65

75

85

95

Females

Deaths per 1,000 (log scale)

1,000

100

10

1

0.1 0

5

15

25

35

1971

SOURCE

Jain 1994; and author’s calculations.

45

1981

1995

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 111

Figure 5.4 Absolute and relative reductions in age-specific death rates by sex, 1971–95 80

Absolute reduction per 1,000

80

60

60 40 40 20

20

0

0 0

SOURCE

Relative reduction (% of 1971)

100

1

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Absolute-female

Absolute-male

Relative-female

Relative-male

Author’s calculations.

rates, reductions are greatest at very young ages and least at very old ages, with the exception of males aged 25 to 34 for whom very little change occurred. Both measures should be borne in mind when examining change. For the decomposition of changes in life expectancy, however, absolute changes in age-specific rates are of interest since these constitute the additive components of the overall change (Pollard 1982). In the past, changes in overall mortality were influenced significantly by reductions in infant mortality. This has become less so as infant mortality rates have reached lower levels. In 1971, the infant mortality rate was 15.0 per 1000 live births for females and 19.5 for males. By 1995, these rates had decreased to 5.1 and 6.4 respectively. About 60 per cent of this decline took place during 1971–81, indicative of the inevitable levelling off in this measure as rates approach zero. In 1971, 88 per cent of infant deaths were due to ‘other’ causes, which for this age mostly comprises certain conditions originating in the perinatal period and congenital anomalies. A further 10 per cent were due to pneumonia and influenza and to other

112 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

accidents. By 1995, 95 per cent of infant deaths were due to ‘other’ causes. Improvements in infant mortality must thus now be gained almost exclusively through continued reductions in perinatal deaths (defined in Australia as stillbirths and deaths within the first 28 days of life) and congenital causes of death. The low levels of mortality during the childhood ages 1 to 14 mean that absolute changes at these ages are small with little room for improvement. The female death rate at 5–9 was as low as 0.132 per 1000 in 1995. The main causes of death at these ages are accidents (MVA and other), cancer and other causes (principally congenital anomalies at age 1–4). The steep incline in mortality from age 10–14 to 15–19 reflects the risk-taking behaviour of youth, especially among males. It is seen in Figure 5.3 that in 1971, this behaviour gave rise to a distinct hump at ages 15 to 24 in the male mortality pattern, characterised by a local maximum at 20–24 and a local minimum at 25–29. This hump emerged after World War II and reached a peak in the late 1970s (Australian Government Actuary 1999). Since more than half of all deaths at 15 to 24 were due to motor vehicle accidents and a further 15 per cent due to other accidents, this hump has been labelled the ‘accident hump’. By 1981, the hump was less distinct but extended to age 25–29 and by 1995 the local maximum had disappeared. Its decline was partly due to a reduction in mortality at 15 to 24 and partly to the absence of a significant decline at ages 25 to 34. This was due to a change in the distribution of causes of death: whereas in 1971 accidents were the main causes at these ages, in 1995 suicide and other causes were also prominent, especially at 25 to 34. Indeed, rates due to suicide and other causes increased over the period. The effect of these increases was to mask the accident hump, which in fact still existed in that local maxima occurred at 20–24 for MVA and at 25–29 for other accidents despite decreased rates. The increase in suicide among male youth and young adults is a well-known feature of Australian mortality (Ruzicka and Choi 1999). Between 1971 and 1995 suicide rates at ages 15 to 34 increased by 56–90 per cent. The effect was to reverse the 1971 pattern where suicide rates increased with age (to 50–54), such that 1995 rates declined from a peak at 20–24 to age 50–54 at least. Suicide accounted for between 6 and 14 per cent of male deaths at ages 15 to 34 in 1971, and between 21 and 28 per cent in 1995. The main contributing factor to the increase in male deaths from

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 113

‘other’ causes is the emergence of HIV/AIDS. While death rates from ‘other’ causes decreased at most ages over the period 1971–95, rates at ages 20 to 35 increased and at ages 20–24 and 25–29 more than doubled. The annual number of male AIDS-related deaths increased from 558 in 1986, when national data on AIDS mortality first became available, to 721 in 1994 and decreased for the first time in 1995 when 629 deaths were recorded (ABS 1995, p. 9). In 1995, 60 per cent of all AIDS-related deaths were of males aged 35 to 54 (ABS 1995, p. 9). The increase between 1981 and 1995 in the overall male death rate at 30–34 (Figure 5.3) can be attributed to AIDS. Death rates for young adult females demonstrate a lower level of risk-taking behaviour than is evident for males (see Figure 5.3). To some extent the emergence and decline of the considerably less-pronounced female accident hump reflects male experience, though the female local maximum occurs at younger ages. In 1971, MVA accounted for 47 per cent of female deaths at 15–19 and 33 per cent at 20–24, with an additional 6 per cent due to other accidents at both ages. By 1995, corresponding percentages were 34 and 25 for MVA and 10 for other accidents. Death rates due to other accidents have remained fairly constant. In contrast to male rates, female death rates at ages 15 to 34 declined for suicide and for other causes, although both of these causes increased in relative importance. AIDS is not a significant contributor to female mortality; in 1991–95 only 5 per cent of AIDS-related deaths were female (ABS 1995, p. 9). At older adult ages, the decline in mortality has been more regular with respect to both age and time. Nevertheless, Figure 5.4 shows that absolute reductions increased considerably from about age 50 for males and 60 for females, though relative to the high death rates at these ages the reductions decreased from age 50 for females and from 60 for males. This pattern is heavily influenced by heart disease, for which the greatest reductions occurred (see Figure 5.2), and to a lesser extent by stroke. As noted above, standardised death rates due to cancer remained almost constant between 1971 and 1995. This is due to the combined effect of modest improvements at ages less than 60 (for males) or 65 (for females) and increased rates at older ages. For most other causes of death, rates have declined at most adult ages. A notable exception is the increase in rates due to COPD among females aged 55+, reflecting past increases in smoking among females, though rates remain well below those of males. Death rates due to diabetes and liver disease have declined relatively slowly, while

114 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

rates due to kidney disease have declined rapidly at younger adult ages, though for all three causes rates at older ages have increased. The increase in death rates at older ages from these causes and from cancer reflects the fact that those who would have died at relatively young ages in 1971 now survive to older ages. Since we must all die, death rates at older ages for certain causes are increased, and given the current state of medical knowledge these causes are cancer, diabetes, liver disease and kidney disease. The distribution of cause of death at ages 40+ is not dissimilar to that at all ages, but there is a slightly greater concentration on heart disease, cancer, stroke, COPD and diabetes. Age-sex cause-specific death rates at ages 40+ in 1995 show that though death rates due to heart disease are highest at very old ages, at younger adult ages (40 to about 75) the highest rates are due to cancer. In 1995, cancer was the second most common cause of death among females aged 75 to 84 and males aged 70 to 94. At older ages, cancer was third after heart disease and stroke. These patterns are reflected in the median ages at death. For females in 1995, the median ages at death from heart disease, cancer and stroke were 83.2, 72.3 and 84.2 years respectively. Corresponding values for males were 76.0, 71.7 and 79.5 years. These compare with median ages for all causes of 80.4 years for females and 73.6 years for males.

C O N T R I B U T I O N S TO I N C R E A S E S I N L I F E E X P E C TA N C Y O V E R T I M E The importance of age-specific reductions in mortality can be assessed by considering their individual effects on life expectancy (Pollard 1982). Figure 5.5 shows the contribution of reductions in mortality at each age to the increases in life expectancy over the period 1971–95. It should be recalled that the increase for males is greater than the increase for females (7.2 years compared with 6.3). It is immediately seen that the major contribution is at older ages with the female distribution peaking later than the male. For females, 41 per cent of the contribution is attributed to ages 65 to 84 and a further 27 per cent to ages 40 to 64. For males, 43 per cent is attributed to ages 55 to 74 and a further 30 per cent to ages 40 to 54 and 75 to 84 combined. The overriding importance of adult mortality to increasing life expectancy is seen in the fact that 25 per cent of the increase in female life expectancy is attributable to ages 77+, 50 per cent to ages 65+ and 75 per cent to ages 44+. For males, the corresponding ages are 71+, 59+ and 41+.

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 115

Figure 5.5 Contribution of each age to increase in life expectancy by sex, 1971–95 1.0

Contribution (years)

0.8

0.6

0.4

0.2

0.0 0

5

15

25

35 Females

SOURCE

45

55

65

75

85

95

Males

Author’s calculations.

Mortality improvements in the first 40 or so years of life thus accounted for only 25 per cent of the gain in life expectancy. About half of this can be attributed to declines in infant mortality which amounted to 12 per cent of the female gain and 13 per cent of the male gain. Given the current low levels of infant mortality, future contributions at this age will be very small. Contributions for females at ages 1 to 39 are already small reflecting their low mortality with little scope for improvement. For males, the decline in death rates at 15 to 24 contributed 5 per cent of the overall increase in life expectancy, while ages 25 to 34 contributed less than 1 per cent. For females, increases in life expectancy were considerably greater in 1971–81 than in 1981–95, whereas male increases are more evenly distributed over the entire period (see Figure 5.6). Over time, the contribution of older ages to the gain in life expectancy has increased as a proportion of total gain. For example, 50 per cent of the female gain in 1971–81 is attributable to ages 64+, compared with ages 66+ in 1981–95. For males, a greater increase occurred, from 57+ years to 61+ years. This increase in the median age of contributions to the gain in life expectancy is mainly due to the gradual shift of the peak in the later part of the distribution towards older ages, accompanied by a wider spread at these ages. An additional factor is the fact that

116 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

the greater part of the overall decrease in infant mortality took place in 1971–81. Age contributions to gains in life expectancy are clearly related to cause of death. The contribution of each of the main causes of death to increases in life expectancy is shown in Figure 5.6 by five-year period (note that the final period 1991–95 covers only four years). For all periods, the largest contribution to increased life expectancy is attributable to reduced deaths from heart disease, especially for males. Reductions in deaths from stroke also contributed significantly, especially for females in the earlier periods. The age distributions of the contributions made by these two causes to increases in life expectancy largely determine the overall age distribution of contributions at ages 40+ (Figure 5.5). The important contribution of other causes in the earlier periods is partly attributable to reduced perinatal deaths and deaths due to congenital anomalies. For both sexes, small negative contributions to gains in life expectancy occurred in all periods. For females, COPD negatively affected life expectancy in each period and in 1981–86 a relatively large negative contribution occurred owing to increased deaths from cancer, notably lung and other smoking-attributed cancers (Peto et al. 1994). Taking the period as a whole, these negative contributions occurred at ages 55+ for COPD and at 60+ for cancer. Minor negative effects also occurred for various other age-cause combinations including kidney disease among females at ages 70+, but over the period as a whole these were offset by positive effects at younger ages. For males, negative contributions due to cancer occurred in the first two periods, again due to smoking-attributed cancers (Peto et al. 1994) while increased rates of suicide negatively affected life expectancy between 1976 and 1991. Over the whole period, cancer had a negative effect at ages 65+ and suicide at ages 15 to 39 and marginally at 80+. The negative effect of increased deaths from other causes, principally due to AIDS, occurred at ages 20 to 39 but was more than offset by positive effects at other ages, notably infancy. Other minor negative effects also occurred, but over the entire period these were offset by positive effects at other ages. For both sexes, liver disease, kidney disease and diabetes contributed negatively in different individual periods. The negative effects of diabetes in the most recent period were due to increased rates at very old ages.

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 117

Figure 5.6 Contribution of selected causes of death to increases in life expectancy by sex, 1971 to 1995 Males 2.0

Contribution (years)

1.5

1.0

0.5

0.0

-0.5

1971–76

1976–81

1981–86

1986–91

1991–95

1986–91

1991–95

Females 2.0

Contribution (years)

1.5

1.0

0.5

0.0

-0.5

1971–76

Heart

SOURCE

Author’s calculations.

1976–81

1981–86

Cancer

Stroke

All other

118 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

SEX DIFFERENTIALS It is evident from Figure 5.1 that females enjoy a considerable advantage in overall mortality. In 1971, the female advantage in life expectancy was 6.7 years, but this had narrowed to 5.8 years by 1995. In fact, the differential widened to 7.1 years in 1981 and narrowed thereafter. Throughout the entire period, male mortality exceeded female mortality at every age. The age pattern of the absolute sex differential is roughly similar (though on a smaller scale) to the age pattern of male death rates since larger differentials can occur in higher rates. Little change in this pattern occurred between 1971 and 1981, but by 1995 the differential had narrowed at all ages except 25 to 39 where it had widened because of increased mortality from AIDS and suicide among males. In relative terms, the greatest differentials (ratios of 2.5–3.5) occurred at ages 15 to 24, widening to ages 15 to 34 in 1995, and the smallest differentials (ratios of 1.0–1.5) at young and very old ages. In the older adult ages, 50 to 74, the relative differential was approximately 2.0 with some diminution over time. The standardised death rates in Figure 5.1 show that the absolute sex differential narrowed in each five-year period, from 5.1 in 1971 to 3.1 in 1995. In relative terms, however, the trend is similar to that already observed in the absolute difference in life expectancies: the ratio of male to female overall standardised death rates increased from 1.64 in 1971 to 1.76 in 1981 and then decreased to 1.66 in 1995. These overall trends are dominated by trends in heart disease. The convergence of male and female standardised death rates due to heart disease is shown in Figure 5.7 to be less rapid in the first two periods to 1981, while the ratio of male to female rates increased slightly during these periods. This is due to the fact that during the first two periods the female rate declined more rapidly than the male but from a smaller base resulting in a smaller absolute decline, whereas in later periods the male rate declined more rapidly (see Figure 5.2). A similar but less significant convergence of rates also occurred for several other causes of death. This includes other causes, despite the effect of AIDS, because of the reduced absolute differential in infant mortality as rates fell. A more notable convergence, clearly seen in Figure 5.7, occurred in COPD owing to both a decreasing male rate and an increasing female rate. In contrast, the divergent rates for suicide result from an increasing male rate and a decreasing female rate. A divergence of

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 119

Figure 5.7 Absolute and relative sex differentials in standardised death rates for selected causes of death, 1971–95 Absolute differentials

Difference (deaths per 1,000)

2.5

2.0

1.5

1.0

0.5

0.0 1971

1976

1981

1986

1991

1995

1991

1995

Relative differentials 6

Male:female death rates

5 4 3 2 1 0 1971

SOURCE

1976

1981

1986

Heart

Stroke

MVA

Cancer

COPD

Suicide

Jain 1994; and author’s calculations.

Other

120 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

rates also occurred for cancer in the first two periods, due mainly to an increased rate for males, contributing significantly to the overall trend. While most ratios range from 1 to 2, ratios of between 2 and 3 occur for MVA, other accidents and liver disease and even higher ratios occur for COPD and suicide. For all causes where the ratio is greater than 2, psychosocial and lifestyle factors are involved.

C O N T R I B U T I O N S TO S E X D I F F E R E N C E S I N L I F E E X P E C TA N C Y The individual contributions of causes of death to sex differences in life expectancy for each five-year period are shown in Figure 5.8. In 1971 and 1976, heart disease accounted for a sex differential of almost three years amounting to over 40 per cent of the overall sex difference. As rates converged, the contribution of heart disease declined to 1.6 years in 1995, constituting less than 30 per cent of the total. The widening sex differential in life expectancy in 1971–81 was due only very marginally to heart disease and mainly to the divergence in cancer rates over this period. Though standardised cancer rates converged slightly after 1981, the contribution of cancer to the overall sex differential in life expectancy continued to increase over the entire period. By 1995, cancer accounted for 25 per cent of the total sex difference. However, after 1981 the increased contribution of cancer was more than outweighed by the decreased contribution of heart disease. The narrowing sex differential after 1981 can be attributed almost entirely to heart disease and would in fact occur at a faster rate if cancer were not increasingly important in determining the sex differential. The third largest contribution to the sex difference is attributable to ‘other’ causes. In 1971 this was due largely to perinatal and congenital conditions but by 1995 a major factor was AIDS. Declines in the absolute and relative contributions of MVA and other accidents over the entire period, and in COPD after 1981, contributed to the downward trend in the overall sex differential. Partly offsetting these were increases in the contribution of suicide which in 1995 accounted for 7 per cent of the total difference. These cause-specific contributions to the sex difference in life expectancy are reflected in the contributions made by each age, shown in Figure 5.9. As in the case of contributions to increases in life expectancy over time, the major contributions to the sex difference are at older adult ages. In 1995, 58 per cent of the sex

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 121

Figure 5.8 Contribution of selected causes of death to sex difference in life expectancy, 1971–95 8 7

Contribution (years)

6 5 4 3 2 1 0

1971

1976

Heart SOURCE

Cancer

1981 COPD

1986 MVA

1991

1995

Suicide

All other

Author’s calculations.

Figure 5.9 Contribution of each age to sex difference in life expectancy, 1971 and 1995 1.0

Contribution (years)

0.8

0.6

0.4

0.2

0.0 0

1

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 1971

SOURCE

Author’s calculations.

1995

122 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

differential in life expectancy was attributable to ages 60 to 84. There has also been a shift over time towards even older ages. In 1971, 25 per cent of contributions were at ages 71+, 50 per cent at ages 62+ and 75 per cent at ages 46+. By 1995, corresponding ages were 77+, 67+ and 50+. This shift to greater contributions at older ages reflects the shift to older contributions to gains in life expectancy given older female contributions to these gains (Figure 5.9). At older ages, heart disease and cancer were by far the main causes contributing to the sex difference in life expectancy. At ages 50+, these two causes accounted for 35 and 34 per cent respectively of the contribution to the sex difference in 1995. A further 9 per cent is attributable to COPD. These percentages have changed considerably since 1971 when 51 per cent of the contribution at 50+ was due to heart disease, 17 per cent due to cancer and 12 per cent to COPD. The shift to older contributions to the sex difference in life expectancy is also due, but to a limited extent, to reduced contributions at younger ages. Infant mortality contributed 5 per cent in 1971 but only 2 per cent in 1995. The reduced sex differential in MVA and other accidents as rates declined over time is seen in the smaller contribution at ages 15 to 24 in 1995, while the effect of AIDS is seen in the larger 1995 contribution at ages 25 to 39. As male suicides have increased, the contribution of this cause of death has intensified at ages 15 to 49. While male mortality is higher than female mortality at every age and for each cause, this is not universally the case where age-causespecific rates are concerned. Notable exceptions occur for cancer at ages 30 to 49, owing to the relatively young age at which breast cancer rates increase, and for stroke at the same ages in 1971 and 1976. Smaller negative contributions have also occurred at various times for heart disease in childhood and at various ages for COPD, diabetes, nephritis and ‘other’ causes.

I N T E R N AT I O N A L C O M PA R I S O N S Australian mortality is among the lowest in the world. Both female and male life expectancy ranked sixth in 1995 (United Nations 1998, 1999). Since 1971, Australia’s mortality ranking has improved significantly, from 12th for females and from 14th for males. Table 5.2 shows life expectancies in 1971 and 1995 for the 20 populations with the lowest female mortality in 1995. It is seen that the improved

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 123

Australian rankings are associated with relatively large gains in life expectancy over time. For the period 1971–95 the Australian female gain ranked sixth while the male gain ranked second. Table 5.2 also shows that the relative position of the sex differential in life expectancy has changed in line with the narrowing of this differential over time: Australia ranked seventh in 1971 and 13th in 1995. Standardised death rates (standardised to the World Standard Population, WHO 1995) show a less balanced situation between the sexes. Among 20 comparison countries, those listed in Table 5.2 excluding Austria, Belgium, Finland and Iceland and including Denmark, Ireland, Israel and Singapore, female rates ranked ninth in 1992 and male rates ranked fifth (AIHW 1998). These ranks have improved from 15th and 17th respectively in 1970–74 with gains that ranked third and second largest respectively in absolute size (AIHW 1998). A large part of the improved ranking of Australian mortality can be attributed to improvements in mortality from cardiovascular disease including stroke (ICD 390–459). Between 1970–74 and 1992, Australia experienced the largest absolute decrease in both female and male standardised death rates from this cause among the 20 comparison countries in the AIHW study. However, Australia still compares unfavourably with several other populations with respect to levels of cardiovascular disease, with males ranking ninth and females ranking 11th in 1992. Making similar comparisons for cancer (ICD 140–208), the decline in Australian death rates ranked fifth for males and tenth for females. Though improvements in mortality from cancer are small, it is worthy of mention that several countries experienced a worsening of mortality from this cause. The overall level of female mortality from cancer was tenth lowest while that from breast cancer, the most common cause of cancer death among females, ranked eighth. For males the overall level ranked sixth lowest, but the level of deaths due to prostate cancer, the most common cause among male cancer deaths, ranked 14th (AIHW 1998). A sizable portion of the mortality toll from cardiovascular disease and cancer can be attributed to the smoking of tobacco. Smoking is estimated to account for 17 per cent of all male and 9 per cent of all female deaths in 1996, with the largest single cause being lung cancer followed by COPD (AIHW 1999b, p. 107). While the number of male smoking-attributed deaths has declined since 1985, for females the number continues to increase (Peto et al. 1994).

124 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 5.2 Life expectancies by sex for low mortality countries, 1971 and 1995 Female 1995 1971 1971–95 1995 Country Rank e(0) Rank e(0) Rank Gain Rank e(0) Japan

1

82.9

8

75.6

1

7.3

2

76.4

France

2

81.9

7

76.1

10

5.8

13

73.9

Switzerland

3

81.9

6

76.2

11

5.7

5

75.7

Hong Kong

4

81.5

10

75.0

3

6.5

4

76.0

Sweden

5

81.5

3

77.3

15

4.3

1

76.5

Australia

6

81.2

12

74.9

6

6.3

6

75.0

Norway

7

81.1

2

77.4

19

3.7

7

75.4

Canada

8

80.9

5

76.4

13

4.5

9

74.6

Italy

9

80.7

13

74.9

9

5.8

11

74.3

Belgium

10

80.6

16

74.2

4

6.4

14

73.9

Iceland

11

80.6

1

77.5

20

3.1

3

76.2

Spain

12

80.5

11

75.0

12

5.5

16

73.4

Netherlands

13

80.4

4

76.7

18

3.7

10

74.6

Greece

14

80.2

20

73.6

2

6.6

8

75.0

Finland

15

80.2

17

74.2

7

6.0

19

72.8

Austria

16

80.1

19

73.7

5

6.4

15

73.5

Germany

17

79.7

18

73.8

8

5.9

18

73.3

UK

18

79.3

9

75.2

17

4.1

12

74.1

New Zealand

19

79.1

15

74.6

14

4.5

17

73.4

US

20

78.9

14

74.8

16

4.1

20

72.5

Australia 1970–72 and 1994–96 from this analysis; Belgium 1970 and 1994; Canada 1992; France 1968; Germany 1994–96, West Germany only in 1971; Greece 1970; Iceland 1973 and 1995-96; Italy 1994; Netherlands 1994–95; New Zealand 1992–94; Norway 1996; Spain 1990–91; Sweden 1996; Switzerland 1995–96; UK 1972. All ranks are from the highest (1) to the lowest (20). SOURCE United Nations 1998, 1999. NOTES

Smoking patterns in Australia are not as heavy as in most of the 20 countries included in the AIHW study. In 1995, 27 per cent of Australian males aged 16+ reported that they smoked regularly, ranking third from the lowest in surveys undertaken in 1990–95 (AIHW 1998). The corresponding percentage for females is 23, ranking sixth. Apparent consumption of tobacco is also relatively low: in

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

Male 1971

1971–95 Rank Gain

• 125

Sex differential 1995 1971 Rank F–M Rank F–M

Rank

e(0)

5

70.2

5

6.2

6

6.5

17

5.4

13

68.5

9

5.4

1

8.0

3

7.6

6

70.2

8

5.6

11

6.2

13

6.0

18

67.4

1

8.6

16

5.5

2

7.7

1

72.0

17

4.5

19

5.0

19

5.3

14

68.2

2

7.2

13

5.8

7

6.7

3

71.2

18

4.2

14

5.7

11

6.2

9

69.3

11

5.3

10

6.3

6

7.0

11

69.0

10

5.3

7

6.4

14

5.9

15

67.8

6

6.1

4

6.7

9

6.4

2

71.6

16

4.6

20

4.4

15

5.9

8

69.7

19

3.7

3

7.1

18

5.3

4

71.0

20

3.6

12

5.8

16

5.7

7

70.1

14

4.9

17

5.2

20

3.5

20

65.9

4

6.9

2

7.4

1

8.3

19

66.6

3

6.9

5

6.6

5

7.2

16

67.4

7

5.9

8

6.4

8

6.4

10

69.0

12

5.1

18

5.2

10

6.2

12

68.6

15

4.9

15

5.7

12

6.1

17

67.4

13

5.1

9

6.4

4

7.4

1995–96 per capita consumption of persons aged 15+ was 1480g, ranking third from the lowest. The level of consumption has decreased steadily from 3500g per capita in 1960–61, with the rate of decrease being more rapid among males than among females, though significant increases had occurred after World War II. These relatively low levels of smoking are reflected in death rates from lung

126 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

cancer (ICD 162), though there is a lengthy time lag involved. In 1992, males ranked fifth from the lowest and had experienced the second largest decline since 1970–74, while females ranked 12th after experiencing the sixth largest increase in rates reflecting earlier increases in tobacco consumption. Death rates from bronchitis, emphysema and asthma (ICD 490–493) ranked 11th for males and 13th for females in 1992, following declines ranking fifth and tenth respectively (AIHW 1998). The increase in suicide death rates among young Australian males is a phenomenon shared with many other developed countries leading to increased overall rates. In 1992, the male death rate from suicide ranked fifth highest with an increase since 1970–74 that ranked seventh. Female suicide death rates in 1992 were relatively low, ranking 13th, and had undergone the fourth largest decline since 1970–74 (AIHW 1998).

F U T U R E D I R E C T I O N S I N M O R TA L I T Y Making detailed projections of future mortality by sex, age and cause is extremely complex because it depends on many factors. These include medical advances and the provision of healthcare, as well as lifestyle factors such as nutrition, exercise, occupation, living standards, smoking habits and other activities that affect health. Changes in many of these factors are fairly predictable but this is not always the case. Advances in medical science, perhaps the principal factor involved, are now occurring apace and on many fronts. It is impossible to predict with any certainty where important scientific advances will be made and how accessible such advances might be for the treatment of disease. Other factors such as epidemics, natural disasters and the outbreak of war are by their nature even less predictable and almost impossible to take into account. The emergence of HIV/AIDS as a leading cause of death in certain age-sex groups was entirely unforeseen two decades ago. The most common approach to mortality projection is to base future expectations of age-sex-specific death rates on past trends. Such projections are made on a regular basis by both the Australian Bureau of Statistics (ABS) and the Australian Government Actuary (AGA). Both examine the rates at which age-sex-specific death rates have improved in the past and make projections, with varying degrees of subjectivity, of future rates of improvements. These projections differ in that the ABS improvement rates are projected by

T H E C H A N G I N G D I M E N S I O N S O F M O R TA L I T Y

• 127

cause of death whereas the AGA projections are based on death rates for all causes. The recent projection from ABS (1998d) covering the period 1997 to 2051 is based on the experience of the previous 30 years. Specifically, projected improvement rates by age, sex and cause of death for July 1997 to June 2006 are assumed equal to observed rates of improvement between the two five-year periods 1987–91 and 1992–96. Projected rates for the period from July 2006 are assumed equal to the average improvement rates experienced in successive five-year periods from 1967–71 to 1992–96. Since recent improvements in mortality have been relatively rapid, this results in a brief continuation of rapid improvement followed by more gradual change. By the end of the projection period, 2051, projected life expectancy is 86.1 years for females and 82.0 years for males. The AGA (1999) produces two projections of mortality from 1996: one based on experience over the last 25 years (referred to here as AGA–25) and the other on experience over the last 100 years (AGA–100). In each case, constant rates of improvement by age and sex are used over the entire projection period. Since average improvements in mortality have been greater during the last 25 years than during the last 100 years, the former produces a higher life expectancy than the latter. By 2050, the 25-year improvement rates produce life expectancies of 90.4 years for females and 86.4 years for males while the 100-year improvement rates produce life expectancies of 86.9 years for females and 81.1 years for males. These life expectancies imply a sex differential of 4.1 years in 2051 if ABS improvement rates are used, and of 4.0 and 5.8 years in 2050 if AGA 25-year and 100-year improvement rates are used. The sex differential in 1995 was 5.8 years. The shorter-term improvement rates thus suggest a continued narrowing of the sex differential. This narrowing is more rapid in the earlier than later years of the ABS projection while in the AGA projection a fairly constant change occurs. The longer-term improvement rates suggest no change in the sex differential over the entire period. Given the fairly rapid narrowing of the sex differential in recent years, a constant differential seems unlikely in the near future. These projections may be compared with new forecasts of allcause mortality made using recently developed methodology (Lee and Carter 1992; Booth, Maindonald and Smith 2002). Based on data for 1968–2000, forecast life expectancy in 2027 is 88.1 years for

128 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

females and 82.9 for males (Booth and Tickle n.d.). These are similar to the AGA-25 projections for 2030 of 87.5 years for females and 83.0 for males (AGA 1999), though they imply a less rapid narrowing of the sex differential. It is noted that, despite referring to a much earlier date, the new forecasts exceed the above ABS and AGA-100 projections for the mid-21st century. Whatever method of projection is used, it has been shown above that it is increasingly the case that the largest absolute improvements in death rates and hence the greatest contributions to life expectancy occur at older ages. At the same time, infant mortality which has contributed significantly to gains in life expectancy in the past will contribute very little to gains in life expectancy in the future since rates are already levelling as they approach zero. Mortality at ages 1 to 39 is also at such low levels that future improvements will be minimal. Thus the main area of interest in mortality in future years will be concentrated on the old and increasingly the very old. It is reductions in mortality at these ages that will determine the length of life in years to come.

6

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M BOB BIRRELL

This chapter provides both a retrospective and a prospective analysis of changes in Australia’s immigration policies and their demographic implications. The prospective part presents unique problems for a discussion about immigration because of the variety of push and pull factors which shape immigration movements. To make sense of what has happened over the past couple of decades and what might happen in those ahead requires a theoretically grounded explanatory framework.

T H E O R I E S O F M I G R AT I O N The approach discussed here reflects the work of Freeman (1995), whose analysis begins with the way the costs and benefits of immigration are distributed in liberal democratic societies. The normal pattern is that the benefits tend to be concentrated with employers interested in a lower cost and more flexible labour market, businesses whose sales are linked to an expanding domestic market and migrant groups who want more open entry rules for their kinsfolk or fellow countrymen. On the other hand, the costs, which may include the exacerbation of environmental problems and increased property prices where immigration adds to competition for housing in high amenity areas, tend to be diffused widely among the voting public. Freeman argues that in liberal democracies politicians tend to respond to pressures from well-funded and determined lobby groups, but do not take much notice of voter concerns if they are

130 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

not focused politically. Therefore, there is a bias towards migration levels higher than most voters would approve; this seems to describe the situation in Australia since 1980. All opinion polls on migration since that time have shown majorities in favour of lower migration, whatever the level of migration at the time of the poll (Betts 1999, p.114). These tendencies seem strong in settler societies like Canada, the United States and Australia where there is some commitment to ‘melting pot’ or ‘multicultural’ ideals. It is widely held that cultural concerns deriving from diverse migrant flows are likely to be most acute in Europe (and much of Asia) where cultural identity is more ‘primordial’ (Money 1999, pp. 27–30). The argument is that societies with an identity revolving around a distinctive language and heritage are more likely to experience ethnically diverse migrant streams as a challenge than those like Australia, with more open conceptions of identity. Another important thread in recent scholarly research concerns the capacity of governments to restrict immigration movements even if they want to do so. This capacity may be problematic in a context where the rights of residents and non-residents alike have expanded in relation to state agencies and where non-residents can call on international human rights treaties to legitimise their claims to stay on. As Hollifield has argued, liberal democracies have to deal with migratory movements in a setting where there is an ‘ongoing extension of rights to individuals who are not full members of the societies in which they reside’ (Hollifield 1992, p. 216). Some analysts take this point further to argue that in an increasingly globalised world, where the international movement of people is an inevitable concomitant of expanding cross-national organisations, the nationstate’s capacity to control these movements is declining. Castles and his colleagues assert that ‘there is a certain unstoppable momentum’ to high migration in Australia. Writing at the end of the 1980s, a decade of high migration, they thought there would be no radical change to the migration program ‘regardless of negative publicity on the levels of immigration or the economic arguments about the ideal composition of the programme’ (Castles et al. 1992, p. 169). At the time it appeared they might be right. In both Australia and Canada, high migration had become linked with broader social justice campaigns in which ethnic and humanitarian advocates had joined with green, feminist and other minority groups to

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 131

advance their cause. The solidarity amongst these groups, which was coupled with broad intelligentsia support, provided both ideological justification for migration and, from the perspective of politicians, the basis for a supporting ‘rainbow’ alliance of voters. In Australia’s case, Castles et al.’s prediction turned out to be incorrect. During the 1990s the Australian government cut back the immigration program, in part by enforcing tough control measures inconsistent with the theories outlined above. Freeman (1995) accommodates this response by arguing that there is a tendency for liberal democracies to sweep evidence of the social and cultural costs of migration ‘under the carpet’. However because these problems tend to mount up, they eventually provoke protest movements. The resulting ‘immigration cycles’, Freeman argues, are inherent in the normal bias of liberal democratic societies towards expansive immigration (Freeman 1995, p. 886). With the advantage of hindsight it is easy to find evidence of this process beginning in the 1980s in Australia (Freeman and Birrell 2001). The hypothesis explored below is that the high intakes of the 1980s brought in their wake a fundamental change in the ethnic and skill characteristics of the migration flow. The change from a predominantly European to a predominantly Asian program exacerbated the cultural tensions associated with migration. But perhaps just as important in shaping the restrictive policies of the 1990s were the increased welfare needs of the newer flow of migrants. This is a thorny issue for rich societies, especially those with a developed welfare state, because since the 1980s the migrant flow to Australia has been from relatively poor countries. This means that Australians have been in effect sharing their accumulated wealth with newcomers, who for the most part pay no entry price. Where many recent arrivals need sustained welfare, educational or medical assistance, as proved to be the case in Australia by the end of the 1980s, this threatens the economic legitimacy of continued high migration.

T H E I M M I G R AT I O N D E PA R T M E N T AND CONTROL ISSUES In the narrative that follows the main theme is the management of immigration flows, as successive Australian governments have sought control over who gains permanent residence in Australia. The actions of the Department of Immigration are inevitably at the centre of the

132 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

analysis. Australia is distinct amongst Western societies in the influence this bureaucratic agency has had over immigration policy. The department played a central role in organising and legitimating the huge post-war immigration intake. Though it is answerable to the government, its long experience, control of information on the issue and location of its minister in Cabinet has given it considerable independence in the development of policy. There is no parallel to this situation in Europe or the US. The Immigration Department has an obvious interest in the maintenance of a significant migration program. But its role near the heart of the political process makes its officers sensitive to the task of maintaining public acceptance of the program. Its officers are aware of the need to allay any concerns that the program may be ‘getting out of control’. The memoirs of John Menadue, who headed the Department from 1981 to early 1983, provide a telling example. Menadue was an enthusiastic public advocate for a high immigration. But as he recalled later, both he and the minister at the time, Ian Macphee, kept a close eye on public opinion: We both strongly believed that an important reason why Australians supported immigration was confidence that the Australian Government controlled the program. If the Australian community, then or now, believed that there was no real control at our borders, that people entered and stayed illegally, there would be a serious loss of confidence in the program (Menadue 1999, p. 217).

THE LEGACY OF AUSTRALIAN I M M I G R AT I O N B Y T H E 1 9 7 0 S The 1970s saw the end of the era of sustained high migration to Australia. During the 1950s and 1960s population growth was seen as fundamental to Australia’s national security and economic fortunes. This was also an era of European migration. Australia’s Immigration Department was set the task of attaining high annual population building targets; the department’s first priority in meeting these targets throughout this period was British settlers. They were the dominant source, and even in the 1960s when economic conditions in Europe improved, some 43 per cent of all settlers arriving in Australia came from the UK (Bureau of Immigration Research 1991). But as interest in migration to Australia from

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 133

Britain diminished, the department turned first to Western Europe (especially the Netherlands and Germany) to Displaced Persons from Eastern Europe, then increasingly to Southern Europe (including the former Yugoslavia). This meant that it had to accept settlers whose English language and occupational skills were limited. But until the end of the 1960s when new sources had to be found, including Turkey and Egypt, the intake was overwhelmingly European. Cultural tensions were kept in check because of the European background of most migrants and because this was a period in which the ethos of assimilation predominated. Public concerns about loss of the British focus were deflected by insistence on migrants becoming ‘new Australians’. Similarly, the social costs of incorporating lowerskilled migrants were manageable because the expansion of manufacturing and construction industries during the 1950s and 1960s meant there was a buoyant demand for unskilled blue-collar workers. Nevertheless, the variety of European source countries during this era left a legacy of community diversity (discussed in the next chapter) which was to contribute to subsequent controversy about identity issues in the 1980s and later.

I M M I G R AT I O N PAT T E R N S S I N C E 1 9 7 0 The Whitlam Labor Government (1972–75) was the first since World War II to challenge the maintenance of large-scale migration programs. Labor cut the migration program from 170 000–180 000 in the late 1960s to just 50 000 in 1974–75. This was largely because of the onset of the 1974 recession; but in addition there were concerns about the capacity of Australia’s cities to accommodate population growth on the scale of the 1960s, and about the need to focus the nation’s scarce capital on internationally competitive industries rather than promoting growth ‘on all fronts’. When a new Liberal-National Party Coalition Government came to power in 1975 it sought to revive the immigration program. It did so in the late 1970s in the context of a mineral boom and the changes in economic thinking just noted. The policy of building up manufacturing industries tied to a protected domestic market was under attack. The increasingly influential idea was that Australia should build on its potential strengths in the global market place, which at that time were seen as linked to export of commodities, particularly those like alumina and aluminium which embodied Australia’s huge

134 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

reserves of energy, as with coal and uranium. Immigration was seen to have a role in this process, but less through its contribution to the expansion of the domestic market. Rather, the emphasis was to be on the skills migrants could contribute to making the new industries more internationally competitive. The priority was on skilled tradesmen, technicians and professionals. A new and complex system of selecting migrants on account of their skills was instituted in 1979, which reflected these priorities. Called the Numerical Multi-Factor Assessment Scheme (NUMAS) it attempted to assess migrants on criteria which would recruit those possessing ‘qualities of initiative and adaptability coupled with excellent work records and sound financial backgrounds who would be an asset to any community’ (Department of Immigration 1978, p. 2). The more stringent controls over selection can be read as the first serious attempt to limit the entry of low-skilled workers. In the early post-war period such migrants were welcomed, and indeed according to some immigration analysts they were the major priority. The notion was that capitalist societies needed an infusion of workers at the lowskilled end of the workforce (or an industrial ‘reserve army’) in order to get the ‘dirty’ jobs done cheaply (Collins 1975). However, there is no evidence to support this argument since the 1970s. The selection criteria introduced in the late 1970s were poles apart from the ‘reserve army’ perspective. The Minister for Immigration at this time, Ian Macphee, and his departmental head John Menadue were both reformers keen to make Australia a more outward looking economy and society. Immigration was seen as playing a role in this transformation. Menadue said publicly at the time (August 1982) that ‘An immigration program is the only tool readily at hand to challenge complacency and parochialism’ (Menadue 1999, pp. 218). As it turned out, this focus on skilled immigration came to an abrupt halt with the deep recession of the early 1980s. It was revived again in the mid-1980s along lines consistent with the Menadue vision. In the meantime, new developments in family reunion policy contributed to growth in the numbers of relatively low-skilled migrants. These developments were not a consequence of Immigration Department advocacy (quite the contrary) but of the changing political environment by the end of the 1970s. By this time the Fraser Government was in open competition with the Labor Party for the political support of Australia’s ethnic communities. Partly as a

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 135

consequence the government embraced the notion of Australia as a multicultural community. By the late 1970s the Greek and Italian communities in particular were forcefully articulating their grievances about the disdain they felt the majority Anglo community directed at their communities. Multiculturalism addressed these feelings with its emphasis on the ideal that migrants from all cultures and countries of origin should receive equal respect in Australia. This setting created a favourable political environment for ethnic communities to lobby for the liberalisation of family reunion rules.

T H E DY N A M I C S O F F A M I LY R E U N I O N A brief account of the US experience with the liberalisation of family reunion rules will help set the context for analysis of the Australian case. In 1965 the United States Congress reformed its immigration regulations so as to make family reunion the dominant selection criterion. At the same time the preceding country-of-origin restrictions on non-Europeans were largely abolished. The result was a rapid growth in numbers, particularly from Asian and Latin American countries, despite the small numbers of people from most of these countries actually resident in the US at the time. The number of ‘immediate relatives of US citizens’, (spouses, children and parents), for which there was no numerical limitation, grew from 60 010 in 1969 to 223 468 in 1986 and 322 440 in 1997 (US Immigration and Naturalization Service 1997, p. 19). This growth occurred despite the built-in delay factor of US citizenship, which requires a five-year minimum residency. It was an outcome of the great interest in migration from Asian and Latin American countries, the pull of opportunity in America, and the willingness of fellow countrymen resident in the US to sponsor relatives, particularly spouses, from the former homeland. The result of this and other family category movements, which dominate the US program, is that the skill level of migrants has diminished significantly since the 1960s (Borjas 1999, pp. 105–26).

THE AUSTRALIAN EXPERIENCE WITH F A M I LY R E U N I O N S I N C E T H E 1 9 7 0 S In 1973, the Whitlam Labor Government established three family reunion categories; one for immediate dependent family members (spouses, dependent children and dependent parents), a second for close non-dependent relatives (including non-dependent parents and

136 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

siblings) and a third for more distant relatives. The original intention was to allow relatively unrestricted entry to the first two categories and qualified entry to the third. However, when economic conditions deteriorated in the mid-1970s, the government restricted the intake to the first category. But before this happened there was a brief window of opportunity during which Australian residents could sponsor relatives in all three categories, with some expectation that the nomination would succeed. The results, as detailed in a 1977 Green Paper prepared by the Fraser Government’s advisory body on immigration, were startling. There were 99 000 personal nominations for the year 1973–74, when it seemed that sponsorships would be accepted in all three categories. The majority was for relatives living in Europe. But sponsorship rates, expressed as nominations per 1000 residents, from the small Asian and Middle Eastern communities then resident in Australia, proved to be very high. They reached 354 per 1000 Filipino residents, 402 for Turkish-born residents and 360 for the Lebanese (Australian Population and Immigration Council 1977, p. 35). The council commented: Those national groups where extended family relationships tend to be stronger also tend to have lower levels of occupational skill, lower incidence of fluency in the English language and larger family sizes. The implications of this are that in a policy emphasising family reunion, the immigration program would very quickly be dominated by such countries and its composition would be predominantly non-English-speaking with low levels of occupational skill (Australian Population and Immigration Council 1977, pp. 36–37). Despite these concerns, the Fraser Government relaxed the tough stand it had inherited by restoring some sponsorship rights for relatives in the second family category noted above (Birrell 1990, pp. 2–3). The Hawke Government which took office in 1983 introduced further concessions for those sponsored by Australian-resident siblings by minimising the skill requirements of those sponsored, and allowed virtually unrestricted entry to those sponsored as parents, regardless of their age or number of children living in Australia. Meanwhile, in the aftermath of the deep recession of the early 1980s, the Hawke Government restricted skilled migration. As a result, by

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 137

the first half of the 1980s the Australian migration program looked quite like that of the US at the time, with its emphasis on family reunion. During the 1980s there was a rapid growth in the family reunion intake. Table 6.1 shows the pattern since 1985–86. Comparable earlier numbers were not available. The peak year was 1987–88, but by 1989–90 the number of visas issued for family members was nearly double that of the early 1980s. All the major categories grew sharply. In each case most of the growth was attributable to sponsorships from the numerically small non-European source communities. In the case of parents, by 1989–90 the largest source country was Vietnam. For the ‘Concessional’ category the largest source country was the UK followed by Malaysia, the Philippines, Hong Kong and India (Bureau of Immigration Research 1991). Table 6.1 Migration program visas granted by category, 1985–86 to 1989–90 Migration category/ Component a Family Preferential Spouses/fiancées Parents Dependent children Other Preferential Concessional Total Family Skill Employer nominations Business migration Special talents Independent Total skill Special eligibility Total non-humanitarian program Humanitarian Total program NOTE a SOURCE

85–86

86–87

87–88

88–89

89–90

33 14 8 1 9 29 63

38 15 11 1 10 34 72

40 16 10 1 12 38 79

46 18 11 2 13 26 72

44 19 9 1 13 22 66

900 300 900 300 400 500 400

200 500 200 300 200 400 600

600 000 700 500 400 900 500

200 700 600 000 900 500 700

000 200 900 900 000 600 600

6 900 1 800 100 6 300 16 200 400 80 000

8 400 3 500 100 15 200 28 500 600 101 700

8 300 7 300 100 23 900 42 000 600 122 100

9 200 10 200 100 28 900 51 200 800 124 700

11 900 10 600 200 30 000 52 700 900 120 200

11 700 91 700

11 300 113 000

11 400 133 500

11 300 136 000

12 400 132 600

Includes estimate for people granted resident status after arrival in Australia. Department of Immigration and Ethnic Affairs 1994.

138 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The growth in sponsorships for siblings prompted the Labor Government to tighten the skill requirements for those sponsored (hence the fall in visa numbers for this category shown in Table 6.1 after 1987–88). This it could do without too much political pain. But the control of other family reunion privileges proved to be more difficult. Even the most restrictive of European societies are reluctant to proscribe the rights of residents to sponsor a spouse. Spouse migration is therefore a crucial component of family reunion since it has an assured or ‘as of right’ entry point. It also serves to introduce potential new family chains through the spouse to parents and siblings. Thus a crucial determinant of the extent of family reunion is the propensity of migrant communities, and other Australians, to look overseas for spouses. Research on this issue suggests spouse sponsorship patterns are linked to the degree of migrant social integration in the host society. If an ethnic community maintains a sense of solidarity amongst its members and of social distance from other Australian residents, whether because the community itself encourages separation or because of host society prejudice, then marital endogamy is likely. In such circumstances there may be strong social pressures from parents and friends on an unmarried community member to select a spouse from ‘home’. The benefits of doing so may include a wide choice of willing partners, given the interest in migration from poor or politically unstable countries like Lebanon, the Philippines, China or Vietnam, and sometimes a dowry. Coleman has shown how this process has worked to sustain a substantial flow of spouses from the Indian subcontinent to the UK (Coleman 1995). By the end of the 1980s spouse and fiancée migration had increased substantially, with most of the growth being from nonEuropean source countries. The largest country-of-origin group was Vietnam followed by the UK, the Philippines, Lebanon and a number of other Asian countries. Most of the sponsors for UK partners were not born in the UK or the children of UK-born migrants. That is, they were not part of a chain connecting migrants with permanent-residence status, or their children born in Australia, back to the homeland in search for a spouse through family or locality connections. The number of UK spouse sponsorships has remained high because they reflect the extent of travel between the UK and Australia (and the ‘boy meets girl’ consequences). Most of the Asian and Middle Eastern spouse or fiancée sponsor-

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 139

ships derive from a different source. Some involve the reunion of spouses who have been split by migration, as with some refugees. Another component derives from the return of recently arrived migrants to find a partner. But by the late 1980s and early 1990s there was evidence of more long-lasting chains linking Australian sponsors resident since their early years, or born in Australia of parents from countries such as Lebanon and Turkey, back to partners living in the homeland. Analysis of their marriage and sponsorship patterns in the 1980s and 1990s showed that some of the second generation from such communities were returning to the home country to find a spouse or fiancée (Birrell 1995; Khoo 2001). This experience helps explain why the spouse-fiancée flow grew so strongly during the 1980s and has continued to do so into the 1990s. On the other hand, by the end of the 1980s, spouse chains back to Europe had largely attenuated for the Southern European communities who arrived in the early post-World War II era, despite a relatively strong pattern of marital endogamy amongst second-generation Italians in Australia. The difference from the Lebanese and similar communities was that where second-generation Italians or Greeks chose partners of the same country-of-origin background they did so amongst persons who, like themselves, were raised in Australia.

T H E R E V I VA L O F S K I L L E D M I G R AT I O N DURING THE 1980S In the wake of the changing source country patterns just described came the first serious debate about what they implied for Australia’s ethnic make-up and identity. Geoffrey Blainey brought these issues into focus in 1984 with his claim that the scale of Asian migration exceeded the level acceptable to the general public. The Hawke Government responded to Blainey’s challenge by seeking a new basis to legitimise the migration program. It did this by highlighting the contribution skilled and entrepreneurial migrants could make to Australia’s economic vibrancy. The argument, which echoed that of Menadue and MacPhee a decade earlier, was that if Australia was to compete as a ‘clever country’ in Asia it needed both to enhance Australia’s professional and technical skills base, and to improve trading links to the region. Skilled migrants were seen as contributors on both these fronts, as were business migrants. Opportunities for business migrants were expanded in the second

140 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

half of the 1980s, with the majority entering under this category coming from Hong Kong. In addition, a new ‘Independent’ category was established in 1986 which gave priority to young applicants with work experience, English language skills and tertiary education. The response was very strong, with tens of thousands of applications per year by the end of the 1980s, mainly from graduates of Asian universities. The Independent program tapped into the rapidly expanding ranks of such graduates, many of whom faced a struggle to find well-paid jobs within their professional field. Immigration to an affluent Western country seemed to provide a solution. With Australia, Canada and the US the only locations with the door ajar, Australia proved to be a popular destination. The scale of the resulting migrant contribution to Australia’s professional workforce is indicated in Table 6.2. By 1991 some 31 per cent of all persons with degree qualifications in Australia were born overseas. In medicine, computer science, electrical engineering and mechanical engineering the figure was considerably higher at 40, 43, 48 and 49 per cent respectively. Just over half of these professionals arrived in Australia before 1980, mainly from Britain. But a remarkable proportion arrived during the 1980s, including over 20 per cent of all persons with degrees in engineering who were resident in Australia in 1991. Most came under the Independent category, though a significant minority entered under the Concessional and other family categories.

T H E I M M I G R AT I O N P O L I C Y S E T T I N G AT T H E E N D O F T H E 1 9 8 0 S By the end of the 1980s Castles et al.’s view that high migration was ‘unstoppable’ looked plausible. The Labor Government at the time appeared intent on maintaining the program. It had embraced migration as a positive element in its objective of incorporating Australia into the booming Asian marketplace. There seemed widespread support for this policy within elite political, economic and cultural circles. There was controversy about the social and cultural costs of high immigration, which manifested in substantial majorities in opinion polls at the time in favour of reduced migration. But there seemed little scope for voters to express their concerns, given that during the 1980s all the major political parties supported the main lines of the government’s immigration agenda.

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 141

Table 6.2 Degree-qualified workforce by field, birthplace and time of arrival in Australia for the overseas-born, 1991 (%)

Field of qualification

Birthplace overseas

Australia

Not stated

Up to 1980

1981–1985

1986–1991

Not stated

Total overseas

Total

Number

Year of arrival

Accounting

69.3

0.1

16.2

4.6

9.6

0.2

30.6

100.0

77 365

Medicine

59.4

0.3

26.6

3.3

10.0

0.4

40.3

100.0

46 924

Nursing

74.1

0.2

16.5

2.9

6.0

0.3

25.7

100.0

46 883

Dental science

64.9

0.1

23.7

3.3

7.9

0.2

35.0

100.0

8 467

Law

76.2

0.2

15.0

2.5

5.9

0.3

23.7

100.0

39 321

Computer science

56.5

0.2

19.3

8.2

15.3

0.3

43.2

100.0

23 684

Civil engineering

61.3

0.1

18.1

5.2

15.1

0.3

38.6

100.0

16 917

Electrical & electronic engineering

51.6

0.2

19.9

7.1

20.8

0.4

48.2

100.0

19 972

Mechanical engineering

50.6

0.2

21.6

7.7

19.5

0.3

49.2

100.0

12 271

Other engineering

57.6

0.2

22.5

5.6

13.7

0.4

42.2

100.0

21 750

Building design

66.8

0.1

19.9

4.2

8.8

0.2

33.1

100.0

12 710

Visual & performing arts

68.9

0.3

17.8

4.1

8.6

0.5

30.9

100.0

23 191

All other 69.8 qualifications

0.2

17.8

3.4

8.4

0.3

30.0

100.0 647 112

Total

0.2

18.2

3.8

9.2

0.3

31.5

100.0 996 567

SOURCE

68.3

Birrell and Hawthorne 1997.

142 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

There was also evidence consistent with Hollifield’s theory about the difficulties of controlling immigration in a context of increased attention to human rights. During the 1980s the government faced an increased number of claims for permanent residency from visitors and other temporary entrants, including some who had overstayed their visas. The Immigration Department experienced great difficulty establishing the bona fides of many of these claims and, especially in the case of those based on de facto marriage, in enforcing decisions through the courts when the application was rejected (Birrell 1992, pp. 31–32). This apparent loss of control was in part due to the granting of legal standing in Australia’s courts to persons who were not permanent residents. This meant that such persons could press their case in the courts against any decisions on the part of the Australian Government to exclude them. This reflected worldwide trends in legal opinion, but in Australia’s case was legitimated in part by the Australian Government’s signing of the International Covenant on Civil and Political Rights. Applicants also benefited from parallel changes in administrative law at the time which extended the rights of citizens in dealing with state agencies. Another major challenge to the Australian Government’s control capacity came from the 27 000 Chinese students who came to Australia before the Tiananmen massacre in Beijing in June 1989. These students were initially issued short-term (usually six months) visas intended for the study of English. Few had left Australia by mid1989. Many had overstayed their visas and thus were illegal residents. Almost all indicated their wish to stay in Australia and were granted temporary residence while the government pondered the situation. In mid-1989 another 20 000 student visa applications from China were being considered. Most of these applications were granted and when their visas lapsed, most of these students, too, remained in Australia, some applying for refugee status, others staying on illegally. At about the same time another challenge to government management of immigration movements arose over the arrival of some 652 unauthorised boat people between November 1989 and October 1991, mostly from Cambodia. They too sought asylum status as refugees.

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 143

A U S T R A L I A N I M M I G R AT I O N P O L I C I E S IN THE 1990S The 1990s saw a sharp turnabout in successive Labor and Coalition government immigration policies. Since the late 1980s there has been a contraction in program numbers, brought about by tighter management of the immigration selection process and of claims to permanent residence made from within Australia. By the end of the 20th century Australia looked more like western Europe in the strictness of its control regime than its settler-society counterparts, the US and Canada. The origin of this turnabout is complex and continues to be the subject of scholarly controversy. There is space for only a brief analysis here. Put in the terms of Freeman’s theory, what happened was that both major political parties concluded that public concerns about the social and economic tensions resulting from the high migration intakes of the 1980s were becoming electorally hazardous. These concerns included the apparent ‘loss of control’ of change-ofstatus claims and issues linked to the changing ethnic make-up of the migrant intake, particularly its increasing Asian component. A major catalyst to this political response derived from the report of a commission of inquiry into Australia’s immigration policies in 1988, which the Labor Government itself appointed. The Committee to Advise on Australian’s Immigration Policies (CAAIP) concluded that poor management of Australia’s immigration policies had resulted in ‘Widespread mistrust and failing consensus [which] threaten community support of immigration. The program is not identified in the public mind with the national interest, and must be given a convincing rationale’ (CAAIP 1988, p. xi). It also concluded that part of this mistrust stemmed from the perception that immigration was serving the interests of migrant communities rather than the national interest, and that multicultural policies were driving immigration policies (CAAIP 1988, p. xii). The debate following this report drew in the then Opposition leader John Howard, whose subsequent critical remarks about the level of Asian migration further raised the temperature. Howard was roundly condemned for his intervention; it appears to have been an important factor in his temporary loss of his party’s leadership in 1989. Nevertheless, the public interest aroused by the debate indicated the potential for the immigration issue to become a divisive electoral issue and thus of the need for tighter management and control. This conclusion was confirmed in the early 1990s when the

144 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

new Liberal leader John Hewson breached the bipartisanism of the 1980s by challenging Labor’s migration and multicultural policies. He was tempted to take this stand in part by evidence collected about public opinion on the issue. In an internal Liberal Party review of this evidence it was concluded that worries about migration and multicultural issues were widespread (Liberal Party 1991, pp. 56–57). During the 1990s immigration became a mainstream political issue. It was no longer the province of contending interest groups battling for influence within bureaucratic circles. In the process the ethnic lobby, which had been very important in shaping policy in the 1980s, lost much of its influence. The emergence of Pauline Hanson as a national figure after her election to Parliament in 1996 and the formation of the One Nation Party in April 1997 ensured that the major parties paid close attention to mainstream voters’ concerns about immigration and multiculturalism. One Nation’s main electoral flag was its restrictive, nationalistic stance on immigration. It was rewarded with 8.4 per cent of first preference votes in the House of Representatives at the 1998 federal election. A final impetus to a retreat from the high migration levels of the late 1980s was the deep recession of the early 1990s and the associated mounting evidence of labour market disadvantage among recently arrived migrants. The recession dramatised the fact that a high proportion of the recently arrived professionals described above were unable to convert their credentials into professional or managerial positions. Those with limited Australian job experience, poor English and backgrounds in countries of which Australian employers had little knowledge tended to go to the back of the queue when job competition intensified (Birrell and Hawthorne 1997, pp. 71–73). Opinions varied as to why employers preferred local graduates or those from Britain or other Commonwealth countries. Some believed the outcome was a result of Australian professional protectionism, as manifested in a reluctance to accept the credentials of persons from non-European source countries. Other commentators were critical of the Australian Government’s skilled-selection system, particularly the absence of any formal testing of English-language capabilities (applicants simply ticked a box to indicate their skill level) in the Independent category selection system; and no assessment of English at all for the Concessional Family category. Criticism from major engineering, medical and other professional

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 145

associations about continued intakes of people surplus to current needs added to the case for tighter control over selection.

T H E M A N A G E M E N T O F I M M I G R AT I O N POLICY IN THE 1990S There was a decisive cutback in the migration program for 1992–1993. As shown in Table 6.3 this decision set the pattern for the rest of the 1990s. Program numbers during the 1990s were maintained well below those of the late 1980s with the exception of the spouse-fiancée category. The task of managing the reduction fell to the Immigration Department. The achievement of the required reductions was a major task given ethnic community leaders’ concerns about maintaining the family and humanitarian programs and the difficulties of enforcing policy changes due to the legal situation described above. It is doubtful whether the task could have been accomplished in the absence of a strong bureaucratic commitment to do the job. As noted earlier officers were well aware that a politically viable program depended on the department itself being the dominant determinant as to who was selected, rather than migrants themselves (through family reunion) or the courts. The new political situation of the 1990s enabled the Immigration Department to implement control measures which officers had long believed were needed, as indicated in the observations from the 1977 Green Paper cited earlier, but could not be implemented under the circumstances of the 1980s. An important administrative measure which was to lay the groundwork for the control measures of the 1990s was established in the late 1980s. In order to reduce the influence of the courts in the implementation of immigration policy the Immigration Department transformed its regulatory structure in 1989 so as to minimise the scope for applicants, their lawyers and the courts to evade the policy’s intent. It did this by as far as possible removing any scope for administrative discretion on the part of decision-making officers or of the courts when interpreting these officers’ decisions by detailed specification of all departmental regulations. These regulations which covered all facets of the department’s operation including humanitarian, family and skilled entry were given the status of law by incorporating them into Acts of Parliament. This administrative revolution removed much of the previous power of the courts to influence policy.

146 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 6.3 Migration program visas granted: 1990–1991 to 2001–2002 Migration category/Component

90–91

92–93

94–95

95–96

Family Spouses/fiancées Parents Dependent children Other

24 10 2 2

500 300 000 000

27 800 5 300 2 700 1.700

26 5 2 3

33 8 2 3

22 500 61 300

7 700 45 300

7 700 44 500

8 000 56 700

7 500 7 000 100 35 100

4 800 3 300 200 13 000

3 300 2 400 100 15 000

4 640 4 900 200 10 600

— — 49 800 12 500 123 500

— — 21 300 13 200 79 700

9 30 14 89

Concessional familya Total family Skill Employer nominations Business skills Distinguished talents Independent Skilled-Australian Linkeda 1 November Onshore Total Skill Humanitarian Total

100 100 500 100

— 600 400 870 770

3 24 16 97

550 890 830 450

— 800 100 750 550

NOTE a From July 1997 the Concessional Family category was replaced by the SkilledAustralian Linked category and transferred from the Family to the Skill stream. On 1 July 1999 it was renamed the Skilled-Australian Sponsored category. SOURCE Department of Immigration and Ethnic Affairs 1999.

ILLEGAL ENTRY AND ASYLUM C ASES In order to reduce the number of on-shore humanitarian claims the government narrowed the circumstances under which such claims were acceptable. With these circumstances written into law, the scope for court action to accept claims outside these circumstances largely disappeared (Birrell 1992, p. 34). The situation for refugee claimants was different because the Australian Government was a signatory to the International Convention on Refugees and therefore bound to consider all claims for asylum according to the criteria specified in the convention. The experience with the Chinese students had already shown how asylum claims could threaten the government’s control of the program. The arrival of unauthorised ‘boat people’ and the subsequent difficulties experienced in managing their claims brought the matter to a head.

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 147

96–97

97–98

98–99

99–00

00–01

01–02

25 7 2 2

25 1 2 2

24 3 2 2

26 1 2 1

28 1 2 1

360 070 120 910

32 740 560 2 160 2 620

130 580 200 330

790 080 190 250

740 120 070 100

330 900 160 600

7 340 44 580

— 31 310

— 32 040

— 32 000

— 33 470

— 38 090

5 560 5 820 190 15 000

5 950 5 360 180 13 270

5 650 6 080 210 13 640

5 390 6 260 110 15 610

7 510 7 360 230 22 380

9 560 7 590 210 29 880

— 980 27 550 13 640 85 810

9 540 370 34 670 13 165 79 155

9 240 180 35 000 12 250 79 260

7 900 60 35 330 11 940 79 270

7 200 60 44 730 13 750 91 950

6 250 20 53 520 12 000 103 610

In response to this situation, the Keating Labor Government, with Opposition support, passed the 1992 Migration Reform Bill. This introduced a mandatory requirement that persons who entered Australia without legal authority (whether by boat or air) were to be detained. If they applied for asylum they were not to be released until their claims had been approved. The purpose was largely to deter prospective unauthorised entrants. In the case of the Chinese students discussed above, their situation was resolved through a series of quasi-amnesties. Such was the number of claimants that to have heard all on an individual basis would have clogged the refugee review and court system. The government has since sought to avoid a repetition of this incident by tightly managing entry of visitors or students from countries like China where there is a strong chance of a subsequent change-ofstatus claim.

148 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The upsurge in asylum seekers arriving by boat since 1999, mainly of persons of Middle Eastern origin, proved that the control system could still be breached. The reaction to these arrivals was to convulse the Australian political scene in the next few years. The numbers arriving by boat was small relative to the hundreds of thousands of unauthorised asylum seekers arriving in Europe each year during the 1990s. However they presented a major symbolic challenge to the Australian Government’s longstanding determination to control the migration flow. This challenge was met by unprecedentedly tough management measures. In 1998–99 there had been 700 unauthorised boat arrivals. However by 1999–2000 their numbers had increased to 4175. In response, in late 1999 the Coalition Government passed legislation designed to deter future arrivals. This introduced ‘safe third country’ rules which stipulated that persons who had passed through countries where they were not in danger of persecution were not eligible to apply for asylum in Australia (Kinslor 2000, p. 58). Furthermore from December 1999 those unauthorised entrants who succeeded in their asylum claims after entering Australia were granted three-year protection visas (rather than permanent residence). These visas made them ineligible to sponsor family members for permanent residence and restricted their access to welfare benefits. Despite this action, a further 4141 boat people arrived in 2000–01. In the month of August 2001 when the Tampa arrived, there were 1212 arrivals. This growth in the scale of the problem created a sense of crisis within the government. It seemed there was to be no end to the flow. Australia was an attractive target for ‘people smugglers’ because 80–90 per cent of arrivals from Iraq and Afghanistan were succeeding in gaining refugee status. It was also proving impossible to return most of the minority whose asylum claims were rejected because the countries of origin would not accept their repatriation. Thus there was a build-up of desperate people in detention prepared to take aggressive action to call attention to their plight. Concern about the situation was growing within Australia from all sides of the political spectrum. Some were appalled with the detention regime; others about the apparent inability of the government to stop the flow of arrivals. The arrival of the Tampa with 433 asylum seekers on board at Christmas Island in August 2001 brought events to a head. The government responded by refusing to allow the asylum seekers to land in

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 149

Australia. Legislation was rushed through the Parliament which declared that Australia’s offshore islands were not part of Australia’s migration zone and that the Australian courts were prohibited from any jurisdiction over the asylum claims of the people involved. The Tampa asylum seekers along with those from several other boats were shipped to Nauru and other Pacific destinations where their claims were assessed. In effect Australia had unilaterally limited its obligations under the United Nations refugee convention by refusing to allow unauthorised asylum seekers to present themselves in Australia so that their claims could be heard. Any doubt about the stringency of this response was removed after August 2001 when Australian naval vessels began implementing a policy of interdicting ships carrying asylum seekers. These policies appear to have had the intended consequences since no more boat people arrived over the ensuing year. These policies were welcomed by most Australians (Newspoll 2002) and appeared to have contributed to the Coalition Government’s electoral success in the November 2001 federal election. They illustrate the paradox of Australia’s continued commitment to immigration at the same time of its determination to control the migration selection process. Through all this tumult the overall humanitarian intake has not varied over the 1990s. It has remained at about 12 000 per year.

F A M I LY R E U N I O N Family reunion presented a particular challenge to successive governments during the 1990s because of the growth impetus deriving from the chain migration process and the political difficulties of constraining programs with strong support within the ethnic communities. On the other hand, the budgetary settings faced by successive governments during the 1990s, which involved continual pressures to reign in expenditures, prompted attention to any source of spending that might be curtailed. The family reunion program was delivering a substantial flow of low-skilled, non-English-speaking and relatively capital-poor migrants who often needed welfare assistance. During the 1990s there was a much publicised debate about the extent of welfare benefit use on the part of newly arrived migrants. By the mid-1990s almost a third of migrants intending to enter the labour market received Commonwealth benefits during their first three years in Australia (Healy 1994, p. 49; 1996, p. 23). Related

150 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

studies of migrants arriving in the aged parent category also showed a high degree of reliance on welfare benefits (Evans 1994). This kind of evidence fed public concerns that the nation’s tax system was being bled by new arrivals who had not contributed to the nation’s wealth. One attitude survey in 1997 showed that 70 per cent of a national sample indicated they were ‘very concerned’ about ‘immigrants getting welfare on arrival’. This concern was higher than that recorded for any other of the wide range of issues put to the sample (White 1997, p. 13). Successive governments during the 1990s responded by winding back access to welfare benefits and by tightening the rules governing family reunion. As to the first, the Labor Government began the process in 1991 when it imposed a two-year bond on the sponsors of parents ($3500 for the principal applicant) which was repayable if no social security benefits were paid out during the time of the bond. Then in 1994 the Labor Government implemented a moratorium on the payment of most social security benefits for the first six months of residence for all migrants except those entering under the humanitarian categories. In 1996, the new Coalition Government, with the support of the Labor Opposition, though not the Democrats, extended this moratorium to two years. In 1999, without dissent from any party, this two-year moratorium was extended to New Zealanders. These budgetary measures were justified in a fiscal context and were implemented by the Department of Social Security. They were not presented as part of the immigration control process but were nevertheless bitterly opposed by ethnic community leaders who saw them as part of such an agenda. These leaders were probably right, because the Coalition Government has made it plain that it wanted to reduce the inflow of migrants thought to constitute a financial burden to the Commonwealth. This objective could be achieved either by reducing access to benefits or by contracting family reunion rights. In practice both options have been pursued by successive governments since the late 1980s. Labor began the process of tightening sponsorship rules with the introduction of a Balance of Family ruling in 1989 for all parent sponsorships: half or more of the children had to be living in Australia. This measure prevented parent sponsorships for most recently migrant communities where the normal number of siblings per family was three or more and where, as was usually the case, insufficient time had elapsed for more than one of the siblings to

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 151

have migrated. As Table 6.3 shows, the number of parent visas did fall sharply in the early 1990s. However the numbers began to rise again in the mid-1990s (despite the bond requirement) partly as a consequence of the high numbers of Chinese students who achieved permanent-residence status at the time. Because of China’s family policies most of these students had few siblings and thus were not affected by the Balance of Family provision. When the Coalition came to power in 1996 it sought to tighten the Balance of Family rules again, by requiring more than half of the children to be resident in Australia. It was unable to get this legislation through the Parliament and instead sought the power to put a quota on the number of parents allowed to enter in any program year. This measure passed into law in March 1997 with Labor support (Birrell 1997, p. 24). Thereafter, the Coalition Government cut the number of visas issued to parents from 7580 in 1996–97 to 1080 in 1997–98. Since that time the parent program has remained at about 1000 per year. Table 6.3 shows that there was an upward trend in visas issued to spouses and fiancées, which peaked in 1995–96, the last year of the Labor Government. It contracted thereafter for several years. A variety of factors are involved in this contraction. One is a decline in the number of spouses or fiancées sponsored from China: by the late 1990s most of the Chinese students granted permanent residence had reunited with former spouses or returned for a partner if not married. Another factor is that the Coalition Government has attempted with varying success to implement new control measures. It unsuccessfully sought the power to cap spouse visas (Birrell 1997, pp. 12–13) but had used the powers already in existence when it came to power in 1996 to cap the issuing of fiancée visas to about half the previous peak of around 6000. In addition, it implemented tougher bona fide tests for assessing whether the claimed relationship is genuine. Finally, in 1997, new rules were introduced, again with the Labor Opposition’s support, to limit spouse visas to two years, pending subsequent proof that the partnership is ‘genuine and continuing’. This temporary visa meant a spouse could not receive Australian welfare or educational benefits during the first two years of residence here. The combination of these measures appears to have culled some applicants whose marriage intentions were not simply a reflection of the personal attractions of their partners. Since 1998–99 there has been a further rise in the number of

152 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

spouses and fiancées visaed to 32 740 in 2001–02, particularly amongst those visaed on-shore. This development in part reflects the parallel increase in visitors, working holiday makers and students to Australia. This increases the chance of Australian residents meeting prospective migrant partners. For the migrants, marriage is the one secure visa option open to visitors interested in permanent residence in Australia. The Concessional Family category which applies to adult siblings and other relatives has effectively been eliminated as a family reunion right. In 1997, the Coalition Government incorporated the Concessional Family category within its skilled migration program and renamed it the Skilled Australian Linked (and since mid-1999 the Skilled Australian Sponsored category). Applicants were assessed according to the same English language and other skill requirements as Independent applicants. They continue to be granted a points concession for their sponsorship, which makes the assessment process less stringent than is the case for Independent applicants. Nevertheless, selection is restricted to English-speaking applicants who must all hold professional, associate professional or trade occupations. Furthermore, those seeking this visa must first establish that their occupational credentials are approved by the relevant Australian authority. A notable feature of these control measures is that they were not distilled from any considered public discourse or government-sponsored public inquiry. Rather they emerged without warning from within the Immigration bureaucracy. They reflected departmental officers’ concern to bring the program back ‘under control’ and their knowledge that in the 1990s governments were likely to support such initiatives. All welfare payments are now coming under closer scrutiny. Government awareness of popular resentment towards people having access to such benefits if they have made no previous contribution means that it is unlikely that there will be any significant response to ethnic community lobbying to reopen the family reunion program.

S K I L L E D M I G R AT I O N Reform of the skilled migration program followed the pattern described above. The numbers granted visas were cut sharply in the early 1990s and the selection system tightened so as to target migrants with skills needed by Australian employers. The measures

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 153

introduced have not been the subject of political party contestation, implying a high measure of consensus about the need to improve the selection system. The reform has focused around tightening English-language and qualifications requirements. In 1992, the Labor Government stipulated that all Independent applicants had to have their English-language skills assessed by a professionally constructed English test (Hawthorne 1995). At this time applicants who did poorly on this test were not automatically failed; however for some occupations labelled ‘occupations requiring English’, a minimum standard of English was required. These occupations included teaching, engineering and all of the health professions, but not accounting, computing or most of the trades. In 1997, the latter occupations along with almost all other professional and associate-professional occupations were designated as ‘occupations requiring English’. In 1999 this precedent was extended by the Coalition Government. Minimum English standards were made a compulsory requirement for all applicants under both the Independent and Skilled Australian Linked categories. Applicants had to be able to speak, read, write and comprehend English at a ‘vocational’ level to be selected, regardless of the points they scored on other selection criteria (Department of Immigration and Multicultural Affairs 1999). In the case of qualification requirements, the reform process began in 1989, when the Labor Government incorporated changes to its skilled selection system recommended by the CAAIP report. The principal change was that applicants had to base their skill claims on credentials recognised by accrediting professional and trade agencies in Australia. It was no longer enough to be well educated: the education had to be related to job requirements in Australia. In the late 1990s business interests began agitating for an expansion in the skilled intake. The context was one of rapid growth in demand for professional workers, particularly those with IT and accounting expertise. Simultaneously, universities had been expanding their overseas full-fee student enrolments in the same fields. Meanwhile, domestic enrolments were stabilising largely because, since the mid-1990s, the number of subsidised places for local students had been capped by successive governments. From the business point of view, it seemed that these overseas students provided an instant solution to the shortages in IT and accounting. The government responded with some major reforms.

154 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Beginning in mid-1998 applicants under the points-tested categories who had completed at least a year’s full-time training in Australia were given two important concessions. One was an additional five points. The other was that the requirement of several years’ work experience in the occupation (mandatory for other applicants) was waived. This reform was based on evidence showing that migrants with Australian qualifications have better employment records than their counterparts from the same countries who had been trained overseas (Department of Immigration and Multicultural Affairs 1999, p. 43). Furthermore, in mid-1999, overseas student graduates were permitted to apply for immigration from within Australia rather than from overseas (as had previously been the case). Visas issued in the Independent category nearly doubled between 1999–2000 and 2001–02 (see Table 6.3), with much of the increase derived from former overseas students. By 2001–02, nearly half of all those who were granted visas in the Independent category had received points for Australian training. This experience illustrated the strong interest in migration to Australia from Asian countries. Enrolments in university courses which offer good prospects for subsequent permanent migration, like the one-year graduate diploma or Masters courses in IT, have grown dramatically. To the consternation of critics, these courses were deemed to be sufficient to obtain the necessary professional accreditation from the Australian Computer Society (Kinnaird 2002, p. 60).

TEMPORARY ENTRANTS At the same time as the formal migration program has been reduced and tightened, there has been a major expansion in the number of persons coming to Australia on a temporary basis. These include skilled migrants employed on short-term contracts, business visitors, students, working holiday makers and short-term visitors (mainly tourists). In every case, arrival numbers have tended to increase. Another important source is New Zealanders, including persons who have migrated to New Zealand and have taken out New Zealand citizenship, which is available after three years’ residence. They are permitted to move to and from Australia without any of the visa restrictions applying to other overseas-born persons. By the late 1990s, New Zealand was the largest source country for persons stating that they were coming to Australia either on a permanent or a long-term (a year or more) basis.

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 155

Except for the New Zealanders, these temporary entrants are a revolving group. The distinguishing feature of most of these programs is that they are designed to meet specific Australian needs with a minimum of cost to Australian taxpayers and no long-term commitment as settlers. An important example is the temporary skilled program. Since August 1996, as a result of a Labor Government initiative, Australian employers have been able to recruit as many skilled workers on short-term contracts as they want. The previous requirement that the sponsor first prove that the skills in question are not available in Australia was abolished. By 1998–99 about 16 000 visas to principal applicants under this program were issued. The workers in question are the responsibility of the sponsoring firms. They have no welfare entitlements and they are normally not allowed to switch employers.

D E M O G R A P H I C I M P L I C AT I O N S During the second half of the 1980s the numbers of migrants granted visas under the immigration program, including the humanitarian categories, averaged around 120 000 per year. As Table 6.3 shows, the program levels for the 1990s were well below this figure, although they increased at the end of the 1990s. A program level of 100 000 would be expected to yield a net permanent intake of around 70 000 given past departure levels. This implies a quite low annual percentage addition to Australia’s population by comparison with the contribution from net migration in the 1950s, 1960s and 1980s. However, the Australian Bureau of Statistics (ABS 2000b) has estimated that during the second half of the 1990s the net overseas migration intake (defined to include net long-term as well as permanent movements) was around 100 000 per annum. Should this annual level continue, net immigration would add about 6.5 million to Australia’s population by 2051 (ABS 2000c, p. 39; see also Chapter 11 for alternative projections). The very high recent net movements from New Zealand, which are not included in the immigration program figures, are an important component of this figure. Another contribution is the current high net long-term movements of visitors: that is, the difference between those who say they are arriving for a visit of over 12 months and those who say when they leave that their visit was 12 months or more. The ABS has estimated this figure at 30 000 or more per year during the second half of the 1990s. Much of this gain is due to the

156 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

steady increase in temporary long-term arrivals since the mid-1990s and the lag between arrivals and departures. Visitors are a revolving group whose movements in and out of Australia should balance out eventually. There is an increasing number of temporary visitors who succeed in changing their status to permanent resident through marriage or the skilled categories as with the overseas students discussed earlier. But their demographic effect is not additional to the visa issued numbers in Table 6.3 because these figures include those who were granted visas onshore and offshore.

CONTEMPORARY DILEMMAS At the beginning of the 21st century, public debate about the future of immigration to Australia has been regenerated by business interests. These interests had already been influential in the expansion of the skill categories discussed earlier. Powerful organisations, including the Business Council of Australia, want sustained higher program numbers. They argue that higher migration will slow the rate of decline in labour force growth and of population ageing that will occur should fertility continue at sub-replacement levels and migration stabilise at 100 000 per annum or less. The Federal Australian Labor Party has indicated some sympathy with this cause. The future of debate about immigration is clouded by the emergence of new issues and new combatants. As has been argued, the migration issue has, and probably will, continue to engage the wider electorate. Interest groups like the Business Council are unlikely to be able to unilaterally determine migration outcomes. One development fanning public interest stems from the settlement patterns of migrants (see Chapter 8 for a discussion of the impact of immigration on population distribution). Sydney’s role as the dominant interface between Australia and the global economy has made it the focus of permanent and temporary migration movements in the business and skilled categories. Sydney has also been the main settlement point for Australia’s Asian and Middle Eastern communities (see Chapter 7). Recent migrants settling in Sydney and the next largest settlement location, Melbourne, particularly the lower skilled arrivals who have entered on family reunion or humanitarian visas, tend to concentrate near existing ethnic communities (see Chapter 8 on the impact of immigration on Sydney’s and Melbourne’s growth). This is partly because of their need to access services directed at their com-

T H E M A N A G E M E N T O F I M M I G R AT I O N : PAT T E R N S O F R E F O R M

• 157

munities, but also because most are low-income earners with little choice but to settle in areas with relatively low-cost housing. The result is high concentrations of struggling non-English-speakingbackground (NESB) migrants (Birrell and Seol 1998). In 2001, 73 per cent of all males aged 25–64 living in Fairfield and Auburn in Sydney were overseas-born, mostly migrants from NESB countries, compared with 42 per cent of all Sydney male residents in the same age category (ABS 2001 Census). Opinions among scholars differ as to whether these settlement patterns are a temporary phenomenon and about their implications for the welfare of the people involved. Burnley (2001), for example, sees many positives. However, the effect of concentrating many struggling migrant families is to strain the capacity of the local welfare and educational institutions. Youths frustrated by poor economic prospects tend to see their predicament as reflecting their ethnic status, thus promoting ethnic solidarity. Where they become involved with criminal activities such as drug smuggling, this contributes to mainstream negative stereotypes. Even with the modest assumption that net overseas migration to Australia will continue at 70 000 per year, the NSW Government is projecting a rise in Sydney’s population of one million, from 3.9 million in 1996 to 4.9 million in 2026 (Culpin Nugent and Truscott 2000, p. 14), most of which will be due to overseas migration. These numbers have drawn the NSW Government into the debate, led by the Premier, Bob Carr. He has put a strong case for low migration in order to save Sydney from the environmental and amenity consequences, notably congestion, that he says will follow from an increase in overseas migration. Ethnic tensions in Sydney add to this volatile mix. Intervention from such a source is unprecedented. Hitherto, state governments have always sought bigger numbers in order to boost their states’ economy. This is only one example of the controversial nature of immigration issues in Australia. As the Tampa affair showed, such issues can readily engage public attention and influence political outcomes. In these circumstances, predictions about future migration patterns, crucial though they are for any assessment of Australia’s demographic future, remain uncertain.

7

A G R E AT E R D I V E R S I T Y O F O R I G I N S S I E W- E A N K H O O

There was already an awareness in the 1970s that Australia’s population would become more diverse in its ethnic composition after the adoption of an immigration policy of non-discrimination on the grounds of race, colour or nationality. The National Population Inquiry’s report (1978, p. 185) noted that ‘Australia is more multicultural today than it was a generation ago, and continuing immigration will tend to increase the multicultural mix in the future’. The report did not go into further details as to what the future population would be like in terms of its ‘multicultural mix’. While the inquiry considered and discussed projections of the population’s age and education profiles and labour force growth, it did not make any projections of the population’s ethnic profile. At that time, Australia’s population already included many distinctive immigrant communities: people from Italy, Greece, Malta, Lebanon, Turkey and a number of Eastern European countries. Immigration from Asia had begun to increase in the early 1970s and a number of refugees had arrived from Vietnam after 1975. There had also been some immigration from South America. There were indications of increasing ethnic diversity compared to the immediate post-war years, but the demographers then were too cautious to extrapolate from these beginnings of diversity to say what the population at the end of the 20th century would look like. Perhaps caution was justified because the 1970s decade was a pivotal era in Australia’s changing ethnic composition. A new immigration policy, one that had the potential to admit new immigrants from anywhere in the world, had just been implemented. It would have

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 159

been difficult then to project the outcomes of such a policy. Now, a generation later, the population has certainly become more diverse ethnically; indeed the growth in ethnic diversity must be considered one of the most important transformations of Australian society during the past 30 years. While immigration has been the driving force behind the increase in ethnic diversity, another contributing factor has been ethnic intermixture through intermarriage. This is also likely to become more important in the future, as immigrants and their Australian-born offspring integrate into Australian society (Price 1994). Australia is not the only country whose population has become more diverse ethnically over the past 30 years. Other countries of immigration such as Canada, the United States and New Zealand have also experienced increased diversity through immigration over the same time period, due to similar reforms in their immigration policies. However, there are some differences between Australia and these other countries; one is that the level of immigration to the other countries on a per capita basis has been less than that to Australia. Thus the foreign-born components in the US and Canada for example are smaller than that in Australia: 9 per cent and 18 per cent respectively in the mid-1990s compared to 23 per cent in Australia. In New Zealand it was 19.5 per cent according to the 2001 Census. Also, immigration to the US is dominated by its proximity to Latin America so that a very large proportion of its immigrants is Hispanic; no single non-English-speaking group predominates in the same way in recent immigration to Australia. Australia and the other countries of immigration are similar in some ways but different in others in their ethnic composition, as will be shown later. This chapter examines the changes in ethnic composition of Australia’s population from several perspectives. First it looks at the origins of the population by place of birth. However, birthplace has some limitations as an indicator of ethnic origin for some people such as second-generation Australians or immigrants from ethnically diverse countries. The recent 2001 Census asked people to state their ancestry. While these data might be a better indicator of ethnic origin, a significant proportion of the population stated their ancestry as ‘Australian’, which did not provide sufficient information about their ethnic background. The measures of ethnic origins developed by Charles Price (1988, 1996) attempt to overcome these limitations and they are reviewed here to provide more details about ethnic diversity. Diversity in language and religion is also examined to complete the picture. Since increasing ethnic diversity has not occurred uniformly across Australia (see Khoo and Price 1996), it is also important to look at

160 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

regional differences. It also seems important to consider the extent to which the different immigrant groups are integrating into Australian society. In this regard, the issues of language shift and intermarriage are also examined. Finally the chapter examines some projections of Australia’s ethnic composition and compares them with similar projections for some of the other countries of immigration.

CHANGES IN BIRTHPLACE COMPOSITION AND ETHNIC ORIGINS If one looks at only the proportion of overseas-born people in the total Australian population at the beginning of the 20th century and at the end, there did not appear to be much change. In 1901, 22.9 per cent of the population was born outside Australia. A hundred years later in 2001, it was 23.1 per cent. There was, however, considerable change during the intervening years. There was a decline in the overseas-born component of the population during the first half of the century to less than 10 per cent by mid-century. The percentage of overseas-born people then increased again after 1950 to its 2001 level. Table 7.1 Distribution of Australia’s population by birthplace, 1901–2001 Birthplace

1901 %

%

%

%

%

%

%

%

%

Australia

77.1

84.5

90.2

85.7

81.6

79.9

78.9

77.2

76.9

New Zealand

0.7

0.7

0.6

0.5

0.5

0.7

1.6

1.7

2.0

UK, Ireland

1921

1947

1954

1966

1976

1986

1996

2001

18.0

12.4

7.1

7.4

7.9

8.2

7.3

6.5

6.1

Other Europe

2.0

1.3

1.5

5.5

8.5

8.1

7.1

6.4

5.9

Africa and Middle East

0.1

0.1

0.1

0.2

0.4

0.5

1.5

1.8

2.0

Asia

1.2

0.6

0.3

0.6

0.9

1.8

2.6

5.0

5.5

America

0.3

0.2

0.2

0.2

0.2

0.6

0.8

0.8

0.9

Other overseas

0.6

0.2

0.1

0.1

0.0

0.2

0.2

0.6

0.7

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

3.8

5.4

7.6

9.0

11.6

13.5

15.6

17.8

18.8

Total Total population (million)

Totals exclude persons not stating their birthplace.Totals for 1996 and 2001 also exclude overseas visitors. SOURCES Population censuses. NOTE

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 161

More significantly, the birthplace origins of the overseas-born in 2001 were quite different from and much more diverse than in 1901 (Table 7.1). At the beginning of the century, most overseasborn people were from the United Kingdom and Ireland. They made up 18 per cent of the total population and nearly 80 per cent of the overseas-born. Other European-born people were only 2 per cent of the total population in 1901 and people born in Asia were just over 1 per cent: mainly Chinese who had arrived during the gold rush of the 19th century. The increase in diversity by birthplace among the overseas-born began after World War II with the arrival of immigrants from Greece, Italy and other Southern European countries and also from Eastern Europe. In 1966, people born in the UK and Ireland had decreased to just under 8 per cent of the total population while those from other European countries had increased to over 8 per cent. The proportion born in Asia was still less than 1 per cent. Over the 30-year period from 1971 to 2001, the main changes were an increase in the proportion of the population born in nonEuropean countries and a decrease in the proportion of Europeanborn people. While people born in the UK were still the largest birthplace group in 2001, they accounted for just over 6 per cent of the total Australian population. People born in other European countries also made up about 6 per cent of the total population in 2001. The Asian-born showed the largest increase to more than 5 per cent of the 2001 population. The other regional groups showing significant increases were those from Africa and the Middle East, and from New Zealand, each constituting 2 per cent of the total population in 2001. Figure 7.1 illustrates the growing diversity among the overseasborn population in Australia between 1971 and 2001. Whereas in 1971, the British and other European-born were dominant, forming 85 per cent of the overseas-born population, their share of the overseas-born had declined to 51 per cent in 2001. The non-Europeanborn had increased their share from 15 per cent to 49 per cent, with the Asian-born being the largest non-European regional group. The increase in the Asian-born population has also been characterised by considerable diversity. People born in the various countries of Asia differ in many ways including race, language, religion, culture and socio-economic status. People from North-East Asia are of a different race from those from South Asia; people within each Asian

162 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 7.1 Birthplace of the foreign-born population, 1971 and 2001 1971 5%

2001

1% 2% 4%

9%

3%

3% 26%

4% 42% 24%

43% 25% 9%

SOURCES

UK, Ireland

New Zealand

America

Other Europe

Asia

Africa & Middle East

Other

1971 and 2001 censuses.

subregion also have different languages and religions. The diversity among the Asian-born groups has been highlighted in a number of studies (eg. Khoo et al. 1994; Coughlan and McNamara 1997; Jayasuriya and Kee 1999; Burnley 2001; Jupp 2001). These studies have also shown that migrants from the various countries in Asia differ from one another in their level of English-language proficiency, labour force participation and settlement patterns. Table 7.2 shows the ten largest overseas birthplace groups in 1971 and 2001. In 1971, the ten largest birthplace groups were all of English-speaking or other European origins. In 2001, the two largest groups were of English-speaking origins (from the UK and New Zealand) but of the remaining eight, four were European and four were Asian. Six of the top ten birthplace groups were also among the top ten in 1971: those from the UK, Italy, Greece, Germany, Netherlands and New Zealand. However while the UK-born maintained its number one position, the other European birthplace groups dropped down the rank order between 1971 and 2001. All of them including the UK-born also decreased in size, reflecting a decline in European

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 163

migration to Australia after 1971. In contrast, the New Zealand-born group increased in number and moved from seventh to second place, due to the increase in trans-Tasman migration during the 1990s (as discussed in Chapter 6). Because of the diversity of origins of recent immigrants, none of the ten largest overseas birthplace groups in 2001, aside from the UK-born, formed more than 2 per cent of the total Australian population. Table 7.2 The ten largest overseas birthplace groups, 1971 and 2001 Birthplace

UK Italy Greece Yugoslavia Germany Netherlands New Zealand Poland Malta Ireland NOTE a Excludes SOURCES

1971 Number of % of persons total pop. 1 046 356 8.2 289 476 2.3 160 200 1.3 129 816 1.0 110 99 80 59 53 41

811 295 466 700 681 854

0.9 0.8 0.6 0.5 0.4 0.3

Birthplace

2001 Number of persons

UK New Zealand Italy Vietnam

1 036 355 218 154

245 765 718 831

% of total pop. 5.5 1.9 1.2 0.8

142 116 108 103 95 83

780 431 220 942 452 324

0.8 0.6 0.6 0.6 0.5 0.4

Chinaa Greece Germany Philippines India Netherlands

Hong Kong and Taiwan.

1971 and 2001 censuses.

Information from the 1986 and 2001 censuses on people’s ancestry provides another indication of the changes in Australia’s ethnic composition during the last decade-and-a-half of the 20th century. Ancestry provides a better estimate of the size of the various ethnic groups than birthplace as it includes both the overseasborn first generation as well as the Australian-born second and later generations. However, a significant proportion of the population (22 per cent in 1986 and 36 per cent in 2001) stated their ancestry as ‘Australian’, and it was not possible to ascertain the ancestry of these people according to the origins of their ancestors who were the migrants to Australia. People could also state more than one ancestry and the ancestry counts included people who were partly of that ancestry.

164 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 7.3 Number of people according to ancestry,a 1986 and 2001 Ancestry

1986 Number

Australian Other Australian ancestriesb New Zealander Maori Other Pacific Islander English Irish Scottish Italian German Greek Dutch Maltese Other European

2001

3 402 407 198 909

% of pop. 21.8 1.3

Number 6739594 106441

% of pop. 35.9 0.6

75 085 26 035 19 833

0.5 0.2 0.1

123314 72956 91739

0.7 0.4 0.5

6 358 880 1 919 727 540 046 880 256 742 212 375 703 268 754 136 754 1 196 164

33.9 10.2 2.9 4.7 4.0 2.0 1.4 0.7 6.4

6 607 902 740 620 510 336 231 125 1 021

228 679 522 227 402 782 148 797 070

42.3 5.8 4.7 4.0 3.3 2.2 1.5 0.8 6.5

Lebanese Turkish Other Middle Eastern Total Middle Eastern

92 36 119 249

428 903 702 033

0.6 0.2 0.7 1.6

162 54 147 363

239 596 032 867

0.9 0.3 0.8 1.9

Chinese Indian Vietnamese Filipino Other Asian Total Asian

201 71 64 38 107 483

331 185 998 698 363 575

1.3 0.5 0.4 0.2 0.7 3.1

556 156 156 129 339 1 339

554 628 581 821 481 065

3.0 0.8 0.8 0.7 1.8 7.1

Othersc

268 424

1.7

243 972

1.3

Percentages do not add to 100 because people could state more than one ancestry. Total responses; b Includes Aboriginal, Torres Strait Islander and Australian of South Sea Islander descent; c Includes ‘mixed’ ancestry. SOURCES 1986 and 2001 censuses. NOTES a

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 165

Between 1986 and 2001, the general pattern was that the percentage of Australians stating a European ancestry decreased while the percentage stating a non-European ancestry increased. Exceptions were the increases in the proportion of Australians claiming Irish, Italian or German ancestry. There appeared to be an understatement of Irish ancestry in the 1986 Census (Australian Bureau of Statistics 1990a) and it seemed that people had become more aware of their Irish heritage in 2001. The increase in the number and proportion of people stating Italian or German ancestry was due primarily to a growing third or later generations in these two communities. The number of people stating an Asian ancestry increased from less than 500 000 in 1986 to 1.3 million in 2001, or from 3 per cent to 7 per cent of the total population. Large increases were seen in all the major Asian ancestry groups such as Chinese, Indian, Vietnamese and Filipino. The number of people of Middle Eastern origins also increased but not to the same extent as the number of Asian origins. However, none of the non-English-speaking groups formed more than 5 per cent of the total population. The largest group—those of Italian ancestry—comprised just under 5 per cent of the total population. People of Chinese ancestry were the largest group of Asian origin, numbering over half a million and making up 3 per cent of the total population in 2001. They included many migrants from countries such as Malaysia, Singapore, Hong Kong, Taiwan and Vietnam, and their Australian-born children. Price (1988, 1996) has developed other measures of ethnic origin using data going back as far as 1947 on birthplace of individuals and their parents, race, language and religion, and taking account of ethnic intermixture and rates of intermarriage by birthplace. Table 7.4 shows his estimates of Australia’s ethnic composition for 1987 and 1999. Between 1987 and 1999, the Anglo-Celtic component of Australia’s population declined from 75 per cent to 70 per cent. The total European component (excluding Anglo-Celts) remained unchanged at just over 18 per cent; however the proportions of North, West and South European origins declined slightly, while the proportion of East European origins increased due to renewed migration from Eastern Europe following the breakup of the former Soviet Union and the former Yugoslav republic. All the Asian groups showed significant increases during the period, and the total Asian

166 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

component increased from 3 per cent to 6.5 per cent—which was very similar to the ancestry figures. The Middle East and North African component went from 2 to 3 per cent, slightly higher than the ancestry figures. These figures also show that aside from the predominance of the Anglo-Celt majority, no non-English-speaking group comprised more than 5 per cent of the population. Table 7.4 Australia’s ethnic composition, 1987 and 1999 Ethnic origin English, Cornish Irish Scots Welsh Manx, Breton, others Total Anglo–Celt

% of total population 1987 1999 43.9 45.0 17.2 12.1 11.9 11.4 1.4 1.3 0.1 0.1 74.5 69.9

German Dutch Scandinavian French Austrian Swiss Other Total North and West European

3.8 1.4 1.0 0.5 0.3 0.2 0.2 7.4

3.5 1.4 0.9 0.5 0.2 0.2 0.2 6.9

Former Yugoslav Polish Russian Ukrainian Belo-Russian Baltic Czech and Slovak Hungarian Romanian Total East European

1.7 0.8 0.2 0.2 0.1 0.3 0.2 0.3 0.1 3.9

1.8 0.9 0.4 0.2 0.1 0.3 0.2 0.4 0.1 4.4

Italian Greek Maltese

3.7 2.0 0.8

3.4 1.8 0.8

A G R E AT E R D I V E R S I T Y O F O R I G I N S

Spanish Portuguese Total South European

0.6 0.2 7.3

0.7 0.2 7.0

Lebanese Egyptian Other Arab Turk Armenian Assyrian Jew Other Total Middle East and N. African

0.7 0.1 0.1 0.3 0.1 0.1 0.7 0.0 2.1

1.4 0.2 0.4 0.3 0.1 0.1 0.7 0.1 3.3

Indian Sri Lankan Other Total South Asian

0.4 0.1 0.1 0.6

0.7 0.3 0.3 1.3

Filipino Indochinese Indonesian Malay Other Total South-East Asian

0.3 0.6 0.1 0.1 0.1 1.2

0.8 1.1 0.3 0.1 0.2 2.5

Chinese Japanese Korean Total North-East Asian

1.3 0.1 0.1 1.5

2.3 0.2 0.2 2.7

African

0.1

0.1

Maori Samoan Tongan Total Pacific Islander

0.2 0.1 0.0 0.4

0.2 0.1 0.1 0.5

Aboriginal and Torres Strait Islander

1.0

1.0

100.0

100.0

Total SOURCES

Price 1989, 2000.

• 167

168 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

These estimates of ethnic composition are based on the measure of ‘ethnic strength’ (Price 1996). They are not equivalent to actual counts of the number of people because they are based on fractional contributions to the ‘strength’ of an ethnic group by people of mixed origins. They reflect the proportionate contribution of each ethnic group to the total population rather than numbers of people. Ethnic intermixture has resulted in many people of mixed ethnic origins who would contribute proportionately to more than one ethnic group. With the increase in ethnic diversity over the past 30 years, this trend is likely to become more significant in the future. Price (1994) has indicated that people of mixed ethnic origins will be the largest element in Australia’s population if current trends in intermarriage continue. As noted earlier, Australia is not the only country to have experienced increasing ethnic diversity. The other main countries of immigration—Canada, US and New Zealand—also have significant immigrant communities although they differ in ethnic composition from one another and from Australia, owing to their different geographic location and sources of immigration. The main minority groups in the US are Blacks and Hispanics; in New Zealand, they are Maoris and other Pacific Islanders. However, as in Australia, all the three countries have received many immigrants from Asian countries during the 1980s and 1990s. Both Canada (Statistics Canada 1999) and New Zealand (Statistics New Zealand 1999) reported that about 5 per cent of their total population in 1996 were born in Asian countries, a figure that was similar to Australia’s, while in the US the proportion was about 3 per cent (US Bureau of the Census 1999).

LANGUAGE AND RELIGIOUS DIVERSITY The increased diversity of origins of Australia’s population is also reflected in the diversity of languages spoken at home and the various religions and Christian denominations being practised today. An indication of this is the increase in the number of languages and religions being coded by the Australian Bureau of Statistics with each census since 1966. The 1966 Census did not have a question on language spoken at home; in the 2001 Census, it was possible to code up to 238 different languages. In 1966, 20 religious categories, 15 of which were Christian denominations, were specified in census publications, with Hebrew (sic) the only specified non-Christian religion. In the 2001 Census there was provision to list 123 religions and religious denominations.

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 169

The 2001 Census counted over 2.8 million people—or 15 per cent of the total population—speaking a language other than English at home. This proportion has not changed much since the 1970s, but there is much greater diversity in the languages being spoken at home now than before. At the time of the 1976 Census, there were 17 languages other than English that were spoken by at least 10 000. In 2001 there were at least 34. Most of the additional languages being spoken at the end of the 1990s decade that were not prevalent before 1980 were Asian languages such as Hindi, Khmer, Korean and Thai. There was also a large increase in the number of people speaking Chinese, Arabic, Vietnamese and Tagalog, due to immigration from China and Hong Kong, the Middle East, Vietnam and the Philippines during the 1980s and 1990s. Between 1976 and 2001, Chinese moved from being the 12th most common language spoken at home to the second most common, after English (Table 7.5). In 1996, it replaced Greek as the third most common language spoken at home after English and Italian, and in 2001 it replaced Italian as the second most common language. In 1976 the ten most common languages were all European; in 2001, they included four nonEuropean languages. European languages such as Dutch and German are being spoken by fewer people today than in the 1970s because the overseas-born communities have aged and the language retention rate has been low among the second generation (Khoo 1995). While there are many more languages spoken in Australia today, it is also significant that the proportion of the total population speaking only English at home is still at 80 per cent. Chinese, the second most common language, was spoken by just 2 per cent of the population in 2001. A large proportion of the overseas-born population of non-English-speaking origins is proficient in English and speaks it at home. Work undertaken by the Department of Immigration and Multicultural Affairs shows that of the overseas-born people from 243 non-English-speaking countries, only those from five countries had less than 60 per cent who could speak English well in 1996: these countries were Vietnam, China, Turkey, Cambodia and Laos. Almost 100 per cent of migrants from 160 other countries and more than 75 per cent of those from another 78 countries were able to speak English well (Department of Immigration and Multicultural Affairs 1997). All people immigrating in the skill category, although not those in the family reunion or humanitarian categories, have to meet selection criteria on English proficiency (see Chapter 6).

170 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 7.5 Main languages spoken at home, 1976 and 2001 1976 Language English only Italian Greek German Slovene

No. of speakers 10 282 359 449 521 262 177 183 636 91 763

2001

% of populationa 83.3 3.6 2.1 1.5 0.7

Language

French Dutch Polish

80 986 64 767 62 946

0.7 0.5 0.5

Spanish Maltese Arabic Chinese Aboriginal languages Serbo-Croatian Hungarian

48 36 34 29 27

344 036 613 902 571

0.4 0.3 0.3 0.2 0.2

English only Chinese Italian Greek Arabic (Lebanese) Vietnamese Spanish Tagalog (Filipino) German Macedonian Croatian Polish Turkish

26 964 19 618

0.2 0.2

Serbian Hindi

NOTES a SOURCES

No. of % of speakers populationb 15 013 965 80.0 401 357 2.1 353 605 1.9 263 717 1.4 209 372 1.1 174 236 93 593 78 878

0.9 0.5 0.4

76 71 69 59 50

443 994 851 056 693

0.4 0.4 0.4 0.3 0.3

49 203 47 817

0.3 0.3

Aged 5 years and older; b Excluding overseas visitors. 1976 and 2001 censuses.

Greater religious diversity over the past 30 years is indicated by the increase in the variety of Christian denominations and in the number of non-Christian religions. The number of people who were of non-Christian religions and their proportion of the population increased significantly. In 1971 less than 1 per cent of the population was of non-Christian religions, with most being Jewish. In 2001 nearly 5 per cent were of non-Christian religions (Table 7.6). The two largest non-Christian religious groups were Buddhists (357 813 or 2 per cent of the population) and Muslims (281 578 and 1.5 per cent of the population). They were followed by Hindus (95 473) and then Jews (83 993) (Australian Bureau of Statistics 2002a). The increase in the number of people in non-Christian religious groups is closely related to the increase in immigration from Asian and Middle Eastern countries. Most people who are Buddhist are from South-East Asian countries, while Hindus are mainly from India.

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 171

Table 7.6 Australia’s population by religious affiliation, 1971 to 2001 1971 % 15.0 31.0 8.6 8.1 2.7 1.4 1.5 17.9 86.2

1981 % 26.0 26.1 4.9 4.4 2.9 1.3 1.4 9.4 76.4

1991 % 27.3 23.9 8.2 4.3 2.8 1.7 1.5 4.3 74.0

2001 % 26.7 20.7 6.7 3.4 2.8 1.7 1.3 4.7 68.0

Buddhist Hindu Islamic Judaic Other non-Christian Total non-Christian

— — 0.2 0.5 0.1 0.8

0.2 — 0.5 0.4 0.2 1.3

0.8 0.3 0.9 0.4 0.2 2.6

1.9 0.5 1.5 0.4 0.5 4.8

Inadequately described No religion Not stated

0.2 6.7 6.1

0.5 10.8 10.9

0.3 12.9 10.2

1.9 15.5 9.9

100.0 12 755.60

100.0 14 576.30

100.0 16 850.2

100.0 18 769.2

Catholic Anglican Uniting Church Presbyterian Orthodox Baptist Lutheran Other Christian Total Christian

Total Population (‘000) NOTE

1996 and 2001 totals exclude overseas visitors. Censuses, 1971–2001.

SOURCES

Muslims are mostly from Middle Eastern countries, with smaller numbers from Indonesia and Malaysia. People born in Australia who are of these religions are mostly the Australian-born children of immigrants although there may have been a small number of Australians who have converted to these religions. Recent immigration is also likely to have contributed to the size of a number of Christian groups. Immigrants from the Philippines are almost all Catholics and many people from Korea are Presbyterians. There are also significant numbers of Christians among people who have migrated from Vietnam, Lebanon, Hong Kong, Malaysia, India and Sri Lanka (Department of Immigration and Multicultural Affairs 2000).

172 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

REGIONAL DIFFERENCES IN DIVERSITY The ethnic diversity in birthplace, language and religion, as discussed above, is not observed uniformly across the country. There are considerable differences among the states and territories and, more significantly, between the cities and country areas. Immigrants, especially those of nonEnglish-speaking origins, tend to settle in the large cities, particularly Sydney and Melbourne, as noted in the previous chapter (see also Burnley 2001). Also, different immigrant groups seem to prefer certain states or territories because of the presence of relatives and friends, employment opportunities and preferences for a particular climate or lifestyle. Western Australia had the highest percentage of overseas-born people in 2001 (27 per cent) while Tasmania had the lowest (11 per cent). Queensland (17 per cent) and the Northern Territory (14 per cent) also had relatively low percentages of foreign-born residents. More interestingly, the states and territories also differ quite considerably in the origins of their overseas-born population (Figure 7.2). New South Wales and Victoria had more diverse overseas-born populations than the other states, with significant proportions of people from Europe, Asia, Middle East and Africa; this is mainly because of the popularity of Sydney and Melbourne among recent immigrants as places to live. While Western Australia had the largest proportion of overseas-born people, they were not as diverse, nearly half being from the UK and Ireland. Victoria with its large Greek and Italian communities in Melbourne had the largest percentage of people born in Europe other than the UK; while Queensland, because of its proximity to the Pacific, had the largest percentage of people born in Oceania (4 per cent). The percentage of Asian-born people was highest in New South Wales (7 per cent) and lowest in Tasmania (1 per cent). Tasmania has the least diverse population among the states in both the proportion of people born overseas and their region of origin. The difference in population diversity between cities and country areas is shown in Figure 7.3. There is a striking contrast between Sydney and Melbourne and the rest of New South Wales and Victoria in both the percentage of overseas-born and their diversity of origins. Each of these cities had three times the proportion of overseas-born people compared with the rest of the state. The proportion from nonEuropean countries was eight times as high in each city as in the rest of the state. The contrast between the capital city and the rest of the state in the other mainland states was less because of the presence of other cities included with the rest of the state in the case of Queensland, and also because of their less diverse overseas-born population.

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 173

Figure 7.2 Percentage born overseas by region of birth: states and territories, 2001 30

% born overseas

25 20 15 10 5 0

SOURCE

NSW

Vic

Qld

SA

WA

Tas

Africa

Asia

Other Europe

America

Middle East, N. Africa

UK, Ireland

NT

ACT Oceania

2001 Census.

Figure 7.3 Percentage born overseas by region of birth: mainland state capital cities and rest of states, 2001 35 30

% born overseas

25 20 15 10

SOURCE

Africa

Asia

Other Europe

America

Middle East, N. Africa

UK, Ireland

2001 Census.

Australia

Rest of WA

Perth

Rest of SA

Adelaide

Rest of Qld

Brisbane

Rest of Vic.

Melbourne

Rest of NSW

0

Sydney

5

Oceania

174 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The regional differences in ethnic diversity are not unique to Australia: similar immigrant concentrations in the large cities are observed in the US, Canada and New Zealand. In the US, New York and Los Angeles are home to many recent immigrant groups and have a much more diverse population than other cities and non-metropolitan areas. In Canada, Vancouver and Toronto also have more diverse populations, with larger proportions of recent Asian immigrants than elsewhere in the country. In New Zealand, most immigrants from Asian and Pacific countries have settled in and around Auckland, the largest city. The regional differences described are based on overseas birthplace only and do not include the second generation by ethnic origin. Also, the population censuses do not differentiate the overseas-born population by their residency status. Therefore, the number of overseas-born people counted in the census includes temporary residents such as students from a number of Asian countries who are also more likely to be located in the capital cities where most tertiary education institutions are found. Notwithstanding these observations, there is no doubt that the major urban areas have a more ethnically diverse population than the country areas. These regional differences can have important social, cultural and political implications.

I N D I C ATO R S O F I N T E G R AT I O N The increase in ethnic diversity in Australia has raised questions of social cohesion and national identity. The multicultural approach in facilitating immigrant settlement that has replaced earlier policies of assimilation is sometimes seen as seeking to maintain differences rather than encourage interaction and integration with the wider society. While birthplace, language and religion data undoubtedly indicate increasing diversity in Australia’s population structure, other information provides some indications of integration and convergence with the majority population. It is important to look at these measures of integration and convergence to obtain a more balanced view of the implications of ethnic diversity.

E N G L I S H P R O F I C I E N C Y A N D L A N G UAG E S H I F T While 15 per cent of the population in 2001 spoke a language other than English at home, most of them could also speak English well or very well. In spite of the diversity of birthplace and languages being

• 175

A G R E AT E R D I V E R S I T Y O F O R I G I N S

spoken at home, a total 97 per cent of the population assessed themselves as proficient in English in the 1996 Census. Only 3 per cent of the total population were reported as not being able to speak English well or at all. Although this was the equivalent of about 500,000 people, most of them were either older people or very recently arrived immigrants. Among the latter group, the ability to speak English usually improves with length of residence. While there may be increasing diversity of origins by birthplace and language, English remains the predominant language and most overseas-born residents in Australia speak it fluently even if they continue to use their native language at home. There are also clear indications of a marked shift to using English at home among the second generation of non-English-speaking parentage. Figure 7.4 compares the percentage speaking English only at home among the first and second generations of some of the larger groups of non-English-speaking origins at the time of the 1996 Census. The intergenerational shift to English was quite large for a number of birthplace groups: for example those of Chinese, Italian, Polish and Filipino origins. Figure 7.4 Percentage speaking English only, first and second generation by country of origin, 1996 100

% speaking English only

80

60

40

First generation SOURCE

1996 Census.

Second generation

Netherlands

India

Germany

Sri Lanka

Malta

Philippines

Poland

Italy

Croatia

Hong Kong

Greece

Lebanon

China

0

Viet Nam

20

176 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

However, there are also differences among the groups. The shift to English is occurring more slowly among people of Vietnamese and Lebanese origins than the other major groups. The proportion speaking English at home was still less than 10 per cent among both first and second generations of Vietnamese origin. The Greek community, which is a much older community, also has a relatively small proportion of the second generation speaking English only at home. Some of the factors affecting the shift to speaking English at home were discussed by Clyne and Kipp (1995) in their study of the extent of community language maintenance in Australia. These include the size of the community group and its geographic concentration. Large groups are better able to maintain their language than smaller groups. This is particularly so when the members also live in close proximity to one another. The Vietnamese, Lebanese and Greek communities all share these characteristics which contribute to community language maintenance. Recent census data also showed that the proportion of the second generation speaking English only at home was much higher among those over age 25 than under age 25 (Khoo 1995). Two-thirds of all the second generation over the age of 30 who have both parents born overseas spoke only English at home compared with less than onethird of those under age 25. This pattern suggests that one of the reasons for maintaining the ethnic language by the second generation may be to communicate with non-English-speaking parents, and that once they leave the parental home, they may be more likely to speak English only at home. If so, this implies a further contraction in the maintenance of the community language in the third generation many of whom will grow up in homes where only English is spoken.

I N T E R M A R R I AG E Intermarriage between immigrants and local residents is often seen as an indicator of immigrant integration into the local community. This is because marriage is seen as the outcome of close interaction between members of the immigrant community and the local community. Studies of intermarriage between people born overseas and in Australia have shown that there is much variation in the intermarriage rate by birthplace. Some overseas-born groups have a higher rate of intermarriage with the Australian-born population than others (Price 1993, 1994; Penny and Khoo 1996). People from Western and Eastern European countries tend to have a higher rate of

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 177

Table 7.7 Overseas-born men and women who married Australian-born spouses or spouses born in the same country, 1996–98 Selected birthplace groups

Vietnam China Korea Cambodia Laos Hong Kong Philippines Thailand Romania Indonesia Iran Fiji Japan Sri Lanka Malaysia Chile India Singapore South Africa New Zealand

Intermarriage ratea Male Female % % 2.8 6.2 2.8 15.0 3.8 18.1 3.9 9.0 5.3 11.9 11.5 20.6 15.4 43.5 16.5 50.9 18.9 19.6 20.6 29.3 23.1 21.4 27.2 28.2 28.9 51.5 33.1 26.6 36.0 41.2 38.0 37.0 40.2 32.4 48.2 48.2 62.5 59.0 65.8 61.3

In-marriage rateb Male Female % % 86.4 79.2 75.7 60.9 84.4 65.4 72.8 65.2 55.3 53.5 47.7 43.5 78.2 30.7 58.4 14.3 55.6 54.8 59.7 41.2 44.4 55.0 55.5 46.1 57.3 17.8 48.5 56.7 26.8 22.1 31.6 29.4 30.5 40.3 14.0 11.1 15.3 15.9 16.6 19.9

NOTES a

Percentage marrying Australian-born spouse; b Percentage marrying spouse born in the same country. SOURCE Marriage statistics 1996–98.

intermarriage with the Australian-born. In contrast, people from Southern European or Middle Eastern countries are more likely to marry within their own community, even among the second generation (Price 1994). Among birthplace groups whose immigration has occurred mostly over the past 30 years, people from Vietnam have the lowest rate of intermarriage with Australian-born persons (Table 7.7). Less than 3 per cent of Vietnam-born men marrying in 1996–98 married Australian-born women, and less than 7 per cent of Vietnam-born women married Australian-born men. Other groups that also have

178 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

low intermarriage rates are those from China, Korea, Cambodia and Laos. People from these countries, especially the men, tend to find marriage partners from within their own community. Groups with higher rates of intermarriage with the Australian-born are those from Malaysia, Singapore, India and Sri Lanka. Between 30 and 50 per cent of men and women born in these countries who married in 1996–98 married Australian-born spouses. However, it is also possible that some people born in these countries are of Australian or British parentage because of the presence of expatriate communities of Australians and Britons in these countries in past years. Among people born in the Philippines, Thailand and Japan, women have a much higher intermarriage rate with the Australian-born than men. In the case of the Philippines, this reflects the sponsored migration of many Filipino women to Australia for marriage to Australian men. Intermarriage rates also continue to vary by ethnic origin among the second generation. Following the approach used by Price (1993, 1994), Table 7.8 shows that of those groups with a sizable second generation of adult age, those of Middle Eastern (Lebanese or Turkish) origins have the highest rate of in-marriage, followed by those of Southern European (Greek or Italian) origins; these are also the groups that are noted in the previous section as being slower to shift from the ethnic language to speaking only English at home. In contrast, men and women of Eastern European or Western European origins have much lower rates of in-marriage and higher rates of intermarriage. Many factors affect the rate of intermarriage. In the past the size of the ethnic community and the ratio of single men to single women of marriage age were important in determining the availability of marriage partners within the community. For example, if the birthplace group was small and there were more single men than women, this would encourage the men to marry local women. Nowadays, when it is simple to sponsor marriage partners from overseas, the availability of marriage partners of the same ethnic background in Australia is no longer a factor in determining the intermarriage rate with members of the local population. Immigrants who prefer a spouse from the same ethnic origin can and often do sponsor one from their country of origin (Khoo 1997; see also discussion on family migration in Chapter 6). Social, cultural and religious factors as well as geographic concentration are also likely to affect the rate of intermarriage (Price 1994).

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 179

Table 7.8 Second generation marrying spouses who are first or second generation of the same country of origin, 1996–98 Country of origin Lebanon Turkey Greece Italy Former Yugoslavia Malta United Kingdom Hungary Poland Baltic states Netherlands Germany

Brides

Grooms

% marrying within the groupa 74.1 61.1 65.3 61.0 57.1 56.3 45.6 42.7 38.9 38.7 23.1 23.2 13.1 12.5 7.3 7.3 7.0 7.7 6.3 8.5 6.3 5.8 4.4 4.3

NOTE a Percentage of Australian-born brides or grooms with mother born in the country of origin who married men or women who themselves were born in, or who have both parents born in the same country of origin. SOURCE Marriage statistics 1996–98.

There also appears to be a strong correlation between community language maintenance and the preference for intra-group marriage. Birthplace groups that have a low rate of intermarriage with the Australian-born population are also those that have a low proportion speaking English only at home. Of the five birthplace groups that have less than 60 per cent of their population speaking English well (DIMA 1997), four (Vietnam, China, Cambodia and Laos) also have the lowest intermarriage rates with the Australian-born population (as shown in Table 7.7). The correlation between the intermarriage rate and level of English proficiency is not surprising. Intermarriage is more likely to occur where there is interaction between the immigrant community and the wider Australian society and the ability to speak English is important in facilitating that interaction. The importance of intermarriage in facilitating integration with the wider community is underscored by data showing that the shift to speaking English only at home is greater among the second generation when one parent is overseas-born and the other is Australianborn than when both parents are overseas-born (Clyne and Kipp

180 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

1995; Khoo 1995). A much higher proportion of children who were born in Australia and had one overseas-born parent and one Australian-born parent were reported to speak only English at home than children whose parents were both born overseas. Therefore, the second generation of those groups that have higher intermarriage rates with the Australian-born will experience the shift to English much faster than those that have low intermarriage rates.

PROJECTIONS OF FUTURE ETHNIC COMPOSITION Projections of Australia’s population by birthplace have been published by the Department of Immigration and Multicultural Affairs (Khoo and Price 1996). These show that if net overseas migration is at 100 000 people per year, 79 per cent of the population in the year 2031 will be Australian-born and 21 per cent will be overseas-born. The proportion of overseas-born will be less than the present level of 23 per cent because net overseas migration which is fixed at 100 000 will make an increasingly smaller contribution to total population over time. The projections also show that if the birthplace origins of immigrants do not change from that observed during the early 1990s, the proportion of the population who are born in the UK and Ireland will decline to less than 4 per cent and people born in other European countries will also decline to less than 4 per cent. On the other hand, people born in Asia will increase to 9 per cent in 2031. The Australian-born component in these projections will include the Australian-born children of immigrants of recent years. The projections by birthplace do not indicate the ethnic origins of the second and later generations and therefore do not provide an indication of future ethnic composition. There are no official projections by race or ethnicity in Australia, unlike in the US, Canada or New Zealand, because of data unavailability. However, Price (1996) has projected the ethnic composition of Australia’s population to the year 2025 based on his estimate of ‘ethnic strength’. According to Price (1996), Australia’s population in 2025 will be 62 per cent of Anglo-Celtic origins, a decline from the present 70 per cent (Figure 7.5). A further 15 per cent will be of other European origins, giving a total of 77 per cent of European background. People of Asian background will constitute 16 per cent of the population and those of Middle Eastern origins will make up 4 per cent. These projections are based on the fertility and mortality assumptions

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 181

of the ABS population projections and assume that the net overseas migration patterns of the 1991–95 period will continue for the next 30 years. The projected percentages by ethnic strength do not translate directly into numbers of people by ethnic origin. This is because Price takes into account fractional contributions to each ethnic group resulting from intermarriage, meaning that a person with mixed ethnic origins will be counted in more than one ethnic group, according to his or her fractional contribution. Since Price (1994) has also predicted that people of mixed ethnic origins will be the largest group in Australia in the future, there will be many people who will be included in more than one ethnic group in 2025. Figure 7.5 Projected population by ethnicity, Australia and the United States, 2025 Australia 2025

USA 2025

1% 4% 2%

6% 13%

16%

18%

15%

SOURCES

1%

62%

62%

Anglo-Celtic

Middle East

White

Asia

Other European

Aboriginal

Hispanic

Native American

Asian

Other

Black

Price 1996; US Bureau of the Census 1996.

Australia is not the only country whose population is projected to be more ethnically diverse in the future. Projections of the US population by race and ethnicity also show that the percentage of people of Hispanic and Asian backgrounds will increase in the next 30 years while the percentage of non-Hispanic whites will decrease (US Bureau of the Census 1996). The percentage of the non-Hispanic

182 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

white population in the US is projected to decline from 74 per cent of the total population in 1995 to 62 per cent in 2025. Because of the Hispanic dominance of immigration to the US, people of Hispanic origins will make up 18 per cent of the US population in 2025, up from 10 per cent in 1995. People of Asian background are projected to increase from 3 per cent to 6 per cent of the US population in 2025. This is less than the 16 per cent projected for Australia’s population. A larger proportion of immigrants to Australia than to the US has been and is projected to be from Asian countries. Population projections for New Zealand also indicate that the percentage of people of Asian origins will increase in that country, from 5 per cent in 1996 to 9 per cent in 2016 (Statistics New Zealand 1999). It is important to note that categories such as European and Asian refer to people of diverse backgrounds in race, ethnicity, language and culture. It is also entirely possible that 30 years from now, such categories may not be adequate or relevant in describing the people of countries such as Australia and the US because of intermixing and changing perceptions of race and identity. The American population’s increasing diversity has led to a recent review of race and ethnic classification and a decision to allow for multiple racial identification. Future changes may call for another rethink of racial and ethnic classification. It is also important to point out that these projections are not predictions. They are usually based on a continuation of current demographic trends and assume no change in the birthplace origins of immigrants from those of recent years. This assumption can be unrealistic. Changes can occur, just as they have in the past decades, owing to policy changes and political and economic conditions in the world. However, projections are useful in giving an indication of what the future looks like if present trends continue.

CONCLUSION Since 1970, there has been net immigration of about 2.5 million people to Australia. The origins of these immigrants have become more diverse since the change in immigration policy in the late 1960s to one of non-discrimination on race or ethnic background. While before 1966 most immigrants had come from the UK or other European countries, since then many people from Asia, the Middle East, Africa and Oceania have settled in this country. As a result the

A G R E AT E R D I V E R S I T Y O F O R I G I N S

• 183

Australian population today has a greater diversity of origins than before, with more than 130 ethnic groups from all regions of the world (Jupp 2001). This transformation in Australia’s population has led to the replacement of a policy of immigrant assimilation with one that recognises the cultural diversity of Australian society and emphasises its positive aspects. The recognition of cultural diversity received a strong emphasis in the Federal Government’s 1989 National Agenda for a Multicultural Australia which encouraged institutions to be more responsive to the needs of a multicultural society and had a number of initiatives on improving immigrant access to government services. This access and equity approach to addressing cultural diversity, which received particular promotion during the first half of the 1990s, also led to some perceptions that it tended to favour minority groups over the majority. Such sentiments were evident among supporters of the One Nation political party that rose to prominence in the mid-1990s. The National Multicultural Advisory Council in its 1999 report had sought to address some of these concerns by emphasising the inclusive aspects of multiculturalism, the benefits of cultural diversity and the importance of promoting social harmony among people of different ethnic background. While the various community groups may differ from one another in race, language and religion, a more significant issue in terms of cultural and ethnic diversity is the existence of regional differences in population composition. The multi-ethnic population of the cities in contrast with the more homogeneous population of the regional areas can exacerbate the divisions between urban and rural areas that have arisen because of other social and economic differences. There have been attempts by both Federal and state governments to encourage immigrants to settle in regional areas, but these efforts have invariably been less than successful. There have been initiatives such as the Regional Sponsored Migration Scheme that enables employers in designated regional areas to sponsor immigrants; however, the number of immigrants going to regional areas is still small (Department of Immigration and Multicultural and Indigenous Affairs 2002). The extent to which cultural diversity will persist over the next 30 years will depend on a number of factors. There is no doubt that the proportion of the population that is of British, Irish or other European origins will decrease and the proportion of the population

184 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

that is of non-European origins will increase. On the basis of current demographic trends a large majority of the population in 2025 will still have Anglo-Celtic or other European ancestry; a large proportion is also likely to have mixed ethnic origins. However, changes in the level of net overseas migration, the main sources of future immigration and the rate of intermarriage can increase or decrease the extent of cultural diversity in the longer term. It is also clear that some immigrant groups, particularly those from Indochina, the Middle East or Southern Europe, are more likely than others to continue to maintain their ethnic and cultural identity. They have much higher rates of marriage within the group and relatively low rates of intermarriage with the Australian-born. However, among the second generation of Middle Eastern or Southern European origins, about a quarter and a half respectively of all men and women are now marrying outside the ethnic group. The extent of intermarriage and language shift among the second generation raises the question of how long minority languages and cultures can be maintained, unless the communities are continually being replenished by new arrivals. Many of the older Southern and Eastern European immigrant communities are now ageing and declining in size because immigration from these countries has declined or ceased altogether. The number of the second generation has exceeded the number of the first generation in many communities, and some are well into their third generation. In the forthcoming years, many of the non-European immigrant communities will also become older in their age structure. If immigration from these countries declines and there is another shift in the sources of immigration, the communities will undergo the same transition as the European ones are now experiencing. As the first generation dies out, so may the community language and cultural traditions. To the second and third generations, having grown up with the experience of ethnic diversity in Australia, concepts of race or ethnicity may become less important or relevant to their lives.

8

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N GRAEME HUGO

Demographers analyse the changing size, structure and spatial distribution of human populations. In Australia as elsewhere, however, the last of the three has attracted less attention from researchers and policy makers than the other two. This is despite the fact that the United Nations (1998, p. 10) has found in 20 years of surveying national governments that ‘population distribution has consistently been the area of most concern to Governments’. In 1996 only 27 per cent of nations indicated that their spatial distribution patterns were considered satisfactory, but population distribution remains the area least studied by demographers. Like population size and composition, however, the spatial distribution of populations is in a constant state of change and it affects, as well as being affected by, national and regional economic and social developments as well as demographic processes. In Australia, population distribution issues have figured among discussions about the national population but have not been subject to effective direct intervention by governments. This chapter examines the contemporary patterns of population distribution in Australia and the demographic processes shaping that distribution. Australia’s population distribution has been characterised as having an overall pattern of stability. Griffith Taylor (1947, p. 44) working over half a century ago contended that Australia’s future population distribution would see the population concentrated in the areas settled by 1860. In many ways he has been proven correct, but another feature of the nation’s population distribution

186 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

has been that within the fundamental pattern of spatial distribution along the lines suggested by Taylor, there is a considerable degree of dynamism and change. As Rowland (1982, p. 33) points out in discussing Taylor’s contention, the details of the Australian settlement pattern are changing while the overall structure is relatively fixed. It seems very unlikely that there will be any large-scale fundamental redistribution of the population of the nation within the foreseeable future, but there are important shifts occurring at the regional and local levels. The understanding of these changes is of basic significance for planning and needs to be examined.

P O P U L AT I O N D E N S I T Y Australia’s population is strongly concentrated, as shown in Figure 8.1, with a strong clustering in the east coast, south-east and southwest regions. Australia is one of the least densely populated countries in the world (2 persons per square kilometre) but it also has one of the most spatially concentrated populations as Figure 8.2 indicates. In 2001 some 84 per cent of the population lived within 50 kilometres of the coast. This uneven distribution has long been a point of debate in Australia (Rowland 1982, pp. 23–24) and raises a number of important policy issues in both the closely and sparsely settled areas. In the former, issues such as negative environmental impacts, overcrowding, and diseconomies in service provision abound, while in the latter, questions of economic and social viability and lack of access to services loom large. The density of population in Australia is one of the lowest in the world with Western Sahara, French Guinea and Namibia being the only nations with fewer people per square kilometre. The density is only a tenth that of all more developed countries (23 persons per square kilometres) and a 20th that of less developed countries (58 persons per square kilometres). The distribution of population in non-metropolitan areas as depicted in Figure 8.2 shows that there is a sharp decline in population density as one moves inland from the east, south-east and south-western coasts of the continent. The majority of the continent, however, has fewer than one person per square kilometre. Table 8.1 shows the stark comparison between the percentages of population and land in each of the population density categories depicted in Figure 8.2. Only 0.8 per cent of the population live in an area which has less than 0.1 person per square kilometre and this covers 70 per cent of the continent. On the other

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 187

hand, 76 per cent of Australians live in areas with 100 or more people per square kilometre but this covers around 0.3 per cent of the area. An interesting feature in non-metropolitan areas is that there is a strong positive relationship between population density on the one hand and the rate of population growth on the other. This relationship has been observed in the state of South Australia by Smailes (1996). One of the most distinctive features of the national population of Australia is the high degree of urbanisation, and it is relevant, also, to examine the pattern of population density within Australia’s major cities. These cities have low population densities compared with those of many other countries, although state and local government policy has attempted to increase the density of the built-up areas through urban consolidation programs. In the past there has been an almost inverse relationship between population density and population growth in Australian cities but as is explained later in this chapter, the pattern is now much more complex than this. Table 8.1 Australia’s population density, 2001 Persons per km2

% of population

Less than 0.1 0.1 – 1 1 – 10 10 – 100 100+

0.81 3.03 9.12 11.01 76.04

SOURCE

Number of people 153 574 1 728 2 087 14 418

778 752 610 031 042

% of land 70.54 20.18 7.97 0.98 0.33

Calculated from 2001 Census data.

P O P U L AT I O N D I S T R I B U T I O N I N N O N - M E T R O P O L I TA N A R E A S A feature of the uneven population distribution is the high degree of concentration in large urban areas. Australia has adopted 1000 persons as the minimum size of a settlement to qualify as an urban area and since the 1960s has adopted a population-density-based system to define urban areas (Hugo et al. 1997). The Australian Bureau of Statistics divides the nation into the following Sections of State:

188 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 8.1 Distribution of the total population, 2001

DARWIN

BRISBANE

PERTH SYDNEY CANBERRA

ADELAIDE

1 dot = 100 people

MELBOURNE

HOBART

SOURCE

2001 Census data.

Major Urban—urban centres with a population 100 000 and over. Other Urban—urban centres with a population of 1000 to 99 999. Bounded Rural Locality—population clusters of 200–999 population. Rural Balance—the remaining areas. Migratory—offshore, shipping and migratory persons. While it is difficult to make comparisons between censuses because of changing boundaries of urban places and population centres moving between categories, Table 8.2 shows the patterns of population change in each Section of State category over the 1954–96 period. Over the period up to 1971 there was an increase in the proportion of the total population living in major urban areas. There was a small decrease between 1971 and 1976 heralding the ‘turnaround’ trend or a sharp reversal of previous patterns of population concentration. Thereafter, comparisons are made more difficult because of boundary changes and

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 189

Figure 8.2 Population density, 2001

SOURCE

Calculated from 2001 Census data.

‘bracket creep’, but using 1991 boundaries there has been at least a stabilisation in the proportion of population living in cities with 100 000 or more residents. On the other hand there was a consistent pattern of decrease in the proportion of Australians living in rural areas up to 1971 and a decline in numbers of rural residents in some periods up to that year (Figure 8.3). However, since then there has been a stabilisation of the rural population at around 14 per cent and an increase in the number of rural dwellers. The late 1990s saw some controversy about non-metropolitan Australia with the success of the One Nation party in the 1998 Queensland elections. Since then considerable attention has been focused upon the economic and social situation of people living in rural and regional Australia. Analyses of population change in nonmetropolitan areas in Australia have shown that while overall population growth outside of Australia’s capital cities has exceeded that in capital cities there are some stark contrasts in population trends in

190 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

non-metropolitan Australia (Hugo 1994b). Structural change in the large metropolitan areas has seen manufacturing replaced with tertiary and quaternary industries, especially the knowledge-based, information and tourist industries, as the main employers in large cities. However these structural changes, together with the revolution in transport and communication technologies, shifts in lifestyle preferences, and the growth in the number of people whose income is based on transfers of various kinds which are not tied to living in a specific location, have made it possible for some parts of the nonmetropolitan sector to experience sustained population growth through in-migration. These areas are quite restricted in their distribution and are typified by one or more of a number of specific characteristics: proximity to large metropolitan areas; attractive, scenic environment (coastal, riverfront, mountainous area); some regional centres; areas of tourist potential; and some mining areas. However, other parts of the non-metropolitan zone, especially the wheat-sheep belt and pastoral zones, have continued to experience the rural depopulation, dominated by school leavers, which characterised the early post-war years. There are substantial areas in the better watered and more accessible parts of non-metropolitan Australia which are continuing to experience significant and sustained net in-migration and population growth. Hence population change in non-metropolitan Australia is likely to become more diverse and perhaps much less predictable in the next decade or so. Indeed it would appear that a divergence is occurring in non-metropolitan Australia between areas of population growth in the areas indicated above and of stability or decline in the dry farming and pastoral zones. There may be an increasing spatial polarisation occurring in non-metropolitan areas which is not only demographic but is similar to the social polarisation occurring in the major cities.

S TAT E – T E R R I TO R Y C O M PA R I S O N S There have also been some changes in the distribution of population between Australia’s states and territories. Table 8.3 shows that over the 1976–2001 period the proportion of Australians living in the south-eastern states (New South Wales, Victoria, South Australia and Tasmania) has decreased from 75 to 69 per cent. On the other hand, Queensland has increased its share of the national population from 15 to 19 per cent and Western Australia from 8 to 10 per cent. The

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 191

PERCENT

Figure 8.3 Australia’s rural population, 1911–96 50

50

40

40

30

30

20

20

10

10

0 1911

21

33

47

54

POPULATION NUMBERS

YEAR

61

66

71

76

81

86

91

96

0

SOURCE : Population Census Data.

2,600,000

2,600,000

2,500,000

2,500,000

2,000,000

2,000,000

1,500,000 1911 YEAR

SOURCES

1,500,000 21

33

47

54

61

66

71

76

81

86

91

96

Australian censuses, 1911–96.

territories have experienced a smaller increase in their share of the national population. Thus there has been a northward, and to a lesser extent, western shift in the centre of gravity of the Australian population distribution. However, these shifts have not seen a massive redistribution of the national population between states.

192 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 8.2 Population growth by section of state, 1954–96 Census year a. 1954 1961 1966 1971 1976 1981 1986 1991 1996 b. 1976 1981 1986 1991 1996 % change a. 1954–61 1961–66 1966–71 1971–76 1976–81 1981–86 1986–91 1991–96 b. 1976–81 1981–86 1986–91 1991–96

Metropolitan number %

Non-metropolitan number %

4 5 6 7 8 9 9 10 11

813 885 730 695 093 202 817 563 221

121 121 663 194 138 318 933 364 393

53.6 56.0 58.0 60.4 59.7 63.2 63.0 62.7 62.8

4 4 4 5 5 5 5 6 6

173 623 871 094 454 364 765 278 659

409 065 231 444 820 012 875 205 821

46.4 44.0 42.0 39.6 40.3 36.8 37.0 37.3 37.2

8 9 9 10 11

654 202 817 461 221

328 318 933 964 393

63.9 63.2 62.9 62.1 62.7

4 5 5 6 6

900 364 784 338 671

703 012 223 576 030

36.1 36.8 37.1 37.9 37.3

22.3 14.4 14.3 5.2 5.0 15.6 7.6 6.2

10.8 5.4 3.7 8.0 11.5 –5.2 8.9 6.1

6.3 6.7 6.6 6.2

9.5 7.8 10.4 4.4

a. Each section of state as defined in the report of each census; b. Based on the section of state as defined in the report of the 1981 Census. Non-major urban includes migratory population. SOURCES Bowie 1987; Australian censuses 1976–96. NOTES

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

Other urban number %

Rural number

%

Australia number

• 193

%

2 2 2 3 3 3 3 3 4

214 696 887 184 532 287 499 776 161

280 147 299 617 868 438 012 550 498

24.6 25.7 24.9 25.0 26.1 22.6 22.5 22.4 23.3

1 1 1 1 1 2 2 2 2

959 926 983 864 921 063 266 501 498

129 918 932 827 952 600 863 655 323

21.8 18.3 17.1 14.6 14.2 14.2 14.5 14.9 14.0

8 10 11 12 13 14 15 16 17

986 508 601 744 547 576 583 841 881

530 186 894 638 958 330 808 569 214

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

2 3 3 3 4

997 287 517 877 161

043 438 360 950 498

22.1 22.6 22.5 23.0 23.3

1 2 2 2 2

888 063 266 510 509

602 600 863 626 532

13.9 14.2 14.5 14.9 14.0

13 14 15 16 17

555 566 602 850 892

031 330 156 540 423

100.0 100.0 100.0 100.0 100.0

21.8 7.1 10.3 10.9 12.9 –12.2 7.9 10.2

–1.6 3.0 –6.0 3.1 9.0 8.2 10.4 –0.1

16.9 10.4 9.8 6.3 7.6 6.9 8.1 6.2

9.7 6.4 10.3 7.3

9.3 9.8 10.8 –0.04

7.5 7.1 8.0 6.2

194 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 8.3 Population of states and territories, 1971–2001 1971 NSW Vic Qld SA WA Tas NT ACT Australia NSW Vic Qld SA WA Tas NT ACT Australia SOURCES

4 3 2 1 1

960.8 811.4 091.7 274.6 178.9 412.4 98.3 207.4 14 035.7

1981 234.9 946.9 245.2 318.8 300.1 427.2 122.6 227.6 14 923.3

1986 ‘000s 5 531.5 4 160.9 2 624.6 1 382.6 1 459.0 446.5 154.4 258.9 16 018.4

898.7 420.4 961.0 446.6 636.1 466.8 165.5 289.3 17 284.0

204.7 560.2 338.7 474.3 765.3 474.4 181.8 308.3 18 310.7

6609.3 4822.7 3635.1 1514.9 1906.1 472.9 200.0 321.7 19 485.3

35.1 26.4 15.7 8.8 8.7 2.9 0.8 1.5 100.0

% of total 34.5 26.0 16.4 8.6 9.1 2.8 1.0 1.6 100.0

34.1 25.6 17.1 8.4 9.5 2.7 1.0 1.7 100.0

33.9 24.9 18.2 8.1 9.6 2.6 1.0 1.7 100.0

33.9 24.8 18.7 7.8 9.8 2.4 1.0 1.7 100.0

5 3 2 1 1

35.3 27.2 14.9 9.1 8.4 2.9 0.7 1.5 100.0

1991 5 4 2 1 1

1996 6 4 3 1 1

2001

ABS 1998c, 2002b.

P O P U L AT I O N D I S T R I B U T I O N A N D C E N T R A L I S AT I O N : A U S T R A L I A A N D T H E U N I T E D S TAT E S C O M PA R E D It is interesting to compare changes in population distribution in Australia and the United States. The period of European occupation of the United States has seen a substantial westward and, to a lesser extent, southward shift in the population centroid.1 There was a symbolic shift between 1970 and 1980 when the centre of population crossed the Mississippi River. In Australia, however, there have not been such substantial shifts in the relative population distribution across the continent. Figure 8.4 shows the shifts which have occurred during the 20th century in the centre of gravity of the Australian population. It will be noted that there has been very little change

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 195

over most of the period since Federation. There has been a slight movement to the west and north reflecting the fact that population growth rates in Western Australia and especially Queensland have been greater than in the south-east of the country in the last two decades of the 20th century. This reflects the pattern of stability in the population distribution although it must be explained that it is very much a ‘dynamic stability’ in that there are substantial flows of population. Indeed around 40 per cent of Australians move house over a five-year period and around 17 per cent move each year (Bell and Hugo 2000). However, the bulk of the movement is compensating so that net redistribution is limited. Figure 8.4 Shifts in the Australian population centroid, 1911–96

Queensland Brisbane

South Australia New South Wales 1996 1986 1981 Adelaide

1991 1947 1911

Sydney

Canberra

Victoria Melbourne

Statistical Division Boundary

Tasmania Hobart SOURCES

Australian censuses.

0 kms

250

500

196 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

I N T R A - U R B A N P O P U L AT I O N D I S T R I B U T I O N Population distribution within the major cities is changing to some degree. While it is an over-generalisation, the main pattern of population change in the post-war years in many industrialised countries, at least until the 1980s, was the classical ‘doughnut’ pattern with population decline in inner and middle suburbs grading to moderate population growth in the middle suburbs and rapid growth on the urban fringe. Population change in Australia’s major cities shows a different pattern. Certainly, areas of population growth are found on the expanding urban fringe but there is also growth in several inner suburbs and in a scattering of older inner and middle suburbs, especially along main transport routes and coastal areas. Both demand and supply factors are driving these changes. There is movement of young people into older established housing areas due to gentrification which has meant the movement of well-to-do, often two-income, couples into attractive older housing areas and inner and middle suburbs; urban consolidation activities of state, local and city governments which have caused development of land in established suburbs formerly occupied by factories, schools and other extensive uses developed for medium density housing; and the ageing of the massive cohort which moved into new housing in the 1950s and 1960s, with many old people dying or moving into specialised accommodation, which has led to unprecedented numbers of houses in the middle suburbs coming on to the housing market. This has offered possibilities for younger people to move in as individual house blocks or groups of them are redeveloped for housing. It has been argued that these trends will continue and in the case of the last will become more evident in the next decade or so. This will lead to vast areas of the middle suburbs developed in the 1950s and 1960s coming onto the land and housing market, increasing the opportunities for housing development in established areas to an unprecedented level and reducing pressure on the fringes of our cities. While to some extent our suburbs have had a degree of age homogeneity within them in the past, they are more diverse now and this will increasingly be the case in the future. This of course makes small-area population projections in cities even more problematical than in the past, but does raise the possibility of less dramatic swings over time in the demand for age-specific services like schools in suburbs, which characterised the later half of the 20th century.

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 197

EX-URBAN AREAS There has long been a recognition of a distinctive zone around large cities in developed countries, which is rural or non-metropolitan in appearance but contains many functions which are strongly associated with the nearby metropolis, and whose residents maintain strong regular, often daily, contacts with the metropolitan area (Spectorsky 1958; Friedmann and Miller 1965; Pahl 1965). This zone is distinctive in function and population composition but is also seen by some as being transitional between metropolitan and non-metropolitan areas. Others consider it to be part of the metropolitan region focused on the major city because of the strong functional links maintained with the city proper. This region has been referred to as ‘ex-urban’ (Davis 1990; McKenzie 1996), ‘peri-metropolitan’ (Burnley and Murphy 1995), ‘technoburbs’ (Fishman 1990) and ‘exurbia’ (Nelson and Dueker 1990). Burnley and Murphy (1995, p. 245) have described these areas as comprising urban centres set in a matrix of rural land where the traditional agricultural and associated service functions have been invaded by uses associated with the nearby metropolitan area. This includes the development of low-density residential areas for commuters to metropolitan areas and retired people. In the United States this has been recognised as a new settlement form, housing some 60 million people or a quarter of the total population (Nelson and Dueker 1990). In the US the term rural-residential has been used to describe new developments of housing of various densities of people predominantly commuting to the metropolitan area (but not necessarily the central business district or inner city), with some having hobby farms and others residential allotments only. Thus it is a distinctive landscape with particular land-use and planning issues and problems often arising from clashes of urban and rural functions and values. In Australia, Burnley and Murphy (1995) have argued that exurban areas are not as well developed as they are in the US because of the stronger planning system in Australia. Nevertheless, increasing attention is being focused on the ex-urban regions surrounding major cities in Australia. Maher and Stimson (1994) and McKenzie (1996), for example, have shown that these regions were the fastest growing in the nation in terms of expansion of population in the 1980s. O’Connor and Stimson (1996) show that these areas have been attracting increasing shares of dwelling construction investment as well as commercial construction investment. Undoubtedly these areas will increase in significance as a distinctive and important type of settlement in Australia over the next 20 years.

198 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

DISTRIBUTION OF SUBGROUPS I N T H E P O P U L AT I O N Thus far we have examined the changing distribution of the total population. However, it is important to recognise that there is some variation in the spatial distribution of some subgroups in the population. In order to indicate the differences which exist between subgroups in their overall distribution, Figure 8.5 shows the population centroids of a number of subgroups in the Australian population; it can be seen that, in fact, there are only small differences in the overall distribution. It will be noted, however, that the centroid of the indigenous population is in south-west Queensland while that of the total population is in central NSW. The total Australia-born population, as would be expected, is close to the total population but slightly north of it and the mainly English-speaking overseas-origin population is in a similar position. However, the non-English-speaking origin group is somewhat south of the total population reflecting their concentrations in Sydney and Melbourne. The map also depicts the centroids of various age categories in the population and there are only minor differences in evidence with older people having a centroid more south and east of the total population and children aged 0–14 years being distributed in almost exactly the same way as the total population. People who moved in the five years before the census have a slightly more northward orientation than is the case for the total population. These differences in population centroid only point to variations which exist in the distribution of various subgroups in the population. The differences are also evident in the Social Atlases prepared by ABS for Australian cities and the Social Atlas of Rural and Regional Australia (Haberkorn et al. 1999).

DEMOGRAPHIC PROCESSES INFLUENCING P O P U L AT I O N D I S T R I B U T I O N Changes in population distribution can occur through a number of demographic processes: natural increase levels (excess of births over deaths) varying between regions; internal migration whereby Australian residents leave some areas and concentrate in others; and international migration whereby arrivals from overseas concentrate in particular areas and adopt a settlement pattern different from that of the established population. All three elements have played a role in producing varying levels of population growth in different parts of

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 199

Figure 8.5 Population centroids for sub-groups in the population, 1996 Queensland Aboriginal & Torres Is. Brisbane

South Australia recent internal migrants

New South Wales

MES Born+ Australian Born Total Population + aged 65+ + + aged 75+ Overseas Born aged 0 - 14 yrs NES Born Adelaide

Sydney

Canberra

Victoria Melbourne

Statistical Division Boundary

Tasmania Hobart

SOURCE

0 kms

250

500

Calculated from 1996 Census data.

Australia and shifts in population distribution. There are variations in age structure, fertility and mortality between areas, which influence regional population growth although these effects are less substantial than the other two processes in shaping population distribution. There certainly are fertility and mortality differentials between groups and areas within Australia. These are influenced partly by the variations between areas in the composition of the population and the fact that mortality and fertility differentials exist between various groups in Australian society (see Jain 1992; Jain and McDonald 1997; Carmichael and McDonald, Chapter 3 of this book). As a result there are variations across Australia in levels of fertility. For

200 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

example, across Australian cities there are substantial differences in levels of fertility. Outer suburban areas tend to have higher levels of fertility than inner and middle suburbs even after standardising for differences in age structure. As Carmichael and McDonald shows in Chapter 3, rural-urban fertility differentials remain in evidence. Similarly it is possible to identify differences in mortality between particular parts of the country; however mortality variations between areas are generally smaller than fertility differences although there are some exceptions. Areas with a high proportion of their population of indigenous origin, for example, have significantly higher mortality than elsewhere (see Chapter 2 for a discussion of Indigenous mortality rates). While differences in natural increase levels between areas exist in Australia, the major demographic forces causing differences in population growth between areas relate to migration. First it is important to stress that international migration not only affects the size and rate of growth of the national population but also influences population distribution because overseas migrants to Australia have not settled across Australia in proportion to the existing distribution of the population. They have settled disproportionately in particular parts of the country, as noted in the previous chapter, and thereby influenced the distribution of the population. As discussed in Chapter 7, New South Wales, Victoria, Western Australia and the Australian Capital Territory have a higher proportion of overseas-born residents than the national average, while Tasmania is the least affected by immigrant settlement, and the Northern Territory also has a below-average presence of migrants. Most striking, however, is Queensland which, despite being by far the most rapidly growing state over the last two decades, has a significant ‘under’ representation of overseas-born people, indicating clearly that the bulk of that state’s rapid growth has been fuelled by interstate, rather than international, net migration gains. On the other hand, NSW and Victoria have been growing at well below the national average but they have continued to receive a disproportionate share of immigrants. This is due partly to Sydney and Melbourne being important ports of arrival of immigrants and also to many immigrants being chain migrants attracted by, and joining, settlers from their country of origin who moved into NSW and Victoria in the earlier post-World War II period. In South Australia, substantial industrial development in the 1950s and 1960s attracted a disproportionate share of immigrants but economic restructuring

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 201

and the decline of manufacturing over the last two decades has resulted in a much smaller share of immigrants settling there. One of the most distinctive features of post-war immigration to Australia has been the tendency for migrants to settle in the largest urban areas. Over the period 1947–96 the number of Australia-born persons living in cities with 100 000 or more inhabitants more than doubled so that in 1996, 58 per cent lived in such cities (Table 8.4). On the other hand, the overseas-born population in the largest urban areas increased more than six times so that by 1996, 80 per cent lived in those cities. Hence the effect of immigration has been felt more in the major cities than in the provincial cities or rural areas, as shown in Chapter 7. Over the 1947–96 period, the proportion of the population in cities with more than 100 000 inhabitants who were born overseas increased from 12 per cent to 29 per cent. The effect of immigration upon the growth of these cities is underestimated by these figures, since the children born in Australia to overseas-born people are included with the Australia-born. The proportion of the total overseas-born population living in provincial cities declined from 14 to 13 per cent over the 1947–96 period. However, the overseas-born in such cities increased almost fivefold so that the proportion of residents who were overseas-born increased from 7 to 12 per cent. In rural areas there was a substantial change. In 1947 a quarter of all overseas-born persons lived in rural areas but this was reduced to 7 per cent by 1996. Nevertheless the proportion of rural residents who were overseas-born increased from 8 to 12 per cent. Thus although the presence of overseas-born has increased in all three urban-rural sectors, the impact has been greatest in major urban areas. This contrasts with a great deal of preWorld War II settlement of immigrants of non-English-speaking origins, which was strongly focused upon rural areas (Borrie 1954). It is noticeable in Table 8.5 that among the overseas-born of mainly English-speaking origins, the deconcentration away from the major cities is occurring. This is consistent with a pattern of ‘counterurbanisation’ or decentralisation among the Australia-born that has been recognised for the last two decades (Hugo 1994b) and suggests that over time there may be some convergence in the internal migration patterns of the overseas-born toward those of the Australiaborn. Bell (1992) identified increased out-migration of longstanding overseas-born older people from major urban areas during the 1981–86 period.

202 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 8.4 Distribution of Australia-born and overseas-born population between major urban, other urban and rural areas, 1947–96 Australia-born 1947 1996

No.

%

No.

%

Major urban Other urban Rural

3 390 591 1 263 724 2 173 068

49.7 18.5 31.8

7 627 194 3 485 125 2 108 236

57.7 26.4 15.9

Totalª

6 827 383

100.0

13 220 555

100.0

NOTE

ª Excludes migratory.

SOURCE

ABS 1947 and 1996 censuses.

International migration has been of critical importance in the post-war growth of Sydney and Melbourne. Table 8.6 shows that over the first two post-war decades, more than half of the cities’ growth was attributable to net gains of overseas migrants, and that net gains of people from elsewhere in Australia were minor. Table 8.5 Number and percentage of overseas-born persons resident in capital cities by origin and length of residence, 1986 and 1996 1986 1996 Birthplace 0–5 Years 5+ Years 0–5 Years 5+ Years No. % No. % No. % No. % English181 747 76.8 877 266 73.0 119 614 75.6 944 892 70.6 speaking origin Non-English- 291 044 88.6 1 236 518 83.5 376 446 90.3 1 588 030 85.1 speaking origin Total 472 791 83.7 2 113 784 78.8 496 060 86.3 2 532 922 79.1 overseas-born SOURCE

ABS 1986 and 1996 censuses.

Moreover, it will be noted that the net gain in Melbourne was larger than in Sydney and indeed that overall growth in the southern city was greater. If we focus on the 1976–86 period, however, a

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 203

Overseas-born % change

1947

1996

1947–96

No.

%

+125.0 +175.8 -3.0

453 368 98 284 181 180

61.8 13.5 24.7

+93.6

733 372

100.0

No.

% change %

1947–96

3 126 263 489 550 290 275

80.0 12.5 7.4

+589.6 +395.4 +60.2

3 906 088

100.0

+432.6

different pattern is in evidence. Overall growth is substantially lower than in the first two post-war decades and natural increase (births minus deaths) is equivalent to almost all (99 per cent) of Melbourne’s growth, and 72 per cent of that of Sydney. However, if we disaggregate net migration into its international and internal components it is apparent that international migration has maintained a significant role (indeed, in the case of Sydney, an enhanced one) in the growth of the cities. This has been counterbalanced by a net outflow of the Australia-born population equivalent in size to half the overseas-born flow in Sydney and almost equal in size to that in Melbourne. The dominance of Melbourne in the early post-war decades both in population growth and in receiving overseas-born settlers had been reversed by 1976–86. In the late 1980s and early 1990s there was a further interesting development, Sydney’s overall population growth outpacing that of Melbourne. Sydney also retained its dominance of the overseas intake with a net gain of 123 000 over the 1986–91 period compared with 46 000 in Melbourne. In 1991–96 the comparative numbers were 173 000 and 102 000 and in 1996–2001, 142 000 and 84 000. Table 8.6 shows that the net internal migration loss in Sydney gathered pace while that in Melbourne declined somewhat. Over the 1996–2001 intercensal period the net increase in population in both cities, 273 000 in Sydney and 205 000 in Melbourne, was substantially higher than the previous intercensal period when the increase was 174 000 and 155 000 respectively. It is

204 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 8.6 Sydney and Melbourne: estimated components of population change, 1947–66, 1976–86, 1986–91, 1991–96 and 1996–2001

1947–66 Sydney Melbourne 1976–86 Sydney Melbourne 1986–91 Sydney Melbourne 1991–96 Sydney Melbourne 1996–2001 Sydney Melbourne

SOURCES

Natural increase

Total

Net migration Total International Internal population increase

‘000s % ‘000s %

379 45.3 366 42.7

457 54.7 491 57.3

441 52.7 485 56.6

17 2.0 6 0.7

836 100 857 100

‘000s % ‘000s %

237 72.0 205 98.6

92 28.0 3 1.4

184 55.9 91 43.8

–92 –28.0 –88 –42.3

329 100 208 100

‘000s % ‘000s %

144 70.9 119 102.6

59 29.1 –3 –2.6

123 60.6 46 39.7

–65 –32.0 –49 –42.4

203 100 116 100

‘000s % ‘000s %

135 77.6 109 70.3

39 22.4 46 29.7

173 99.4 102 65.8

–134 –77.0 –56 –36.1

174 100 155 100

‘000s % ‘000s %

149 54.6 109 53.2

124 45.4 96 46.8

142 52.0 84 41.0

–18 –6.6 12 5.9

273 100 205 100

Hugo 1989, p. 68; ABS 1990b, p. 10, unpublished data.

noticeable that there was a dramatic reduction in net internal migration loss in both Sydney and Melbourne. Indeed, Melbourne experienced a small net gain. This appears to represent a substantial change. The latest intercensal period data, however, were calculated before the 2001 Census data were available so that they need to be checked against the census. Despite the visible shifts of the latest intercensal period, it is

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 205

apparent that over much of the last three decades there has been a ‘switch-over function’ (Maher and McKay 1986) in Sydney and Melbourne, whereby a net loss of migrants in exchange with other parts of Australia is more than counterbalanced by an inflow of overseas migrants. This has been an important feature of these two cities in the post-war period and part of the phenomenon of the ‘turnaround’ in Australia (Hugo 1989). A key point is that net international migration gains have directly accounted for more than half of Sydney and Melbourne’s net population growth over the post-war period, and if their indirect contribution through the children born to migrants since settling in Australia is taken into account, that contribution is closer to two-thirds of net growth. Figure 8.6 shows the growth of Sydney and Melbourne’s population over the post-war period and parts of that growth which have been made up of the overseas-born. It can be seen that the overseasborn have increased faster than the total population. In Melbourne they doubled between 1947 and 1954 and almost doubled again between 1954 and 1961, while the total population increased from 1.2 million to 1.9 million. Between 1961 and 2001, the overseas-born population more than doubled while the total population increased to 3.4 million. In Sydney the growth of the overseas-born over the 1947–61 period was somewhat slower than in Melbourne with an increase of 133 per cent while the total population increased by 47 per cent. However, in the 1961–2001 period Sydney’s overseas-born population increased by 184 per cent compared with 115 per cent in Melbourne. Sydney’s total population increased by 83 per cent. Figure 8.6 shows that the growth of the overseas-born from nonEnglish-speaking countries is especially striking. In Melbourne there was an almost fivefold increase between 1947 and 1954, a more than doubling between 1954 and 1966, and a 98 per cent increase between 1966 and 2001. Again in Sydney the growth was a little less rapid initially with an increase of 269 per cent between 1947 and 1954, and 149 per cent between 1954 and 1966. However, between 1966 and 2001, the increase of 202 per cent was more than twice as rapid as that in Melbourne.

206 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 8.6 Growth of total population, and total and non-English-speaking overseas-born population, Sydney and Melbourne, 1911–2001 4,000,000

SYDNEY

Total

3,500,000 3,000,000 2,500,000

POPULATION NUMBERS

2,000,000 1,500,000 Overseas Born 1,000,000 500,000 Overseas Born (NES)

0 1911

33

47

54

61 66 71 76 81 86 91 96 ‘01

CENSUS YEAR 4,000,000

MELBOURNE

3,500,000

Total

3,000,000 2,500,000

POPULATION NUMBERS

2,000,000 1,500,000 Overseas Born

1,000,000 500,000

Overseas Born (NES)

0 1911

33

CENSUS YEAR

SOURCES

Australian censuses.

47

54

61 66 71 76 81 86 91 96 ‘01

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 207

Table 8.7 Sydney and Melbourne Statistical Divisions: Proportion of population overseasborn, 1947–2001 Sydney Statistical Melbourne Statistical All Australia Division Division No. of % of all No. of % of all No. of overseasoverseasoverseasoverseasoverseasborn born born born born 1947 191 107 25.7 125 258 16.8 744 187 1954 308 778 24.0 261 470 20.3 1 286 466 1961 434 663 24.4 444 479 25.0 1 778 780 1966 558 236 26.2 568 365 26.7 2 130 920 1971 681 313 26.4 687 266 26.6 2 579 318 1976 736 754 27.1 706 331 26.0 2 718 855 1981 834 280 27.8 754 117 25.1 3 003 833 1986 912 578 28.1 788 266 24.3 3 247 381 1991 1 070 627 28.5 893 445 23.8 3 755 554 1996 1 148 869 29.4 915 449 23.4 3 908 213 2001 1 233 487 30.0 954 037 23.2 4 105 444 SOURCE

ABS Australian censuses.

Table 8.7 shows the growth of the overseas-born population in the two cities between 1947 and 2001. While Sydney gained huge numbers of immigrants during the long boom period and its overseas-born population more than doubled between 1947 and 1961, the effect was less than had occurred in Melbourne. The table shows the significance of this immigration with Melbourne’s overseas-born population trebling between 1947 and 1966, and its share of the nation’s total overseas-born increasing by 10 percentage points to 27 per cent. By 1961, Melbourne had surpassed Sydney in having the largest overseas-born community in the nation, but in the last two decades Sydney has reasserted itself as the major focus of immigrant settlement in Australia, so that at the 2001 Census it had 30 per cent of the nation’s overseas-born compared with Melbourne’s 23 per cent. These fluctuations have been in concert with shifts in the changing roles of the two cities. The major demographic process causing differences between areas within Australia in relation to population growth, however, is internal migration. Figure 8.7 shows migration between states over the 1996–2001 period and it is clear that Queensland has dominated the

208 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

pattern of net gains. It is apparent from Table 8.8 that interstate migration has varied in its importance to the growth of population at the state level. It has clearly been a major factor in the growth of Queensland and, to a lesser extent, Western Australia. It has also contributed to the slow growth in South Australia and Tasmania and dispersed the growth that occurred in NSW and Victoria owing to net gains of overseas migrants. Figure 8.7 Migration between Australian states 1996–2001

SOURCES

1996 and 2001 censuses.

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 209

Table 8.8 Natural increase, net overseas migration, net interstate migration and total population growth,Australian states and territories, financial years 1996–2001 State

Natural increase

Net overseas migration

Net interstate migration

NSW Vic Qld SA WA Tas NT ACT

No. 244 414 166 298 149 510 39 745 84 107 14 184 16 662 17 510

% 60.9 53.6 41.0 118.9 47.6 385.1 87.4 199.7

No. 243 869 141 572 88 129 19 621 79 144 1 550 4 172 –453

% 60.8 45.6 24.2 58.7 44.8 42.1 21.9 –5.2

No. –86 925 2 332 126 659 –25 950 13 361 –19 417 –1 773 –8 287

% –21.7 0.8 34.8 –77.7 7.6 –527.2 –9.3 –94.5

Australiaa

732 649

56.0

576 221

44.0

—

—

NOTE a SOURCE

Total population growth No. 401 358 310 202 364 298 33 416 176 612 –3 683 19 061 8 770 1 308 870

Includes other Territories. ABS 2002b.

In non-metropolitan Australia, net gains of internal migrants had been restricted to a few areas: coastal localities, especially along the east coast and in the south-west; areas around major cities—or exurban areas; tourist and retirement destination areas; and a few regional centres. On the other hand, the dry farming and pastoral areas are experiencing net loss of population. For the greater Sydney area, net migration gains are still being experienced mostly in the outer suburban area although there are smaller net gains being experienced in some inner and middle suburbs, reflecting the effects of residential in-fill, gentrification and urban consolidation.

I S S U E S R E G A R D I N G P O P U L AT I O N DISTRIBUTION L O C AT I O N A L D I S A DVA N TAG E There has been in the 1990s considerable debate in Australia about the extent to which people living in certain areas suffer disadvantage purely as a result of their location. Studies have suggested that cities are becoming more polarised than before and persons living in some areas suffer low levels of accessibility to goods, services, employment and utilities. This applies both within large cities (Maher et al. 1992)

210 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

and in non-metropolitan Australia. Figure 8.8 shows the distribution of the ARIA (Accessibility/Remoteness Index of Australia) index values. This index is a generic measure of remoteness across all nonmetropolitan parts of Australia. The measure was developed using a Geographical Information System to calculate an index of remoteness at each of over 11 000 populated places around Australia. This involved measuring the road distance of each point to the nearest centres in each of four levels of the urban hierarchy, and developing an aggregate measure. The scores are divided into five categories with the lightest shading indicating the highest degrees of accessibility and the darkest shading the most remote areas. The map shows the increasing levels of remoteness the farther one moves away from the major capital cities. The high accessibility of the south-eastern part of Australia and areas around Adelaide, Perth, and to a lesser extent, Darwin and Tasmania, is evident. The bulk of the dry farming, wheat-sheep areas come into the intermediate levels of remoteness categories while virtually all of the rangelands are in the most remote category. A crucial question here is to establish the extent to which population redistribution is leading to increasing polarisation of rich and poor areas and populations within large urban areas or within nonmetropolitan areas. To what extent is there a migration of the poor which is contributing to a dichotomisation of prosperous and unprosperous areas in both rural and urban areas?

T E M P O R A RY M I G R AT I O N Much of our analysis of population relates to that of the population according to their usual place of residence. However, Australia is an extremely mobile population. Not only do 43 per cent move their place of residence during a five-year period but there is an increasing amount of non-permanent movement. On the night of the 1996 Census, some 972 780 people or 5.5 per cent of the total population were at a location other than their usual place of residence (Bell and Ward 1998). There has also been an exponential increase in the extent to which overseas visitors have entered Australia. What is clear is that temporary movers within Australia and overseas short- and long-term visitors are not distributed in the same way as the population enumerated at the place of usual residence. Thus it is possible to conceive of a number of different spatial distributions of the Australian population: day time versus night time, summer versus

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 211

Figure 8.8 ARIA values interpolated to 1 km grid, with contours

SOURCE

GISCA.

winter, school holidays versus non-school holidays, working week versus weekend. Each of these different distributions will create different levels and types of demand for services, impacts on the environment, and so on. However, our data sets and conceptual understanding limit our ability to measure these different distributions and assess their effect.

D I S T R I B U T I O N E F F E C T S O F I N T E R - R E L AT I O N S H I P S B E T W E E N I N T E R N A L A N D I N T E R N AT I O N A L M I G R AT I O N An important area of research relates to the inter-relationships between internal and international migration. In the US, Frey (1995) has established a strong correlation between large-scale immigration

212 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

into cities such as Los Angeles and an exodus of native-born and long-standing local populations. He has partly attributed this to a new form of ‘white flight’. In Australia, Figure 8.9 would suggest that there is a similar inverse relationship between the two types of migration. But it has been argued that this is due more to structural change in the economy, which has seen an outflow of former bluecollar and clerical workers who have been displaced after decades of employment often with redundancy packages, and after capitalising on their homes in Sydney (Hugo 1996). On the other hand, the incoming migrants tend to have different skills in service, industry, information technology, and so on and, unlike the people leaving, are prepared to work in non-maximised, fractional, non-standard-hours situations. The extent and nature of this relocating needs to be researched along with its effects on population distribution.

U R B A N C O N S O L I DAT I O N Local and state governments in Australia’s major cities have embraced urban consolidation policies for the last decade or so. These have led to an infilling and ‘thickening’ of population distribution within built-up areas of the major cities. The extent to which such initiatives will cause an increasing proportion of new households in the cities to choose an established rather than a new dwelling on the periphery of the metropolitan area is an important issue over the next decade or so. Complicating this issue is the fact that an unprecedented number of houses are going to come onto the market in Australian cities from elderly people dying or moving to appropriate accommodation. An additional factor here is the bulge of the postwar baby boom generation, now in their 40s and 50s and entering the empty nest stage of the lifecycle. Some have suggested that the housing preferences of the baby boomers in their 50s and 60s may differ significantly from those of earlier generations in these age groups. For example, the baby boomers entering the empty-nest stage may be forsaking family houses in the outer suburbs for smaller units in the inner and coastal suburbs of the major cities (Hugo 1999).

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 213

Figure 8.9 Net internal and international migration, Sydney Statistical Division, 1971–2000 60 55

Overseas migration

50 45 40 35 30 25 20

NUMBER (thousands)

15 10 5 0 1971/72 1975/76 -5

1980/81

1985/86

1990/91

-10 -15 -20 -25 -30 -35 -40 -45 -50 -55 -60

SOURCE

Internal migration

NSW Department of Urban Affairs and Planning.

1995/96

214 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

REGIONAL DEVELOPMENT Not since the 1970s has there been such an interest in development in non-metropolitan areas of Australia. The alienation from major political parties in recent state and Federal elections has drawn attention anew to the ‘third face’ of the Australian political scene: regional governments. Regional development has been given a new prominence. And population distribution will be influenced by the extent to which this translates into policies and programs which arrest rural depopulation, enhance local potential in non-metropolitan areas, result in reversing the reduction of services in isolated areas, and harness new developments such as information technology. Certainly there is a favourable political context for such developments and some would argue that modern developments in information technology and transport as well as structural change in the economy have greatly reduced the necessity for enterprises to be located in major cities like Sydney and Melbourne. E N V I R O N M E N TA L I S S U E S The nature of the relationship between population and the environment has been the subject of a number of government and other high-level inquiries (National Population Council 1991, 1992). However, there remains little understanding of this complex and important relationship beyond the fact that environmental degradation is a function not only of population growth but also of resources consumed per capita and the technology used in the exploitation of the environment. We know little of the effects of changing population distribution on the environment. To what extent is the continued growth of Sydney contributing to, for example, increased air and water pollution and congestion? (NPC 1991). To what extent are ex-urban development and the development of low-density communities in non-metropolitan south-west Queensland and the north coast of NSW leading to increased fuel consumption and alienation of scarce, high-quality agricultural land? To what extent is tourism, national and international, producing pressure on fragile environments? To what extent is there a mismatch between the distribution of the Australian population on the one hand and the distribution of exploitable water reserves in the cities? We have not begun to address the complex environmental impacts of current patterns of population distribution or of possible future population distributions.

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 215

P O L I C I E S I N F L U E N C I N G P O P U L AT I O N DISTRIBUTION Debates about Australia’s population distribution have a long history, as do the policies to ‘decentralise’ the nation’s population (e.g. Neutze 1963). However, for much of the period since Federation, decentralisation has been ‘everyone’s policy but nobody’s program’. Lip service has been given to the concept but, with minor exceptions (e.g. during the early 1970s), there has been little follow-up with programs to operationalise it. In an era where development of communication systems has greatly reduced the need for business and people to be located in major urban areas, and decentralisation is occurring in other Euro-American societies, this issue needs revisiting. In earlier periods it was argued that economies of scale and agglomeration militated against regional development, and there was an ever-increasing concentration of economic activity in larger urban centres. However, the situation has changed dramatically. First, structural change in the economy has seen the focus of employment shift away from manufacturing to tertiary and quaternary activities. For many of the latter activities, location in a major city is not mandatory. Growing industries like tourism and information technology are frequently locatable outside of large cities, or relatively footloose. Developments in information technology have released the bonds tying many activities to downtown locations and opened up the options for location. Moreover, in some large city areas, especially Sydney, the diseconomies of large city locations in terms of high costs of operation, congestion and pollution are being felt. Accordingly, the current era is quite different from that prevailing in the 1970s when the Australian Government made a major attempt to foster regional development. Moreover, the recent backlash against major parties, especially those in power in regional areas, has revealed a political environment favourable to regional development. There is also strong evidence of regional areas becoming more organised in creating a ‘third force’ in Australian politics. There are, therefore, likely to be more sustained efforts than in the past in the area of development outside of Australian major cities. This development would basically include focusing more activity on the peripheral states some of which have experienced slow economic growth in the 1990s; and encouragement of development in selected parts of the non-metropolitan area which have the potential for sustainable economic activity.

216 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The above raises the question as to whether efforts currently being expended in trying to attract newly arrived immigrants to areas of Australia perceived to have too low population growth (Department of Immigration and Multicultural Affairs 2001) may be more productively directed at the Australian resident population in areas perceived to be experiencing pressures of population such as diseconomies of scale, environmental pollution, spiralling land and home costs, congestion and accelerating overhead costs. It can also be legitimately questioned that states and regions need to have population growth if they are to become prosperous. But given that some areas wish to reverse current net migration losses or increase net migration gains, it may be more productive to attempt to attract established Australians from elsewhere than to focus only on newly arrived migrants. To take the case of South Australia, for example, one could make the following argument for adopting this strategy: other Australians are much more likely to have information about South Australia than newly arrived migrants; some will have visited Adelaide and may even have originated from South Australia or have relatives and friends there. They are more likely to be aware of the advantages of living in South Australia such as lower house and land prices, lower operating costs for companies, less congestion and life style advantages; and the development of modern transportable information technology is making it less necessary for businesses to be physically located in downtown Sydney or Melbourne to interact effectively with others located in those cities. Thus a program aimed at attracting particular groups, especially small and medium scale entrepreneurs from areas in the eastern states experiencing some stress, may be more effective in increasing net migration gains if this is considered necessary. There is support from the experience of contemporary North America and Europe where there has been substantial decentralisation of people and economic enterprises away from large cities. One aspect of regional development relates to the provision of basic services to people living in isolated areas. Over most of the dry farming and pastoral areas of Australia, population has been either static or declining and there has been a withdrawal of services, often in association with rationalisation, full cost recovery and privatisation activities in the public and private sectors. There is a growing awareness by policy makers that Australians living in rural areas should have reasonable access to a range of services. Policies such as those to

C H A N G I N G PAT T E R N S O F P O P U L AT I O N D I S T R I B U T I O N

• 217

subsidise general practitioners in remote and unattractive areas may become an increasingly important part of maintaining the economic and social viability of parts of the nation.

CONCLUSION It is argued here that while Australia’s population distribution in macro terms conforms to the broad patterns of coastal concentration established more than a century ago, a number of dynamic elements are altering the population distribution within this matrix. While the distribution is unlikely to undergo a massive change over the next few decades, there are likely to be important changes within this structure which are going to influence the lives of many Australians. Population distribution can be influenced by policy both directly and indirectly, and policy makers have to be aware both that they can influence population distribution and that many of their actions actually do influence population distribution even though this may not be their intention. The Australian economy and society are undergoing transitions with structural change, globalisation, the effect of the revolution in information technology and the ageing of the population. The implications for the distribution of population are unclear and need to be investigated. There has never been greater potential for analysing changing population distribution patterns and their implications. The development of spatial information system methodology and technology, the revolution in the capacity of computers and the explosion of data availability at the small area level not only from the Australian Bureau of Statistics but from administrative data sources, provide us with an unprecedented opportunity to investigate and answer many of the pressing questions regarding population distribution. However, there do remain some problems. We have not yet agreed to geocoding of our census and other major data collections which would allow researchers and policy makers to fit population data to the regions and areas most relevant to them. The Australian Standard Geographical Classification (ASGC) needs a major revision so that the basic spatial units used to collect and disseminate information have economic and social reality in the 21st century (ABS 1996, 1997; Hugo et al. 1997). In fact, one could argue that the spatial dimension of data collection in this country is not given a high enough priority and so the availability of information for small area, local and regional planning is not as advanced as it is in many other industrialised nations.

218 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

A number of important forces will impinge on Australia’s population distribution in the future. Demographically, the scale and composition of international migration, the ageing of the population (especially the Baby Boom generation) and changing patterns of family and household formation will have an influence. Internal migration will be shaped by shifts in economic structure, changes in tastes and preferences for different lifestyles, political developments and the advance of information technology. It is crucial that these developments be monitored and their implications for environmental degradation, equity, economic development and the well-being of all groups of Australians are established.

9

T R A N S F O R M AT I O N S I N T H E LABOUR FORCE1 B R U C E C H A P M A N A N D C E Z A RY A . K A P U S C I N S K I

Projections of the Australian population have been regularly produced by the Australian Bureau of Statistics (ABS) since the late 1970s (see Adam 1992). Labour force participation is one of the most stable economic indicators and this has encouraged the ABS to provide projections in this area. The first such projections were published in 1992 (ABS 1992) with the most recent set covering the period from 1999 to 2016 (ABS 1999c). These demographic projections have been used to alert researchers, policy-makers and the wider community to fundamental changes in the structure of the population and the labour force, due to the ageing of the population. However, almost nothing has been written about the likely future profile of either employment or unemployment. While aggregate macro-economic models such as the Treasury’s TRYM model, or general equilibrium models such as the MONASH model, provide some indications of short-term trends, or future deviations from extrapolated current trends, they usually only span several quarters and rarely go beyond a decade. The story of unemployment cannot be studied in isolation, and in particular, without consideration of employment growth. While the analysis of population and labour force provides a picture of longterm trends in economic activity of the population, examination of employment and unemployment is much more important from a policy perspective since it deals with the behaviour of the economically active subset of the population. Demand-side policies include a number of instruments available

220 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

to the government, such as interest rates and labour market training programs, but the range of instruments relevant to population dynamics is much weaker: for example, the size and composition of immigration. The small size of the Australian economy, its internationally open character and the importance of the agricultural sector are all factors which are beyond the control of the government but which nevertheless have an influential role in the shaping of the business cycle in Australia. With this acknowledged, there remains room for counter-cyclical policies which affect employment growth and thus unemployment. This chapter could have been concerned solely with the implications of alternative population projections on the profile of the future labour force, but it is much more interesting to consider also employment and unemployment. To do this properly we use the identities which link labour market components and a number of assumptions regarding the behaviour of employment growth to derive alternative profiles of future unemployment. These profiles are not statistical forecasts of future employment but they nevertheless have great value as comparisons of potential outcomes in the labour market. The chapter begins by explaining the basic relationship of population to the labour force and employment, then describes how unemployment fits into the picture, and explores the mechanics behind the generation of both population and labour force projections. The profile of the labour force over the last three decades is also considered. After this we turn to employment growth over the last three decades and ask what happens to the employment profile as a result of just a couple of poor years of growth. Finally, we use a number of simple scenarios of future employment growth, derived from the outcomes of past cycles, to investigate the significance of employment on unemployment outcomes.

METHODOLOGICAL ISSUES It is useful to begin with some basic definitions. First, the analysis of labour force issues is usually restricted to the adult population, that is, persons aged 15 and above. The labour force is defined as the sum of all people who are either employed or unemployed. To be employed a person must work in the market place, that is, receive income for work; and a person is classified as unemployed if he or she is not employed and is both available and looking for work.2

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 221

To provide indicators of movements in the labour force, economists usually quote relative measures of the sizes of the appropriate segments of the labour market. Thus, the labour force participation rate is the proportion of the adult population which is in the labour force while the unemployment rate is the proportion of the labour force which is in the state of unemployment. A different summary of the labour market is the employment-population ratio which provides a picture of the labour market which is independent of participation. Reliance on a single measure is ill-advised since it provides an incomplete picture of developments in the labour market. For example, during early periods of recovery even rapid job creation may have little impact on unemployment if labour force participation rates increase in response to job growth. Because of these and other issues a combination of measures is presented in this chapter. The population projections produced by the ABS are based on the cohort component method, which in essence ages the population year by year and applies pre-specified fertility and mortality rates to each cohort (subset) of the population. Traditionally, these cohorts are single-year sex-specific subpopulations. In addition, the cohort method accounts for overseas migration by using pre-specified rates of immigration and emigration. The procedure is relatively simple: the main area of discussion is the specification of assumptions regarding fertility, mortality and immigration. For example, the latest set of ABS population projections includes population scenarios which span fertility rates from 1.6 to 1.75 births per woman. Presentation of a number of scenarios provides an indication of the sensitivity of projected trends of population in relation to the underlying assumptions (see Chapter 11 for a discussion of alternative future population scenarios). The generation of labour force participation rates and, consequently, labour force levels starts with evaluation of the past trends of labour force participation for sex and age group cohorts. Simple linear or logistic models are then fitted to the data and used to extrapolate participation rates into the future (see ABS 1999c). Since such modelling and derivation of participation rates do not incorporate socio-economic factors, they are limited to providing only very broad trends in labour force participation rates, but do provide a number of variants of the labour force (by multiplying the extrapolated participation rates by the population projections) determined by underlying population projections.

222 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

L A B O U R F O R C E A N D E M P L OY M E N T : A N H I S TO R I C A L OV E RV I E W Modern statistical measurement of the labour markets began in Australia in 1964 with regular quarterly surveys of labour markets in the capital cities. Later these were extended to cover the whole country and in 1978 were transformed into monthly labour force surveys. Many facets of labour market measurement had not been standardised until the onset of the monthly surveys. In particular, the measurement of long-term unemployment (that is, of people who have been unemployed for 12 months or longer) was introduced only in February 1978. Figure 9.1 plots the size of the Australian population over the last three decades with labour force participation. It is evident from this figure that population has been increasing at a very steady rate since 1964. On the other hand, participation rates have varied somewhat over the same period, ranging from 58 per cent at the beginning of the sample to 64 per cent in 1990. Population changes are essentially due to the long-term secular movements in fertility and mortality rates, and international migration, while short-term cyclical vagaries of the economic business cycle influence the labour force participation rate.

65

16,000

64

14,400

63

12,800

62

11,200

61

9,600

60

8,000

59

6,400

58

4,800

57

3,200

56

1,600

55

'000 persons

Per cent

Figure 9.1 Population and labour force participation, 1964–99

0 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 Labour force participation rate (%)

Population (persons)

ABS, Labour Force, Australia, Catalogue No. 6203.0, various issues; Reserve Bank of Australia, Australian Economic Statistics, 1964–78, Sydney.

SOURCES

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 223

Labour force participation has also exhibited a relatively stable long-term trend. Increased entry into the labour force by women has more than compensated for the decrease in labour force participation by younger males because of longer participation in education, and older males through the trend towards early retirement. The longerterm perspective adopted in what follows allows us to separate cyclical changes from underlying long-term movements in population and its components. Figure 9.2 presents the relative sizes and trends of labour force and employment since 1964. The upward trend in labour force participation, and population growth, are the reasons that the size of the labour force has also been rising over this period, on average by about 2 per cent per year. A somewhat different picture emerges with respect to employment: first, employment is much more variable than the labour force with distinct troughs visible at major recessions such as those of 1981–82 and 1990–92. Second, despite significant employment growth after each economic recession since the first oil shock of 1974, employment has become increasingly different from the labour force. The difference between the two is, of course, unemployment, which has been high since the mid-1970s and continues to be one of Australia’s major policy challenges. Figure 9.2 Labour force and employment, 1964–99 10,000 9,000

'000 persons

8,000 7,000 6,000 5,000 4,000 1964

1967

1970

1973

1976

1979

Labour force SOURCES

As for Figure 9.1.

1982

1985

1988

1991

Employment

1994

1997

224 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

It is worthwhile at this point to consider more closely the growth of employment, as illustrated in Figure 9.3. By comparing employment growth with the growth of the Gross Domestic Product (GDP), the acknowledged yardstick of the performance of an economy, we can relate developments in employment to wider changes in the economy. Thus, as shown in Figure 9.3, employment exhibits a very cyclical pattern of growth with negative rates during recessions, that is, when GDP exhibits negative growth rates, and positive rates over the remainder of the business cycles. Employment is dependent on economic conditions (aggregate demand in the economy) rather than on the labour force (aggregate supply). Figure 9.3 Employment growth and GDP growth rate, 1965–99 6 5 4

Per cent

3 2 1 0 -1 -2 -3 1964

1967

1970

1973

1976

1979

1982

Employment growth

1985

1988

1991

1994

1997

GDP growth

NOTE GDP growth rates are calculated as quarter-to-quarter changes with June values plotted in the graph. SOURCES Employment as in Figure 9.1; GDP from ABS, National Income and Expenditure, Australia, Catalogue No. 5206.0 (various issues).

Over the last three decades, two years, 1983 and 1991, are significant for the contraction of the stock of jobs. The number of jobs fell respectively by 2.1 and 2.3 per cent, with an annual average growth of employment over the sample period of 1.9 per cent. Such large declines in the employment stock are rare events and thus provide a clue to the possible importance of counter-cyclical policies. They are also pivotal for the design of our counterfactual simulations, presented below.

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 225

To put these outcomes in perspective, we can compare average growth rates of employment with average growth in the labour force. Figure 9.3 suggests that, apart from the very few periods of negative growth, the majority of years in the 1964 to 1999 period looked reasonably healthy. However, as Figure 9.2 implies and Table 9.1 illustrates, the average growth of employment since the mid-1960s has been marginally less than the average growth of the labour force: the economy has not been able to generate enough jobs to stop the rate from rising. Table 9.1 Average annual growth of employment and labour force, 1964 to 1999 Period Employment growth (%) Labour force growth (%) 1960–1969 3.12 3.17 1970–1979 1.54 2.03 1980–1989 2.22 2.35 1990–1999 1.27 1.31 1964–1999 1.93 2.13 SOURCES Authors’ calculations based on data from ABS, Labour Force, Australia Catalogue No. 6203.0 (various issues); Reserve Bank of Australia, Australian Economic Statistics, 1964–1978, Sydney.

The above relationships and outcomes are reflected in Figure 9.4, which provides a picture of Australia’s unemployment history over the last three decades. The most pronounced aspect of this graph is the substantial and rapid rise in unemployment at the onset of the recessions of the mid-1970s and the early 1980s and 1990s. There is, however, a noticeable lack of symmetry: during the expansionary phases of the business cycle the unemployment rate has declined very slowly and has not come down fully to the pre-recession level. That is, increasing unemployment seems to be associated with more pronounced difficulties associated with its reduction, a phenomenon known as hysteresis. To illustrate how unemployment has increased over the period it is useful to look at simple period averages: during the 1960s the unemployment rate averaged 1.5 per cent; in the 1970s the average was 4.2 per cent; in the 1980s, 8.1 per cent and in the 1990s, 9.6 per cent. In fact, the early 1990s was a consistent period of double-digit unemployment rates with 11.6 per cent being recorded in 1993, the highest joblessness rate since the Depression of the 1930s. The impact of recessions on employment can also be assessed by

226 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 9.4 The unemployment rate, 1964–99

14 12

Per cent

10 8 6 4 2 0 1964

SOURCES

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

As for Figure 9.1.

comparing the employment-population ratio and employment growth. This is a useful extension of our analysis because a fall in the employment-population ratio unambiguously indicates a deterioration in economic conditions (job availability) irrespective of whether it is accompanied by changes in the unemployment rate; as noted previously a fall in the unemployment rate can come about through a lower participation rate which is a possible indication of a ‘discouraged-worker’ effect. However, if accompanied by a rising employment-population ratio, a fall in the unemployment rate necessarily indicates an improving labour market situation. Figure 9.5 shows clearly the periods of significant contraction of employment growth and decreases in employment-population ratios. The very marked decreases in both measures in the early 1980s and 1990s is particularly striking and apposite for our later exercises.

LABOUR MARKET AND ECONOMIC POLICY We now consider a number of counterfactual experiments in order to illustrate the potential gains of anticipating an emerging problem and taking appropriate counter-cyclical action. The obvious way to start is to consider the two outliers in the profile of employment growth: the troughs of the recessions of the early 1980s and early 1990s. A

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 227

6

61

4

59

2

57

0

55

-2

53

-4

Per cent

Per cent

Figure 9.5 Employment-population ratio and employment growth, 1965–99

51 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 Employment growth (left scale) Employment–population ratio (right scale)

SOURCES

As for Figure 9.1.

counterfactual question is: how would employment and unemployment profiles have looked if instead of the 1983 and 1991 decreases in employment growth the economy experienced different situations? To address this question we consider three counterfactual experiments or scenarios. Instead of the high negative employment growth values recorded in 1983 and 1991, the following assumptions are made: (1) employment growth was stagnant during the two years (a growth rate of zero per cent was experienced) (2) employment growth in the two years was higher than it was by the annual average of employment growth in the lower half of the years between 1966 and 1999, that is, 0.38 per cent is added to the actual employment growth in 1983 and 1991, giving new figures of –1.72 and –1.92 (3) employment growth in the two years was higher than it was by the annual average employment growth in the period 1966 to 1999, that is, 1.93 per cent is added to the actual employment growth in 1983 and 1991. Thus the new employment growth rates for 1983 and 1991 become –0.24 and –0.36 respectively.

While these scenarios are ‘what if ’ questions it should be realised that during an international recession it is very difficult to imagine

228 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

that the government could implement policies which would generate an extra growth equal to a sample average growth of employment. Thus, the third scenario provides what is seen to be an upper bound on employment growth, which can be contrasted with the actual experience. The second scenario provides an exposition of the effect of a mildly successful counter-cyclical policy. The first scenario, while at first sight suggesting a hypothetical policy inaction (a stagnant economy with a zero employment growth), does imply a very strong intervention equal to generating over 2 per cent additional job growth in both 1983 and 1991. Given that, as indicated above, the unemployment rate is a result of interactions between the supply of and the demand for labour, one cannot simply derive the profiles of the simulated unemployment rate on the basis of the new employment growth rates without taking account of the changes in labour supply. To handle a pro-cyclical participation response we have estimated an auxiliary equation linking the growth of the labour force with the current and lagged growth of employment and a time trend.3 This equation is then used to predict labour force participation rate increases, and, hence the resulting growth in labour force, for each of the three scenarios of employment growth. Once both employment and the labour force are derived, we can then calculate the unemployment rates which reflect both changes in employment (the demand side) and changes in labour force participation (the supply side) under the three scenarios of employment growth. Our calculations of the resulting unemployment rate profile under the three scenarios are presented in Figure 9.6. Given that we influence only two points, the resulting profiles are all approximately of the same pattern. The major difference, however, is the magnitude of the generated unemployment rates. First, scenario 1 translates into an average level of unemployment of 3.2 percent, with the 1999 value of the unemployment rate being just 4.2 per cent. Scenario 2 produces unemployment which is, on average, about 0.5 percentage points lower than actual. Finally, for scenario 3 this discrepancy rises to 2.8 percentage points on average with the final values of unemployment rate being some 3.7 percentage points lower. As is well known (see Junankar and Kapuscinski 1992), under-utilisation of labour resources results in significant economic and social costs to the society, ranging from individual loss of income through loss of present and future productive capacity to disruption of the social fabric of the society.

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 229

Figure 9.6 Policy effects on unemployment rate, 1964–99 14 12

Per cent

10 8 6 4 2 0 1965

SOURCES

1968

1971

1974

1977

1980

1983

1986

1989

1992

UER: actual

UER: scenario 2

UER: scenario 1

UER: scenario 3

1995

1998

As for Figure 9.1; and authors’ calculations.

The exercises can be made more interesting through considering one component of the unemployment stock that involves people more disadvantaged than others: the long-term unemployed. With techniques presented by Chapman, Junankar and Kapuscinski (1992), it is a straightforward matter to translate these generated unemployment-rate scenarios into outcomes of long-term unemployment (LTU). The methodology involves estimating a reduced-form model of LTU on a subset of the available data, and forecasting the levels of LTU using the generated unemployment rates under different scenarios. Given the similarity of the unemployment rates for scenarios 1 and 2, our discussion focuses on scenarios 2 and 3 which represent small and large changes respectively. The results of these forecasts are given in Figure 9.7. The first point about these forecasts is that they still obey a general pattern of cyclical variations: the profile of LTU still increases quickly over a recession and falls only slowly during a recovery. However, even a small decrease in unemployment, scenario 2 (A), has significant repercussions: the average annual reduction in the number of longterm unemployed in absolute terms is around 55 000 people, the final value being approximately 60 per cent of the actual stock in 1999.

230 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 9.7 Policy effects on long-term unemployment, 1983–99 400 350

'000 persons

300 250 200 150 100 50 0 1978

1980

1982

1984

LTU: actual SOURCES

1986

1988

1990

LTU: scenario A

1992

1994

1996

1998

LTU: scenario B

As for Figure 9.1; and authors’ calculations.

The forecasts for scenario 3 (B), which are for marked increases in employment growth, suggest a reduction of LTU by over 100 000 people by 1999. These results highlighted how critical it can be to avoid just a few very poor years of labour market performance. This section closes with a brief consideration of the policy significance of LTU. There are two basic points which relate essentially to issues of distribution and macro-economic efficiency. With regard to distribution and equity there is an overwhelming case for attention to be paid to the long-term unemployed. Members of the group are some of the least advantaged in the labour market, and are disproportionately made up of those with low formal skills and education, Aboriginals and Torres Strait Islanders, and immigrants from nonEnglish-speaking backgrounds (see Junankar and Kapuscinski 1991; ABS 1994). Moreover, those with high unemployment duration are by definition not accumulating labour market experience, one of the most important determinants of future wage income. Long-term unemployment is also fundamental to policy because of its impact on macro-economic efficiency. A labour supply pool with a large proportion of long-term unemployed will be a labour supply pool characterised by structural mismatch, with employers tending to bypass people with high unemployment duration in job

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 231

hiring. Such an outcome results in a waste of society’s resources since the labour market works more efficiently when the unemployed are able to fit easily into vacancies when they emerge. Piggott and Chapman (1995) have estimated that long-term unemployment in the early 1990s cost Australian taxpayers about $1 billion a year, from lost taxes and high welfare payments. In addition, there is considerable evidence that the associated structural mismatch decreases the potential for rapid recovery, because the skill losses associated with long-term unemployment mean that employers will be bargaining over a relatively small pool of ‘relevant’ labour. This, in turn, implies higher prospects for wage inflation, even in periods of high measured unemployment; lower potential output, budgetary waste, and continuing inequities for those with lengthy unemployment duration (Budd, Levine and Smith 1988; Fahrer and Pease 1993).

L A B O U R F O R C E , E M P L OY M E N T A N D U N E M P L OY M E N T : T H E F U T U R E The projections of the population based on the medium scenario (which assumes a fertility rate of 1.75, current trends in mortalityrate changes and net annual migration of 70 000) and the projections of the labour force published by the ABS (1999c) are shown in Figure 9.8. The figure shows that because of decreasing participation rates, the growth rate of the labour force is expected to average about 0.8 per cent per annum, compared to 2.1 per cent over the last 30 years, and 1.3 per cent during the 1990s; which results in a labour force of 10.8 million by 2016.4 These medium-term projections involve people who are already alive, since to be in the labour force in 2016 a person must be at least one year old in year 2000; consequently, variations in fertility and mortality are not likely to change the overall profile of the labour-force projections. On the basis of these projected population and labour-force series it is straightforward to derive employment levels consistent with unemployment following a certain profile, or vice versa. While we are unable to predict future economic conditions, it is sensible to ask what the required rates of growth of employment would be, given target levels of unemployment assuming there is no participation-rate response. Figure 9.9 provides four scenarios of the future growth of employment (conditional on labour-force projections) if a target unemployment rate is to be achieved over the next 15 years. The four target values of the unemployment rate are 9, 7, 5 and 3 per cent.

232 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 9.8 Labour force and population projections, 2000–2016 20,000

65 64

16,000

63

61 60

8,000

59

Per cent

'000 persons

62 12,000

58 4,000

57 56

0 55 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 2016 Population: actual

Labour force: actual

Population: projected

Labour force: projected

Labour force participation rate: actual Labour force participation rate: projected SOURCE

ABS 1999c.

Figure 9.9 Projected employment growth rates with different target unemployment rate scenarios, 2000–2016 1.8 1.6 1.4

Per cent

1.2 1.0 0.8 0.6 0.4 0.2 0.0 2000

SOURCE

2002

Authors’ calculations.

2004

2006

2008

2010

2012

2014

Target UER = 9%

Target UER = 5%

Target UER = 7%

Target UER = 3%

2016

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 233

The data of Figure 9.9 can be understood as follows. If target rates of unemployment are to be reached by 2016, average annual employment growth rates illustrated would achieve this in a smooth fashion. There is an infinite array of possible time paths that would achieve these outcomes; those being shown of relatively large employment growth rates at the beginning of the period are illustrative only. From these projections, achieving an unemployment rate of 5 per cent by 2016 requires an average employment growth of just under 1 per cent, which ranges from 1.4 per cent in 2000 to 0.5 per cent in 2016. This target should not be regarded as optimistic; it is worth remembering that employment growth of 1.4 per cent is much less than the sample average over the last three decades. Furthermore, in the light of historical profiles of the growth rates of employment and labour force, and given the projected growth rates of the labour force of 0.8 per cent, the average annual growth rate of employment of 0.98 per cent should be regarded as a credible target. Finally, we also consider the longer-term future of the labour force over the next three decades. Over such time frames, the demographic factors tend to assume much more important roles in driving the labour force than do cyclical (economic) factors. For example, the ABS publishes only one set of labour force projections (discussed above and covering a short term, up to 15 years) which assume a total fertility rate of 1.75, a net migration of 70 000 and the continuation of current trends in mortality rates. Unpublished projections based on alternative assumptions indicate very little change in the population series and, hence, the projected labour force series. Longer time periods, however, increase the importance of such underlying assumptions on the labour force projections. Using unpublished projections of labour force series by McDonald and Kippen (2002b), we can illustrate such an effect. Thus, Figure 9.10 compares two labour force projections, one based on a higher fertility rate (1.75) and one based on a lower fertility rate (1.65). Both series assume 100 000 net migration and current profile of mortality rates. The results indicate that projections of the labour force conditional on a lower fertility rate imply a slightly slower growth rate of the labour force, 0.58 per cent per annum, than under the higher fertility rate, 0.61 per cent per annum. As a result, the projected size of the labour force in 2030 with the lower-fertility scenario, is 11.3 million and 11.4 million with the high-fertility scenario.

234 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 9.10 Projected size of the labour force, 2000–2030 11.6

Persons (million)

11.2 10.8 10.4 10 9.6 9.2 2000

2005

2010

Based on low fertility rate (1.65) NOTE

2015

2020

2025

2030

Based on high fertility rate (1.75)

Both projections assume a migration program of 100 000 immigrants per annum.

SOURCE

McDonald and Kippen (2002b), private communication.

The unpublished projections of the labour force series by McDonald and Kippen (2002b) also allow a look into the age decomposition of the future labour force. While the basic trends and in particular the ageing of the population are well known facts, little discussion has so far been accorded the compositional changes in the labour force. However, Figure 9.11 indicates that over the next three decades there will be some compositional changes in the workforce: a decrease in the share of 15–24 year-olds from around 20 per cent to 17 per cent; an increase in the share of 55–64 year-olds from 8 per cent to 12 per cent; and an almost doubling of the share of people aged 65 and over from 1.4 per cent to 2.5 per cent. The data also suggest that people aged 25 to 34 will account for a decreasing proportion of the labour force: from 24 per cent in 2000 to just under 22 per cent 30 years later. The two remaining age groups will experience somewhat smaller changes: the 35–44 age group will decrease its share by 1.3 percentage points while the 45–54 age group will grow by 0.9 of a percentage point. These changes are also presented in relative terms in Figure 9.12.

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 235

Figure 9.11 Projected composition of the labour force by age, 2000–2030 100

80

Per cent

60

40

20

0

2000

2005

2010

2015

2020

2025

Aged 65+

Aged 45–54

Aged 25–34

Aged 55–64

Aged 35–44

Aged 15–24

2030

NOTE Projections are based on a net migration of 100 000 per annum and a fertility rate of 1.65. SOURCE McDonald and Kippen (2002b), private communication.

Finally, it is also instructive to look at the effects of different migration scenarios on the relative composition of the labour force. Based on recent immigration levels, the population projections include a low migration scenario of 60 000 immigrants per annum and a high migration scenario of 140 000 immigrants per annum. Such scenarios appear to have relatively modest effects on the age composition of the workforce over a 30-year period. In particular, Figure 9.13 indicates that shares of all groups change by less than one percentage point with larger impacts occurring towards the end of the 30-year projection period. These projections also show that a larger immigration intake has a more favourable impact on the relative shares of the younger age groups, especially 25–34 year-olds. It also tends to reduce the shares of older age cohorts.

236 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 9.12 Projected changes in the age composition of the workforce, 2000–2030 200

Index number (2000 = 100)

180 160 140 120 100 80 2000

2005

2010

2015

2020

2025

2030

Aged 15–24

Aged 35–44

Aged 55–64

Aged 25–34

Aged 45–54

Aged 65+

NOTE All shares have their 2000 value as the base. Projections are based on a net migration of 100 000 per annum and a fertility rate of 1.65. SOURCE McDonald and Kippen (2002b), private communication.

Figure 9.13 The difference between high and low migration scenarios on the projected composition of the labour force by age, 2005–2030 0.8 0.6

Percentage points

0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8

2005

2010

2015

2020

2025

Aged 15–24

Aged 35–44

Aged 55–64

Aged 25–34

Aged 45–54

Aged 65+

2030

NOTE The low and high migration scenarios assume 60 000 and 140 000 immigrants per annum respectively. Both projections assume a low fertility rate of 1.65. SOURCE McDonald and Kippen (2002b), private communication.

T R A N S F O R M AT I O N S I N T H E L A B O U R F O R C E

• 237

CONCLUSIONS Population and labour force projections indicate that in the next three decades the rate of growth of the labour force will slow, as a consequence of current and projected declines in fertility, the subsequent slowing of the population growth rate, and the overall fall in participation rates. While aggregate female participation rates are slowly increasing, male rates are declining. Even though the former developments are bigger than the latter, the overall changes in the composition of the population, and in particular, the relative increase of the age groups with lower participation rates, explain the overall decline in the participation rates. Our illustrations of the projected age composition of the labour force also indicate that future compositional changes, as well as relative declines of younger age cohorts and increases in older age cohorts, may well have important implications for the structure and nature of employment. The analysis presented here also indicates that understanding of labour force movements provides relatively few policy-relevant implications. However, our focus on unemployment provides much more meaningful input into issues of policy. Our scenarios and the comparisons of the unemployment projections, even though derived in a hypothetical way, indicate that the avoidance of just two bad employment-growth years can substantially affect subsequent outcomes up to three decades later. This implies that avoidance of downturns and the possible role of pre-emptive active counter-cyclical policies are arguably critical to the long-run health of the labour market, perhaps much more so than are policies designed to promote recovery. This is a particularly apposite point with respect to longer-term unemployment, which remains the major labour market policy challenge.

10

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ? D O N R OW L A N D

Population ageing commonly raises concerns about problems relating to health and government expenditure. Yet such problems make up only part of the total range of changes arising. Population ageing is also creating a new social structure with regard to the relative numbers of people at different stages of life. A focus mainly on negative consequences misrepresents the characteristics of the new structure. This chapter provides an account of changes due to population ageing that emphasises the more prevalent positive outcomes. The main characteristic of the population in later life is the capacity for independent living. This is made possible through the population’s resources for sustaining personal independence, especially good (or at least adequate) health together with family support. To identify trends in independent living, the chapter compares the characteristics of the birth cohorts that span the older ages today or will do so during the next few decades. Thus the cohorts considered are those born between 1901 and 1951. The chapter first discusses the development of a new social structure in terms of the transformation of Australia’s age profile and the emergence of the ‘Third Age’ as a major new stage of life. This provides the setting for the account of independent living in later life as a key feature of the contemporary social structure. The prevalence of the capacity for independent living is then examined from survey data on the health of older people. Later sections compare cohorts in terms of survival patterns and further resources for independent living—including family support, education and housing.

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 239

AGE STRUCTURE CHANGE In the 25 years from 1971 to 1996, the decline of the birth rate transformed Australia’s long-term outlook from a rapidly growing, young population to a declining, old one, as illustrated in the stable population models (Figure 10.1). These show the ultimate form of the country’s age structure if the birth and death rates for a given year remained fixed, yielding a constant rate of growth called the intrinsic growth rate. In 1971, Australia’s intrinsic growth rate was 1.3 per cent per annum. As the National Population Inquiry (1975) pointed out, maintenance of this growth rate would have resulted in a stable population with a young triangular age structure and a relatively low representation of older people. By contrast, in 1996, Australia’s population had an intrinsic growth rate of –0.5 per cent per annum. The long-run outlook, about 60 years ahead, from such a rate is a top-heavy age structure with lower percentages in each younger age group. The 1996 model describes a future population in which the aged outnumber children and 23 per cent are 65 years and over, nearly double the 1996 figure of 12 per cent. Although based on data for just one year, the 1996 situation was by no means an exception: demographic rates throughout the 1980s and 1990s would have produced similar results. The reversal in Australia’s demographic outlook after the early 1970s, from growth to decline and from demographic youth to demographic old age, underlines the magnitude of the impetus for change inherent within Australia’s present age structure. The ageing of the population is one of the major transformations that Australia’s population is experiencing. Only an increase in fertility could alter substantially the outlook shown by the 1996 stable population model. The main process in the transformation is the movement or ‘flow’ of birth cohorts—groups born in the same years—through the age structure. Cohort flow changes the numbers, as well as the characteristics, of the population in older ages. This is illustrated in the comparison of the 1996 and 2031 age structures. Assuming a total fertility rate of 1.8 children per woman, much of the future growth in older ages would be due to the movement of larger cohorts from the middle to the top of the age structure. In the projections for Figure 10.2, zero net migration results in a total population of 19.7 million in 2031 and 22.8 per cent aged 65 years and over. Annual net migration of 80 000 produces 23.3 million in 2031 and 20.1 per cent aged 65 and over. Migration intakes within the bounds of past

240 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 10.1 Stable population models, Australia 1971 (shaded) and 1996 Males

Females

90-94 80-84

Age Group

70-74 60-64 50-54 40-44 30-34 20-24 10-14 0- 4 6

4

2

0

2

4

6

% Total Population

Summary of stable population models

SOURCES

Age 0–14 15–64 65+ Total Intrinsic growth rate

1971 28.9% 61.7% 9.3% 100.0% 1.3%

1996 16.0% 60.8% 23.3% 100.0% –0.5%

1996 and 1971 censuses; Australian life tables for 1996 and 1970–72.

experience would therefore make only a small difference to the percentages in older ages, but could prevent population decline in the long term. The concept of population momentum summarises the prospects for growth from the progression of cohorts into later ages. Momentum measures the potential for cohort flow to increase population numbers as larger cohorts grow older. Falling death rates add to cohort size and increase momentum. Falling birth rates decrease cohort size and reduce momentum. This produces a bell-shaped curve in momentum over the demographic transition (Rowland 1996a). Falling birth rates also increase the percentages in older ages: hence lower momentum further signifies population ageing in percentage terms.

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 241

Figure 10.2 Australia’s age structure in 1996 and 2031, assuming annual net migration of zero and 80 000

Males

Australia 1996 (shaded) and 2031

Females

80-84 70-74 60-64 50-54 40-44 30-34 20-24 10-14 0- 4 800000 600000 400000 200000 TFR 1.8; ANM 0

Males

0 200000 400000 600000 800000 Numbers

Australia 1996 (shaded) and 2031

Females

80-84 70-74 60-64 50-54 40-44 30-34 20-24 10-14 0- 4 1000000

500000

TFR 1.8; ANM 80,000

0

500000

1000000

Numbers

ANM = annual net migration. The total fertility rate (TFR) is constant in both projections at 1.8 children per woman. SOURCE 1996 Census and author’s projections. NOTE

242 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 10.3 Population momentum in Australia, 1881–2001

60

Momentum %

50 40 30 20 10

2001

1991

1981

1971

1961

1951

1941

1931

1921

1911

1901

1891

1881

0

Year SOURCE

Rowland 1996a.

Australian statistics show the level of total population momentum in the age structure since 1881, that is from about the time when fertility decline began (Figure 10.3). While the typical trend from high to low momentum is clear, the protracted baby boom after World War II caused an interruption, initiating a period of demographic rejuvenation between 1946 and 1971. The momentum curve shows that 1971 was a turning point marking the end of population rejuvenation and the return of population ageing. The baby boom added considerably to the momentum of growth, foreshadowing heightened increases in the numbers in middle and later ages in the future. In 1971, Australia’s age structure had inherent growth potential of 34 per cent; by 1996 this had fallen to 15 per cent. Zero momentum, which is likely to be reached early this century, is not a limit, however, as a further continuation of below replacement fertility would create negative momentum. The population could eventually

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 243

decline owing to the progress of successively smaller cohorts through the age structure. The 1996 stable population model represents a future population with negative momentum (Figure 10.1). It is also a population characterised by ‘hyper ageing’ or ‘super ageing’: there are relatively high proportions in the older ages, because younger cohorts are progressively smaller and the total population is declining.

INDEPENDENT LIVING Birth cohorts moving through the age structure vary considerably in numbers and characteristics; trends in marriage, family formation, immigration and survival have affected each of them differently. For example, some cohorts have substantial family resources arising from marriage and childbearing, while others have more limited family resources because of childlessness or marriage breakdown. Cohort characteristics that help or limit personal independence are a particular focus of this chapter. Independence is a significant theme in accounts of population ageing, because it is the preferred situation for the great majority of older people. It is also a condition that contemporary social policies seek to reinforce. Much writing about ageing focuses on dependence because of questions about meeting the needs of the frail and containing the costs of health care and income maintenance; critics of this preoccupation see it stifling human potential. In his book, A Fresh Map of Life: The Emergence of the Third Age, Peter Laslett wrote: Instead of so rearranging our affairs, and so dividing our lives, that we can begin to realize the full potential human experience for the first time in history, we have taken fright. In our own country [United Kingdom] at the moment all that we seem to be able to see is the ever growing number of failing elderly people who weigh upon the individuals who support them (Laslett 1989, p. 1). Laslett used the concept of ‘the Third Age’ as a basis for a more positive discussion of later life. The phrase ‘the Third Age’ was first introduced in France in the 1970s in reference to Universities of the Third Age (Laslett 1989, p. 3), and subsequently in Britain and other countries from the early 1980s. The term Third Age recognises the transformation of the social structure through the emergence of a new stage of life.

244 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Laslett (1989, p. 4) described four lifecycle stages: First Age Dependence, socialisation, immaturity, education Second Age Independence, maturity, responsibility, earning, saving Third Age Personal fulfilment Fourth Age Dependence, decrepitude, death His Third Age is ‘the age of personal fulfilment’ or the age of ‘personal achievement’ (Laslett 1989, pp. 4, 152). This, he said, can be lived simultaneously with the First and Second Age. For example, some athletes begin their Third Age in their First (Laslett 1989, p. 77), but it is not clear whether they remain in the Third Age. Laslett stated, however, that for most people the Third Age becomes a possibility only at retirement (p. 78). Laslett (1989, p. vii) also identified a Fourth Age of ‘true dependency and decrepitude’, which this chapter refers to as the age of later-life dependency. The main problem arising is that there are difficulties in attributing a particular state of mind to a lifecycle stage or an age range. Laslett’s Third Age does not appear to accommodate individuals who are retired and healthy but do not have a strong sense of personal achievement and fulfilment; consequently, some re-define the Third Age simply as the stage between retirement and the Fourth Age (Siegel 1990). The disadvantage of such an alternative is that it neglects the positive view of the Third Age that Laslett was seeking, justifiably, to establish. Other possibilities are: the Third Age of ‘active leisure’ (Blaikie 1999, p. 69); the Third Age of ‘leisure and contentment when one can seek wisdom and a resolution of life’s purpose’ (Mason and Randell 1997, p. 213); or the perspective of the University of the Third Age in Australia, where the Second Age constitutes ‘working life and home-making’ while the Third Age is ‘the age of active retirement’. This chapter re-defines Laslett’s concept of the Third Age as a life-cycle stage of independent living after age 65. At a minimum, this entails living with little or no reliance on instrumental support in activities of daily living. People in the Third Age typically live in their own homes and pursue an active retirement, although there is no requirement here that either of these characteristics be present. Many retire early, but the threshold of 65 years gives comparability with national and cross-national studies of later life. At ages 55–64 in 1993, only 63 per cent of men and 28 per cent of women were in the

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 245

labour force as employed or unemployed persons (ABS 1993, p. 19). There is no comparable chronological marker for an upper limit of the Third Age, since many people remain independent throughout long lives. The practice of distinguishing between broad age ranges, such as the young-old (65–74 years), the old-old (75–84 years) and the oldest old (85 years and over) is a convenient means of summarising information, but it is unsatisfactory as a basis for distinguishing between the Third Age and Fourth Age. Age-based ‘dependency ratios’ are similarly misleading because of their implicit false assumptions that everybody between 15 and 64 years is economically active, while those 65 years and over are all dependent on younger people. The capacity for independent living is the distinguishing feature of the Third Age: there is no other universal characteristic. The challenge for society is to realise the potential of this large group and maximise the prevalence of other positive characteristics in the Third Age, such as socially and personally beneficial uses of time and abilities. Table 10.1 Percentages in the Third and Fourth Ages, Australia 1993 Age group

65–69 70–74 75–79 80+ Total 65+ SOURCE

Third Age

92.6 88.1 83.9 58.8 82.8

‘Transitional’ Handicap: Severe 3.3 3.7 4.4 7.1 4.4

Fourth Age Handicap: Profound 4.0 8.2 11.7 34.1 12.8

Total

100.0 100.0 100.0 100.0 100.0

ABS 1993.

Survey statistics on the presence or absence of ‘handicaps’ (1993 Disability Survey terminology), or ‘core activity restrictions’ (1998 Disability Survey terminology), provide a first approximation for identifying people in the Third and Fourth Ages (Table 10.1). Table 10.1 omits early starters in the Third Age, who retired before 65. Figures in the table confirm that the great majority of older people are in the lifecycle stage described here as the Third Age. Any disabilities they may have are no impediment to living independently. A small proportion of the Third Age population are in circumstances described in Table 10.1 as ‘transitional’. This subgroup has

246 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

severe handicaps such that they sometimes need personal help or supervision with self-care, mobility or communication. The great majority of them still live in the community (AIHW 1999a, p. 171). The transitional group has access to varied sources of support, especially the family, community services, or sheltered accommodation. The transitional subgroup also includes some frail older people who manage with just minimum assistance from community services and relatives. Receiving one meal a day from Meals on Wheels, for instance, hardly constitutes a high level of dependency. The Fourth Age of dependency is approximated here by persons with profound handicaps, who always need personal help or supervision with self-care, mobility or communication. Studies of ageing and disabilities usually present statistics combining information for the severely and profoundly handicapped into one category. Such figures overstate the size of the population with ‘true dependency and decrepitude’. As a result there is little published information available on the extent to which the Fourth Age population per se is cared for at home rather than in institutions. Across all age groups, 71 per cent of the profoundly handicapped population lived in households in 1993 (ABS 1993, p. 31); in 1998, 70 per cent of the severely or profoundly handicapped population aged 65 and over lived in households (AIHW 1999a, p. 171). Recent estimates suggest that only about a quarter of men and a third of women reaching age 65 will ever be admitted to a nursing home (Liu 1998). The Fourth Age is a fairly rare stage among the under 80s, and even at more advanced ages the majority of the population remain independent: the Third Age spans the greater part of later life. Most people never reach the Fourth Age, except perhaps briefly before death. The transition to a protracted Fourth Age usually requires prolonged survival, which is more frequent among women because of their greater longevity. Finally, although people 65 years and over have almost always been present in society, the emergence of the Third Age is a fairly recent development. Laslett specified several prerequisites for the Third Age to constitute part of the social structure. Australia now meets most or all of these: 1

2

More than 10 per cent of the population are 65 or over, which Laslett (1989, p. 85) said denotes an ‘adequate size’ for a Third Age society. This proportion makes the Third Age a conspicuous life stage in the society. Average life expectancy is such that a majority of the population survive beyond retirement. Laslett attached some importance to a 50 per cent

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

3

4

• 247

chance of survival from the Second to the Third Age (measured between the 25th and the 70th birthday), since he believed this leads people to have greater confidence that they will live longer and will plan accordingly. A related point is that sufficient time has passed for the presence of the Third Age to be recognised. Laslett (1989, p. 78) considered that there is a time lag in the emergence of the Third Age ‘because it takes some time for people to become aware that they can expect this longer future and start to conduct their lives with such a future in view’. He thought that, in the past, people tended to discount the possibility of becoming old (p. 87). Thus while he dated the appearance of the Third Age to the 1950s in Britain, he did not see it as a settled feature of the social structure until the 1980s (p. 79). Even then, the potential of the Third Age to benefit society as a whole remained unrealised (Blaikie 1999, p. 11). In Australasia, Laslett dated the establishment of the Third Age no later than the 1960s. Other requirements are protracted healthy life expectancy, sufficient disposable wealth for the society and the individual, attitudes supportive of the participation of the aged in society, and adequate cultural and educational facilities (Laslett 1989, p. 85).

S U R V I VA L O F G E N E R AT I O N S Families with aged relatives were largely a product of the 20th century. Improvements in survival increased the prevalence of families with three or four surviving generations. By the late 20th century, the multigenerational family was probably typical, with family networks long including surviving grandparents and sometimes greatgrandparents. Conventional measures of survival, from official life tables, understate improvements because they record changes only to a given year, rather than over the whole lifetime of a cohort. The statistics in Table 10.2 are based on the observed and projected mortality of cohorts, and thus more closely represent actual lifetime experience. The life expectancies at birth of cohorts born between 1901 and 1951 differ from the cross-sectional life expectancies for their birth year by more than six years for males, and more than eight years for females. The statistics in Table 10.2 reveal appreciable increases from cohort to cohort in life expectancy and the proportions surviving to older ages, especially for women. The estimates of cohort life expectancy at age 65 show gains of four years for both males and females in the cohorts born between 1901 and 1951. A striking feature is that male life expectancy at older ages for the 1946–51 cohort was much the same as for the 1901–06 female cohort. Like the life expectancy figures, the proportions surviving at older ages in the

248 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 10.2 Life expectancy and survival of cohorts born 1901–51, Australia Cohort birth years

Life expectancy (ex) 65

75

85

Males 1901–06 1906–11 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

13.8 14.4 14.9 15.4 16.0 16.4 16.7 17.1 17.4 17.7

8.5 8.9 9.2 9.4 9.7 9.9 10.1 10.3 10.4 10.7

4.9 5.0 5.1 5.1 5.2 5.3 5.3 5.4 5.5 5.6

Females 1901–06 1906–11 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

17.3 18.1 18.7 19.2 19.7 20.0 20.3 20.6 20.9 21.2

10.7 11.1 11.5 11.8 12.0 12.2 12.4 12.6 12.8 13.0

5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.6

NOTE a SOURCE

Per 100 000 born. Rowland 1996b.

male cohorts trail well behind those of females, although appreciable increases are apparent. In the 1946–51 cohort, 33 per cent of men and 52 per cent of women are projected to reach age 85, compared with 14 and 28 per cent respectively for the 1901–1906 cohort. The main implication is that, from cohort to cohort, people are living more years per person in the Third Age. Figures on cohort life expectancy actually understate the extent of the revolution in length of life: increases in total person-years lived in later stages of life far outpace increases in life expectancy. In Australian female cohorts for example, 30 per cent increases in life

• 249

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

Proportion surviving (lx)a 65

75

Person-years liveda 85

65–74

75–84

85+

60 61 63 65 68 71 73 75 77 78

290 922 921 910 634 325 003 088 351 728

38 41 44 47 50 53 56 58 61 63

977 476 130 221 661 872 307 802 503 509

14 16 18 20 22 24 26 28 30 32

213 319 366 468 830 993 878 860 950 774

502 523 548 572 605 634 655 678 703 720

673 928 289 187 158 428 603 939 496 574

263 286 313 340 371 400 422 447 472 493

046 261 376 429 618 331 993 186 599 054

69 80 92 105 118 131 143 155 169 183

283 920 880 047 588 634 226 854 911 734

70 72 74 77 79 82 83 85 86 88

211 396 925 177 805 118 594 355 958 214

54 58 61 65 68 71 73 75 77 78

169 158 822 027 270 074 109 191 138 813

28 31 35 38 41 43 45 47 50 52

050 802 210 181 195 722 789 984 257 506

629 660 692 719 749 774 791 811 828 843

856 402 424 189 138 469 952 350 702 097

419 459 498 530 563 592 613 635 656 677

289 710 267 507 538 000 006 336 857 447

162 188 212 234 256 276 293 312 331 347

150 023 688 639 350 036 219 261 343 649

expectancy at age 65 have accompanied increases of around 70 to 100 per cent in the number of person-years lived after 65. Table 10.3 shows, per 100 000 males or females in each cohort, the number of years lived in the young-old, old-old and oldest-old ages. The figures indicate substantial growth in the number of years successive cohorts will live at each age range of later life. In the young-old ages, 65–74, the years of life lived, per 100 000 born, are projected to be 43 and 34 per cent greater for the 1946–51 male and female cohorts than for their 1901–06 counterparts. Such gains denote a greater resource for society of people in the Third Age.

250 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Whether these trends also translate into more years of life in the Fourth Age depends on the duration of life of the profoundly handicapped. Available estimates of disabled life expectancy provide only combined figures on years lived with severe and profound handicaps (Mathers 1996). The estimates, for persons aged 65 in 1993, indicate an expectation of 2.4 years of life with severe or profound handicaps for males and 4.6 years for females (Mathers 1996, p. 10). The figures were equivalent to 15 per cent of total life expectancy at 65 for males and 24 per cent for women. The proportion of later life lived in the Fourth Age is likely to be appreciably lower.

F A M I LY R E S O U R C E S As previously noted, information on health serves only as a first approximation of the numbers in the Third Age. Independent living also requires other personal resources. People who cannot speak English, for example, may be reliant on relatives for help with shopping and other everyday transactions. Beyond health, the resources of greatest general relevance, to the Australian-born and the overseas-born alike, are the family and economic resources. Older homeowners with a spouse and offspring are well placed to obtain informal support, or to provide for themselves financially as the need arises. By contrast, others without close relatives or the financial security of home ownership are more likely to become reliant on formal services or institutional care at a lower level of disability. In 1986, childless older women were 66 per cent more likely than older mothers to be living in a hostel or nursing home (Rowland 1998b). The Third Age population has varying degrees of what may be termed ‘supported independence’. At one extreme is the transitional group (Table 10.1), since they would require others to make continual interventions to enable them to live in the community. At the opposite extreme are people who receive little practical help from others or who could manage on their own if necessary. Probably the greater number of people in the Third Age, as well as in the Second Age, are not fully self-reliant but regularly give and receive support that enhances their lives and expands their opportunities in independent living. The interdependence between husband and wife, such as where one can drive and the other can cook, is the type of mutual assistance that enables people to live with fewer constraints. Family relationships with a spouse or offspring rank highly as personal resources enabling people to live with minimum restrictions in the Third Age.

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 251

M A R I TA L S TAT U S Support within, rather than between, generations is the prime source of family assistance in sustaining personal independence throughout later life. Accordingly, marital status is an indicator of family resources for care and support: increases in the proportions married particularly denote greater access to such resources. The marital histories of the early 20th century cohorts span much of the ‘marriage revolution’, entailing a trend to younger and more universal marriage (McDonald 1975). Later developments, notably the 1970s ‘marriage bust’ (Carmichael 1987) and the rise of consensual unions, largely occurred too late to affect them. Table 10.3 Marital status of cohorts born 1901–51, at ages 65–69 and 75–79, Australia (%) Cohort birth years

Never married 65–69

Married

75–79 65–69

Widowed

75–79

65–69 75–79

Separated or divorced 65–69 75–79

Males 1901–06 1906–11 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

9 8 7 7 7 7 6 6 6 6

8 7 6 7 6 6 6 5 5 6

77 78 79 79 79 75 73 69 66 63

68 70 73 70 68 67 65 62 58 56

10 9 8 7 7 6 6 6 6 6

21 18 16 16 16 16 15 14 14 14

5 5 6 7 7 12 15 19 23 26

4 4 4 7 9 11 15 19 22 25

Females 1901–06 1906–11 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

10 8 6 5 5 4 4 5 5 6

9 8 6 6 5 5 5 5 5 6

50 52 55 58 61 60 59 55 51 46

28 32 36 36 38 38 37 33 29 26

36 35 34 31 27 27 26 25 25 24

60 57 53 53 52 51 51 52 52 50

4 4 5 6 7 9 11 15 20 24

3 3 4 4 5 6 8 11 14 17

The figures for each age group sum to 100 within male and female cohorts. Until the 1961 Census, separated persons were included in the ‘married’ population. SOURCE Rowland 1994. NOTES

252 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Lacking a spouse and offspring, the never-married elderly generally have the most limited family resources; this being so, statistics on the marital status of cohorts suggest that potential family resources have increased through time. The never-married have become a smaller minority from cohort to cohort, although the proportions have declined less for men (Table 10.3). In the 1901–06 cohort, 10 per cent of women were never-married at ages 65–69 compared with less than half this figure among women born in the 1920s and 1930s. Because of differences between the ages of husbands and wives, there are relatively high proportions of men still married in later life: they are likely to have younger wives who will outlive them. Nevertheless, despite the marriage revolution, the proportion of men currently married at ages 65–69 has shown only moderate growth and is projected to decline (Table 10.3). This seems to be because the higher the proportion marrying in a cohort, the greater the subsequent attrition due to divorce. For example, 85 per cent of men in the 1931–36 cohort were married by ages 35–39, but by ages 55–59 the proportion was 80 per cent. The latter figure is lower than corresponding figures for previous male cohorts at the same age, which had lower proportions ever married. More recent cohorts have experienced even greater attrition in their proportions married as they have advanced through their 30s and 40s. Thus the rise in divorce rates, dating from the mid-1970s, will contribute to higher proportions entering later life without a marriage partner. Among women, the proportions married differ from those for men, particularly through the effect of relatively early deaths of husbands. For a time, improvements in the survival of husbands were mainly responsible for raising the proportions of women still married when they reached their late 60s: from 50 per cent in the 1901–06 cohort to 60 per cent in the 1920s cohorts. Separations and divorces are now reversing this trend. Also, because of female longevity, previously married women are becoming more evenly distributed between the statuses of widowed and divorced. Previously married men are more clustered in the latter group, because they are less likely to be widowed. While the long-run effects of high divorce rates are apparent in older age groups, early widowhood is at least on the wane. The observed and projected figures show a continuing decrease in the proportion of men and women widowed at all older ages, with the largest falls occurring among women in their late 60s and 70s (Rowland

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 253

1994). Among the three main components of change in the marital status of the aged—the incidence of marriage, the incidence of marriage breakdown, and the incidence of premature death—the last is now the only factor strengthening potential family resources. Being legally married for life has never been more prevalent than in the present cohorts of the aged. They are the most conspicuous products of the marriage revolution. Their years of family formation coincided with a time of strong societal endorsement of marriage and domestic life and accompanying expectations of conformity to supposedly ‘traditional’ family values of marriage, childbearing and lifelong commitment to one partner. Yet they could be both the first and the last of a line. Previous cohorts never had such high proportions marrying and staying married. Future cohorts are also unlikely to match their experience. In the next few decades, the composition of the older population will begin to change as birth cohorts with disrupted marital histories advance into later life. Australia is on the threshold of a decline in family resources as higher proportions experience marriage breakdown and live their later years without spouses, the main carers and supporters of the aged. From the point of view of the marital status data, the Third Age reached its zenith in the late 20th century. Nevertheless, a substantial proportion of older people will continue to have the significant family resource of a spouse.

CHILDLESSNESS Daughters and sons are the second most important sources of family support. Over time there has been little change in the average number of offspring per older parent. Since the 19th century, most parents have had two or three children surviving in their later years. This is mainly because large average family size formerly offset the effects of high infant mortality, whereas now low death rates ensure prolonged survival of the two-child family (Rowland 1992, pp. 40–41). Childlessness, however, can be an important deficit in the family resources of older people. Expressive, or emotional, support is not necessarily dependent on children, because the widowed and childless often meet such needs through contact with friends, neighbours and other relatives (Bachrach 1980; Rempel 1985; Connidis and McMullin 1993). Overseas evidence suggests that these networks are less effective in providing instrumental, or practical, support when the need is continuing (Bachrach 1980).

254 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

In 1996, 12 per cent of women aged 65 years and over were childless, because they never married, or they had childless marriages, or their children had died. Some of the ‘childless’ had adopted children, or stepchildren, and parent-child relationships equivalent to those of people with their own offspring. Conversely, others with offspring were ‘functionally childless’ through lack of contact with them. Figure 10.4 Total proportions childless at ages 45–49 years, female cohorts born 1851–1957, Australia

Proportion of All Women Childless 35.0

% Total Cohort

30.0 25.0 20.0 15.0 10.0 5.0

SOURCE

1952-57

1942-47

1931-36

1921-26

1911-16

1901-06

1891-96

1881-86

1871-76

1861-66

Cohort

1851-56

0.0

Rowland 1998a.

The overall proportion of women childless in later life is quite high in some cohorts. Starting at a level of over 20 per cent for the 1851–56 cohort, the proportion peaked above 30 per cent for cohorts born in the 1890s and early 1900s, then declined continuously to 9 per cent for the early 1940s cohort (Figure 10.4). The peak figures were at least partly due to high levels of involuntary childlessness when events disrupted family building, especially during World War I, the Great Depression and World War II. The fertility of younger cohorts is incomplete, but there is clearly a new increase in the levels of childlessness in the post-war cohorts. Estimates of the percentage remaining childless, among all women of reproductive

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 255

age in 1996, range from 20 (Merlo and Rowland 2000, pp. 29–30) to 28 (ABS 1999a, p. 43). In most cohorts born before the 1890s, more than half the childless were never married. Subsequently, the ‘marriage revolution’ reduced this proportion. In cohorts born in the first half of the 20th century, the married have composed up to two-thirds of the childless at ages 65–69. Mortality of children has had but a very small part in overall childlessness (Rowland 1998a). While the childless are more vulnerable to reliance on formal sources of support, only a minority of older people is faced with such a situation. Satisfactory health, together with the support of a spouse, enables most to live independently. Retirement villages also offer security for those who have financial resources at their disposal to offset deficits in family resources. Nevertheless, the high levels of childlessness in cohorts born in the 1890s and early 1900s seem to provide some rationale for the former emphasis on institutional care in government policies. For the widowed childless, ‘supported independence’ has been less attainable. In the late 1990s, cohorts with relatively high levels of childlessness were at advanced ages, but the proportions childless were declining as the 1920s and 1930s cohorts took their places at the apex of the age structure. The increasing prevalence of childlessness in cohorts born since World War II will bring childlessness among the aged into heightened prominence again after 2011. Overall, the analysis of the family characteristics of cohorts reaching later life illustrates the emergence of the Third Age, together with the varied developments arising as contrasting cohorts advance through the older ages. The cohort data further provide a glimpse into the future, since the characteristics of the aged are more settled and their numbers are more predictable than those of the young. From the perspective of family resources, the analysis suggests that Australia has reached, for the time being, the limits of progress towards independent living. Nevertheless, the next section shows that there is still latitude for advancement in other resources for independent living, such as towards achieving a greater degree of cohort self-sufficiency in retirement income.

OT H E R R E S O U R C E S While family resources and health have a pivotal role in enabling people to live independently in later life, other personal resources are

256 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

also important. These include education and economic resources. Education is a personal resource that may contribute to both the quality and the quantity of life. It can improve occupational skills, increase lifetime earnings and retirement income, enhance adaptability, and better equip people to manage their finances as well as exercise their rights. Higher levels of education are also thought to be associated with living longer (Jorm et al. 1998). Yet resource deficits do not have uniform outcomes. For example, some older people in ill health are ‘body transcendent’ while others are ‘body preoccupied’, which can lead to differences in activity levels among people with the same physical health status (Shanas et al. 1968, p. 67). Similarly, lack of formal education is an obstacle to economic prosperity in the lives of some, but not of others. In the cohorts born in the first half of the 20th century, ages at leaving school indicate substantial changes in levels of education (Table 10.4). Among people born before the mid-1920s, less than half stayed at school until they turned 15. In later cohorts progressively higher proportions stayed at school longer, such that nearly 90 per cent of the 1946–51 male and female cohorts left school aged 15 or more; there were corresponding increases in the percentages leaving school at all later ages. Among the same cohorts, there was also a marked rise in the proportions with formal post-school qualifications, including university degrees, diplomas and vocational qualifications (Table 10.4). Economic resources, including income and assets such as the family home, further influence outcomes at turning points in later life. In the Third Age, economic resources determine material living standards and the latitude for discretionary expenditure. Paying rent or making mortgage repayments from a retirement income can impose severe financial constraints. For people in the transitional phase of the Third Age, economic resources further affect health-related outcomes. Ability to afford private health insurance, for example, influences the nature and timing of medical interventions. Some of the aged with private means can also afford to purchase retirement village accommodation or private support services such as transport, shopping, housekeeping, and home nursing. The less well-off are necessarily more reliant on relatives and community services. For people living at home in the Fourth Age, government-funded safety net arrangements include day care, respite care, community services and ‘aged care

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 257

University degreeb

Diplomab

Vocational qualificationb

No qualificationb

Total

Females 1906–11 or earlier 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

Left school aged 15+a

Males 1906–11 or earlier 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

Age in 1996

Cohort birth years

Table 10.4 Schooling and qualifications of cohorts, Australia 1996 (%)

85+

41

6.2

2.7

16.4

74.7

100.0

80–84 75–79 70–74 65–69 60–64 55–59 50–54 45–49

48 49 49 54 66 71 78 87

5.9 5.4 6.1 7.2 8.4 9.8 13.2 15.8

4.5 4.9 4.6 5.1 5.2 6.3 7.2 7.4

14.6 19.8 21.4 22.8 23.6 27.6 27.1 27.1

75.0 69.9 67.8 64.9 62.9 56.4 52.4 49.7

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

85+

43

2.3

2.5

1.4

93.8

100.0

80–84 75–79 70–74 65–69 60–64 55–59 50–54 45–49

48 49 52 58 66 72 80 87

1.7 2.8 2.9 3.8 4.2 8.1 11.1 14.3

3.4 3.7 4.0 5.2 6.0 8.0 9.8 10.2

1.7 3.2 4.0 4.9 5.0 6.0 6.7 8.1

93.2 90.3 89.2 86.1 84.8 77.9 72.4 67.4

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

NOTES a

Denominator excludes not stated and still at school; Denominator excludes not stated. SOURCE 1996 Census, Household Sample File. b

packages’, which provide intensive domiciliar y support. Nevertheless, for most of those at home in the Fourth Age, without the financial means to purchase long-term security, continuing assistance from relatives is still essential.

258 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

The principal exception to fully independent living in the Third Age is the widespread reliance on a government pension for retirement income. While the majority of older people have little need of instrumental support in their activities of daily living, a high proportion are dependent on the government for their income. Over time, rising proportions of successive cohorts are likely to have saved for their own retirement income, especially through superannuation (AIHW 1999a, p. 180-81). The present latitude for change is considerable, however, more so because unemployment and early retirement have reduced the ability of private sources of income to sustain economic self-reliance throughout later life. Increasing life expectancy also means that retirement incomes are needed longer on average from cohort to cohort. In 1997, a government pension or benefit, mainly an age pension or a veteran’s pension, was the main source of income for 71 per cent of retired people aged 65 and over (AIHW 1999a, p. 180). Table 10.5 Weekly individual income of cohorts by age and sex, Australia 1996 Weekly income ($)a Cohort birth years Males Pre-1911 1911–16 1916–21 1921–26 1926–31 Females Pre-1911 1911–16 1916–21 1921–26 1926–31

NOTE a SOURCE

Age in 1996

<200

200–299

300–499

500+

Total

10.0 9.3 10.1 10.6 11.8 10.8

100.0 100.0 100.0 100.0 100.0 100.0

3.1 3.7 5.0 4.7 4.6 4.4

100.0 100.0 100.0 100.0 100.0 100.0

85+ 80–84 75–79 70–74 65–69 Total 65+

52.9 57.4 55.4 53.3 55.5 54.9

24.1 22.9 20.9 21.2 19.5 20.8

Percentage 13.1 10.4 13.6 14.9 13.1 13.5

85+ 80–84 75–79 70–74 65–69 Total 65+

60.5 65.1 65.7 66.4 68.6 66.1

29.3 23.6 21.6 20.5 17.7 21.2

7.1 7.6 7.7 8.4 9.1 8.2

Excludes income not stated. 1996 Census, Household Sample File.

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 259

Despite differences in education levels, older cohorts varied little in their income distributions in 1996. Male-female differentials were more marked than those by age or cohort (Table 10.5). About 55 per cent of men and 66 per cent of women had individual incomes of less than $200 a week in 1996; thus a majority received no more than the equivalent of a single-rate age pension. The age pension in June 1996 was $171.30 weekly for single persons and $142.90 each for married couples (Department of Social Security 1996). Gender differences in lifetime earnings and access to superannuation account for the underrepresentation of women in all the higher income groups. Many older Australians are ‘asset rich and cash poor’ (Dowling 1997); in retirement they have limited income for discretionary expenditure, but they have attained the goal of home ownership. This is an important achievement since the home is a key influence on lifestyles and life satisfaction among the aged (Kendig 1981, p. 85). Home ownership is also one of the most important prerequisites for economic security in retirement; it is a major investment that can be used to purchase alternative accommodation if circumstances change. Freedom from mortgage repayments and rent also leaves more income for other essentials. In 1996, 77 per cent of people aged 65 years and over were householders, or joint householders, in homes that they owned outright (Table 10.6). The proportions owning their home peaked at 80 per cent for age group 65–69 in 1996; at later ages there were somewhat higher proportions of householders who were renting, as well as more ‘non-householders’ who lived in another’s home or in a nonprivate dwelling such as a hostel or nursing home (Table 10.6). Nevertheless, even at ages 85 and over, 84 per cent of people where still householders and 69 per cent of them owned their home. The housing situation of the aged is largely an outcome of decisions and economic achievements at younger ages. The boom in home ownership after World War II benefited those who are now in the older ages; they were able to purchase in a time of rising real incomes and low interest rates (Kendig 1984). Thus, while the housing situation of today’s aged is favourable to independent living, it is no guide to the future. Cohorts have had varied experiences of housing costs, interest rates and unemployment, as well as of other social circumstances that influence abilities to achieve home ownership, such as the rise in female labour force participation and the emergence of the dual-income family. In future cohorts of retired people,

260 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 10.6 Dwelling tenure of householdersa by age, Australia 1996 Dwelling tenure Cohort birth years

Age in 1996

Pre-1911 1911–16 1916–21 1921–26 1926–31 1931–36 1936–41 1941–46 1946–51

85+ 80–84 75–79 70–74 65–69 60–64 55–59 50–54 45–49

69.4 72.5 76.0 78.4 79.9 76.6 68.1 57.2 45.4

1.8 2.4 3.1 4.9 4.6 7.7 14.8 24.2 34.9

15.1 13.5 12.2 10.9 11.4 12.1 13.6 15.4 16.6

65+

77.2

4.0

11.9

Total NOTE a SOURCE

Fully Being Rented Other owned purchased

Not stated

Total

Householders as % total pop. 65+

5.8 5.1 3.6 2.1 1.9 1.6 1.6 1.6 1.7

7.8 6.5 5.1 3.8 2.2 2.0 1.9 1.6 1.5

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

84.4 90.0 91.8 93.3 93.4 93.9 93.7 93.8 93.4

2.9

4.1

100.0

92.1

Owners or renters of a dwelling, or the spouse of the owner or renter. 1996 Census, Household Sample File.

rather more will have had to confront the problem of declining housing affordability, as well as the impact of marital breakdown and divorce, which shatter plans for saving and investing in a home. At the same time, dual incomes together with inheritances from the current generation of elderly home owners (see King and McDonald 1999) will prevent economic difficulties for some.

LIVING ARRANGEMENTS Information about housing and family and other personal resources give partial indications of outcomes for independent living. Data on living arrangements illustrate more clearly the outcomes of resource levels. The statistics show that independent living arrangements are the setting for the lives of the great majority of older people. In 1996, 80 per cent of those 65 and over lived as couples, lone persons or individual householders with others. The figure was the same in 1986 and a little lower in 1976 (76 per cent). These autonomous living arrangements are generally preferred throughout adult life, in old age as much as in youth and middle age.

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 261

Table 10.7 Living arrangements by age and sex, Australia 1996 (%) Cohort birth year: Age in 1996:

1926–31 1921–26 1916–21 1911–16 Pre-1911 65–69

70–74

75–79

80–84

85+

Total 65+

Males As a couplea As householder with others Lone person Relative of householder Non-family member Visitor in private dwelling Institutionb Other Total

71.5 1.8

69.8 1.5

62.8 2.4

53.0 4.0

39.3 4.6

65.5 2.2

12.0 1.9 3.9 5.5

14.3 1.5 3.5 5.1

18.7 2.3 2.9 4.4

22.6 3.0 2.6 2.8

19.1 4.3 3.3 2.8

15.5 2.1 3.4 4.7

1.5 1.9 100.0

2.4 1.9 100.0

5.2 1.3 100.0

10.0 1.9 100.0

24.9 1.7 100.0

4.7 1.8 100.0

56.3 6.1

44.9 6.9

33.2 6.2

16.9 8.2

6.5 6.9

37.5 6.7

23.0 3.8 2.4 5.2

32.4 4.4 2.6 4.2

39.6 5.8 2.9 4.9

45.2 7.7 1.9 3.2

35.0 8.7 1.9 2.1

33.3 5.5 2.4 4.3

1.4 1.7 100.0

3.2 1.3 100.0

6.3 1.2 100.0

15.3 1.7 100.0

36.5 2.4 100.0

8.8 1.6 100.0

Females As a couplea As householder with others Lone person Relative of householder Non-family member Visitor in private dwelling Institutionb Other Total

NOTES a Includes a small number of couples co-resident in another’s house; b Includes convents, hostels for the homeless, prisons, as well as hospitals, hostels for the disabled and nursing homes. SOURCE 1996 Census, Household Sample File.

As previously noted, the family experience of contemporary cohorts of the aged is notable for the high proportions ever marrying and the long joint-survival of couples. Living with a spouse is the most common living arrangement in later life (Table 10.7), and the one most fitted to maintaining prolonged independence. Mutual support between husband and wife can enable both to continue living with little or no outside help even in the face of pronounced

262 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

disabilities. About 50 per cent of the aged in both 1986 and 1996 were living with a spouse, up from 45 per cent in 1976. In 1996, a third of older women lived alone, with higher proportions for women in their late 70s and early 80s (Table 10.7). The corresponding figures for men were around half those for women, because more lived with a spouse. For those who live alone, supported independence is often possible through occasional assistance from relatives who live elsewhere. Such relatives are part of the older person’s modified extended family, in which different generations live separately but maintain contact and exchange support. The family provides the strongest assurance of continued independent living in the face of changes in health. The final category of independent living arrangements is ‘householder with others’; often such householders are widows. The others in the household include offspring who have not left home, or who have returned home, as well as siblings, other relatives and unrelated persons. Since these aged are the householders—the owners or renters of the dwellings—they are likely to be providing some economic support to others and to have a position of authority and control within the household (Elman and Uhlenberg 1995; Macunovich et al. 1995, p. 18). Nevertheless, the living arrangement ‘householder with others’ undoubtedly encompasses varying forms and degrees of interdependence among household members. Turning to the more dependent types of living arrangements, coresidence of the aged in the home of a son or daughter is mainly evidence of some obstacle to living separately, such as ill-health, lack of English, or low income and meagre assets. Co-residence can arise from positive factors such as close family relationships, traditional values, mutual support and lower living expenses (Da Vanzo and Chan 1994). Generally, however, joint living arrangements are at odds with expressed preferences of the aged in Western societies for privacy and independence and their desire not to be a ‘burden’ on younger relatives. Impediments to independence are probably the decisive factors in the majority of joint living arrangements, although those concerned are also likely to emphasise positive aspects of the situation. Despite the ageing of the aged and the greater policy emphasis on family and community care of the frail, co-residence was no more prevalent in 1996 than it had been in 1991. In 1996, only 4 per cent of the aged were living in a relative’s home. The 1986 figure was

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 263

double this, but around half of the difference was due to the inclusion of visitors (ABS, personal communication, 1999); visitors did not become a separate category of relationship in the household until the 1991 Census. The 1996 figure almost certainly understates substantially the importance of co-residence as a family safety net. American studies point to shorter durations in such living arrangements (from 4 to 1 years on average), but with apparently little change in the overall proportions co-residing at some time in their lives. About a quarter of daughters in one study experienced a parent or parent-in-law moving in with them (Schoeni 1998, p. 311; Weinick 1995). The proportion of the total aged population enumerated in highly dependent living arrangements—in hospitals, aged care hostels and nursing homes—actually declined from about 9 per cent in 1986 to 8 in 1991 and 7 in 1996. Since the mid-1980s, access to residential care (hostels and nursing homes) has become limited to the most severely disabled or to people without access to instrumental support from relatives. The accompanying strategy of enabling people to remain in their own homes for as long as possible has presumably added to the numbers of profoundly handicapped people living in the community. Nevertheless, the absence of an increase in joint living arrangements with relatives in a time of continuing ageing of the aged and diminished access to residential care suggests that dependence on relatives, especially through co-residence, is not preferred and actually widely avoided, except as a last resort. Most probably a range of factors have been operating to maintain independent living and supported independence such as expansion of community services, assisting more to stay in their own homes; growth of retirement village accommodation, offering a supportive environment for people experiencing or anticipating frailty; a more active role for the modified extended family in supporting the disabled and housebound; and improvements in the personal resources of older people. The present generation of older people have greater personal resources for maintaining independent living than earlier generations.

CONCLUSION In the 1990s, about 80 per cent of the aged had good health and/or independent living arrangements indicative of being in the Third Age. Although there was no exact correspondence between these two

264 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

characteristics, for instance because some in poorer health nonetheless had independent living arrangements, it is reasonable to conclude that the great majority of the population aged 65 and over were in the Third Age. Only the income figures present a clear exception to the thesis that independent living is the main characteristic of the older population, although the future is likely to see higher proportions with self-funded retirement incomes. The prevalence of the positive characteristic of independent living is also consistent with findings about the life satisfaction and well-being of older people in Australia. Contrary to negative stereotypes, largescale surveys have shown that older Australians, including the old-old (aged 75+), are no less satisfied with their lives than the adult population in general. While they report fewer positive moods and emotions, since their lives are less turbulent, they also report fewer negative moods, including anxiety and depression (Headey 1999, p. 34). Negative consequences have been a major focus for research on the ageing of the population; this reflects the fact that the economic problems arising are important concerns for planning and policy making (McIntosh 1998; Walker 1998; OECD 1999). Yet this emphasis creates a misleading impression of later life as a time of dependency. Recognition of the prevalent positive aspects of ageing is essential to understanding the nature of changes in society and new features of the social structure. The emergence of the Third Age is also relevant to decisions about how to adapt or develop policies that are consistent with the characteristics of the older population. As Laslett noted: it has to be reckoned that the institutions and instruments which have been created to meet the problem of ageing are in no position to provide us with a policy for that great majority of retired people who present no problem at all. We need a new outlook, a new language and we need above all a new institution, or set of institutions (Laslett 1989, pp. 2–3). Laslett’s ideas about the Third Age represent a necessary counter to preoccupations with economic problems. Better recognition of the Third Age as a positive development seems essential for further progress towards a society that is well prepared for increases in the numbers and percentages in older ages. There is much potential for regarding older people themselves as a resource and for seeking appro-

A N A G E I N G P O P U L AT I O N : E M E R G E N C E O F A N E W S TA G E O F L I F E ?

• 265

priate ways to enhance opportunities for their engagement with and integration into Australian society as a whole. This will be an important challenge for social institutions in the 21st century as increasing numbers of people enter the Third Age of independent living.

11

A U S T R A L I A’ S F U T U R E P O P U L AT I O N : P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y PETER McDONALD

Carmichael and McDonald in Chapter 3 draw the conclusion that Australia’s fertility rate will continue to fall below its 2000 level of 1.75 births per woman and that a return to replacement level fertility (2.07 children per woman) is extremely unlikely in any foreseeable future. Given this long-term future of low fertility, what population policies are desirable for Australia? The Howard Coalition Government and its Labor predecessor have both stated that Australia should not have a specific population policy (Australia 1994; Ruddock 2000). In contrast, the present Labor Opposition has stated on several occasions over the past few years that Australia should have a population policy (Crean 2002), but the precise policy is yet to be revealed. Many business and community organisations (Chadwick 2001; Sustainable Population Australia 2002) have also called for a population policy and many opinion leaders have been quite specific about what they want, but the range of policies advocated has been very wide. Indeed, proposed population futures for Australia have often extended beyond the bounds of what is demographically reasonable. It has not been uncommon for utter nonsense to be put forward as the ideal population policy for Australia (see McDonald and Kippen 1999a). But, what is a population policy? Australia has an immigration policy that is highly specific and rigidly administered. As the social and economic setting in which people live affects their decisions about the number of children that they have and as government policy influences the social and economic setting, Australia has policies

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 267

that indirectly affect the birth rate. Obviously, Australia has health policies that affect death rates, although population policy is not the purpose of these policies. For example, Kippen and McDonald (2000) have shown that there were almost one million people alive in 2000 in Australia who would have been dead if Australian mortality rates had remained at their 1970 levels. This extra one million people, highly concentrated at the older ages, makes a considerable difference to Australia’s demography. Australia has policies in place related to the age structure of the population: in particular, policies to address the future ageing of the population. Australia also has policies about where people live. While Australians are free to live wherever they choose, there are policies, for example provision of infrastructure, that indirectly affect where people live. In combination, these policies could be said to constitute a population policy. However, in the strict sense, this is not the case. The usual definition of a population policy is a policy that is directed at achieving a particular target population size or a particular target rate of population growth. These policies certainly influence the size of the population and the rate of population growth but the outcomes are largely incidental; they do not constitute a population policy in the strict sense. Australia is not alone in not having a specific population policy. No OECD country has a population policy expressed in terms of a target population size or a target rate of population growth (United Nations 2000). Australia would be exceptional if it did have such a policy. Indeed, it is very difficult to socially engineer a particular rate of growth or a particular target population by a given date. The factors that go into these outcomes are complex and variable. One can hardly imagine that the birth rate can be adjusted to compensate for changes in annual net migration, or changes in mortality. Migration and fertility can be traded off against each other in demographic models, but not so readily in the real world. Thus, a population policy could only be expressed in relatively broad terms and the settings would need to be adjusted from time to time as circumstances changed.

A U S T R A L I A N P O P U L AT I O N F U T U R E S : THE DEMOGRAPHIC COMPONENTS In considering what is desirable for policy, it is important to examine projections of Australia’s future population in order to assess the range of possibilities and to identify any potential issues.

268 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Population futures are the net outcome of four components: the known size and age structure of the current population, and estimates of future levels of fertility, mortality and annual net international migration. This section discusses possible futures for fertility, mortality and migration.

F U T U R E F E RT I L I T Y There are several reasons to expect further falls in the Australian fertility rate from its 2001 level of 1.73 births per woman: •

•

•

•

•

•

•

Australian fertility has been falling slowly but steadily from 1992 onwards (Figure 3.1). While the decline levelled off between 1999 and 2000, it recommenced between 2000 and 2001. The analysis in Chapter 3 has shown that fertility rates at younger ages (under age 30) are continuing to fall sharply (Figure 3.2). While the age at first birth continues to rise (Figure 3.4), further falls in the total fertility rate can be expected. This has been the experience of all other countries with low fertility. Fertility rates are lower and continuing to fall in countries that are similar to Australia in culture, economic circumstances and policy. The fertility rate in Canada in 1999 fell to 1.52. It was 1.63 in the United Kingdom in 2001 and 1.47 in Scotland. The fertility rate is higher in New Zealand (1.97 in 2001) owing to the effect of higher Maori fertility, but the rate there is also falling. Fertility rates in many European and East Asian countries have fallen to as low as 1.2–1.3 births per woman, and show little sign of rising (Eurostat 2002). There was a very large fall in the number of first marriages in Australia between 2000 and 2001, and this will almost certainly feed through to lower fertility in the short term. The human capital of young Australian women is continuing to rise through increased participation in education and the labour force and, without improvement in existing policies to support the combination of work and family, this trend will lead to lower fertility (see the discussion of differences in fertility in Chapter 3). Participation in full-time education by women aged 20–24 rose from 10 per cent in 1988 to 25 per cent in 2002. Labour force participation of Australian women aged 25–34 years, the peak childbearing ages, rose from 62 per cent in 1988 to 72 per cent in 2002. Increased human capital and delay of the first birth mean that women’s wage rates at the time that they are considering having their first child are considerably higher than was the case 30 years ago. Thus, the potential loss of earnings if childrearing cannot be readily combined with paid employment stands as an obstacle to having children for both women and their partners. The association between a country’s approach to work and family and its fertility rate is now well established (McDonald 2000a, 2002; Castles 2002). The sharp rise in housing prices in large cities in Australia in the five-year period from 1997 to 2002 may be an obstacle to family formation, at least in the short term.

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 269

How low is Australian fertility likely to fall? This is a question not easily answered because future fertility will be dependent upon future policy. Projections of existing time trends made by Bacon (2000) and Kippen (2001) suggest a Total Fertility Rate (TFR) for Australia by 2011 of between 1.54 and 1.65 births per woman. The Australian Bureau of Statistics has recently suggested that ‘the possibility that Australia’s fertility could fall to 1.3 should not be discounted’ (ABS 2002c, p. 14). For illustrative purposes here, two assumptions about future fertility have been used: (i) a gradual fall from the 1.73 in 2001 to 1.65 at which point fertility levels off and (ii) a gradual fall to 1.40 at which point fertility levels off. These two assumptions provide a reasonable range, barring major changes in policy that would support the fertility rate or social change that would lead to very low fertility rates.

F U T U R E M O RTA L I T Y Future mortality levels can be relatively accurately predicted in the medium term (next 20 years) but are very difficult to predict in the longer term (see Chapter 5). The projections presented here incorporate only one assumption about future mortality. This assumption is that the expectation of life at birth will increase by one year in every decade. This is very similar to the main assumption made by the Australian Bureau of Statistics in their most recent projections (ABS 2000c). However, the result may be better than this. In the past 30 years, expectation of life has been rising at around 2.5 years every decade. If this much more favourable trend were to continue, it would mean that there would be many more people at ages 70 years and over than the projections made here indicate, and that ageing of the population would continue throughout the 21st century. F U T U R E A N N UA L N E T I N T E R N AT I O N A L M I G R AT I O N The level of net international migration for Australia is shown in Figure 11.1 for the period 1950–2000. It is evident from the graph that net migration has fluctuated fairly widely around a mean between 80 000 and 90 000 per annum. Note that net migration does not include short-term movements (12 months or less) into and out of Australia, but it does include long-term temporary migration as well as permanent migration. There is a theory that the level of net migration follows the rises and falls in the economic cycle and this is generally confirmed by the movements shown in the figure with troughs appearing around the early 1960s, the mid-1970s, the early

270 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

1980s and the early 1990s. Nevertheless, in the entire period, net migration has been positive and almost always above 40 000 per annum. In terms of peaks, net migration has rarely risen above 120 000 per annum. Since 1995, the level of net migration has been relatively stable, fluctuating around the long-term mean of about 85 000. However, the government has announced an increase in its official migration program that will apply for four years from 2002–03. This will mean that net migration will be close to 100 000 per annum for this period. Figure 11.1 Net overseas migration, Australia 1950–2000 180 160 140

Thousands

120 100 80 60 40 20 0 1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

SOURCES ABS, Australian Demography Bulletins (various years), Overseas Arrivals and Departures, Catalogue No. 3401.0 (various issues).

There is a debate about whether migration should be projected forward as a number or as a rate or percentage of the total population (McDonald and Kippen 2002a; Withers 2002). In the 1950s, a target rate of 1 per cent of population was set for net migration. However, in the 50 years from 1950 to 2000, experience has been more consistent with a constant number for net migration than with a constant rate. In recent years, the rate has fallen to 0.43 per cent of population and this is one of the highest rates of net migration of any country in the world. To be at 1 per cent in 2002, Australia’s net migration would have to be 197 000 per annum, a numerical level well above any ever experienced in Australia. Furthermore, if

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 271

constant rates of migration at each age rather than for the total population were assumed, there would be a result not far different from a constant number, because most of the future increase in population will be at ages 50 years and over, at which ages net migration rates are close to zero (McDonald and Kippen 2002). For most of the years shown in Figure 11.1, net migration is roughly equivalent to net permanent migration, that is, net longterm temporary migration has been close to zero. In recent years, however, net permanent migration has been around 50 000 per annum while total net migration has fluctuated between 80 000 and 100 000. This has occurred because the number of long-term arrivals in Australia has been rising sharply (see Chapter 6 on temporary entrants). While the number of long-term departures has also been increasing, there is an average three-year lag between arrival and departure for long-term temporary immigrants. This lag creates an increasing stock of long-term visitors in the population of Australia so long as the number of temporary arrivals increases. For net migration to remain at around 80–90 000 per annum while net permanent migration remains at 50 000 per annum, the upward trend in longterm visitor arrivals would have to continue indefinitely to reach very large numbers in the longer term. As such a trend is unlikely, continuation of net migration at 80–90 000 per annum or higher implies long-term increases in future levels of net permanent migration. This is consistent with the increase in the government permanent migration program from 2002/03. To cover the range of possible future levels of net migration, this chapter assumes a future minimum level of 40 000 per annum. This is based on the history shown in Figure 11.1 and the argument of the last paragraph that the emphasis on long-term temporary migration from the mid-1990s leaves open the possibility of somewhat lower future levels of net migration. There are strong arguments, however, that future governments would not accept net migration at this low level over a long period of time. Thus, with one proviso, this is an unlikely outcome. The proviso is that the Australian economy remains vibrant. In recent years, there has been a quantum leap in the level of international competition for skilled immigrants. In 2001, both the United States and the European Union had net migration levels that were around ten times the level of net migration to Australia (Eurostat 2002). Thus, Australia has already become a small player in the market for international migrants. Because of low fertility rates,

272 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

many more countries will be seeking immigrants in the future. However, relative to its population size, Australia remains a highly favoured destination and a change in this situation is only likely to occur if there is a major, long-term downturn in the Australian economy. Nevertheless, similar countries like New Zealand and Canada are struggling to maintain their net migration levels as their workers are attracted away to the neighbouring economies. For example, the estimated emigration from Canada increased from 25 000 in 1994/95 to 62 000 in 1999/2000 (Statistics Canada 2001, p. 39) and, in the same period, permanent and long-term emigration from New Zealand increased from about 40 000 in 1994/95 to 79 000 in 2000/01 (Statistics New Zealand 2002). New Zealand has now had three successive years of net loss from permanent and long-term migration despite having had a large-scale immigration program. Permanent departures from Australia stood at about 30 000 per annum for many years but had risen to 41 000 by 2000. As a future maximum level, net migration of 120 000 is assumed. The level of 120 000 per annum is very high as an assumed longterm average because this level has only been exceeded in a few years in the past 50 years. Rises above this level would take Australia well beyond the bounds of its past experience. As a comparator, a level of 80 000 net migration is taken. This is roughly the average of the 1990s and of the period 1950–2000.

THREE PROJECTIONS Three projections are compared. All assume the same future level of mortality. The three projections are labelled Low, Standard and High. The standard projection is so labelled because, effectively, it is the extension of existing trends. The low projection is so labelled because it is difficult to see the Australian population following a trajectory lower than this unless fertility falls to the very low levels now evident in some European and East Asian countries. It is assumed that government policies to support the fertility rate would be implemented in Australia before fertility fell to these very low levels. The high projection is so labelled because it is difficult to see the Australian population following a trajectory higher than this unless fertility were to remain around or above the 1.75 level. If fertility were to remain around 1.75, it seems unlikely that policy would support a continuation of the very high migration level assumed under the high projection. While results outside the bounds of the high and

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 273

low projections are possible, they are unlikely given the present state of knowledge. The assumptions of the three projections are: Low projection:

Fertility falls to 1.4 births per woman Annual net migration is 40 000

Standard projection:

Fertility falls to 1.65 births per woman Annual net migration is 80 000

High projection:

Fertility falls to 1.65 births per woman Annual net migration is 120 000.

P R O J E C T I O N R E S U LT S The standard projection produces a beehive-shaped age distribution with a population of around 25 million by 2050 (Figure 11.2). After 2050, the population size and age structure would hardly change at all. This scenario has been described as a ‘soft landing’ at zero population growth. With the low projection, the total population would rise to a peak of 21.6 million in 2030 and decline thereafter. However, by 2050, the projected total population of the low projection (20.7 million) would still be one million higher than the population in 2001. For Australia’s population to be lower in 2050 than its 2001 level of 19.5 million people, the fertility and/or the net migration level would have to be lower than the levels assumed for the low projection, both very unlikely outcomes. By 2100, the low projection produces a population of 15.6 million. The most important difference between the standard and the low projections, however, is not the size of the total population but the resulting age structure. In contrast to the beehive-shaped age structure produced by the standard projection, the low projection produces a coffinshaped age structure, wide at the top (the old ages) and narrow at the bottom (the young ages). I have argued (McDonald 2000b) that a coffin-shaped age structure is not sustainable because the population size spirals downwards, there is a great shortage of young workers and the population ages dramatically. The additional 40 000 net migration in the high projection also leads to a beehive-shaped age structure, very similar in shape to that produced by the standard. McDonald and Kippen (1999b) have shown that once the level of net migration has passed the level that, in combination with low fertility, produces zero population growth, further increases in migration have little effect on the age structure

274 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Figure 11.2 Projected total population, Australia 2000–2100 33 31 29

Millions

27 25 23 21 19 17 15 2000

2010

2020

2030

2040

TFR = 1.4, NOM = 40,000

2050

2060

2070

2080

2090

2100

TFR = 1.65, NOM = 120,000

TFR = 1.65, NOM = 80,000 SOURCE

Author’s projections.

of the population. All that is produced with high migration is a larger population. However, the population under the high projection reaches only 27.8 million by 2050, just 2.8 million above the result for the standard projection. This indicates the constraint that is put upon future population growth by low fertility. Between 1950 and 2000, Australia’s population grew by 250 per cent; by 2050, even under the high projection, the increase of population would be less than 50 per cent above the level in 2000. If, as I argue, Australia should have policies in place that avoid the emergence of a coffinshaped age structure, the standard projection is the minimum outcome to which Australia should aim. Thus, the range of potential population futures for Australia by 2050 is relatively narrow, somewhere in the range of 25–28 million. Figure 11.3 indicates the extent of ageing of the Australian population under the three projections. As already discussed, the low projection produces much more substantial ageing than the other two projections, but substantial ageing of the population is evident from all three scenarios. The figure also indicates that the extent of ageing is similar in the standard and high projections. Ageing of the

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 275

population under the standard and high projections is a ‘one-off ’, with most of the ageing occurring before 2050. This is not the case if mortality were to fall more sharply than is assumed in these projections. With a sharper fall in mortality, say expectation of life at birth increasing by two years per decade, substantial ageing of the population would continue throughout the 21st century (McDonald and Kippen 1999c). Figure 11.3 Projected percentage aged 65 years and over, Australia 2000–2100 36 32

Percentage

28 24 20 16 12 2000

2010

2020

2030

2040

TFR = 1.4, NOM = 40,000

2050

2060

2070

2080

2090

2100

TFR = 1.65, NOM = 120,000

TFR = 1.65, NOM = 80,000

SOURCE

Author’s projections.

Figure 11.4 shows the growth of the Australian labour force in the next 50 years under the three scenarios, assuming that age- and sex-specific labour force participation rates remain at their level in 2000. Under the standard projection, the Australian labour supply would grow by just 10 per cent between 2000 and 2030 after which there would be no further growth. This is in sharp comparison to the 80 per cent increase in the Australian labour force between 1970 and 2000. However, there is considerable scope to increase the labour supply through increases in labour force participation rates. By international standards, the labour force participation rates of both men and women in Australia are relatively low. If it were assumed that,

276 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

over the next 30 years, labour force participation rates for men aged 35 years and over returned to what they were in Australia in 1970 and labour force participation rates for women moved to the level applying in Sweden in 2000 (or in the Australian Capital Territory in 2000), the resulting labour supply with the standard population projection is indicated by the dashed line in Figure 11.4. This shows a 34 per cent increase in the labour supply between 2000 and 2030 without any increase in the total population size, a desirable outcome. Of course, higher labour force participation rates for women imply much higher levels of support for the combination of work and family. They also imply that the high levels of human capital of Australian women would be used more effectively than is presently the case. Figure 11.4 Labour force, Australia 2000–2050 14 13

Millions

12 11 10 9 8 2000

2005

2010

2015

2020

TFR = 1.65, NOM = 80,000

2025

2030

2035

2040

2045

2050

TFR = 1.65, NOM = 120,000

TFR = 1.40, NOM = 40,000

NOTE

The upper dotted line represents a high labour force participation (see text).

SOURCE

Author’s projections.

In summary, the projections indicate that low fertility is an issue because of the impact that it has upon the future age structure of the population. In the longer term, falling population size may also become a problem. Low fertility means more rapid ageing of the population and future falls in the number of new entrants to the labour

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 277

force, that is, low fertility leads to a problem of demographic sustainability. For example, low fertility in Japan means that Japan is facing a fall in the size of its labour force of 20 million workers between 2000 and 2040. Italy is facing a fall of 11 million workers in the same timespan. Simultaneously, both populations will be ageing dramatically. This is an unsustainable future (McDonald and Kippen 2001).

HOW LOW C AN FERTILITY FALL BEFORE T H E R E I S A P R O B L E M O F S U S TA I N A B I L I T Y ? There is no single answer to this question. The answer depends upon a country’s capacity to attract and absorb immigrants. To some extent, immigration can compensate for low fertility, and this is particularly the case in a country like Australia where there is a culture of immigration and where the total population size is relatively small. Australia can easily produce a sustainable demographic future with a fertility rate somewhat lower than the 2001 level of 1.73 births per woman. Fertility could fall as low as about 1.50 in Australia with the low fertility being able to be offset by higher migration (annual net migration of 105 000, for long-term zero population growth). This approach has become known as ‘replacement migration’, that is, using a combination of migration and below-replacement fertility to produce at least zero population growth. Table 11.1 indicates for Australia the levels of net migration (column 2) that in combination with particular levels of fertility (column 1) would lead to sustained zero population growth in the long term. The table suggests that if fertility were to fall below 1.5, even Australia would have difficulty in offsetting low fertility with a replacement migration strategy because the level of net migration required would be well beyond the level of its experience (135 000 with TFR = 1.4). The third column of Table 11.1 shows, for each combination of fertility and migration, the total population size that would be reached before zero population growth was achieved. This indicates the very interesting result that a combination of relatively high fertility with low migration achieves zero population growth at a much lower population size than is the case in the reverse circumstance of very low fertility and very high migration. This result is interesting because it indicates that those who favour a low population future for Australia should not be arguing for lower fertility. The final columns of Table 11.1 show two measures of the age structure of the populations resulting from the particular combinations of fertility and

278 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Table 11.1 Combinations of fertility and migration leading to population near-stationarity within 50 years: population outcomes for 2100 TFR ANM Population Average % aged (‘000) (millions) age of 20–64 20–64 years population 1.0 285 35 44 53 1.1 235 32 44 53 1.2 200 30 44 53 1.3 165 28 44 52 1.4 135 27 43 52 1.5 105 25 43 52 1.6 85 25 43 52 1.7 60 24 43 51 1.8 40 23 43 51 1.9 20 22 43 51 2.0 0 21 43 50 ANM = Annual net migration. SOURCE Author’s projections. NOTE

migration that lead to zero population growth. Again the result is interesting in that all combinations produce the same age structure. Australia is somewhat exceptional, however, in its capacity to employ the replacement migration approach. Almost all other countries would be looking more to fertility rates of 1.7–1.8 in order to produce a more sustainable demographic future. To conclude, a fertility rate below 1.5 is likely to cause sustainability problems in any country. This conclusion is confirmed by the 1999 United Nations (2000) survey of government views of fertility rates. All 26 countries that had a fertility rate under 1.5 reported that they considered their fertility rate to be too low. Thus, applying the precautionary principle, Australia needs to be applying the brakes to the fall in its fertility.

CONCLUSION For at least two decades, the central debate about population in Australia has been between those who favour a low population in the interests of preservation of the natural environment and those who favour a high population in the interests of economic growth. What

A U S T R A L I A ’ S F U T U R E P O P U L AT I O N P O L I C Y I N A L O W- F E R T I L I T Y S O C I E T Y

• 279

conclusions can be drawn about this debate from this analysis? First, while the population is ageing, Australia should avoid population decline because all of the decline would necessarily be at the young, productive ages. A gradual movement to zero population growth should be set as the minimum in any population policy for Australia. The only substantial threat to this outcome is further falls in fertility. Consequently, Australia should revolutionise its policies to support childbearing and childrearing. No element of population policy is more important than this, but, while the rhetoric is there, neither the business nor environmental ends of the population debate are in the forefront of debate about family support policy in Australia. The protagonists in this debate are much more comfortable with immigration as the traditional battleground. If the fertility rate can be held at a level that will sustain the population in the longer term and maintain a reasonable age structure (say, around 1.6–1.9 births per woman), the next issue for policy is the level of immigration. This is a question of population size, not a question of age structure. Here, a rigorous debate between those who see population growth as a threat to the environment and those who see high population growth as the foundation of economic growth in Australia is not problematic. Australia urgently needs to conserve and rehabilitate its natural environment but, at the same time, it needs to ensure that its economy is viable, that living standards rise and that international competitiveness is maintained. The history of this debate in Australia is that neither side has been willing to concede that the other side might have a point, again, aside from the rhetoric. Ultimately, however, as observed in this chapter, the likely population outcomes for Australia by 2050 appear to be relatively narrow, a total population ranging somewhere between 25 and 28 million. In this circumstance, there is a question as to whether the protagonists in the population debate should recognize that future population size is not the solution to their issues. There would seem to be more to be gained by focusing energy directly upon the issues of protecting the natural environment and raising economic productivity.

NOTES

CHAPTER 3 FERTILITY TRENDS AND DIFFERENTIALS 1 Expressing each standardised value as a percentage of the standardised value for the standard population (i.e., indexing them to a standard population value of 100) converts values to a scale that renders comparison more straightforward. It does, however, need to be appreciated that differences in age-standardised mean CEB reflect differences in the timing of fertility as well as in its level. If two populations have the same completed fertility, but one begins childbearing younger, that population will record a higher agestandardised mean CEB figure because at younger reproductive ages its agespecific mean CEB will be higher. CHAPTER 4 TRANSFORMATIONS IN THE AUSTRALIAN FAMILY 1 Originally published as ‘Family Relationships in Australia: The Conservative-Liberal-Radical Debate’, Review of Population and Social Policy, 2000 (9), pp. 171–94, National Institute of Population and Social Research, Tokyo, Japan. Permission from the National Institute to publish in this volume is acknowledged. 2 Same-sex relationships in which the couple lives together are a very small fraction of all intimate couple relationships. In the Australian National University’s Negotiating the Life Course Survey which asked people ‘Are you married, or in a relationship?’, respondents who answered ‘yes’ were asked the sex of the partner. The survey showed that less than one per cent (0.9 per cent) of all relationships were same-sex relationships. The 1996 Census also had an approach to identifying same-sex relationships. The incidence identified by the Census was much lower than that found in the Negotiating the Life Course Survey. 3 The percentages ever married in the survey are a little higher than those recorded at the 1996 Census. The equivalent percentages ever married in this age range from the 1996 Census were 90 per cent for men and 94 per cent for women. CHAPTER 8 CHANGING PATTERNS OF POPULATION DISTRIBUTION 1 Plane and Rogerson (1994, p. 31) define this as follows: ‘The population centroid, also called the mean centre, the mean point, the centre of gravity, or sometimes simply the centre of population. Conceptually, if the mythological Atlas were to hold up the entire area for which a centre is being

282 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

computed—let’s say the United States—and assuming that people were the only objects contributing to the weight (and also assuming everyone weighs the same!), the point where he would have to stand to balance the country would be the centroid.’ CHAPTER 9 TRANSFORMATIONS IN THE LABOUR FORCE 1 We are grateful to Rebecca Kippen and Peter McDonald for providing their unpublished data on population and labour force projections. We would also like to thank Tony Salvage for research assistance and Stephanie Hancock for production. The views and opinions expressed in this chapter are our own and are not necessarily those of our employer institutions. 2 See ABS 1991 for precise definitions of both the activity test and the availability test. 3 Details of this equation are available on request from the authors. 4 The detailed breakdown shows that one-third of the growth of the labour force will occur in the 45–54 age group while only 20 per cent of the growth will occur in the 20–44 age group.

BIBLIOGRAPHY

Adam, AY (1992) The ABS Population Projections: Overview and Evaluation, Journal of the Australian Population Association, 9(2): 109–30. Altman, JC (ed.) (1991) Aboriginal Employment Equity by the Year 2000, Academy of the Social Sciences in Australia and Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. Altman, JC and Nieuwenhuysen, J (1979) The Economic Status of Australian Aborigines, Cambridge University Press, Cambridge. Altman, JC and Sanders, W (1994) From Exclusion to Dependence: Aborigines and the Welfare State in Australia. In J Dixon and R Scheurell (eds.) Social Welfare for Indigenous Populations, Routledge, London, pp. 206–29. Altman, JC, Gillespie, D and Palmer, K (1998) National Review of Outstation Resource Agencies, unpublished report to the Aboriginal and Torres Strait Islander Commission, Canberra. Anderson, I (1997) I, the ‘Hybrid’ Aborigine: Film and Representation, Australian Aboriginal Studies, 1: 4–14. Anderson, I and Brady, M (1995) Performance Indicators for Aboriginal Health Services, CAEPR Discussion Paper No. 81, Centre for Aboriginal Economic Policy Research, Australian National University, Canberra. Australian Bureau of Statistics (ABS) (annual-a) Deaths, Australia, Catalogue No. 3302.0, Canberra. —— (annual-b) Estimated Resident Population by Marital Status, Age and Sex, Australia, Catalogue No. 3220.0, Canberra. —— (1976) Labour Force Status and Other Characteristics of Families, November 1974, Canberra. —— (1990a) Census 86: Data Quality – Ancestry, Catalogue No. 2603.0, Canberra. —— (1990b) Population Growth and Distribution in Australia, Catalogue No. 2504.0, Canberra. —— (1991) Measuring Employment and Unemployment, Catalogue No. 6279.0, Canberra. —— (1992) Labour Force Projections, Australia, 1992–2005, Catalogue No. 6260.0, Canberra. —— (1993) Disability, Ageing and Carers, Australia, 1993, Catalogue No. 4430.0, Canberra.

284 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

—— (1994) Australia’s Long-Term Unemployed, ABS Catalogue No. 6255.0, Canberra. —— (1995) Causes of Death, Australia, Catalogue No. 3303.0, Canberra. —— (1996) Australian Standard Geographical Classification (ASGC) Review Phase One Report, Australian Government Publishing Service, Canberra. —— (1997) Australian Standard Geographical Classification (ASGC) Review Draft Phase Two Report, Australian Government Publishing Service, Canberra. —— (1998a) Experimental Projections of the Aboriginal and Torres Strait Islander Population: 1996–2006, Catalogue No. 2035.0, Canberra. —— (1998b) Labour Force Status and Other Characteristics of Families, June 1998, Catalogue No. 6224.0, Canberra. —— (1998c) Population Growth and Distribution, Australia, 1996, Catalogue No. 2035.0, Canberra. —— (1998d) Population Projections, Australia, 1997–2051, Catalogue No. 3222.0, Canberra. —— (1999a). Births, Australia, 1998, Catalogue No. 3301.0, Canberra. —— (1999b) Children, Australia: A Social Report, Catalogue No. 4119.0, Canberra. —— (1999c) Labour Force Projections, Australia, 1999–2016, Catalogue No. 6260.0, Canberra. —— (2000a) Deaths, Australia, 1999, Catalogue No. 3302.0, Canberra. —— (2000b) Migration, Australia, 1998–99, Catalogue No. 3412.0 Canberra. —— (2000c) Population Projections, Australia, 1999–2101 Catalogue No. 3222.0. Canberra. —— (2001) Census Customised Matrix, Centre for Population and Urban Research, Monash University, Melbourne. —— (2002a) 2001 Census of Population and Housing, Selected Social and Housing Characteristics, Catalogue No. 2015.0. —— (2002b) Australian Demographic Statistics, December Quarter 2001, Catalogue No. 3101.0, Australian Government Publishing Service, Canberra. —— (2002c) Australian Social Trends, Catalogue No. 4102.0, Canberra. —— (2002d) Population Distribution: Aboriginal and Torres Strait Islander Australians, Catalogue No. 4705.0, Canberra. —— (2002e) Year Book Australia 2002, Catalogue No. 1301.0, Australian Government Publishing Service, Canberra. Australian Bureau of Statistics and Australian Institute of Health and Welfare (AIHW) (1999) The Health and Welfare of Australia’s Aboriginal and Torres Strait Islander Peoples, Catalogue No. 4704.0, Canberra. Australian Government Actuary (AGA) (1999) Australian Life Tables 1995–97, Canberra. Australian Institute of Health and Welfare (AIHW) (1998) International Health: How Australia Compares, Canberra. —— (1999a) Australia’s Welfare 1999: Services and Assistance, Canberra. —— (1999b) The Burden of Disease and Injury in Australia, Canberra. —— (2000) Australia’s Health 2000, Canberra. Australian Population and Immigration Council (1977) Immigration Policies and Australia’s Population: A Green Paper, Australian Government Publishing Service, Canberra. Bachrach, CA (1980) Childlessness and Social Isolation Among the Elderly, Journal of Marriage and the Family, 42: 627–37. Bacon, B (2000) Projecting the Decline in Fertility, Paper presented at the Workshop on Declining Fertility in Australia: Policy and Research Issues, Australian Institute of Family Studies, 30 May, Melbourne. Barnes, A (1996) Recent Developments in the Collection of Indigenous

BIBLIOGRAPHY

• 285

Statistics. In JC Altman and J Taylor (eds.) The 1994 National Aboriginal and Torres Strait Islander Survey: Findings and Future Prospects, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra, pp. 13–28. Baxter, J (1996) Gender Equality and Participation in Domestic Labour: A Cross National Perspective, Journal of Comparative Family Studies, 28(3): 220–47. Beggs, J and Chapman, B (1988) The Forgone Earnings from Childrearing in Australia, Centre for Economic Policy Research Discussion Paper 190, The Australian National University, Canberra. Bell, M (1992) Internal Migration in Australia, 1981–1986, Australian Government Publishing Service, Canberra. —— (1995) Internal Migration in Australia, 1986–1991: Overview Report, Australian Government Publishing Service, Canberra. Bell, M and Hugo, GJ (2000) Internal Migration in Australia 1991–96: Overview and the Overseas-Born, Australian Government Publishing Service, Canberra. Bell, M and Ward, G (1998) Who are the Visitors? Measuring Circulation in Australia, Paper presented to the Ninth Biennial Conference of the Australian Population Association, University of Queensland, Brisbane, 29 September – 2 October. Betts, K (1999) The Great Divide, Duffy and Snellgrove, Sydney. Birrell, B (1990) The Chains that Bind: Family Reunion Migration to Australia in the 1980s, Bureau of Immigration Research, Australian Government Publishing Service, Canberra. —— (1997) Immigration Reform in Australia: Coalition Government Proposals and Outcomes since March 1996, Centre for Population and Urban Research Monash University, Melbourne. Birrell, B and Hawthorne, L (1997) Immigration and the Professions in Australia, Centre for Population and Urban Research, Monash University, Melbourne. Birrell, B and Seol, BS (1998) Sydney’s Ethnic Underclass, People and Place 6(3): 16–29. Birrell, R (1992) Problems of Immigration Control in Liberal Democracies: The Australian experience. In G Freeman and J Jupp (eds.) Nations of Immigrants, Oxford University Press, Oxford, pp. 23–40. —— (1995) Spouse Migration to Australia, People and Place, 3(1): 9–16. Blaikie, A (1999) Ageing and Popular Culture, Cambridge University Press, Cambridge. Bogle, J (1996) Feminism and the Family. In B Muehlenberg et al. (eds.) The Family: There is No Other Way, Australian Family Association, Melbourne. Booth, H and Tickle, L (n.d.) The Future Aged: New Projections of Australia’s Elderly Population. Booth, H, Maindonald, J and Smith, L (2002) Applying Lee-Carter under Conditions of Variable Mortality Decline, Population Studies, 56(3): pp. 325–60. Borjas, G (1999) Heaven’s Door, Princeton University Press, Princeton. Borrie, WD (1954) Italians and Germans in Australia: A Study of Assimilation, Cheshire, Melbourne. —— (1994) Progress in Australian Demography, Journal of the Australian Population Association, 11(1): 1–9. Bowie, IJS (1987) The Urban and Non-Urban Populations of Australia 1800–1981, Mitchell Occasional Paper in Geography, Mitchell College of Advanced Education, Bathurst. Budd, A, Levine, P and Smith, P (1988) Unemployment, Vacancies and LongTerm Unemployed, Economic Journal, 98: 1071–91.

286 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Bureau of Immigration Research (BIR) (1991) Australian Immigration Consolidated Statistics No. 16 1989–90, Australian Government Publishing Service, Canberra. Burnley, IH (2001) The Impact of Immigration in Australia: A Demographic Approach, Oxford University Press, Melbourne. Burnley, IH and Routh, N (1985) Aboriginal Migration to Inner Sydney. In IH Burnley and J Forrest (eds.) Living in Cities: Urbanism and Society in Metropolitan Australia, Allen & Unwin, Sydney, pp. 199–215. Burnley, IH and PA Murphy (1995) Ex-Urban Development in Australia and the United States: Through a Glass Darkly, Journal of Planning Education and Research, 14: 245–54. Caldwell, JC (1994) New Challenges for Demography, Journal of the Australian Population Association, 11(1): 9–21. Carlson, A (1996) What is the Family? In B Muehlenberg et al. (eds.) The Family: There is No Other Way, Australian Family Association, Melbourne. Carmichael, GA (1987) Bust After Boom: First Marriage Trends in Australia, Demography, 24(2): 245–64. —— (1988) With This Ring: First Marriage Patterns, Trends and Prospects in Australia, Department of Demography, The Australian National University and Australian Institute of Family Studies, Canberra. —— (1995) Consensual Partnering in the More Developed Countries, Journal of the Australian Population Association, 12(1): 51–86. —— (1996) From Floating Brothels to Suburban Semi-Respectability: Two Centuries of Non-Marital Pregnancy in Australia, Journal of Family History, 21(3): 281–315. —— (1998) Things Ain’t What They Used to Be! Demography, Mental Cohorts, Morality and Values in Post-War Australia, Journal of the Australian Population Association, 15(2): 91–113. Carmichael, GA, Webster, A and McDonald, P (1997) Divorce Australian Style: A Demographic Analysis, Journal of Divorce and Remarriage, 26(3/4): 3–37. Castles, F (2002) Three Facts about Fertility: Cross-National Lessons for the Current Debate, Family Matters, 63, Australian Institute of Family Studies, Melbourne. Castles, S, Kalantzis, M, Cope, B and Morrissey, M (1992) Mistaken Identity, Third Edition, Pluto Press, Sydney. Chadwick, R (2001) Population: Why Population Policy Matters, BCA Papers, Business Council of Australia, Melbourne. Chapman, B, Dunlop, Y, Gray, M, Liu, A and Mitchell, D (1999) Children and Their Mother’s Earnings. Paper presented to Conference on Labour Market Trends and Family Policies: Implications for Children, Centre for Economic Policy Research, The Australian National University, Canberra. Chapman, BJ, Junankar, PN and Kapuscinski, CA (1992) Projections of LongTerm Unemployment, Australian Bulletin of Labour, 18: 195–207. Clyne, M and Kipp, S (1995) The Extent of Community Language Maintenance in Australia, People and Place, 3(4): 4–8. Coleman, D (1995) Spouse Migration from the Indian Sub-Continent to the UK: A Permanent Migration Stream, People and Place, 3(1): 9–16. Coleman, JS (1993) The Rational Reconstruction of Society, American Sociological Review, 58: 1–15. Collins, J (1975) The Political Economy of Post-War Migration. In E Wheelright and K Buckley (eds.) Essays in the Political Economy of Australian Capitalism Vol. 1, Australian and New Zealand Book Company, Sydney. Committee to Advise on Australia’s Immigration Policies (CAAIP) (1988) Immigration: A Commitment to Australia, Australian Government Publishing Service, Canberra.

BIBLIOGRAPHY

• 287

Commonwealth of Australia (1978) Population and Australia: Recent Demographic Trends and Their Implications, Supplementary Report of the National Population Inquiry, Australian Government Publishing Service, Canberra. —— (1987) Return to Country: The Aboriginal Homelands Movement in Australia, Report to the House of Representatives Standing Committee on Aboriginal Affairs, Australian Government Publishing Service, Canberra. —— (1994) Australia: National Report on Population for the United Nations International Conference on Population and Development, Cairo, Australian Government Publishing Service, Canberra. Connidis, IA and McMullin, JA (1993) To Have or Have Not: Parent Status and the Subjective Well-Being of Older Men and Women, Gerontologist, 33: 630–36. Coombs, HC, Dexter, BG and Hiatt, LR (1982) The Outstation Movement in Aboriginal Australia. In E Leacock and RB Lee (eds.) Politics and History in Band Societies, Cambridge University Press, Cambridge, pp. 427–41. Coughlan, JE and McNamara, DJ (eds.) (1996) Asians in Australia: Patterns of Migration and Settlement, MacMillan, Melbourne. Crean, S (2002) Population Policy, Address to the Population Summit 2002, 25 February, Melbourne. Culpin, A, Nugent, S and Truscott, J (2000) Population Projections for New South Wales—A Regional Analysis, People and Place, 8(1): 12–20. DaVanzo, J and Chan, A (1994) Living Arrangements of Older Malaysians: Who Coresides With Their Adult Children? Demography, 31(1): 95–113. Davis, JS (1990) Ex-Urban Community Patterns: A Case Study of the Portland Oregon Region, PhD Dissertation, Portland State University, Portland, Oregon. Deeble, J, Mathers, C, Smith, L, Goss, J, Webb, R and Smith, V (1998) Expenditures on Health Services for Aboriginal and Torres Strait Islander People, Australian Institute of Health and Welfare and National Centre for Epidemiology and Population Health, Canberra. Department of Immigration (1978) Extension of Migrant Entry Policy, unpublished paper. Department of Immigration and Ethnic Affairs (1994) Population Flows: Immigration Aspects, Canberra. Department of Immigration and Multicultural Affairs (DIMA) (1997) Classification of Countries into English Proficiency Groups, 1996 Census data, Statistical Focus, Canberra. —— (1999) Review of the Independent and Skilled-Australian Linked Categories, Canberra. —— (2001) State/Territory Specific Migration, Fact Sheet No. 26, Canberra. —— (2002) Population Flows: Immigration Aspects, 2001 edition, Canberra. Department of Social Security (1996) Annual Report 1995–1996, Australian Government Publishing Service, Canberra. Dodson, M (1994) The Wentworth Lecture: The End in the Beginning: Re(de)fining Aboriginality, Australian Aboriginal Studies, 1: 2–13. Dowling, P (1997) Asset-Rich and Cash-Poor?: Socio-Economic Contradictions in Retirement Migration, BA Hons Thesis, Department of Sociology, The Australian National University, Canberra. Dugbaza, T (1994) Recent Trends and Differentials in Aboriginal and Torres Strait Islander Fertility: 1981–1991, Demography Working Paper 94/1, Australian Bureau of Statistics, Canberra. Dugdale, AE, Musgrave, IA and Streatfield, K (1990) Causes and Health Policy Implications of Low Birthweights of Aboriginal Children. In A Gray (ed.) A Matter of Life and Death: Contemporary Aboriginal Mortality, Aboriginal Studies Press, Canberra, pp. 25–33.

288 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Elman, C and Uhlenberg, P (1995) Coresidence in the Early Twentieth Century: Elderly Women in the United States and Their Children, Population Studies, 49: 501–17. Eurostat (2002) First Results of the Demographic Data Collection for 2001 in Europe, Luxembourg. Evans, S (1994) Welfare Dependency Amongst Recently Arrived Aged Migrants, People and Place, 2(2): 35–37. Evans, M and Kelley, J (1999) Small families or large? Australia in International Perspective, Australian Social Monitor, 2(1): 13–19. Fahrer, J and Pease, A (1993) The Unemployment-Vacancy Relationship in Australia, Australian Economic Review, 26(4): 43–57. Faulkner, HW and French, S (1983) Geographical Remoteness: Conceptual and Measurement Problems, Reference Paper No. 54, Bureau of Transport Economics, Canberra. Fishman, R (1990) Megalopolis Unbound, Wilson Quarterly, 14: 25–48. Fisk, EK (1985) The Aboriginal Economy in Town and Country, George Allen & Unwin, Sydney. Freeman, G (1995) Modes of Immigration Politics in Liberal Democratic States, International Migration Review, 29(4): 881–902. Freeman, G and Birrell, B (2001) Divergent Paths of Immigration Politics in the United States and Australia, Population and Development Review, 27(3): 525–51. Frey, WH (1995) Immigration and Internal Migration ‘Flight’: A California Case Study, Population and Environment, 16: 353–75. Friedmann, J and Miller, J (1965) The Urban Field, American Institute of Planning Journal, 31(4): 312–19. Gale, F (1972) Urban Aborigines, Australian National University Press, Canberra. Gale, F and Wundersitz, J (1982) Adelaide Aborigines: A Case Study of Urban Life 1966–1981, Development Studies Centre, The Australian National University, Canberra. Gilding, M (1997) Australian Families: A Comparative Perspective, Longman, Melbourne. Gray, A (1985) Some Myths in the Demography of Aboriginal Australians, Journal of the Australian Population Association, 2(2): 136–49. —— (1989) Aboriginal Migration to the Cities, Journal of the Australian Population Association, 6(2): 122–44. —— (1990a) National Estimates of Aboriginal Mortality. In A Gray (ed.) A Matter of Life and Death: Contemporary Aboriginal Mortality, Aboriginal Studies Press, Canberra, pp. 147–59. —— (1990b) Aboriginal Fertility: Trends and Prospects, Journal of the Australian Population Association, 7(1): 57–77. —— (1997) The Explosion of Aboriginality: Components of Indigenous Population Growth 1991–99, CAEPR Discussion Paper No. 142, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. —— (1998) Parentage and Indigenous Population Change, CAEPR Discussion Paper No. 166, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. Gray, A and Tesfaghiorghis, H (1993) Aboriginal Population Prospects, Journal of the Australian Population Association, 10(2): 81–101. Gray, A, Trompf, P and Houston, S (1991) The Decline and Rise of Aboriginal Families. In J Reid and P Trompf (eds.) The Health of Aboriginal Australia, Harcourt Brace Jovanovich, Sydney, pp. 80–123. Gregory, R (1999) An Australian Overview, paper presented to Conference on Labour Market Trends and Family Policies: Implications for Children, Centre for Economic Policy Research, The Australian National University, Canberra.

BIBLIOGRAPHY

• 289

Gregory, RG and Hunter, B (1995) The Macroeconomy and the Growth of Ghettos and Urban Poverty in Australia, Discussion Paper No. 325, Centre for Economic Policy Research, The Australian National University, Canberra. Haberkorn, G, Hugo, GJ, Fisher, M and Aylward, R (1999) Country Matters: A Social Atlas of Rural and Regional Australia, Bureau of Rural Sciences, Canberra. Hawthorne, L (1995) The Politicisation of English Part 2, People and Place, 3(1): 17–28. Headey, B (1999) Old Age is Not Downhill: The Satisfactions and Well-Being of Older Australians, Australasian Journal on Ageing, 18 (3 Supplement): 32–37. Healy, E (1994) Unemployment Dependency Rates Amongst Recently Arrived Migrants: An Update, People and Place, 2(3): 47–54. —— (1996) Welfare Benefits and Residential Concentrations amongst Recently Arrived Migrant Communities, People and Place, 4(2): 20–31. Hogg, RS (1992) Indigenous Mortality: Placing Australian Aboriginal Mortality within a Broader Context, Social Science and Medicine, 35(3): 335–46. Hollifield, J (1992) Immigrants Markets and States: The Political Economy of Postwar Europe, Harvard University Press, Cambridge MA. Holmes, JH (1988) Remote Settlements. In RL Heathcote (ed.) The Australian Experience: Essays in Australian Land Settlement and Resource Management, Longman, Melbourne, pp. 75–102. —— (2002) Diversity and Change in Australia’s Rangelands: A PostProductivist Transition with a Difference?, Transactions of the Institute of British Geographers, 27, forthcoming. House of Representatives Standing Committee on Long-Term Strategies (1994) Australia’s Population ‘Carrying Capacity’: One Nation—Two Ecologies, Australian Government Publishing Service, Canberra. Hugo, GJ (1989) Australia: The Spatial Concentration of the Turnaround. In AG Champion (ed.) Counterurbanization: The Changing Pace and Nature of Population Deconcentration, Edward Arnold, London. —— (1994a) Australia’s Population Growth, Composition and Distribution: Emerging Research Needs, Journal of the Australian Population Association, 11(1): 55–83. —— (1994b) The Turnaround in Australia: Some First Observations from the 1991 Census, Australian Geographer, 25(1): 1–17. —— (1996) The Relationship between Immigration and Internal Migration in Sydney, Australia. Paper prepared for 28th International Geographical Congress, The Hague. —— (1999) Demographic Trends Influencing Housing Needs and Demands in Australia. Paper presented to AHURI Workshop on Innovation in Housing, 15 April, Melbourne. Hugo, GJ, Griffith, D, Rees, P, Smailes, P, Badcock, B and Stimson, R (1997) Rethinking The ASGC: Some Conceptual and Practical Issues, Monograph Series 3, National Key Centre for Social Applications of Geographical Information Systems, University of Adelaide, Adelaide. Hunter, B (1998) Assessing the Validity of Intercensal Comparisons of Indigenous Australians, 1986–96, Journal of the Australian Population Association, 15(1): 51–69. Jain, SK (1989) Estimation of Aboriginal Fertility, 1971–86: An Application of the Own-Children Method of Fertility Estimation, Catalogue. No. 4127.0, Australian Bureau of Statistics, Canberra. —— (1992) Recent Trends in Mortality in Australia: An Analysis of the Causes of Death through the Application of Life Table Techniques, Journal of the Australian Population Association, 9(1): 1–24. —— (1994) Trends in Mortality by Causes of Death in Australia, the States and

290 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

Territories during 1971–92, and in Statistical Divisions and Sub-divisions during 1991–92, National Centre for Epidemiology and Population Health and Australian Bureau of Statistics, Canberra. Jain, SK and McDonald, P (1997) Fertility of Australian Birth Cohorts: Components and Differentials, Journal of the Australian Population Association, 14(1): 31–46. Jayasuriya, L and Kee, PK (1999) The Asianisation of Australia? Some Facts about the Myths, Melbourne University Press, Melbourne. Jordan, DE (1985) Census Categories—Enumeration of Aboriginal people, or Construction of Identity?, Australian Aboriginal Studies, 1: 28–36. Jorm, AF, Christense, H, Henderson, AS, Jacomb, PA, Korten, AE and Mackinnon, A (1998) Factors Associated with Successful Ageing, Australasian Journal on Ageing, 17(1): 33–37. Junankar, PN and Kapuscinski, CA (1991) The Incidence of Long-Term Unemployment in Australia, Australian Bulletin of Labour, 17(4): 325–52. —— (1992) The Costs of Unemployment in Australia, Economic Planning Advisory Council, Background Paper No. 24, Canberra. Jupp, J (2001) The Australian People: An Encyclopedia of the Nation, Its People and Their Origins, Cambridge University Press, Cambridge, UK. Kendig, HL (1981) Housing and Living Arrangements of the Aged. In AL Howe (ed.) Towards an Older Australia: Readings in Social Gerontology, University of Queensland Press, Brisbane, pp. 85–101. —— (1984) Housing Tenure and Generational Equity, Aging and Society, 4: 249–72. Kenen, R (1987) Health Status: Australian Aborigines and Native Americans— A Comparison, Australian Aboriginal Studies, 1: 34–46. Khoo, SE (1995) Language Maintenance amongst the Second Generation, People and Place, 3(4): 9–13. —— (1997) Sponsors of Spouse Migration to Australia, Department of Immigration and Multicultural Affairs, Canberra. —— (2001) The Context for Spouse Migration to Australia, International Migration, 39 (1): 111–132. Khoo, SE, Kee, PK, Shu, J and Dang, T (1994) Asian Immigrant Settlement and Adjustment in Australia, Asian and Pacific Migration Journal, 3(2–3): 339–72. Khoo, SE and Price, CA (1996) Understanding Australia’s Ethnic Composition, Department of Immigration and Multicultural Affairs, Canberra. Kinfu, Y and Taylor, J (2002) Components of Indigenous Population Change, 1996–2001, CAEPR Discussion Paper No. 189, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. King, A and McDonald, P (1999) Private Transfers Across Australian Generations, NATSEM Discussion Paper No. 41, National Centre for Social and Economic Modelling, Canberra. Kinnaird, B (2002) Australia’s Migration Policy and Skilled ICT Professionals: The Case for an Overhaul, People and Place, 10 (2): 55–69. Kinslor, J (2000) Sending Asylum Seekers Elsewhere: Recent Developments in Australian ‘Safe Third Country’ Law, People and Place, 8 (2): 53–68. Kippen, R (2001) Trends in Age- and Parity-Specific Fertility. Paper presented to International Union for the Scientific Study of Population meeting on Low Fertility: Trends, Theories and Policies, 21–23 March, Tokyo. Kippen, R and McDonald, P (2000) Australia’s Population in 2000: The Way We Are and the Ways We Might Have Been, People and Place, 8(3): 10–17. —— (1998) Achieving Population Targets for Australia: An Analysis of the Options, People and Place, 6(2): 11–23. Krugman, P (1995) Development, Geography, and Economic Theory, MIT Press, Cambridge, MA.

BIBLIOGRAPHY

• 291

Kunitz, SJ (1990) Public Policy and Mortality among Indigenous Populations of Northern America and Australasia, Population and Development Review, 16(4): 647–72. —— (1994) Disease and Social Diversity: The European Impact on the Health of Non-Europeans, Oxford University Press, New York. Langton, M (1981) Urbanizing Aborigines: The Social Scientist’s Great Deception, Social Alternatives, 2(2): 16–21. Laslett, P (1989) A Fresh Map of Life: The Emergence of the Third Age, Weidenfeld and Nicolson, London. Lee, RD and Carter, LR (1992) Modelling and Forecasting U.S. Mortality, Journal of the American Statistical Association, 87(419): 659–71. Liberal Party (1991) Australians Speak, Australia 2000, Liberal Party of Australia, unpublished paper. Liu, Z (1998) The Probability of Nursing Home Use Over a Lifetime, Working Paper No. 16, Australian Institute of Health and Welfare, Canberra. Logan, MI, Maher, CA, McKay, J and Humphries, JS (1975) Urban and Regional Australia: Analysis and Policy Issues, Sorrett Publishing, Melbourne. Macunovich, DJ, Easterlin, RA, Schaeffer, CM and Crimmins, EM (1995) Echoes of the Baby Boom and Bust: Recent and Prospective Changes in Living Alone Among Elderly Widows in the United States, Demography, 32 (1): 17–28. Maher, C and McKay, J (1986) 1981 Internal Migration Study Final Report: Internal Migration in Australia, Department of Immigration and Ethnic Affairs, Canberra. Maher, C and Stimson, RJ (1994) Regional Population Growth in Australia, Australian Government Publishing Service, Canberra. Maher, C, Whitlaw, J, McAllister, FR, Palmer, J, Chee, E and Taylor, P (1992) Mobility and Locational Disadvantage Within Australian Cities, BIMPR and Department of Prime Minister and Cabinet, Canberra. Mason, R and Randell, S (1997) Education Policy for an Ageing Society. In A Borowski, S Encel and E Ozanne (eds.) Ageing and Social Policy in Australia, Cambridge University Press, Cambridge, pp. 211–31. Mathers, C (1995) Mortality Patterns of Urban, Rural and Remote Populations in Northern Australia, People and Place, 3(2): 15–24. —— (1996) Trends in Health Expectancies in Australia 1981–1993, Journal of the Australian Population Association, 13(1): 1–15. McDermott, RA, Plant, AJ and Mooney, G (1996) Has Access to Hospital Improved for Aborigines in the Northern Territory?, Australian and New Zealand Journal of Public Health, 20(6): 589–93. McDonald, P (1975) Marriage in Australia: Age at First Marriage and Proportions Marrying, 1860–1971, Department of Demography, The Australian National University, Canberra. —— (1995) Families in Australia: A Socio-Demographic Perspective, Australian Institute of Family Studies, Melbourne. —— (1996) Demographic Life Transitions: An Alternative Theoretical Paradigm. In RM Douglas, G Jones and RM D’Souza (eds.) The Shaping of Fertility and Mortality Declines: The Contemporary Demographic Transition, Supplement to Health Transition Review 6, The Australian National University, Canberra, pp. 385–92. —— (1998a) Changing Families and Changing Relationships in Australia. Paper presented to Conference of the Australian Population Association, Brisbane. —— (1998b) Contemporary Fertility Patterns in Australia: First Data from the 1996 Census, People and Place, 6(1): 1–13. —— (2000a) Gender Equity, Social Institutions and the Future of Fertility, Journal of Population Research, 17(1): 1–16.

292 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

—— (2000b) The Shape of an Australian Population Policy, Australian Economic Review, 33(3): 272–280. —— (2002) Low Fertility: Unifying the Theory and the Demography. Paper presented to Session 73, Future of Fertility in Low Fertility Countries, 2002 Meeting of the Population Association of America, 9–11 May, Atlanta. McDonald, P and Kippen, R (1999a) Population Futures for Australia: The Policy Alternatives, Parliamentary Library Seminar Series, Research Paper 5, 1999, Australian Parliamentary Library, Canberra. —— (1999b) The Impact of Immigration on the Ageing of Australia’s Population, DIMA Discussion Paper, May, Department of Immigration and Multicultural Affairs, Canberra. —— (1999c) Ageing: The Social and Demographic Dimensions. In Policy Implications of the Ageing of Australia’s Population, Productivity Commission and Melbourne Institute of Applied Economic and Social Research, Canberra, pp. 47–70. —— (2001) Labour Supply Prospects in 16 Developed Countries, 2000–2050, Population and Development Review, 27(1): 1–32. —— (2002a) Projecting Future Migration Levels: Should Rates or Numbers be Used? People and Place, 10(1): 82–83. —— (2002b) private communication. McEwin, M (1998) Family Statistics in Australia. In Australian Bureau of Statistics, The Siena Group Seminar, Volume 2, Canberra, pp. 15–20. McIntosh, G (1998) The ‘Boomer Bulge’: Ageing Policies for the 21st Century, Research Paper 4, 1998–99, Parliament of Australia, Parliamentary Library, Canberra, McKenzie, F (1994) Regional Population Decline in Australia, Australian Government Publishing Service, Canberra. —— (1996) Beyond the Suburbs: Population Change in the Major Ex-Urban Regions of Australia, Australian Government Publishing Service, Canberra. McNicoll, G (1993) The Economics of Australian Immigration, with Reference to Trans-Tasman flows. In GA Carmichael (ed.) Trans-Tasman Migration: Trends, Causes and Consequences, Australian Government Publishing Service, Canberra, pp. 315–26. —— (2000) Reflections on ‘Replacement Migration’, People and Place 8(4): 1–13. —— (2001) Government and Fertility in Transitional and Post-Transitional Societies. In RA Bulatao and JB Casterline (eds.) Global Fertility Transition. Supplement to Population and Development Review, Vol. 27, Population Council, New York, pp. 129–59 Menadue, J (1999) Things You Learn along the Way, David Lovell Publishing, Melbourne. Merlo, R and Rowland, D (2000) The Prevalence of Childlessness in Australia, People and Place, 8(2): 21–32. Money, J (1999) Fences and Neighbors: The Political Geography of Immigration Control, Cornell University Press, New York. Murray-Darling Basin Commission (1999) The Salinity Audit of the MurrayDarling Basin: A 100-Year Perspective, Canberra. National Population Council (NPC) (1991) Population Issues and Australia’s Future: Environment, Economy and Society, A Discussion Paper of the Population Issues Committee, Australian Government Publishing Service, Canberra. —— (1992) Population Issues and Australia’s Future: Environment, Economy and Society, Final Report of the Population Issues Committee, Australian Government Publishing Service, Canberra. National Population Inquiry (1975) Population and Australia: A Demographic

BIBLIOGRAPHY

• 293

Analysis and Projection, Australian Government Publishing Service, Canberra. —— (1978) Population and Australia: Recent Demographic Trends and their Implications, Supplementary Report, Australian Government Publishing Service, Canberra. Nelson, AC and Dueker, J (1990) The Ex-Urbanisation of America and its Planning Policy Implications, Journal of Planning Education and Research, 9: 91–100. Neutze, GM (1963) Decentralisation Dialogue, Current Affairs Bulletin, 31(8): 115–27. Newspoll (2002) The Australian Inquirer, 14–15 September, 2002, p. 26. O’Connor, K and Stimson, RJ (1996) Convergence and Divergence of Demographic and Economic Trends. In PW Newton and M Bell (eds.) Population Shift: Mobility and Change in Australia, Australian Government Publishing Service, Canberra, pp. 108–25. Olshansky, SJ and Ault, AB (1986) The Fourth Stage of the Epidemiologic Transition: The Age of Delayed Degenerative Diseases, The Milbank Quarterly, 64(3): 355–91. Omran, AR (1971) The Epidemiologic Transition, The Milbank Quarterly, 49(4): 509–38. —— (1977) Epidemiologic Transition in the United States: The Health Factor in Population Change, Population Bulletin, 32(2): 1–42. Organization for Economic Co-operation and Development (OECD) (1999) Coping With Population Ageing. In OECD. Economic Surveys 1998–1999, Australia, OECD Publications, Paris, pp. 99–163. Pahl, RE (1965) Urbs in Rure: The Metropolitan Fringe in Hertfordshire, Geographical Papers No. 2, London School of Economics and Political Science, London. Passel, JS (1996) The Growing American Indian Population, 1969–1990: Beyond Demography. In GD Sandefur, RR Rindfuss and B Cohen (eds.) Changing Numbers, Changing Needs: American Indian Demography and Public Health, National Academy Press, Washington DC, pp. 79–103. Penny, J and Khoo, SE (1996) Intermarriage: A Study of Migration and Integration, Australian Government Publishing Service, Canberra. Peron, Y, Desrosiers, H, Juby, H, Lapierre-Adamcyk, E, Le Bourdais, C, MarcilGratton, N and Mongeau, J (1999) Canadian Families at the Approach of the Year 2000, Statistics Canada, Ottawa. Peto, R, Lopez, AD, Boreham, J, Thun, M and Clark, H (1994) Mortality from Smoking in Developed Countries 1950–2000: Indirect Estimates from National Vital Statistics, Oxford University Press, Oxford. Piggott, J and Chapman, BJ (1995) Costing the Job Compact, Economic Record, 71(215): 313–28. Plane, DA and Rogerson, PA (1994) The Geographical Analysis of Population: With Applications to Planning and Business, John Wiley, New York. Pollack, DP (2001) Indigenous Land in Australia: A Quantitative Assessment of Indigenous Landholdings in 2000, CAEPR Discussion Paper No. 221, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. Pollard, JH (1982) The Expectation of Life and Its Relationship to Mortality, Research Paper No 247, School of Economics and Financial Studies, Macquarie University, Sydney. Price, CA (1988) Ethnic Groups in Australia, Australian Immigration Research Centre, Canberra. —— (1993) Ethnic Intermixture in Australia, People and Place, 1(1): 6–8. —— (1994) Ethnic Intermixture in Australia, People and Place, 2(4): 8–11.

294 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

—— (1996) Immigration and Ethnicity, Department of Immigration and Multicultural Affairs, Canberra. —— (2000) Australians All: Who on Earth Are We?, Canberra. Rempel, J (1985) Childless Elderly: What Are They Missing?, Journal of Marriage and the Family, 47: 343–48. Ross, K (1999) Population Issues, Indigenous Australians, Catalogue No. 4708.0, Australian Bureau of Statistics, Canberra. Ross, K and Taylor, J (2002) Improving Life Expectancy and Health Status: A Comparison of Indigenous Australians and New Zealand Maori, Joint Special Issue: Journal of Population Research and NZ Population Review, September 2002, pp. 219–38. Rowland, DT (1982) Population Growth and Distribution. In Population of Australia, Volume 1, ESCAP Country Monograph Series No. 9, United Nations, New York, pp. 10–33. —— (1992) Ageing in Australia: Population Trends and Social Issues, Longman Cheshire, Melbourne. —— (1994) Family Change and Marital Status in Later Life, Australian Journal on Ageing, 13: 168–71. —— (1996a) Population Momentum as a Measure of Ageing, European Journal of Population, 12: 41–61. —— (1996b) Cohort Survival in Ageing Populations, Genus, 52: 71–82. —— (1998a) The Prevalence of Childlessness in Cohorts of Older Women, Australasian Journal on Ageing, 17(1): 18–23. —— (1998b) Consequences of Childlessness in Later Life, Australasian Journal on Ageing, 17(1): 24–28. Rowley, CD (1971) The Remote Aborigines, Australian National University Press, Canberra. Royal Commission on the Decline of the Birth Rate and on the Mortality of Infants in New South Wales (1904) Report, New South Wales Government Printer, Sydney. Ruddock, P (2000) Australian Centre for Population Research: Public Seminar Series, People and Place, 8(4): 14–21. Ruzicka, LT and Choi, CY (1981) Recent Decline in Australian Fertility. In Australian Bureau of Statistics, Year Book Australia 1981, Canberra, pp. 114–27. —— (1999) Youth Suicide in Australia, Journal of the Australian Population Association, 16(1/2): 29–46. Sanders, W (1991) Destined to Fail: The Hawke Government’s Pursuit of Statistical Equality in Employment and Income between Aborigines and other Australians by the Year 2000 (or a Cautionary Tale Involving the New Managerialism and Social Justice Strategies), Australian Aboriginal Studies, 2: 13–18. Santow, G (1991) Trends in Contraception and Sterilization in Australia, Australian and New Zealand Journal of Obstetrics and Gynaecology, 31(3): 201–08. Schoeni, RF (1998) Reassessing the Decline in Parent-Child Old-Age Coresidence During the Twentieth Century, Demography, 35(3): 307–13. Shanas, E, Townsend, P, Wedderburn, D, Friis, H, Mihoj, P and Stehouwer, J (eds) (1968) Old People in Three Industrial Societies, Routledge and Kegan Paul, London. Siegel, JS (1990) Review of a Fresh Map of Life: The Emergence of the Third Age, Population and Development Review, 16(2): 363–67. Smailes, PJ (1996) Demographic Response to Rural Restructuring and Counterurbanisation in South Australia, 1981–1991, International Journal of Population Geography, 2: 261–87. Smith, LR (1980a) The Aboriginal Population of Australia, Australian National University Press, Canberra.

BIBLIOGRAPHY

• 295

—— (1980b) Aboriginal Vital Statistics: An Analysis of Trends, Australian Government Publishing Service, Canberra. —— (1980c) New Black Town or Black New Town: The Urbanisation of Aborigines. In IH Burnley, Pryor, RJ and Rowland, DT (eds.) Mobility and Community Change in Australia, University of Queensland Press, Brisbane, pp. 193–209. Snipp, MC (1997) Some Observations about Racial Boundaries and the Experiences of American Indians, Ethnic and Racial Studies, 20(4): 667–89. Spectorsky, AC (1958) The Exurbanites, J.B. Lippincott, Philadelphia. Stacey, J (1993) Good Riddance to ‘The Family’: A Response to David Popenoe, Journal of Marriage and the Family, 55 (August): 545–49. Statistics Canada (1999) Census Nation Tables, Ottawa. —— (2001) Population Projections for Canada, Provinces and Territories 2000–2026, Ottawa. Statistics New Zealand (1999) Asian Population Projections, Christchurch. —— (2002) Demographic Trends 2001, Catalogue Code 03.009.0001, Wellington. Stevens-Long, J and Commons, M (1992) Adult Life: Development Processes, Mayfield Publishing Company, Mountain View, CA. Streatfield, K, Streatfield, RW, Korzy, M and Musgrave, IA (1990) Are Demographic Factors Important in Low Birthweight among Aboriginal Children? In A Gray (ed.) A Matter of Life and Death: Contemporary Aboriginal Mortality, Aboriginal Studies Press, Canberra, pp. 33–45. Sustainable Population Australia (2002) About Us: Mission, Taylor, J (1992) Geographic Location and Aboriginal Economic Status: A Census-Based Analysis of Outstations in the Northern Territory, Australian Geographical Studies, 30(2): 163–84. —— (1993) Regional Change in the Economic Status of Indigenous Australians, 1986–91, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. —— (1997a) The Contemporary Demography of Indigenous Australians, Journal of the Australian Population Association, 14(1): 77–114. —— (1997b) Changing Numbers, Changing Needs? A Preliminary Assessment of Indigenous Population Change, 1991–1996, CAEPR Discussion Paper No 143, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. —— (2002a) Indigenous Enumeration in the Late Twentieth Century: Emerging Issues for Population Analysis. In G Briscoe and LR Smith (eds.) The Aboriginal Population Revisited: 70,000 Years to the Present, Aboriginal History Monograph 10, The Australian National University, Canberra, pp. 93–109. —— (2002b) The Spatial Context of Indigenous Service Delivery, CAEPR Working Paper No. 16, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. Taylor, J and Bell, M (1996) Mobility among Indigenous Australians. In PW Newton and M Bell (eds.) Population Shift: Mobility and Change in Australia, Australian Government Publishing Service, Canberra, pp. 392–412. —— (1998) Making Sense of Census Data: A Components Analysis of Employment Change among Indigenous Australians, Journal of the Australian Population Association, 15(1): 35–51. —— (1999) Changing Places: Indigenous Population Movement in the 1990s, CAEPR Discussion Paper No. 189, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. —— (2002) The Indigenous Population of Cape York Peninsula, 2001–2006,

296 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

CAEPR Discussion Paper No. 189, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra. Taylor, J and Hunter, B (1998) The Job Still Ahead: Economic Costs of Continuing Indigenous Employment Disparity, Aboriginal and Torres Strait Islander Commission, Canberra. Taylor, R, Lewis, M and Powels, J (1998a) The Australian Mortality Decline: All-Cause Mortality 1788–1990, Australian and New Zealand Journal of Public Health, 22(1): 27–36. —— (1998b) The Australian Mortality Decline: Cause-Specific Mortality 1907–1990, Australian and New Zealand Journal of Public Health, 22(1): 37–44. Taylor, TG (1947) Australia: A Study of Warm Environments and their Effect on British Settlement, Methuen, London. Thomson, N (1983) Aboriginal Infant Mortality, 1976–1981, Australian Aboriginal Studies, (1): 10–15. —— (1990) Trends in Aboriginal Infant Mortality. In A Gray (ed.) A Matter of Life and Death: Contemporary Aboriginal Mortality, Aboriginal Studies Press, Canberra, pp. 1–9. United Nations (1998) Demographic Yearbook No.48, 1996, New York. —— (1998) National Population Policies, New York. —— (1999) Demographic Yearbook No.49, 1999, New York. —— (2000) Global Population Policy: Database 1999, Department of Economic and Social Affairs, Population Division, New York. —— (2002) International Migration 2002, New York. US Bureau of the Census (1996) Resident Population of the United States: Middle Series Projections 2015–2030, by Sex, Race and Hispanic Origin, with Median Age, Washington, D.C. —— (1999) Resident Population Estimates of the United States by Sex, Race and Hispanic Origin: April 1 1990 to June 1 1999, Washington, D.C. US Immigration and Naturalization Service (1997) Statistical Yearbook of the Immigration and Naturalization Service, US Department of Justice, Washington DC. US National Research Council (1997) The New Americans: Economic, Demographic, and Fiscal Effects of Immigration, Report of the Panel on the Demographic and Economic Impacts of Immigration, National Academy Press, Washington, D.C. Van de Kaa, D (1998) Postmodern Fertility Preferences: From Changing Value Orientation to New Behaviour, Working Papers in Demography No. 74, Demography Program, The Australian National University, Canberra. Walker, A (1998) Australia’s Ageing Population: What are the Key Issues and the Available Methods of Analysis? NATSEM Discussion Paper No. 27, National Centre for Social and Economic Modelling, Canberra. Weinick, RM (1995) Sharing a Home: The Experiences of American Women and Their Parents Over the Twentieth Century, Demography, 32(2): 281–97. White, S (1997) The Silent Majority: The Everyday Problems of the Average Australian, Clemenger, Melbourne. Withers, G (2002) Projecting Future Migration Levels: Reply, People and Place, 10(1): 84–86. World Health Organization (WHO) (1995) World Health Statistics Annual 1994, Geneva. Young, CM and Ruzicka, L (1982) Mortality. In Population of Australia, Volume 1, ESCAP Country Monograph Series No. 9, United Nations, New York, pp. 160–182. Young, CM and Ware, H (1979) Contraceptive Use in Australia, Australian and New Zealand Journal of Obstetrics and Gynaecology, 19: 1–6. Young, EA (1995) Third World in the First: Development and Indigenous Peoples, Routledge, London.

INDEX

Aboriginal and Torres Strait Islander 60; ethnic origin 167, 181; languages spoken at home 170; selfidentification as 24–25, 26, 38; see also Indigenous population abortion 72, 102 Accessibility/Remoteness Index of Australia (ARIA) 210, 211 Adelaide 35–36, 173 Afghanistan, immigration from 148 Africa, ethnic origin 167; immigration from 160, 161, 162, 172, 173, 177, 182 Africa, North: ethnic origin 166 aged care 246, 256–57 ageing, healthy 15; see also independent living ageing population viii, 1, 12, 75, 76, 156, 184, 219, 238–65, 267, 269, 276–77, 279; and housing 212; and labour force 234, 235, 236, 273; projections 274–75 Americas, immigration from 161, 162, 173 ‘ancestry’ 159, 163–65 Anglo-Celtic, ancestry 9, 184; ethnic origin 165, 166, 180, 181 Arabic, language spoken at home 169, 170 Asia, ancestry 9, 165; ethnic origin 165–66, 180, 181; immigration from 9, 131, 136, 139, 154, 156, 158, 160, 161–62, 170, 172, 173, 180, 182; languages spoken at home 169 assimilation 133, 174, 183

asylum seekers 148; see also ‘boat people’; immigration, illegal; refugees Australian ancestry 165 Australian Capital Territory 24, 173, 194, 200 Austria 43, 44, 124–25, 166 Belgium 44, 124–25 Benelux countries 43 birthplace 159, 160–63, 180 Blainey, Geoffrey 139 ‘boat people’ 142, 146; see also asylum seekers; immigration illegal; refugees Brisbane 28, 29, 173 Buddhism 170, 171 Cambodia, immigration from 142, 177, 178, 179; language spoken at home 169 Canada, employment 99; ethnic diversity 159, 180; fertility 13, 43, 44, 268; immigrant concentration 174; immigration from 272; immigration to 8, 9, 130, 140, 159, 168; life expectancy 124–25; settlement 5 Canberra 5, 31 cancer 105, 107, 108, 109, 113–114, 116–118, 119, 120, 121, 122, 123 Caribbean countries 10 ‘carrying capacity’ 4 cause of death 105 Census questions: children ever born

298 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

1970–2030

52; Indigenous 18; intermarriage 23; language spoken at home 168 ‘centre of gravity’ see centroid centroid 190, 194–95, 198, 199 child care 16, 90, 95, 98, 99–101, 102; maternal 40; targeted 95 childbearing 15–16; Indigenous 23; preferences 40, 50–51, 75; see also fertility childbirth, age at 13, 14, 40, 42–43, 45, 46–47, 48, 50, 51, 55, 74–75, 268; delaying 41, 47, 48, 92–93, 102, 268; and education 23, 55, 56, 63; and employment 23, 63; loss of earnings 93–94, 268; and marriage 52–55, 65, 72–74; number of 2, 13, 40–41, 42, 48–50, 54, 55, 64; return to work after 93, 94–95, 97, 100; spacing of 94, 98 childlessness 2–3, 13–14, 48, 50, 63, 78, 93; and aged support 243, 250, 253–55 ‘children ever born’ (CEB) 51; see also childbirth, number of China, ancestry 165; ethnic origin 167; immigration from 138, 142, 146, 147, 151, 161, 163, 169, 177, 178, 179; language spoken at home 169, 170, 175; population distribution 5 Christian 170, 171 cohabitation 83, 84, 101, 102; dissolution of 85, 86, 91, 92 contraception 41, 43, 73, 85 Czech Republic (former) 13, 44; ethnic origin 166

employment 189–90, 215, 219, 222–26; family-friendly 99; female 90, 92–94, 94–101, 102, 223, 244–45, 268; growth 219, 220, 223, 224, 225, 226, 227, 228, 231–33; of immigrants 172; Indigenous 36–37; male 94, 244–45; part-time 92, 93, 97, 100; projections 220, 231–33 employment-population ratio 221, 226, 227 ‘enclave demography’ 18 England, ancestry 165, 183–84; ethnic origin 166; see also United Kingdom English language, proficiency 174–76; spoken at home 170, 178, 179–80 environmental issues 129, 157, 184, 214, 215–217, 278, 279 ‘error of closure’ 24, 25 ethnic communities 135, 138, 143, 145, 149, 152, 156–57, 176, 178, 179, 183, 184, 207 ‘ethnic origin’ 165–66, 181 ‘ethnic strength’ 168, 180 Europe, ancestry 165, 183–84; ethnic origin 165, 181, 184; immigration from 9, 131, 132, 136, 160, 162–63, 172, 173, 176, 180, 182; policies 6 Europe, Eastern: fertility 43; immigration from 158, 161, 176 Europe, Southern: fertility 43; immigration from 161, 184 European Union, immigration to 10, 271

Darwin 31 death, cause of 11–12, 105, 106, 107–114, 116, 120–21, 127 decentralisation 201, 215, 216 Denmark 44 dependence, of aged 243, 244, 263 dependency ratios 245 disability 12, 245, 246 disease 104, 105 diversity, ancestry 159, 163–65; ethnic 158–84; language 159, 168–70; regional differences in 172–74; religious 168, 170–71 divorce 14, 87, 88, 89, 91, 101, 251, 252, 260

families 1, 13–15, 77–103, 157, 218, 243, 268; definitions of 77–78; multi-generational 247; oneincome 94; size of 42, 253; soleparent 78, 89, 91, 92, 95, 97; support of aged 246, 250–55, 262; theoretical interpretations of 78–80; see also immigration, family reunion fertility viii, 1, 2–3, 13, 14, 15, 23, 40–76, 156, 240, 242, 266–79; age-specific 45–47, 72–73; differential 51–52; and education 14, 51, 55–57, 74–75; and employment 14, 51, 55, 57–58, 59, 75, 233–34; of immigrants 56, 64–70, 75, 180–81; Indigenous 18–19, 22–24, 26; male 40; and marital status 53–55, 59, 63–64, 65, 73, 75–76; projections 268–69, 273;

education 157, 255(56, 257, 259; female 90, 268; male 223 Egypt, ethnic origin 167; immigration from 133

INDEX

regional 58–60, 75, 199–200; and religion 60–63; replacement level 15, 23, 27, 266; of second-generation immigrants 70–72, 75 Finland 44, 124–25 FitzGerald Report 7 France, ethnic origin 166; fertility 43, 44; immigration to 7; language spoken at home 170; life expectancy 124–25 Freeman, G 129–30, 131 Germany, ancestry 165; ethnic origin 166; fertility 43, 44; immigration from 133, 162, 163, 179; immigration to 7; language spoken at home 169, 170; life expectancy 124–25 Gilding, Michael 78–80 governments, Coalition 133, 134, 148, 149, 150, 151, 152, 153; Fraser 136; Hawke 136, 139; Howard 266; Keating 147; Labor 133, 135, 138, 140, 147, 150, 151, 153, 155, 266; New South Wales 157; state 157; Whitlam 133, 135 Greece, ancestry 165; ethnic origin 166; fertility 44; immigration from 135, 139, 158, 161, 162, 163, 172, 178, 179; language spoken at home 169, 171, 176; life expectancy 124–25 handicap see disability Hanson, Pauline 144 health 1, 11–12, 238, 253; Indigenous 15, 19 health care 11, 21, 126 health insurance 256 heart disease 106, 107, 108, 109, 113, 114, 116, 118, 119, 120, 121, 122, 123 Hindu 170, 171 HIV/AIDS 113, 116, 118, 120, 126 Hobart 31 Hong Kong, immigration from 137, 169, 171, 177 housing 98, 102, 129, 157, 196, 212, 216, 259–60, 268 human rights 142 Hungary, ethnic origin 166; fertility 43, 44; immigration from 179; language spoken at home 170 ‘hysteresis’ 225 immigrants, intermarriage 9; nonEnglish speaking 175; second-

• 299

generation 163, 174, 175, 176, 177, 179–80, 184, 201; third-generation 165; value in human capital 8; see also individual countries of origin; ethnic communities immigration viii, 1, 7–10, 129–57, 159, 182, 200, 201, 209, 218, 220, 239–40, 243, 267, 268, 269–72, 273, 277; economics of 7–8, 129; family reunion 8, 134, 135–39, 145, 148, 149–52; illegal 8, 146–49; low-skilled 133, 134, 149; and migration 211–212; selection criteria 139–40, 143, 145, 153–54; settlement 156, 172–74, 183, 198, 200–201, 202, 201–207, 216; skilled 8, 134, 136, 139–40, 141, 144–45, 146, 152–54; sponsored spouse 178; temporary entry 142, 154–55, 156, 174, 210, 269, 271 immigration control 130, 131–35, 180; see also immigration, selection criteria; policies, immigration ‘immigration cycle’ 131, 269–70 Immigration, Department of 131–35, 142, 145 independent living 238, 243–47, 250, 255, 262 India, ancestry 164, 165; ethnic origin 167; immigration from 137, 163, 170, 171, 177, 178 Indigenous population 1, 6, 17–39; distribution 6, 11, 60, 198; history 18–19; research into 17; size 24–26 Indonesia 171, 177 integration 174–80 Intergenerational Report viii intermarriage, within immigrant communities 184; immigrant and non159, 160, 165, 168, 176–80, 181, 184; Indigenous and non- 23, 26; of second-generation immigrants 179 Iran, immigration from 177 Iraq, immigration from 148 Ireland, ancestry 164, 165, 183–84; ethnic origin 166; fertility 43, 44; immigration from 160, 161, 162, 163, 172, 173, 180 Islam 170, 171 Italy, ancestry 164, 165; ethnic origin 166; fertility 13, 44, immigration from 135, 139, 158, 161, 162, 163, 172, 178, 179; immigration to 10; labour force 277; language

300 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

spoken at home 169, 170, 175; life expectancy 122–25 Japan, ethnic origin 167; fertility 43, 44, 277; immigration from 177, 178; life expectancy 122–25; policies 6, 15 Jewish 170, 171; ethnic origin 167 Jones Committee 7 Korea, ethnic origin 167; immigration from 177, 178 Labor Party 134, 156 labour force 1, 8, 37, 129, 219–37, 245; 55+: 12; definition 220–21; and families 79; growth 225, 228, 237; projections 220, 275–76 labour force participation 219, 221, 222, 228, 231, 275; female 237, 259, 268, 275–76; Indigenous 35; male 237, 275–76 language 159, 184; see also diversity, language Laos, immigration from 63, 177, 178, 179; language spoken at home 169 Laslett, Peter 12, 238, 243, 244, 246–47, 264 Latin America 10 Lebanon, ancestry 165; ethnic origin 167; immigrant fertility 67(68; immigration from 138, 139, 158, 171, 178, 179; language spoken at home 176 life expectancy 1, 3, 21–22, 104, 105, 114–118, 247–50, 258; female 21, 105, 114–118, 122, 127, 128, 247, 248–49; Indigenous 21–22; male 21, 105, 114–118, 122, 127, 128, 247, 248–49; projection 269; sex differential 120–23, 127–28 locational disadvantage 209–210 Macphee, Ian 134 Malaysia, ethnic origin 167; immigration from 137, 171, 177, 178 Malta, ancestry 165; ethnic origin 166; immigration from 161, 163, 179; language spoken at home 170 Maori, ancestry 164; ethnic origin 167; fertility 268; life expectancy 22 marriage 14, 23, 72, 77, 80–89, 101, 138, 142, 178, 243, 251, 255, 261, 268; age at 42, 45; breakdown 83, 87, 91, 92, 243, 252, 253, 260; deferred 83, 85, 102; and education 55, 56–57; and eth-

1970–2030

nicity 85, 176–80; of immigrants 143, 151; and religion 85, 178; ‘shot-gun’ 72, 74; see also intermarriage Melbourne 7, 9, 28, 29, 156, 172, 173, 206; immigration to 198, 200, 202–207; migration from 205 Menadue, John 132, 134 Mexico 10 Middle East, ancestry 164, 165; ethnic origin 166, 180, 181; immigration from 136, 139, 148, 156, 160, 161, 162, 169, 170, 171, 172, 173, 177, 184 migration (internal) 1, 6, 30, 60, 198, 203, 204, 207–209, 210, 216, 218; and immigration 211–212; Indigenous 28–31, 32; interstate 208, 209; to regions 209; temporary 210–211 mortality 2, 3, 11, 20, 21, 104–28, 240, 247, 266; adult 11; age-specific 108–114, 127; aged 108, 111, 113, 114, 128; child 11, 20, 112; female 12, 105, 113, 114, 115, 118–20, 122, 123, 127; Indigenous 11, 18–22, 200; Indigenous child 20–21; infant 115, 116, 118, 122, 128, 253; male 11–12, 105, 112–113, 114, 115, 118–20, 122, 123, 127; projections 126–28, 268, 269, 272, 275; regional 199, 200; sex differential 111, 118–20; youth 112–113 mothers, single 97, 98 multiculturalism 130, 135, 143, 144, 158, 174 Muslim see Islam National Agenda for Multicultural Australia (1989) 183 national identity 1, 9, 174 Netherlands 7, 44, 124–25; ancestry 165; ethnic origin 166; immigration from 133, 162, 163, 179; language spoken at home 169, 170 New South Wales 172, 173, 190, 194, 200, 208 New Zealand, ancestry 164; ethnic diversity 159, 180; fertility 43, 44, 268; immigrant concentration 174; immigration from 8, 154, 155, 160, 161, 162, 163, 177, 272; immigration to 168, 182; indigenous populations 22, 39

INDEX

Nordic countries 43, 99 Northern Territory 26, 32, 33, 34, 172, 173, 194, 200 Norway 44, 124–25 One Child Family policy 68, 72 One Nation Party 144, 183, 189 ‘outstations’ 32–33, 34 parenthood, and employment 89, 92–101; and marriage 89 Passel, JS 24 paternity, recording 91–92 people smuggling 148 Perth 28, 173 Philippines, ancestry 165; ethnic origin 167; immigrant fertility 68; immigration from 136, 137, 138, 163, 169, 171, 177, 178; language spoken at home 169, 175 place of birth see birthplace Poland, ethnic origin 166; fertility 44; immigration from 163, 179; language spoken at home 170, 175 policies 15–16, 76; children 279; economic 228, 229; employment 219–20, 230–31; family 100–101, 279; fertility 272; health 267; immigration 1, 8, 15, 129–55, 158, 182, 183, 266, 270–71, 279; Indigenous 20, 36; population 1, 266–67, 279; regional 15, 214; urban 187 population density 4–5, 186–87, 188 population distribution 4–6, 185–218; factors affecting 198–215; Indigenous 27–37; policy effects 215–217; regional 187–94; urban 196–97, 212 population growth 1, 19, 28, 187, 192–93, 209; Indigenous 25–26, 37–38; measuring 24; projections 2, 26, 266–79; regional 29–31, 189–90, 192–93, 199; targets 267; urban 192–93, 197; zero/sustainable 273, 277–78, 279 ‘population momentum’ 240, 242–43 population structure, age 104–105, 109, 199, 219, 238, 241, 242, 243, 268, 273, 279; changing 238–43; and fertility 273–74, 276–78 Population Summit viii Portugal 44 Queensland 32, 172, 173, 190, 194, 195, 200, 207–208

• 301

refugees 10, 139, 142, 146–49, 158 regional population viii, 1, 4, 5, 28, 189, 191; ethnic diversity in 160, 172–74, 183; Indigenous 39 Regional Sponsored Migration Scheme 183 ‘replacement migration’ 277, 278 retirement 256, 259–61; active see ‘Third Age’; early 223 Russia 44; ethnic origin 166 Scandinavia, ethnic origin 166 Scotland, ancestry 164; ethnic origin 166; fertility 268; see also United Kingdom settlement, self-reinforcing 7 smoking 116–118, 123–26 South America, immigration from 158 South Australia 32, 173, 187, 190, 194, 200–201, 208 Spain, ethnic origin 166; fertility 13, 44; language spoken at home 170; life expectancy 124(25 suicide 105, 107, 108, 112–113, 116, 118, 119, 120, 121, 126 Sweden 7, 44, 124–25, 276 Switzerland 44, 124–25; ethnic origin 166 Sydney 7, 9, 28, 29, 156, 157, 172, 173, 206; immigration to 198, 200, 202–207, 213; migration to/from 205, 213 Tampa 148–49, 157 Tasmania 24, 25, 26, 172, 173, 190, 194, 200, 208 Taylor, Griffith 185–86 ‘Third Age’ 12, 238, 243, 244, 245–47, 263–64 Total Fertility Rate (TFR) 40–41, 43, 45; see also fertility Turkey, ancestry 165; ethnic origin 167; immigration from 133, 136, 139, 158, 178, 179; language spoken at home 169 Ukraine, ethnic origin 166 unemployment 219, 220, 221, 225, 226, 228; of immigrants 230; Indigenous 35–36, 230; long-term 229–30, 237; projections 220, 229, 231–33 United Kingdom, fertility 13, 43, 44, 268; immigration from 132–33, 137, 138, 140, 160, 161, 162, 163, 172, 173, 179, 180, 182; immigration to 7; life expectancy

302 •

T H E T R A N S F O R M AT I O N O F A U S T R A L I A ’ S P O P U L AT I O N :

124–25; policy 103 United States, children 10; employment 99; ethnic diversity of 159, 180, 181; fertility 13, 43, 44; immigrant concentrations 174; immigration to 7, 8, 9–10, 130, 135, 137, 140, 159, 162, 271; indigenous populations 38; life expectancy 124–25; policy 103; population distribution 5, 6, 195, 197 urbanisation 6–7, 133, 187; Indigenous 27, 28–29, 39 Victoria 24, 172, 173, 190, 194, 200, 208 Vietnam, ancestry 165, 165; immigrants’ religion 63, 68; immigration from 137, 138, 158, 163,

1970–2030

169, 171, 177, 179, 184; language spoken at home 169, 176 Wales, ethnic origin 166; see also United Kingdom welfare, aged 256–57, 258, 259; family 59–60; immigrants 131, 149–50, 151, 152, 157; sole-parent 101–102; targeted 90; unemployment 231 welfare dependency 103, 150 Western Australia 32, 172, 173, 190, 194, 195, 200, 208 ‘white flight’ 212 Working Nation 36 Yugoslavia, ethnic origin 166; fertility 66; immigration from 133, 163, 179

Also from UNSW Press

Developing Australia’s Regions: Theory and Practice Andrew Beer, Alaric Maude and Bill Pritchard The press and contemporary political debate send conflicting messages about the economic future of Australia’s regions. Developing Australia’s Regions: Theory and Practice is a book that recognises that regions matter — what takes place in our diverse regions fundamentally determines the nation’s quality of life. This practical books draws upon regional development theory, and national and international experience, to set out the principles and strategies that can be used to establish a stronger future for our regions. ISBN 0-86840-548-5 Published May 2003

Land of Discontent: The Dynamics of Change in Rural and Regional Australia Edited by Bill Pritchard and Phil McManus The last decade has been one of great change in rural and regional Australia. The resulting shock waves have resonated across the country as traditional political allegiances and notions of our national identity have been challenged and reshaped. In Land of Discontent, an array of eminent Australian social scientists exposes the complex and often contradictory forces that are affecting the lives of many Australians. ISBN 0-086840-578-7 Published November 2000