Research Paper 5 1999-2000

Population Futures for Australia: the Policy Alternatives

Professor Peter McDonald
Rebecca Kippen
Social Policy Group
12 October 1999

Contents

Major Issues

In recent years, there has been intense debate about the population size to which Australia should aim in the 21st century. Some argue for a much lower population than we have now. The environmentalist, Tim Flannery, for example, has suggested that Australians might opt for a future population of between 6 and 12 million people. Others argue for a considerably larger population. For example, former Prime Minister, Malcolm Fraser, would like to see a population of 45 to 50 million people in the next 50 years. This paper shows that the targets specified by both Flannery and Fraser are unattainable in the next 50 years because such targets can only be reached through levels of immigration which are impossibly negative on one side or impossibly large on the other side.

The demographic reality is that the options for Australia's future population size are much more limited than the options that are considered in popular debate. The limiting factor is our low, and falling, level of fertility. The powerful demographic effects of low fertility have been little appreciated. One of the aims of this paper is to provide such an appreciation. We show that with zero net migration, Australia's projected levels of fertility and mortality would lead to the population rising initially to 20 million and then falling slowly in the first half of next century and rapidly in the second half. The population would age dramatically and the size of the labour force would fall markedly. We argue that these are not desirable outcomes and that, as a consequence, we should seek to have immigration levels which, at the least, lead to zero population growth. A level of net annual migration of around 80 000, in combination with our low fertility rate, would lead to zero population growth in about 25 years and a population size of about 24-25 million.

Given current trends in fertility and mortality, annual net migration to Australia of at least 80 000 persons is necessary to avoid spiralling population decline and substantial falls in the size of the labour force. This level of annual net migration also makes a worthwhile and efficient contribution to the retardation of population ageing. Levels of annual net migration above 80 000 become increasingly ineffective and inefficient in the retardation of ageing. Those who wish to argue for a higher level of immigration must base their argument on the benefits of a larger population, not upon the illusory 'younging power' of high immigration.

The level of about 80 000 is close to the average net migration of the past eight years and of the past 50 years. Thus, if present settings do not change, we are headed towards this outcome. There is an upper limit to annual net migration. We argue that there were difficulties in the late 1980s when net migration rose for just two years to over 150 000 per annum. While it is not possible to be prescriptive, a sustained net migration level of 120 000 per annum is at the high end of what Australia seems to be able to manage. Furthermore, it is little understood that the Australian Government has only a very limited degree of control over the final level of annual net migration. It has almost no control over movements out of the country and little control over long-term (as distinct from permanent) movement into the country. Even major components of permanent movement into the country (New Zealand citizens and spouses or children of Australian residents) are largely beyond the government's control. That is, it is not possible for any Australian government to precisely determine the level of annual net migration. For example, net migration was 72 000 in 1997 and 112 000 in 1998. The increase of 40 000 between these two years was not the outcome of deliberate government policy.

If the desirable range of annual net migration is as narrow as 80 000 to 120 000 and if government is unable to determine this level with any degree of precision, heated debate about immigration levels seems pointless. In these circumstances, the major parties should reconsider the advantages of a bipartisan approach to immigration policy. Our analysis in the paper is based on the assumption that the fertility rate in Australia will not fall below 1.65 births per woman on average in the long term. In our view, population policy for Australia should revolve around ensuring that fertility does not fall below this level.

Introduction

In the 1970s, the Report of the National Population Inquiry (the Borrie Report)(1) placed demography in the centre of discussion about the future of the Australian population. Since the Borrie Report, however, the science of demography has been relatively absent from the population debate.(2) The dominant players in the past two decades have been those concerned with the demographic outputs rather than with the demographic inputs. The demographic inputs (fertility, mortality and migration) in this context have been largely taken for granted, either through an assumption that we can manipulate them at will or, alternatively, that there is nothing we can do to influence their future path. In the absence of science, populism prevails and, as we shall demonstrate, some of the more populist statements made by prominent people about Australia's future population are sheer demographic nonsense.

Like Borrie in his Report of the National Population Inquiry, we are very sceptical of the concept of an optimum population for Australia, or for any other country, simply because circumstances change; the economy changes, the environment changes, the politics change, and demographic components change. Furthermore, most of those calling for a particular target population for Australia consider only the total population size and ignore population composition, particularly age composition. As we demonstrate, age composition cannot be ignored. If the aim is to achieve a particular target size of population, it is insufficient to merely advocate a target number. We must also indicate in demographic terms how that target can be achieved and, more importantly, how that target would be sustained once it has been achieved. To do this, we must re-emphasise the forgotten elements of the population debate, births and deaths, and we must give consideration to the age structure of the population ensuing from our history of births, deaths and migration.

The dominant feature of Australia's long-term population future is not immigration, but the balance between births and deaths and the consequent effects upon the age structure of the population. All industrialised countries now have fertility rates and mortality rates which imply that, in the absence of migration, their populations will fall once the growth effect of current young age structures has passed (about 25 years in Australia's case). The populations of some countries in Europe have already begun to decline. The importance of fertility and mortality is evidenced by the emergence of the ageing of the population as a central demographic issue. Borrie, in 1975, made almost no mention of ageing as a future policy issue. Ageing has recently emerged as an issue because fertility and mortality have both fallen since the mid-1970s to a much greater extent than was envisaged in the Borrie projections. If birth rates were the same at each age today as they were in 1973, there would have been 40 per cent, or 100 000, more births in 1998. If death rates at each age were the same today as they were in 1971-76, there would have been 60 per cent, or 78 000 more deaths in 1998. These are remarkable changes within a short period of time. One hundred thousand less people each year at the young end of the age structure and close to 78 000 more people each year at the old end of the age structure are the reasons ageing of the population has emerged as a policy issue.

During the past 20 years, while these significant changes in fertility and mortality have been altering our demographic future, the policy attention has focused only upon immigration. Debates about the future size of Australia's population and about its age structure have been debates about immigration, with no consideration of possible future variations in levels of fertility and mortality. We consider this to be a very serious error. Immigration can play an important role in shaping population futures, but its impact must be considered in the context of trends in fertility and mortality. The aim here is to consider the demographic feasibility of population paths recommended by various prominent Australians. We conclude that, because of low fertility, the range of options is much more limited than the more populist statements would imply.

Criteria for feasibility

In assessing the demographic feasibility of different future population scenarios, we make use of a set of demographic criteria. The five criteria we use are as follows:

We should aim to avoid excessive ageing of the population

Substantial ageing of the population in the next 40 years is absolutely inevitable and we should be considering policy approaches to deal with it. At the same time, we should avoid future demographic pathways that exacerbate the extent of population ageing.

We should avoid creating a substantial momentum of population decline

At present in Australia, our fertility rate is below the level of generational replacement. That is, those in the childbearing ages today are not having sufficient births to replace themselves. Despite this, there are many more births in Australia than deaths. The reason for this is that, in Australia today, there are large numbers of people in the reproductive ages as a result of the post-war baby-boom, and relatively small numbers at older ages. An age structure configured in this way provides what is termed a momentum for population increase. When a population has a momentum for population increase, it will continue to grow even though its fertility and mortality rates are low.

However, the momentum for population growth inherent in our present age structure will disappear over the next 25 years and from that point, in the absence of immigration, our population would begin to decline. Sustained low fertility and the ageing of our population would then create an age structure that has a momentum for population decline, the reverse of our present situation. As we shall demonstrate, a population with a momentum for decline can fall in size very rapidly, with the decline being difficult to reverse. Thus, a criterion that we specify is that our future population path should avoid creating a substantial momentum for population decline.

We should avoid excessively high or negative numbers of immigrants

Population targets for the future cannot be such that we would have to make people leave Australia in order to achieve the target. Also, there are upper limits to the number of immigrants that we can recruit and settle within any given period of time. Significant problems arose when net migration rose to an average of over 150 000 people per annum in the two-year period 1988-89. Unemployment rates were very high among this group of immigrants relative to those who entered at times of low immigration. Part of the immigration equation has been that, as numbers increase, skill levels decrease. Alternatively, skill levels might be maintained, but the ages of immigrants would have to rise. Sustained levels of net migration above 150 000 per annum could be considered to be excessively high.

As far as possible, we should avoid wide fluctuations in our age structure

Fluctuations in age structure can mean that demand for public services such as education and health services also fluctuate. As these services can involve substantial infrastructure and training costs (schools, teachers, hospitals, doctors, nursing homes), it is more efficient to try to avoid widely fluctuating age structures.

We should avoid a substantial fall in the number of people in the working ages

A substantial fall in the number of people at working ages combined with an increase in the numbers at older ages, depending upon improvements in productivity, could lead to a fall in economic output per capita. Certainly, economic output per capita would be at a considerably lower level than it would have been if the labour force did not decline. There is also an argument that a considerable fall in the size of the economy has features that are undesirable, such as substantial falls in asset values, which, in turn, underpin retirement incomes.

In the discussion that follows, the different scenarios relating to our future population will be tested against these five criteria.

The assumptions of the standard population projection

In this paper, we compare the future populations advocated by various people with a single standard population projection. Our first task is to show the features of this standard. The standard is a standard because it is based upon a projection of current and recent trends into the future. That is, it is the future we will have if there is little change in the demographic paths we are following at present.

The mortality assumption we use in the standard and in all but one of the alternative scenarios is the same as that used by the Australian Bureau of Statistics in the most recent official projections.(3) With this assumption, the expectation of life at birth increases by about 10 years over the next 100 years. At the end of the discussion, we shall consider a scenario in which the expectation of life increases by 20 years over the next 100 years.(4) Immigrants are assumed to take on the fertility and mortality levels of the Australian population after they arrive in the country.(5)

The standard assumes that annual net migration (ANM) will be 80 000 per year, roughly the average of the past eight years (78 000). The age distribution of immigrants is assumed to be the average of the years 1994-97. This age distribution is used in the standard projection and in all of the scenarios. The recent age structure of immigrants is quite young such that a significantly younger age distribution would be most unlikely unless we were to provide preference to immigrants with children.

Finally, the standard projection assumes that the birth rate (technically, the Total Fertility Rate or TFR)(6) will fall to 1.65 births per woman over the next ten years after which it will remain constant at that level. This assumption requires discussion.

Future levels of fertility in Australia

Table 1 shows the Total Fertility Rate in various industrialised countries in 1997, or in the nearest available year. Australia's fertility rate at 1.78 births per woman is comparatively high on this list and there is at least a potential that our fertility rate could fall to the lower levels now prevailing in most other industrialised countries; fertility has already fallen in Canada, a similar country to Australia in many ways, to 1.64 births per woman.

In general, Table 1 shows that relatively higher fertility rates apply in the English-speaking countries and in the Scandinavian countries. The lowest fertility rates apply in the Southern European and Germanic countries, as well as in Japan. McDonald(7) has argued that this international ranking of countries according to their fertility level matches their ranking on the extent to which they facilitate the employment of mothers in the paid labour force and the extent to which a degree of gender equity applies within the family itself. The countries which, through their social institutions, make it difficult or unrewarding for women to combine work and family, or which provide incentives for mothers to stay at home rather than to be employed, are the countries that have very low fertility. Faced with a choice between an uninterrupted career or having a child and withdrawing from the work force for an extended period, women in these countries often make the decision not to have the child. In short, where countries continue to support or promote the male breadwinner model of the family, fertility falls to very low levels. The consequence for the countries that have had very low rates of fertility is that they are now facing faster and higher levels of ageing than is Australia.

Table 1. Total Fertility Rates, Various Countries, 1997 (or nearest available year)

Country	TFR		Country	TFR
Spain	1.15		Scotland	1.58
Italy	1.22		Canada	1.64
Greece	1.32		France	1.71
Germany	1.36		Luxembourg	1.71
Austria	1.36		England	1.74
Japan	1.44		Denmark	1.75
Portugal	1.46		Finland	1.75
Switzerland	1.48		Australia	1.78
Sweden	1.52		Norway	1.85
Belgium	1.55		New Zealand	2.04
Netherlands	1.57		United States	2.06

In Australia, policy seems to have shifted back to a model that is unfriendly to the combination of work and family. Child care has become more expensive, employment conditions favourable to families with young children are being withdrawn through new industrial agreements, the tax-transfer system is being modified to provide greater benefits to parents who do not work, and the sense of job security has fallen giving rise to uncertainty.

It is no surprise to find that, under these circumstances, Australian fertility has fallen in the 1990s (Table 2) and can be expected to continue to fall. Another indicator of the future trend in fertility is the timing of the first birth. We cannot expect a cessation of fertility decline, or an increase in fertility, while the age at first birth continues to rise.

Table 2. Total Fertility Rate, Australia, 1992-1997

Year	TFR
1992	1.89
1993	1.86
1994	1.85
1995	1.83
1996	1.80
1997	1.78
1998	1.74

Finally, Table 3 shows the regional variation in fertility rates in Australia. Some States and Territories already have fertility rates in the region of 1.65 births per woman and this is the level that now already applies in the combined metropolitan areas.

Table 3. Total Fertility Rates, States and Territories and Capital Cities of Australia, 1997

State/Territory	TFR		City	TFR
New South Wales	1.83		Sydney	1.74
Victoria	1.69		Melbourne	1.61
Queensland	1.80		Brisbane	1.67
South Australia	1.70		Adelaide	1.59
Western Australia	1.79		Perth	1.67
Tasmania	1.80		Hobart	1.64
Northern Territory	2.17		Darwin	1.95
ACT	1.60		Canberra	1.60
Australia	1.78		All capitals	1.67

Thus, the assumption that Australia's fertility will continue to fall over the next decade to at least 1.65 births per woman seems justified. It is very difficult to predict the future course of fertility after this time. However, unless there is a shift in policy direction in Australia, further falls in fertility are possible beyond the next decade. Here, we take the somewhat optimistic assumption that policy makers will recognise that the level of fertility cannot continue to be forced downwards by social policy and that, within the next decade, policies halting further decline will be put in place.

Immigrants are assumed to have the same fertility rates as the rest of the population after they arrive in Australia. While there are wide variations in the levels of fertility among different immigrant groups, this assumption is very close to being correct across all immigrant groups. This is because some of the larger immigrant groups, such as those from the United Kingdom and New Zealand, have lower fertility than the Australian average.(8)

The results of the standard population projection

The results of the standard projection (and of the other scenarios) are presented in two main forms: a series of diagrams (Figure 1) which show the projected numbers in five-year age groups at six points in time (1998, 2018, 2038, 2058, 2078 and 2098) and a table showing outcome measures at the same six points in time.

Table 4. 'Standard' projection statistics, Australia, 1998-2098

Year	Total population (millions)	r%	<15%	65+%	DR
1998	18.8	1.0	20.9	12.2	0.73
2018	22.0	0.6	16.7	16.6	0.84
2038	23.9	0.2	15.2	23.1	1.04
2058	24.5	0.0	14.6	25.8	1.15
2078	24.6	0.0	14.2	27.3	1.19
2098	24.7	0.0	14.0	28.5	1.24

In the diagrams, each successive layer from the bottom to the top represents the size of the population in a five-year age group. The male population is depicted on the left side of the diagram and the female population on the right side. The bottom layer is age group 0-4 years, the next is 5-9 years and so on. The second top layer is age group 80-84 years and the top layer is age group 85 years and over. The diagrams relating to years after 1998 all have a grey shadow in the background. This is the size and age structure of the original population in 1998. The shadow allows the projected population to be compared with the original population. The projected population at each age for the given year is indicated by the outer lines of the diagram. The inner lines depict the projected original (1998) population and their descendants, while the gap between the inner lines and the outer lines depicts the immigrants since 1998 and their descendants. Thus, the diagram enables us to see the impact of immigration after 1998 on future population.

The main story of the standard population is that numbers are progressively added at older ages (50 and over) while, at the same time, there is very little change to the size of the population at ages below 50. Thus, the population becomes older, but there is no reduction in the working age population. Immigration has contributed substantially to maintaining the numbers in the labour force ages.

Table 4 shows five outcome measures which relate to each of the six points in time: the size of the total population in millions, the annual rate of growth of the population in the given year (r%), the proportion of the population aged less than 15 years (<15%), the proportion of the population aged 65 years and over (65+%), and an age dependency ratio (DR).(9)

The table for the standard projection shows that the total population increases to just over 24 million by 2058, after which time there is very little further increase in the size of the population. Accordingly, the second column of the table shows that the growth rate would drop close to zero by the middle of next century and remain at zero. Hence, the standard provides a zero population growth scenario from the middle of next century, but one which is achieved through a combination of low fertility and a constant level of migration.

While the numbers aged less than 15 years do not change very much during the 100 years of the projection, the percentage of the population aged less than 15 years falls from 21 per cent to 15 per cent in the next 40 years. Thereafter the percentage of children remains around 14 per cent. The main changes occur at the other end of the age distribution. The proportion aged 65 years and over more than doubles between 1998 and 2058 shifting from 12.2 per cent to 25.8 per cent. After 2058, the proportion at older ages increases at a much slower rate, reflecting the continued fall in mortality rates.

The age dependency ratio also shifts up sharply during the next 60 years after which the changes are more moderate. It is important to realise that the low level of dependency in Australia in 1998 arises from the fact that, in 1998, the age structure is heavily concentrated in the working ages. As we describe below, this is a historically unique situation arising as our age structure shifts from a young to an old age structure. The dependency ratio never has been lower than it is now and never will be again.

The standard projection scores well on all five of the demographic criteria described above. The population ages, but, as we shall see, in comparison with other projections, ageing under the standard is close to the minimum that could be expected. No momentum for population decline is created. The numbers of immigrants, being at the level of recent years, are not excessively high or negative. The age structure is very stable at the younger ages and does not fluctuate, and the numbers in the working ages do not fall. Thus, the standard projection is a good benchmark against which other projections can be compared.

In the following sections, we take quotations of prominent Australians about Australia's future population and examine the demography underlying their statements. We examine the statements of both those who call for a much smaller population and those who call for a much larger population. We have chosen the statements on the basis that they represent a range of viewpoints about Australia's future population. That is, the statements should be seen not so much as representing the views of the person quoted so much as covering a range of views held or expressed in the wider community.

Harry Recher: zero migration and a one-child policy

In this projection, fertility is assumed to fall immediately to 1.0 births per woman and annual net migration is set to zero. The mortality assumption of the standard applies. This projection allows us to demonstrate the power of low fertility, something which we believe is little understood. If, on average, each woman has one child, then the size of each successive generation will be half the size of the previous generation. A demographic generation, the mean age at childbearing, is about 30 years. Literally applied, the Recher approach would reduce Australia's population to 5 million by 2098. Sustained low fertility would then lead to less than half a million by the end of the next century and, soon afterwards, the population would disappear. If we call for a one-child policy, we need to specify how and when it would be reversed as reversed it would need to be.

This projection fails on several of the criteria. Ageing of the population occurs at a much faster rate than for the standard. By 2058, 44 per cent of the population would be aged 65 years and over compared to 26 per cent for the standard. By 2098, 50 per cent of the population would be aged 65 years and over. The undercutting of the age structure also creates an enormous momentum for future population decline. Cessation of the decline at some point would require a very large increase in fertility; if all our daughters do indeed have one child, we would have to ask our grand-daughters to have four children each in order to stop the decline of population. Having a one-child policy for a period and then reversing to a high level of fertility would lead to violent fluctuations in age structure. This projection also leads to a very sharp fall in the numbers at working ages. On the other hand, as a zero migration projection, the number of immigrants is not negative or excessively high.

From a social perspective, the Recher approach tends to assume that fertility can be engineered according to our preference. All the evidence of the determinants of fertility rates is contrary to this view.

Table 5. 'Standard' and 'Recher' projection statistics, Australia, 1998-2098

	Total population (millions)		r%		<15%		65+%		DR
Year	Std	Recher	Std	Recher	Std	Recher	Std	Recher	Std	Recher
1998	18.8	18.8	1.0	0.0	20.9	20.9	12.2	12.2	0.73	0.73
2018	22.0	18.3	0.6	-0.4	16.7	10.6	16.6	19.4	0.84	0.75
2038	23.9	16.0	0.2	-1.2	15.2	7.9	23.1	32.2	1.04	1.22
2058	24.5	12.0	0.0	-1.8	14.6	6.7	25.8	43.7	1.15	2.01
2078	24.6	7.9	0.0	-2.4	14.2	5.9	27.3	47.1	1.19	2.02
2098	24.7	5.0	0.0	-2.1	14.0	5.8	28.5	50.0	1.24	2.20

Con Sciacca and Jeff Kennett: solving our ageing problem through immigration

There is no question that immigration, at least the first 80 000 immigrants, provides a worthwhile reduction in the extent of ageing of the population. However, immigration cannot 'solve our ageing problem'. Substantial ageing of the Australian population over the coming decades is absolutely inevitable. To illustrate the lack of power that immigration has in relation to our age structure, we investigate the levels of immigration that would be required to maintain the proportion of the population aged 65 and over at its present level of 12.2 per cent. In doing this, we maintain the fertility and mortality assumptions of the standard but allow annual net migration to change.

To achieve our aim, enormous numbers of immigrants would be required, starting in 1998 at 200 000 per annum, rising to 4 million per annum by 2048 and to 30 million per annum by 2098 (Table 6). By the end of next century with these levels of immigration, our population would have reached almost one billion. This is obviously not what Mr Sciacca nor Mr Kennett had in mind, but it is important that the message is heard that our population cannot be kept young through immigration.(10) The problem is that immigrants, like the rest of the population, get older and as they do, to keep the population young, we would need an increasingly higher number of immigrants.

This projection obviously has excessively high levels of immigration.

Table 6. Annual Net Migration (ANM) required for the 'Sciacca/Kennett' projection

Year	ANM (millions)	Year	ANM (millions)	Year	ANM (millions)	Year	ANM (millions)
1998	0.2	2023	1.4	2048	4.0	2073	10.9
2003	0.4	2028	1.4	2053	5.2	2078	14.3
2008	0.8	2033	1.8	2058	6.2	2083	18.6
2013	1.0	2038	2.2	2063	7.3	2088	24.0
2018	1.2	2043	3.0	2068	8.7	2093	30.4

Table 7. 'Standard' and 'Sciacca/Kennett' projection statistics, Australia, 1998-2098

Tim Flannery: long-term population target of 6-12 million

In this projection, we aim to achieve a population of 12 million (the upper end of Flannery's target) over a period of 50 years. The fertility and mortality assumptions of the standard are used but net migration is assumed to be an annual constant number to the point where the target is reached. Once the target is reached, net migration is varied to maintain the target population at 12 million.

The main point to note here is that, to achieve the aim of this projection, we would have to remove 100 000 persons net from Australia every year for the next 50 years. However, to keep the population at 12 million once that target had been reached, we would have to switch to positive immigration from 2048 onwards. Thus, the projection fails the criterion that net migration should not be negative. It also fails the test of maintaining the size of the labour force, but this would be expected if we are aiming at a much smaller population than we have now. The age structures of this projection are interesting in that initially the population becomes much older than the standard (36 per cent aged 65 years and over in 2048 compared to 24 per cent for the standard). However, after 2048, the population becomes younger again, ending in 2098 with a very similar age structure to that of the standard. Thus, this projection leads to severely fluctuating age structures.

Substantial population decline is difficult to achieve in anything but the very long term because of the momentum for population increase inherent in Australia's present age structure.

Table 8. Annual Net Migration (ANM) required for the 'Flannery' projection

Table 9. 'Standard' and 'Flannery' projection statistics, Australia, 1998-2098

Statement 4. Malcolm Fraser If we believe we can maintain Australia at 18 or 20 million people without increasing envy, without marginalising ourselves, without challenge, then we are gravely and seriously mistaken...Australia's population has grown 2 1/2 times since 1945. There is no reason at all why we could not grow 2 1/2 times again by the middle of next century. We would then be a nation of 45 million to 50 million people. Weekend Australian, May 3-4 1997

Malcolm Fraser: 50 million in 50 years

In this projection, we assume that fertility and mortality will follow the path of the standard projection, but that migration will be set at a constant annual figure for the next 50 years such that a population of 50 million would be reached in 2048.

The constant annual number of immigrants that would be needed if we were to reach 50 million in 50 years is 463 000 (Table 10). This is clearly above any level of immigration that could be realistically achieved. Once 50 million had been reached in 2048, we would require a period of negative migration in order to avoid overshooting the 50 million mark.

In the first 50 years, this projected population is noticeably younger and less dependent than the standard, but by 2098, it has the same age distribution as the standard. It is interesting to note that, despite vastly different assumptions about immigration, the 'Flannery' projection, the 'Fraser' projection and the standard all end in 100 years time with the same age structure. This is because the level of immigration has little impact on age structure in the long-term. In order of significance, age structure in the long-term is determined by fertility, mortality and the age distribution of immigrants, and these are the same in all three projections.

Our population cannot continue to grow at the rate of the last fifty years. The reason for this is that the fertility rate today is only half of what it was in the first 25 years after the Second World War. If current fertility was double the prevailing rate, we would have an additional 250 000 births each year. Because 250 000 people per year are not added through births in this projection, they have to be added through immigration, resulting in an unsustainably high level of immigration.

Table 10. Annual Net Migration (ANM) required for the 'Fraser' projection

Year	ANM ('000)		Year	ANM ('000)		Year	ANM ('000)		Year	ANM ('000)
1998	463		2023	463		2048	-99		2073	131
2003	463		2028	463		2053	-29		2078	150
2008	463		2033	463		2058	35		2083	167
2013	463		2038	463		2063	79		2088	180
2018	463		2043	463		2068	109		2093	186

Table 11. 'Standard' and 'Fraser' projection statistics, Australia, 1998-2098

	Total population (millions)		r%		<15%		65+%		DR
Year	Std	Fraser	Std	Fraser	Std	Fraser	Std	Fraser	Std	Fraser
1998	18.8	18.8	1.0	2.9	20.9	20.9	12.2	12.2	0.73	0.73
2018	22.0	31.1	0.6	2.0	16.7	18.6	16.6	13.1	0.84	0.73
2038	23.9	43.9	0.2	1.4	15.2	17.1	23.1	16.8	1.04	0.83
2058	24.5	50.0	0.0