Population Dynamics in East and West Germany Projections to 2050

Transcription

1 Population Dynamics in East and West Projections to 2050 In 2003, the population of declined _ albeit only slightly _ for the first time since The decrease was primarily caused by the combination of a continued excess of deaths over births and a fall in immigration. The decline in the population applied to east, in particular, while west continued to show a slight increase. The future demographic trend will largely depend on migration dynamics, but also on changes in life expectancy. As the future development of these two parameters remains uncertain, the DIW Berlin population forecast for east and west presented in this report was calculated using two variants for each case: for net migration, annual average net inflows of (variant I) and (variant II); for life expectancy, an alleviated growth rate (variant A) and a constant growth rate (variant B). According to variant IA, the population of west will increase by around 2 million to 69.5 million in 2020, but subsequently fall to just under 65 million by In east, immigration from abroad will be just sufficient to counterbalance the natural decline in the population and the ongoing emigration to west up to Subsequently, the east German population will decrease _ to 14 million in The population of as a whole will then amount to 79 million. According to variant IIB, which has higher inflows and steadily rising life expectancy, even over the long term the population of will not fall below its current level. According to these calculations, in 2050 the population of east will be only slightly lower than it is today, while west can expect a population increase of around 1 million to 68.5 million by that time. Although the two variants show significant differences in the level of population at the end of the forecast period, the shifts in age distribution are almost identical. East will be more affected by the ageing of society than west. The old-age dependence ratio, i.e. the ratio of the number of persons aged 65 and over to those aged between 20 and 64, will rise from 28% to around 54% in west and from 28% to 62% (IA) or 63% (IIB) in east. Long-term population forecasts are the basis for long-term planning and political decision-making processes in many realms of society. Today, the public debate is dominated not only by family policy, in particular the expansion of state-sponsored child care, but also by developments in the healthcare system and in long-term care insurance. Another related discussion regards the capacity for financing pension schemes over the long term. In addition, there has been heated debate over the content of the new Immigration Act, which has now been ratified and will come into force on 1 January The background to these discussions is the fact that both future migration dynamics and the continued rise in life expectancy will have a significant impact on the size of the population and on the ratio of 'inactive' to 'active' inhabitants. The DIW Berlin periodically calculates population projections in order to illustrate the effects of alternative assumptions regarding these components. 2 In view of the substantial differences that exist with respect to fertility behaviour, life expectancy and migration behaviour, distinctions are made between east and west, and between Germans and foreigners. Past trends and future assumptions At the end of 2003, had a population of around 82.5 million, 67.7 million of whom lived in west and 14.8 million in east (cf. table 1). East and west have experienced contrasting trends since While west has seen massive inflows, which have vastly offset the low natural decline in the population, in east the drastic fall in the number of births has exacerbated the negative trend caused by emigration, and immigration from abroad has not sufficed to counterbalance the downturn. Between 1989 and 2003, east lost around 1.8 million inhabitants, while the population of west increased by around 6 million. Fertility trends The fertility behaviour of German women has changed perceptibly in recent decades both in east and in west. While the 1930 cohort still achieved the maintenance level _ in other words, 1000 women gave birth over the course of their lives to around 2100 children _ the 1950 cohort fell significantly below this level: in east The last population forecast was published in Cf. Erika Schulz: 'Zur langfristigen Bevölkerungsentwicklung in Deutschland _ Modellrechnungen bis 2050'. In: Wochenbericht des DIW Berlin, no. 42/1999, pp

2 Box 1 DIW Berlin population model In the population model of the DIW Berlin, the population is projected forward using the component method. The starting point is the population by single year of age, sex and nationality for east and west. Using data on births, deaths and population inflows and outflows (differentiated for domestic and external migration), the population at the beginning of the year is then extrapolated and aged by one year. The forecast of the individual parameters (birth and death rates, inflows and outflows) is based on analyses of past long-term trends. The fertility behaviour of German women is analysed and projected by year of birth (cohorts). The analysis and projection of external migration is carried out separately for the different population groups (ethnic German immigrants, other Germans, asylum seekers, EU citizens, other foreigners), because these figures are significant, for example with respect to their distribution across the individual federal states (and therefore between east and west ). It is no longer possible to project the population trend separately for Germans and foreigners and for east and west Germans as in previous forecasts because of the new German citizenship law, the growing importance of naturalisation and the reform of Berlin's borough boundaries. Since the Berlin borough reform of 1 January 2001, it is no longer possible to differentiate easily between East and West Berlin. For this forecast, the distribution for the year 2000 was used for the base population at the end of However, this will no longer be possible in the future without creating severe distortions. In other words, this is probably the last forecast which is structured according to the 'old' subdivision of regions. The impact of the new citizenship law is even more significant. Under this law, which came into force on 1 January 2000, children of foreign parents are automatically granted German citizenship if at least one parent has lived legally in for at least eight years. Most of these children also inherit their parents' nationality at birth. Between the ages of 18 and 23, they can then opt for one of the two nationalities. If they fail to act, they forfeit their German citizenship. In other words, while naturalised Germans acquire permanent German citizenship, children with dual citizenship risk 'denaturalisation' between the ages of 18 and 23. The DIW Berlin population model was revised to account for this change by drawing up separate forecasts for these children and listing some of them as foreigners again between the ages of 18 and 23. Naturalisations are taken into account by means of an age- and sex-specific naturalisation rate. In addition, the single years of age of the population were raised to 115 years in view of the increasing number of extremely old people in the population. The model's previous maximum age was 100., 1800 children were born per 1000 women, while in west the figure fell to less than 1700 children (cf. table 2). Not only did the pro-natalistic measures lead to a higher birth rate in the former German Democratic Republic, but the age of women at birth-giving of their first child was also much lower than in the Federal Republic of. Almost every woman born in 1950 had become a mother by the age of 21, and only 8% of women had no children. The political changes in east have led to massive changes in the fertility behaviour. This becomes particularly evident in the 1970 and successive cohorts. Up to around the age of 20, those born in 1970 behaved much like those born in 1950, but subsequently the birth rate plummeted (cf. figure 1). The change in the prevailing social situation, the uncertainty of individuals regarding their personal and professional futures and the spread of alternative lifestyles led to this abrupt drop in the birth rate. Between 1990 and 1993, the period-specific total fertility rate decreased by half in east. The drastic decline in the birth rate was considered partly a consequence of the circumstances of upheaval, and it was expected that some births would just be postponed to a later date, so that some cohorts would show something akin to a 'double birth peak'. 3 However, a catch-up effect of this kind, which was still expected in 1995, never really materialised. While the period-specific birth rates are now gradually recovering, the younger cohorts in east are not bearing any more children than those in west. All in all, recent years have tended to see an alignment with fertility behaviour in west, which has been relatively stable since the 1970 cohort. Little change is now expected in this respect (around 1400 children for every 1000 women), and it is assumed that there will be only a slight upward shift in the age of first-time mothers. It is not possible to carry out a cohort analysis of the fertility behaviour of foreigners because of the massive distortions caused by migration. The period-specific total fertility rate of foreign women has decreased substantially since In 1999, their birth rate was actually lower than that of German women in west. Since the introduction of 's new law on citizenship, the possibilities for analysing the fertility behaviour of foreigners have deteriorated further. According to what is known as the 'legal definition', the newborn children of foreign women who are eligible for German citizenship under the new law are counted as 3 Cf. Erika Schulz: 'Alternde Gesellschaft _ Zur Bedeutung von Zuwanderungen für die Altersstruktur der Bevölkerung in Deutschland'. In: Wochenbericht des DIW Berlin, no. 33/1995, pp

3 Table1 Demographic Trend in, 1970 to Population at beginning of year Net natural population growth Net migration 2 Year FRG 3 or west GDR or east Total FRG 3 or west GDR or east Total FRG 3 or west GDR or east Total In millions In 000s In 000s 1970 to end of to end of Individual years Total change from 1989 to end of to 2004: preliminary results. 2 Not including east-west migration since Data from 1971 to 1986 revised on the basis of 1987 census figures. Sources: Federal Statistical Office; Statistical Yearbooks for the GDR; DIW Berlin calculations. children of German women. This procedure leads to a mathematical increase in the birth rates of the latter and reduced birth rates for foreign women; this applied to around half of all births to foreign women in 2000 and As a result, their total fertility rate fell to half of the 1999 level (cf. figure 2). This forecast uses the old legal definition and assumes a total fertility rate of 1350, which is kept constant over the projection period. Cross-border migration The massive flow of immigrants to at the beginning of the 1990s was characterised by inflows of ethnic Germans, asylum seekers and civil-war refugees. However, the introduction of the regulation on safe third countries in asylum law, the modification of the regulations for ethnic Germans (who must now prove that they have some knowledge of the German language), and the return of some civil-war refugees to their countries of origin have all led to a perceptible decrease in net immigration in recent years. Between 1989 and 2003, a net total of around 5.3 million people migrated to. Of these, 3.1 million were foreigners and 2.2 million were Germans (cf. table 3). The new Immigration Act, which will come into force on 1 January 2005, will introduce new regulations regarding the residence and integration of foreigners in. Under this Act, the non-german spouses (provided the marriage has lasted at least three years) and the children of ethnic German immigrants can be included in the immigrant's admission notice only if they have basic knowledge of the German language and the immigrant expressly requests their inclusion. Inflows began to decrease already in 2003, when around ethnic Germans were admitted, and a further 335

4 Table 2 Birth Rates in Period-specific total fertility rates 1 Cohort-specific fertility rates 2 Year Germans Foreigners 3 Mothers' Germans West East year of birth 4 West East Sum of age-specific birth rates of women aged 15 to 50 in the year observed. 2 Sum of age-specific birth rates (births per 1000 women of each year of age) of women born in the same year. 3 West. According to legal definition (both parents foreigners and illegitimate children of foreign mothers) onwards: DIW Berlin calculations. 4 Partial estimation from 1955 cohort onwards, full estimation from 1985 cohort onwards onwards: new citizenship law. 6 Distribution between east and west estimated. Sources: Federal Statistical Office; Statistical Yearbooks for the GDR; DIW Berlin calculations. decrease can be expected. While the decline will be due partly to the new law, this type of immigration is likely to become less significant anyway in the future. Under the terms of the guideline on family reunification, it is still possible for the spouses and children aged up to 16 of foreigners living legally on German territory to join their families. Children aged over 16 but still not legally of age can be granted admission if they have basic knowledge of German and their prognosis for integration is positive. Citizens of the European Union generally enjoy full freedom of movement, while citizens of the accession countries are subject to the regulations laid down in the Accession Treaty of 16 April In addition, individuals are still admitted to German territory on humanitarian grounds _ including those subject to non-state or gender-specific persecution _ and immigrants facing deportation can be granted a residence permit when there are obstacles to their expulsion; this last rule has been introduced as an alternative to repeated tolerance orders, while tolerance has been maintained as an instrument for exceptional cases. The non-skilled and the low-skilled are still subject to a recruitment freeze, but this can be annulled in exceptional cases for skilled workers. There is no recruitment freeze for the highly skilled, who are immediately granted a settlement permit, which also entitles their family members to take up employment. The selfemployed are issued a residence permit if they invest at least one million euro and create ten jobs. Foreign students are issued a one-year residence permit on conclusion of their studies so as to enable them to seek work in. Little has changed as a result of the new Immigration Act for the main immigrant groups: those seeking family reunification and admission on humanitarian grounds. It remains to be seen how attractive will be for the highly qualified and the self-employed. The more important questions, probably, are whether the freedom of movement that will be granted at a later date to workers from the accession countries will lead to increased immigration to and how the expected decline in the domestic labour force potential will impact on immigration after A net annual average of around people migrated to between 1989 and 2003, but it must be remembered that the situation in the early 1990s was particularly unusual. By comparison, with an annual average of people, the net inflow during the period from 1970 to 2003 was lower. It is assumed here that future inflows will be in around the same order of magnitude. While inflows of ethnic Germans from eastern Europe will weaken significantly and probably be zero after 2030, it is assumed that the inflow of foreigners will increase in separate steps: first due to the 336

5 Figure 1 Cohort-specific Birth Rates of German Women in East Births per 1000 women Ageinyears DIW Berlin 2004 Source: DIW Berlin estimates and calculations. freedom of movement granted to workers from the accession countries after 2010 and, in variant II, once again after 2020 due to the decline in the domestic labour force potential. Overall, variant I shows that an annual average net inflow of immigrants can be expected, compared to in variant II. All in all, a total of between 10.9 million and 13.3 million persons will migrate to between 2002 and It is assumed that external migration will be distributed between west and east at a ratio of 80:20. Figure 2 Period-specific Birth Rates of Foreign Women According to Legal Definition Births per 1000 women up to and more Ageinyears DIW Berlin Under the legal definition, the birth rates are calculated by counting the newborn children with foreign citizenship as births by foreign women. The 'German' newborn children of foreign women (as defined by the new citizenship law) are not taken into consideration. Source: Federal Statistical Office. 337

6 Table 3 Migration Beyond German Borders 1 In 000s of people Year Germans Foreigners Total Of which inflows of Inflows Outflows Balance Inflows Outflows Balance Inflows Outflows Balance Ethnic Germans Asylum seekers to Borders of 3 October Sources: Federal Statistical Office; DIW Berlin calculations. Internal east-west migration Life expectancy The years 1989 and 1990 saw massive migration from east to west (cf. figure 3). All in all, a total of around people left east during this period. The outflows had already declined sharply by 1991, and inflows to east began to increase. East only lost around people through emigration in each of 1996 and Subsequently, however, outflow figures rose again. In 2003, emigration losses from the new federal states (not including Berlin) amounted to around people. The economic situation and the shortage of apprenticeship places are an important factor behind emigration from east. Given that little will change in these respects over the coming years, the forecast assumes that emigration from east will continue and amount to an annual average of persons up to Subsequently, outflows are expected to weaken to a net annual average of (up to 2020) and then of 6000 (up to 2030). The long-term expectation is that net migration will balance out between east and west. This assumption of weakening outflows is also based on the decline in the number of people seeking apprenticeship places due to the 'kink' in the birth curve or _ in general terms _ due to the size of the successive generation. The life expectancy of men has increased over the last four decades by over eight years, while that of women has risen by almost nine years. Newborn males, according to the 2000/2002 mortality table, 4 have a life expectancy of 75.4 years, while women can expect to live for 81.2 years (cf. table 4). The annual average rate of increase in life expectancy was much higher for the older age groups _ especially during the past 20 years _ than for newborn children. Thus, the average life expectancy of a 65-year-old man increased by almost 1% per annum in the 1980s and 1990s in west and by 1.1% in east, but these values were somewhat higher for an 80-year-old man (1.1% and 1.6%, respectively). The dimension of future increases in average life expectancy is heavily discussed among demographic experts. Some assume that _ as observed in the past _ a constant increase in average life expectancy can still be expected. 5 Others believe that as average life expectancy increasingly approaches maximum possible life expect- 4 It is assumed that the death rates for this period remain constant over the life span of a newborn child. 5 Cf. Jim Oeppen and James W. Vaupel: 'Broken Limits to Life Expectancy'. In: Science, 296, 2002, pp

7 Figure 3 East-west Migration in, 1957 to 2001 In 000s Outflows from eastern to western Inflows from western to eastern * DIW Berlin 2004 * Not including Berlin. Source: Federal Statistical Office. ancy, which is generally considered to be around 120 years, the rate of increase in average life expectancy can be expected to weaken. 6 This forecast takes both viewpoints into consideration by projecting mortality rates in two ways. On the one hand, the age- and sex-specific annual rates of change for the period 1987/99 (variant B) will be held constant over the forecast period. On the other hand, weakened rates of change are expected (variant A). In variant A, the life expectancy of men up to 2050 will increase by another 8 years to 83.4, while that of women will increase by 5.6 years to If the rates of increase remain constant, life expectancy will rise to 85.3 years for men and to 88.2 years for women. This procedure reduces the difference between the life expectancies of men and women. This is caused by a higher increase in the life expectancy of men than of women during the reference period _ a development that was partially due to the natural 'dissipation' of the war cohorts. Results Variant IA results in an increase in the German population to 84.5 million up to 2020, followed by a decline to 6 Cf., for example, Herwig Birg: 'An Approach for Forecasting Life Expectancy and its Application in '. In: Zeitschrift für Bevölkerungswissenschaft, no. 1, 2000, p million up to 2050 (cf. table 5). The trend up to 2020 will be more dynamic in west than in east. While the west German population can be expected to expand further to 69.5 million, the size of the population in east (15 million) will hardly change at all. In the remaining period up to 2050, a decrease in the population of around 7% in both regions to 14 million (east ) and 64.9 million (west ) can be expected. If annual immigration were to increase by persons and life expectancy were to rise by around two years, then the population of east would hardly fall below today's level (14.6 million) up to 2050, while the population of west would actually exceed the initial figure (68.5 million) by 1.1 million persons. Thus, according to variant IIB, the population of as a whole would be even higher, at 83.1 million, in 2050 than today. While there is a difference between the two variants of 4.2 million inhabitants in 2050, the shifts in age structure are similar. The share of children and young people aged under 20 will fall from 21% in 2001 to around 16% in 2050 in both variants IA and IIB; the share of the 'active' population (aged 20 to under 60) will fall from 55% to around 47% in both cases, while the share of those aged 60 years and over will increase from 24% to almost 37% (cf. table 6). East will be more strongly affected than west by the ageing of the population. In east 339

8 Table 4 Life Expectancy in East and West Life expectancy of a Abridged mortality tables Newborn child 65-year-old 80-year-old Men Women Men Women Men Women In years West 1960/ / / East 1960/ / / / Variant A Variant B Annual average growth rate (%) West 1961/ / East 1961/ / /2050 Variant A /2050 Variant B Sources: Federal Statistical Office; Statistical Yearbooks for the GDR; DIW Berlin calculations. the share of those aged 60 years and over will increase from 25% in 2001 to 40% in 2050, while in west the share of older people will increase from 24% to 36% over the same period and will thus be four percentage points lower than in east. How will migration and the rise in life expectancy affect the level of population, the number of 'active' people and the age structure? In order to answer these questions, three additional variants of the population forecast were calculated: a variant without further migration and without a further increase in life expectancy (variant 0C), a variant without further migration, but with the same further increase in life expectancy as in variant A (variant 0A), a variant without further migration, but with the same further increase in life expectancy as in variant B (variant 0B). Given that the aim is simply to illustrate the way in which migration and a further increase in life expectancy have an impact on the level of population and on the age structure, the following examples only refer to the results for as a whole. Without further immigration and with life expectancy remaining constant, the population of would decline dramatically _ to 54.5 million in 2050 (cf. table 7). A continued increase in life expectancy would lead, in the absence of further immigration, to an increasingly larger population. In 2050, 7.3 million more people (variant 0A) or 8.7 million more people (variant 0B) would live in than if life expectancy were to remain constant. This increase in the number of inhabitants is mainly due to an increase in the number of persons 60 years old and older (cf. table 8). Variant 0C results in around 20.8 million people in this age group in 2050, as opposed to around 28.9 million people _ 8.1 million more _ with 340

9 Table 5 Demographic Trend in by Age Groups In millions at end of year West East Age groups Variants Variants Variants IA IB IIA IIB IA IB IIA IIB IA IB IIA IIB 2001 Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Sources: Federal Statistical Office; DIW Berlin projections. variant 0B. Their share in the population as a whole amounts to around 38% in variant 0C, and around 46% if life expectancy rises constantly. The rise in life expectancy contributes very substantially to the ageing of society because mortality is reduced mainly in the middle and older age groups and the reduction is not _ as in the past _ based on a significant decrease in infant and child mortality. In addition to the rise in life expectancy, migration will also lead to an increase in the population by 2050 _ of 17 million (migration variant I) or 20 million (variant II). 's immigrants are mainly young and mobile, 7 and this counteracts the ageing of the society caused by rising life expectancy. The number of 'active' people will rise by 11.7 million (variant I) or 13.5 million (variant II) by This will increase the proportion of the 20- to under 60-year-old age group by 7 The inflows reinforce both the group of under 20-year-olds and that of 20- to under 60-year-olds, because family reunifications bring young spouses and children to, while admissions on humanitarian grounds and migration from EU countries bring young workers. 341

10 Table 6 Age Structure of the German Population (%) West East Age groups Variants Variants Variants IA IB IIA IIB IA IB IIA IIB IA IB IIA IIB 2001 Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Sources: Federal Statistical Office; DIW Berlin projections. almost 6 percentage points (I) or 6.5 percentage points (II). Although migration therefore slows the ageing of society compared with variants 0A and 0B, even the high inflows in variant II fail to fully neutralise the ageing of the population _ as already illustrated above. This situation is mirrored by the development of the old-age dependence ratio, which is observed here for three age categories. Normally the old-age dependence ratio is defined as the number of those aged 60 years and over in relation to the number of 20- to under 60- year-olds, because employment levels are relatively low prior to age 20 and then fall sharply after age However, because it will be almost impossible to draw a full pension before age 65 owing to the new pension-law regulations, the trend for an old-age ratio 2, defined as '65- year-olds and older in relation to 20 to under 65 yearolds', is also calculated. Finally, the debate about raising 8 According to the 2003 microcensus, the activity rate, i.e. the labour force (employed and unemployed) as a share of the population, was around 34% for men and 27% for women aged between 15 and 20. The activity rate for 60- to under 65-year-olds was around 37% for men and 18% for women. 342

11 Table 7 Impact of Migration and Life Expectancy on Population Growth in In millions OC OA OB IA IIA IB IIB Variant No migration Net immigration Lower Higher Lower Higher Life expectancy Constant Weakened increase Constant increase Weakened increase Constant increase Effect of rising life expectancy Effect of net immigration added to effect of rising life expectancy Source: DIW Berlin forecasts. the standard retirement age to 67 determines the need for old-age ratio 3, defined as '67 years old and older in relation to 20- to 66-year-olds'. It becomes clear that by 2050 in all variants with migration and rising life expectancy old-age ratio 1 will be almost doubled, while old-age ratios 2 and 3 will more than double (cf. table 9). Without additional immigration, by contrast, the trend would be much less favourable. Most future migrants to will, as assumed, be foreigners. variant I results in net immigration of around 9.6 million foreigners between 2002 and 2050, compared with 12.0 million with variant II. Under the terms of the new Immigration Act, new immigrants are entitled to integration benefits, especially language courses, while integration is also being increasingly promoted for foreigners already living in. It is hoped that these efforts will help them to integrate better into society and the labour market. A large share of foreigners already satisfies the requirements for naturalisation, and this share could increase further in the future. Naturalisations and the new citizenship law mean that the share of citizens with foreign nationality will not rise to the extent assumed in previous forecasts. The share of foreigners will increase to something over 13% in 2050 according to variant I, while variant II results in a share of around 15%, with a somewhat lower share of foreigners in east with the latter variant (cf. table 10). Comparisons with other forecasts Last year, the Federal Statistical Office published the results of its 10th Coordinated Population Forecast. 9 The Federal Statistical Office carries out its calculations on the basis of three variants each for life expectancy and migration (cf. table 11). variant W3, which has the highest level of migration, is comparable with variant II presented here, while variant I is more similar to the middle migration variant W2. Eurostat also published a population projection for in 2000, which assumes there will be an 9 Cf. Bettina Sommer: 'Bevölkerungsentwicklung bis 2050 _ Annahmen und Ergebnisse der 10. koordinierten Bevölkerungsvorausschätzung'. In: Wirtschaft und Statistik, no. 8, 2003, pp For further information, cf. Federal Statistical Office Internet pages: 343

12 Table 8 Impact of Migration and Life Expectancy on Age Structure of Population in OC OA OB IA IIA IB IIB Variant No migration Net immigration Lower Higher Lower Higher Life expectancy Constant Weakened increase Constant increase Weakened increase Constant increase In millions 2001 Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Age structure (%) 2001 Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Total Aged under Aged 20 to under Aged 60 and over Source: DIW Berlin projections. increase in life expectancy to 85 years for women and to 80 years for men, as well as annual net migration of in the standard variant. 10 In addition, Eurostat 10 These are population projections that Eurostat has drawn up specifically for the EU Working Group on Ageing. The results have been published in: EU Economic Policy Committee: Progress Report to the Ecofin Council on the Impact of Ageing Populations on Public Pension Systems. Brussels expects a somewhat higher birth rate than the Federal Statistical Office and the DIW Berlin, namely 1500 children per 1000 women. The DIW Berlin's variant IIB is reasonably comparable with the Federal Statistical Office's variant 9. The trend is similar, but the population in variant 9 is 1.8 million lower in 2050 than the results presented here. As pointed out above, a rise in life expectancy has a sub- 344

13 Table 9 Development of Old-age Dependence Ratio in in Relation to Migration and Life Expectancy Ratios (%) OC OA OB IA IIA IB IIB Variant No migration Net immigration Lower Higher Lower Higher Life expectancy Constant Weakened increase Constant increase Weakened increase Constant increase Old-age ratio Old-age ratio Old-age ratio Ratio of number of those aged 60 years and over to number of 20- to 59-year-olds. 2 Ratio of number of those aged 65 years and over to number of 20- to 64-year-olds. 3 Ratio of number of those aged 67 years and over to number of 20- to 66-year-olds. Source: DIW Berlin projections. stantial influence on the size of the population, so that the higher life expectancy assumed here for men is probably partially responsible for the difference. However, the difference could also be due to a different temporal distribution of the inflows and possibly to the extension of the maximal life span in the model to 115 years. Thus, the Federal Statistical Office's variant 9 has around persons aged 95 years and over, while the corresponding figure is around in variant IIB. The Federal Statistical Office's variant 5 can be compared to the DIW Berlin's variant IA. There is a difference of 3.8 million persons in 2050, which is also probably a result of the higher life expectancy for men and a different distribution of net immigration over time, which is actually only 0.4 million higher in total. The number of persons aged 95 years and over is higher with variant IA than in the Federal Statistical Office's variant 5. Eurostat's standard variant results in a lower number of inhabitants than variant IA, but this figure is higher than with the Federal Statistical Office's variant 5. Although the level of population varies in the individual forecasts, the trend for the old-age dependence ratio (ratio of those aged 60 years and over to 20- to under 60- year-olds) is similar: it amounts to 77-78% in 2050, compared to 75% for Eurostat. This illustrates clearly once again that while alternative assumptions about the mortality and migration trends have an impact on the age structure, the ageing of society seems inevitable nonetheless. Is the solution a higher birth rate? In order to clarify the influence of this variable, it was assumed for 345

14 Table 10 Demographic Trend in by Nationality In millions at year end Nationality Total Aged 20 to under 60 IA IB IIA IIB IA IB IIA IIB 2001 Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) West 2001 Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) East 2001 Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) Germans Foreigners Total Share of foreigners (%) Source: DIW Berlin projections. 346

15 Table 11 Assumptions and Results of Selected Population Forecasts and Model Calculations Assumptions/ results Eurostat 2000 DIW 2004 Federal Statistical Office 2003: 10th Coordinated Population Forecast Birth rates : Total fertility rate will increase to 1500 up to 2018, then remain constant West : Slight rise in average age of first-time mothers; 1995 cohort will bear 1400 children (German) East : Alignment with west German fertility behaviour up to 1995 cohort : Birth rate of foreign women will remain constant at West : Current total fertility rate of 1400 will remain constant East : Alignment with west German fertility behaviour up to 2010 Life expectancy : Increase to 85 (women) and 80 (men) West : Further increase in life expectancy West : Life expectancy of newborn children Men Women Men Women A L B L L Migration 1999 to 2050 East : Sharper increase than in west, same mortality rates in the long term (2020 onwards) Net migration in 000s East : Alignment with west German mortality rates up to 2020, then same trend as in west 2002 to to 2050 Germans Foreigners Total Germans Foreigners Total Variant N/0/W Variant ST/I/W Variant H/II/W Standard variant Variant I Variant II Variant 0 Variant W1 Variant W2 Variant W3 Average annual net migration Results (year end) Standard variant Variant IA Variant IIB Variant 0A Variant 1 Variant 5 Variant 9 Total population in millions to under 60 year-olds in millions Old-age dependence ratio Under old legal definition. 2 Ratio of persons aged 60 years and over to 20- to under 60-year-olds (%). Sources: Federal Statistical Office; Eurostat; DIW Berlin projections. 347

16 variant IA, for example, that all women immediately return to the maintenance level. The fertility behaviour of the 1937 cohort, whose total fertility rate amounts to 2105, was used here and kept constant over the forecast period. In the original variant IA, the old-age dependence ratio 1 increases to 77%, whereas in the event of an immediate return to the maintenance level it would amount to only 61% in The slowing of societal ageing sets in after a time delay when those born in the baby-boom years reach working age. In 2010 (46%) and 2020 (55%) the old-age dependence ratios are the same as in variant IA; but as early as 2030 the slowdown already amounts to 5 percentage points and then reaches 10 percentage points in Thus, the ageing of society follows a significantly different path in the alternative calculation. A change in fertility behaviour as massive as the one assumed here seems highly unlikely, however, from today's perspective. Erika Schulz 348