Estonian population projection according to different migration scenarios

Introduction

A number of industrialised countries are currently experiencing slowing population growth and decreases in the size of their labour forces. Most countries of Europe, as well as China, will be faced with this same situation in the next decade. Estonia has been experiencing this demographic reality already for the quarter of a century. Compared with the beginning of the 1990s, the size of the Estonian population is smaller by more than 250,000 people, out-migration being one of the contributors to the population decline. Recently, the continuing population decline has received vocal attention from entrepreneurs’ associations as well as other organisations monitoring the country’s economy, such as the Bank of Estonia (Tamsar 2016). The main concern from an economic point of view is the pressure on increased wages due to declining numbers of working age adults. Fewer people available to join the workforce create a situation of heightened competitiveness to hire employees thus resulting in higher wages. The second major concern is related to the declining number of taxpayers relative to the groups dependent on tax income.

Increasing immigration has been suggested as one of the solutions to problems arising from declining population. Arguments favouring immigration in order to augment the working age population are based on the Statistics Estonia projections that analyse potential Estonian demographic changes until 2040. The Statistics Estonia projections cover a relatively short timeframe in the population development context, thus not allowing evaluation of long-term implications of various demographic transformations. To fill this gap, we created a population projection for the Estonian Human Development Report, which covers the whole of the 21st century with the purpose of clarifying the effects of demographic developments in the long-term perspective. The uniqueness of these projections lies in differentiating population groups by ethnic self-identification, which is relevant in the migration context, and that has not been provided by any previous population projection about Estonia.

We seek to answer two interrelated questions by using the projection as a tool. Firstly, how would population size, the number of working-age people, age and ethnic structure of the population of Estonia change in the context of different migration scenarios? Secondly, what might be the effect of fertility on population transformations – fertility being the second demographic process that has showed large variation during the period of regained Estonian independence. The principles and methodology of the population projections are explained in the beginning of this sub-chapter.

General principles of a population projection

The basis of modern population projections is the cohort component method that considers population changes coherently (Wattelar 2006). Specifically, the projection shows the step-by-step changes in birth cohorts in the population from one calendar year to the next one. Every year a new birth cohort is added to the projected population, while older cohorts are decreased by the number of deaths. In the case of open populations, the cohort size is affected also by migration. The outcome of the cohort component projection is the size of different cohorts in the future; it also allows the forecasting of changes in population groups by socio-economic or cultural characteristics (Samir et al. 2010, Loichinger 2015).

The timeframe of the projection depends on its purpose. Longer timeframes compared with previous projections can be found in the works of many leading centres conducting such exercises. The United Nations shifted the endpoint of its projection from 2050 up to the end of the 21st century, while Eurostat shifted its projections from 2060 to 2080. The ability of long-term projections to distinguish and present clearly the effects of demographic trends can be considered one of the reasons for lengthening the timeframe of projections. The knowledge of the impact helps societies to develop strategies to better deal with challenges arising from population change.

Contemporary population projections are usually constructed with several scenarios. Comparing scenarios that have different assumptions gives practical information about the outcome that a population may end up with in the event that one or another assumption manifests over time. Given the impact of changes in policies or demographic dynamics, a population projection can indicate possible development paths that a society should strive for with the support of policy measures. Also, comparing different projection variants helps to bring out similar by outcome, meaning that the chances of influencing processes underlying these similar outcomes will yield little or no effect.

The methodology of the Human Development Report population projection

The purpose of the population projection prepared for the Human Development Report is to evaluate the changes in the population size and structure of Estonia in the short-term (up to 2040), mid-term (2040-2070) and long-term (2070-2100) perspective, depending on migration, fertility and mortality developments. On the demographic time scale these timeframes correspond to the lengths of one, two and three generations. We base our interpretations on “normal” development, without assuming any large or unexpected societal disruptions.

The age and sex structure of the total population of Estonia - Estonians as well as other ethnic groups as of beginning of 2015 - is the basis of this projection. Calculations were done using the cohort-component method, in which the changes in size and composition of population groups were projected up to the year 2100. Information for the total Estonian population was acquired by summing projection results obtained for sub-groups (Estonians and other nationalities; men and women).

Around twenty scenarios were played around with, combining various developments regarding the future dynamics of migration, fertility and mortality. Some of the demographic processes correspond to the assumptions made about Estonia in the latest United Nations projections published in 2015. Using United Nations projection enabled fertility and mortality options to be based on an internationally acclaimed methodology (probabilistic projection methodology). This also makes comparison of results with other countries’ projections easier. Next, we briefly explain the migration, fertility and mortality assumptions that form the basis of this population projection’s scenarios.

Migration Scenarios

Projecting future migration developments is considered to be one of the most complicated tasks when creating population projections. Therefore, migration often receives less attention in projections than fertility or mortality. The United Nations projection, used as the basis for our projection, is no exception.

For most countries the United Nations projection expects a continuation of the current migration situation until 2050 (UN 2015). It is assumed that the positive as well as negative net migration (migration balance) will decrease by about 50% by the second half of the century (additionally the United Nations projection also contains a zero-migration scenario). The United Nations migration scenario for Estonia foresees a decrease in negative net migration. In the first half of the 21st century it is expected that out-migration will be on the scale of 1,000 people per year, while for the second half of the century this is expected to halve (-500 people per year). Estonia’s negative net migration between the intercensal period 2000–2011 was actually 3.5 times larger (-3500 persons per year on average). The projection outcome would be more pessimistic if extrapolated, as exemplified by the US Census Bureau’s projection results for Estonia (US Census Bureau 2016). In the case of the Human Development Report projection, this path is taken as the most pessimistic migration scenario (R-). Should it materialise, migration to such an extent would decrease the Estonian population size by 71,300 people over the projection period (-5.4% of the 2015 population).

Additionally, the Human Development Report projection uses three migration scenarios (Table 1.5.1.). The scenario of balanced migration flows (R0) assumes that in-migration and out-migration reach a balance during the current decade and stay at this level over the projection period. The moderate in-migration scenario (R+) foresees the declining dominance of out-migration and reaching a balance of flows by the second half of the 2020s. Starting from the 2030s the moderate in-migration scenario predicts a positive net migration in the scale of 1,000-1,250 persons a year. This scenario would bring in almost 79,000 new people by the end of the projection period (6% of the 2015 population size). A strongly positive in-migration scenario (R++) assumes larger positive net migration (2,000-2,500 persons a year) already starting from the current decade. The latter scenario would increase the direct migration effect on the Estonian population size during the projection period by 186,100 persons (14% of the 2015 population).

Table 1.5.1. Human Development Report population projection migration scenarios.

Table 1.5.1 presents assumed net migration per different migration scenarios for Estonians and other ethnic groups for the life of the projection period. The scenarios of balanced migration flows assume similar in- and out-migration size and structure for both population groups. In-migration scenarios predict a replacement of negative net migration of Estonians with a moderate positive balance (+500 persons a year) by the beginning of the 2030s. Assuming this positive in-migration holds for 30 or 40 years, (return) migration would increase the number of Estonians in Estonia by 11,000 or 16,000 persons, correspondingly. However, most of the positive net migration in the cases of the R+ and R++ scenarios would be based on other ethnic groups.

Fertility and mortality scenarios

The basis of the Human Development Report fertility and mortality scenarios are forecasts that are considered most likely by the United Nations (UN 2015). The United Nations uses stochastic models in determining these future directions, which take into account the specifics of population development in each country as well as the general experience of similar countries (Alkema et al. 2011; Raftery et al. 2014).

For low-fertility countries the United Nations’ main scenario assumes a moderate increase in fertility and its stabilisation by the second half of the 21st century (for European countries at 1.72–1.95 children per woman). For Estonia, the United Nations predicts an increase in the total fertility level from 1.66 to 1.87 children per woman in the last quarter of the century. Such a fertility level (1.87) is close to the actual number of children born to female generations that end their reproductive period in the current decade. According to assessments done by Myrskylä, Goldstein and Cheng (2013) the average number of children for women born in Estonia in 1970, 1975 and 1979 will reach 1.86, 1.89 and 1.90, respectively. Similar results have also been obtained by other authors (Puur, Klesment 2012).

Four additional scenarios are used in the Human Development Report to explain the impact of different fertility levels from the main scenario (S0) (Table 1.5.2). The first one (S++) assumes an increase of the total fertility rate by the 2050s at the level of 2.03–2.04 children per woman and reaching the replacement level (2.07 children) by the last decades of the century. The realisation of this scenario would require much stronger child- and family-friendly policies of the society. A stabilisation of fertility at the current level of 1.67 children per woman is assumed in the case of the unchanging fertility scenario (S–). The unchanging fertility scenario would allow us to see where the current situation would lead Estonia if it remained at that level. The assumed fertility levels of the rest of the fertility scenarios (S- and S+) were between the main scenario and the unchanging-fertility scenario, and the main scenario and the replacement-level fertility scenario levels (Table 1.5.2).

Table 1.5.2. Human Development Report population projection fertility scenarios.

As with migration, the Human Development Report projections’ fertility scenarios differ somewhat by nationality. We expect a slightly higher fertility for Estonians in comparison to the total population and a lower fertility by a tenth for other ethnic groups than for Estonians. The 2011 population census showed that the actual fertility difference between women born in the 1950–1970s was actually bigger, but the report projection considers possibilities of further integration of the foreign-origin population. Complete fertility convergence can be hardly expected due to the foreign-origin population concentrating in cities where fertility is lower as a rule. Second-generation migrants have demonstrated a lower fertility by a tenth also in Sweden (Andersson, Persson 2015).

The report projection’s mortality scenario is based on the assumptions of the United Nations population division (UN 2015). With regard to Estonia the projection expects prolonging life expectancy up to age 85 for men and age 90 for women by the end of the 21st century. It is also assumed that the life expectancy gap will disappear between Estonians and other nationalities quite soon. In 2014 the life expectancy gap between Estonians and other nationalities had decreased to 2.3 years for men and 0.7 for women. Between 2006-2010 it reached over four years for men and over two years for women.

Main scenarios of the Human Development Report projection

Twenty different projection variants were calculated as the result of combining fertility and migration scenarios for the total Estonian population - Estonians and other nationalities. We present and analyse five of these scenarios that are distinct (Table 1.5.3).

• V1 is the base variant of the projection, showing the outcome of the Estonian population in the case of a continuation of the current demographic situation.
• V2 and V3 assume that negative net migration will be replaced with a balance of in- and out-migration. People would move in and out of Estonia, but the numbers of in-migrants and out-migrants would be the same. In terms of fertility, V2 expects the end of postponing childbirth to later ages by the end of the next decade and the gradual increase of fertility to similar levels to that of female generations of the 1970s (1.87 per woman).V3 plays out the scenario of successful contribution into a family- and child-friendly society, which would lead to a replacement-level fertility in the second half of the 21st century.
• V4 and V5 anticipate that out-migration is substituted with increased in-migration. The number of Estonians living abroad would decrease due to a return migration of them; additionally a couple of thousand foreigners would enter into Estonia every year. The fertility scenarios assumed for V2 and V3 are repeated for V4 and V5.

Table 1.5.3. Human Development Report projection analysis variants

Note: The differences in migration and fertility between Estonians and other nationalities is described in Tables 1.5.1. and 1.5.2. The average life expectancy is expected to increase in the case of all projection scenarios up to 85 years for men and 90 years for women until year 2100.

Projection results

Population size

The Estonian population size decreased by a quarter of a million between 1990 and 2015. According to the report population projection, the population decrease is likely to continue also during the next quarter of the century, however the scale of it depends on the chosen development path (Figure 1.5.1). Continuation of the current trends into the future (V1) would lead to a decline in population size of more than 150,000 (-12%) persons by the year 2040. Achieving a balance of migration flows together with a moderate or stronger fertility increase would reduce population decline to 108,000 (V2) or 92,000 (V3) persons, correspondingly. If the negative net migration is replaced with large positive net migration in the next few years, coupled with a fertility increase, the population decline would stay between 45,000 (V4) and 27,000 (V5) in the years 2015–2040 (-3% or -2%).

Grouping projection outcomes by migration scenarios (V2-V3 and V4-V5) reveals that migration could have a larger effect on changing population size than fertility within a 20-year timeframe. The impact of fertility differentials would be visible in the population size later, increasing impact each following decade. Figure 1.5.1 illustrates the comparison between outcomes with similar migration scenarios, but diverging fertility assumptions (V2-V3 and V4-V5).

In the longer term, V1 shows that the prevalence of out-migration and the continuation of low fertility rates would cause a decrease of the Estonian population to levels below the level of 800,000 persons by the end of the century, without any signs of stabilisation. A moderate improvement with regard to migration (balance of the flows) as well as fertility (TFR of 1.87 or 90% of the replacement level) would decrease the scope of the population decline by a few hundred thousand. In case of realisation of the V2 scenario, Estonia would move into the 22nd century with about a million inhabitants – however, that would not be sufficient to stop further population decline.

Figure 1.5.1. Projected population size, Estonia, 2015–2100

Source: authors’ calculations

Two scenarios, V3 and V4 promise an end to the population decline in the second half of the 21st century. These variants differ from each other by the population size stabilisation mechanisms. In case of V3, population decline would stop due to fertility reaching replacement levels, supported by the balance of migration flows substituting negative net migration. V4 would halt the population decline by a constant and rather large surplus of in-migration, supplemented with a moderate fertility increase. In case of replacement-level fertility and balance of migration flows the Estonian population size would be 1.14 million by the end of the century; in case of strong in-migration and a moderate fertility increase, the population size would reach 1.23 million. The different population size in 2100 proceeds from the fact that V3 assumes a step-by-step fertility increase over several decades while V4 foresees dominance of in-migration already in the coming decade. Since these are specific assumptions done for the projection, it should not be deduced from these figures that contributing to in-migration would result in a smaller population decline than contributing to fertility. An examination of the numbers show that if population size would be kept stable solely by in-migration in the second half of the century, then a somewhat larger annual number of immigrants would be required starting from 2090 onwards than assumed in the V4 scenario. To avoid decline, the annual excess of in-migration should be around 3,000 persons per year in the final decade of the 21st century. This would lead to a positive net migration in the amount of 195,000 – 200,000 people for the life of the projection period.

Population growth instead of decline would happen starting from the year 2040, only if scenario V5 is realised. In this case the number of Estonian inhabitants would increase in the second half of the 21st century by more than 2,500 persons a year, and would reach 1.4 million (7% more than currently). The basis of this development is the co-effect of increased positive in-migration and replacement-level fertility.

Working-age population

The size of the working-age population (ages 20–64) in Estonia has declined by 132,000 people in the 1990–2015 period. As the working-age population forms more than half of the total population, it is not a surprise that the change in the working-age population to a significant extent corresponds to the changes in population size described above. In terms of working-age population dynamics, the report projection variants are clustered into three different groups (Figure 1.5.2).

In the case of the realisation of variants V1 and V2, Estonia would face a persistent and very large decrease in the working-age population. If demographic processes would follow the current path, with only slight positive changes (V1) and the parameters of the working age stay the same, then Estonia would have to cope with a workforce that is approximately two-times smaller (395,000 persons) by the end of the 21st century. In the case of balanced in- and out-migration flows and moderate fertility increases, the number of 20–64-year-olds would decrease to half a million by the end of the century. Similar to population size, the realisation of variants V3 and V4 would lead to the stabilisation of the working-age population size in the second half of the century, around 585,000 or 640,000 persons, respectively. Again, variant V5 differs from the rest of the working-age population outcomes – its assumptions would lead to an increase in the number of working-age people beginning in the 2060s. However, unlike population size, the working-age population would not reach similar levels as in the beginning of the projection period.

The dynamics of the working age population and the population size by different scenarios, presented in Figures 1.5.1. and 1.5.2., display divergent variance for these two trends of numbers. The main reason behind this is the inertia of demographic processes, which has its greater effect on the working-age population. Due to this inertia, the scale of decrease of the working-age population is larger in all scenarios than it is for population size. Population decline in 2015–2060 would stay within 2% in case of the most optimistic variant, V5, which assumes strongly positive in-migration and replacement level fertility. The decline of the working-age population would be 17% by 2060 in case of the same scenario. A sudden increase in fertility starts to influence the number of working age people with a 20-year delay – this is the time that is required for the generations born during the fertility increase to reach working age. Figure 1.5.2 illustrates this delay with the similarity in scenarios V2-V3 and V4-V5 up to the 2050s. Such a two-decade-long delay is probably one of the main reasons why debates concerning the workforce focus less on fertility than increasing in-migration.

The Human Development Report projection results for the coming two-three decades indicate that even if negative net migration would be substituted by a positive net migration in the amount of a couple of thousand people annually, supported by replacement level fertility, Estonia would still have to be prepared for a continuous decrease in the working-age population up until the 2050s. According to the most optimistic report projection – scenario V5 – the decrease of the working-age population in the 2015–2040 period would be around 89,000 persons, followed by a further decline during 2040–2060 of 43,000 persons. Keeping the size of the working-age population at the current level, which we do not consider realistic, would translate into accepting about 170,000 new incomers in the next 25 years, and even more in the longer-term perspective. That would translate into a net migration of 7,000 people per year, which is close to the average number in the Soviet era.

Figure 1.5.2. Projected working-age population (ages 20–64), Estonia, 2015–2100

Source: Authors’ calculations

Age structure

In addition to a population decline, Estonia has also experienced a rapid population ageing since the beginning of the 1990s. This has been driven by changes in demographic processes (fertility decline, negative net migration and an increase in average life expectancy), as well as the large cohorts of post-war immigrants reaching old age. As a co-effect of these factors, the relative number of older (65+) people in the population increased between 1990 and 2015 in Estonia over 1.6 times – from 11.6% to 18.8%.

All report projection scenarios indicate a remarkable increase in the proportion of older adults until the end of the 2050s (Figure 1.5.3). This peak would be followed by a moderate decline during the 2060s, caused by the small sizes of cohorts that were born in Estonia after regaining independence reaching retirement age. Depending on the projection scenario, the above-mentioned decline would be followed by a new increase or a stabilisation in the numbers of older people from the 2070s onwards. Population ageing would be largest in the case of continuing negative net migration and the persistence of low fertility (V1). If this scenario is realised, the over-65 people will comprise 31% of the population in the 2060s. The share of older people would not reach levels that high for other projection scenarios (27–29%).

The second half of the century is interesting due to the change in the interrelationship between the V2-V5 scenarios. After the peak of ageing indicators, the indicators of the scenario of balanced migration flows (V2) and that of the positive net migration scenario (V5) start to approach each other. By 2100 the difference in the relative numbers of older people between these scenarios is below one percentage point. The ageing indicators of V3 and V5, both based on different migration assumptions, resemble each other even more. The reason for the regrouping of the projection variants is the fertility difference, which turns out to have a larger impact on the population age structure than migration. Therefore, in the case of a moderate fertility increase (V2 and V4), older people would comprise 27–28% of the Estonian population in the second half of the century. An increase of fertility to replacement level (V3 and V5) would stabilise the shares of the older population to around 25%, resembling the situation of the 2040s.

The pronounced long-term effect of fertility on the population age structure should not come as a surprise, since large in-migration flows can have a rejuvenation effect on the population only for a few decades, when incoming people to Estonia have not yet reached retirement age. The effect of fertility on rejuvenation of the population age structure remains larger as well as lasts longer.

Figure 1.5.3. Projected proportion of older people (ages 65+), Estonia, 2015–2100

Source: Authors’ calculations

The dynamics of the proportions of the working-age population (ages 20–64) in Figure 1.5.4 reveal that there is a very small difference in the projection variants in the proportion of the working-age population compared to the non-working-age population. According to the report projection Estonia is nevertheless headed towards a decreasing share of the 20–64-year-olds from the current 61% to 49–51% level by the end of the 2050s. Compared to the extent of the expected decline in the number of working-age persons, the differences in relative share between scenarios are five-six times smaller. The decreasing trend stops in the second half of the century, replaced by a fluctuating at lower levels trend. The differences in proportions of the working-age population fit within a couple of percentage points also during the last decade of the century.

Similar proportions of the working-age population and non-working-age population in different projection scenarios is possible due to the balance between the proportions of older people and children. Comparing Figures 1.5.3 and 1.5.5 reveals how a larger share of older people goes hand in hand with a smaller share of children and youth in the population (variant VI) while a larger share of the latter becomes evident with smaller proportion of older people (V3 and V5) in the society. Though younger people, often in the ages of family formation, constitute a large part of immigrants, in-migration is not automatically accompanied by a big proportion of children and youth in a population. Examining the variants V2-V3 side by side demonstrates that fertility holds the key to a more sustainable age structure for a population (Figure 1.5.5).

Figure 1.5.4. Projected proportion of working-age population (ages 20–64), Estonia, 2015–2100

Source: Authors’ calculations

Figure 1.5.5. Projected proportion of children and youth (ages 0–19), Estonia, 2015–2100

Source: Authors’ calculations

Ethnic structure

Estonians formed 62% and other ethnic groups formed 38% of the total population of Estonia at the time of regaining independence. The end of immigration from the Soviet Union and, in part, the out-migration of people who had earlier moved here, increased the proportion of Estonians to 68% by the end of the 20th century. By the beginning of the report projection period the proportions had shifted by one additional percentage point to the benefit of Estonians.

From the perspective of the ethnic structure, the report projection variants can be grouped into two (Figure 1.5.6). In the case of large positive net migration (V4 and V5) the proportion of Estonians would slowly decrease. By mid-century it would decrease to 68% in the case of both variants. By the end of the 21st century Estonians would constitute 64-65% of the total population. In case of fertility reaching replacement level (V5), the projected number of Estonians would be larger by 1,000 than in the beginning (909,000) of the projection period, and the number of people belonging to other ethnic groups would increase from 405,000 to 493,000. In the case of large positive net migration and moderate fertility increase (V4) the number of Estonians would decrease to 786,000, while inhabitants of other groups would increase to 445,000.

In the case of other projection variants (V1-V3) the relation between different nationalities would shift slowly to the benefit of Estonians. In the middle of the century the proportion of Estonians would have increased up to 73%, by the end of the projection period Estonians would constitute 77-78% of the total population. The described transformation in proportions would happen both in the case of continuing negative net migration as well as balance of migration flows. The source of this shift is the assumption of a 10% fertility differential.

Figure 1.5.6. Projected proportion of Estonians, Estonia, 2015–2100

Source: Authors’ calculations

Despite the similarly changing proportions of ethnic groups, the number of Estonians and other nationalities would result in diverging outcomes under different scenarios. If current trends would continue (V1) the number of Estonians would decrease to 617,000 persons and the number of other ethnic groups would decrease to 173,000 persons by the end of the projection period. If there would be a balance of migration flows and a moderate fertility increase (V2), the number of Estonians would decline to 767,000 and other groups would decline to 221,000 persons by the year 2100. If a balance in migration would be supplemented with a fertility increase up to the replacement level, then the size of Estonians and other nationalities would be 888,000 and 252,000 persons, respectively.

Comparing the Human Development Report projection variants enables one to assess the scope of the positive net migration, in which case the excess of in-migration would not change the proportions of Estonians and other ethnic groups in Estonia. Such an equilibrium point is located at around up to +1000 people per year (positive net migration), given that the assumption of a 10% difference in fertility levels in favour of Estonians holds. However, if one of the demographic targets were to be slowing the decline of Estonians, then it would be necessary to increase fertility close to replacement level as well as balance the in- and out-migration of Estonians.

Summary

This chapter aimed to clarify the Estonia’s demographic future from the perspectives of different migration and fertility scenarios based on a population projection. The analysis indicated a large variety of possible demographic paths leading up to the end of the 21st century. This holds especially true for the size of the working-age population as well as the total population – in both cases the outcomes of the projection variants differed two-fold by the end of the century. The projection does not give an answer regarding how the future will actually be, but comparison of alternative future paths provides food for thought in terms of which road Estonian society should strive for.

The comparison of the projection scenarios indicates that the least desired outcome from the sustainability point of view would be the continuation of current trends without any notable changes. Reducing the excess of out-migration and increasing fertility to the highest levels of the period of regained independence in 2008–2010 would not be sufficient to stop the declining size of the total population and working-age population.

Somewhat simplifying, there are two different possible ways to stop the decline of population size and the working-age population, and to stabilise the trends. One path assumes a migration pattern where the movement of people across international borders may be intense, but in the long-term is governed by a balance of in- and out-migration flows. This path is supplemented by an assumption of replacement-level fertility where children’s generations would not be smaller than those of their parents’. Second, an alternative path foresees a moderate fertility increase that does not assume equal sizes of children’s and parents’ generations. In order to stabilise the population size, the declining birth cohorts will have to be augmented by rather large in-migration flows in the long-term perspective.

A turn towards the increase in total and working-age populations in the middle of the 21st century requires the fulfilment of at least one of two preconditions – fertility has to increase above the replacement level, or an even larger prevalence of immigration than mentioned above has to occur. This chapter experimented with only the strongly positive net migration scenario out of the possible alternatives leading to an increase in population size and working-age population. In total this scenario expected a positive net balance of almost 190,000 persons between 2015–2100, supported by replacement-level fertility (2.07 children per woman). This would lead Estonia to a population size of 1.4 million, larger than it is at the moment.

The analysis of paths of population size stabilisation (or growth) brings to attention two circumstances. Firstly, the comparison of projection scenarios revealed that the changes in the proportions of children, the working-age population and older people, which are crucial from the perspective of the functioning of the economic and social system, are in the long-term affected more by fertility levels. Therefore bring fertility closer to the replacement level would lead to more sustainable development than advancing immigration. Secondly, the analysis indicated that the choice of migration policy is decisive for the coming changes in the ethnic structure. Concentrating on immigration of foreign origin population would lead to a gradual decrease in the proportion of Estonians in the future, while in the case of balanced migration flows, the outcome would likely be a slow increase of Estonians.

In addition to the changes dependent on demographic pathways presented in these scenarios, the analysis highlighted developments that are not likely to be affected by the choice of demographic paths. One of these outcomes is the decline in the number and proportion of the working-age population in the next few decades – this trend will continue up to the middle of the century in the case of even optimistic fertility and migration scenarios due to the inertia of population processes. In-migration may alleviate this tendency to some extent, but successful strategies for coping with the demographic transformations that lie ahead for Estonia have to be more diverse. These could include investments in the health sector, resulting in longer working life, good quality education, life-long learning opportunities and family policies that advance a child- and family-friendly society. Additionally, coping with demographic changes presupposes developing economic competitiveness which is vital for increasing the standard of living as well as sustaining funding for education, culture, science, state defence and social programs. Developing all these policies wisely and basing them on knowledge-based principles helps directly or indirectly to solve the challenges arising from demographic development.

Acknowledgements

The authors thank Erik Terk and Jaak Valge for their remarks and recommendations. The preparation of this sub-chapter has been supported by a grant from the Ministry of Education and Research (No. SF0130018s11).

References

Alkema, L., Raftery, A., Gerland, P., Clark, S., Pelletier, F., Buettner, T., Heilig, G. (2011). Probabilistic projections of the total fertility rate for all countries. Demography, 48(3): 815–839.

Andersson, G., Persson, L. (2015). Childbearing among the descendant of immigrants in Sweden. FamiliesAndSocieties Working Paper Series 39-2015: 42–66.

Barsotti, O., Bonaguidi, A. (2006). Hypotheses for migration projections. In Graziella C., Vallin, J., Wunch, G. (eds.). Demography: Analysis and Synthesis. A Treatise in Population, 3. New York: Academic Press, 189–196.

Disney, G., Wiśniowski, A., Forster, J.J, Smith, P.W.F., Bijak, J. (2015). Evaluation of existing migration forecasting methods and models. Report for the Migration Advisory Committee. Southampton: University of Southampton.

Goujon, A., Bauer, R. (2015). Demographic forces shaping the religious landscape of Vienna. In Brun, S.D. (ed.). The Changing World Religion Map. New York: Springer, 2145–2165.

Loichinger, E. (2015). Labor force projections up to 2053 for 26 EU countries by age, sex, and highest level of educational attainment. Demographic Research, 32(15): 443–486.

Myrskylä, M., Goldstein, J.R., Cheng, Y.-H.A. (2013). New cohort fertility forecasts for the developed world: Rises, falls, and reversals. Population and Development Review, 39(1): 31–56.

Puur, A., Klesment, M. (2012). Signs of stable or provisional increase in fertility? Reflections on developments in Estonia. Demográfia, 54(5): 31–55.

Raftery, A., Alkema, L., Gerland, P. (2014). Bayesian population projections for the United Nations. Statistical Science, 29(1): 58–68.

Samir, K.C., Barakat, B., Goujon, A., Skirbekk, V., Sanderson, W., Lutz, W. (2010). Projection of populations by level of educational attainment, age, and sex for 120 countries for 2005‒2050. Demographic Research 22(15): 383‒472.

Tamsar, T. (2016). Eestis on tööjõuga päris halvasti. Postimees, 26.05.2016.

UN (2015). World population prospects. Methodology of the United Nations population estimates and projections. New York: United Nations.

US Census Bureau (2016). International Data Base. http://www.census.gov/population/international/data/idb/informationGateway.php

Wattelar, C. (2006). Demographic projections: History of methods and current methodology. In Graziella C., Vallin, J., Wunch, G. (eds.). Demography: Analysis and Synthesis. A Treatise in population, 3. New York: Academic Press, 149–160.