The EU population: past expansions through enlargement and population growth, future shrinking after the Brexit?
The EU population: past expansions through enlargement and population growth, future shrinking after the Brexit?

Despite recurring fears about an impending population implosion in Europe, the European Union has experienced sustained population growth ever since its predecessor, the European Economic Community, was established in 1957. The current territory of 28 EU Member States saw its population increasing gradually through a combination of natural population growth and immigration (green line). The EU-28 population grew by 100 million between 1960 and 2016, reaching a 500 million milestone in 2009 and 506 million in 2016.

Successive waves of EU enlargement were the major driving force of the fast increase in the actual EU population. As the number of EU member countries grew from 6 before 1972 to 28 since 2013, its population doubled from 172 million in 1960 to 344 million in 1991 and almost tripled by 2016 (blue line). Although in 2015 the EU experienced a natural population decline (i.e. an excess of deaths over births) for the first time in its history, continued migration is expected to fuel further population growth. According to our main projection scenario the total population of EU-28 is expected to reach 540 million by 2050 (green line). The contribution of migration to this trend can be compared with a hypothetical case of EU population change in the absence of migration. If no migrants were accepted into the EU after 2015 and no EU residents were to leave, the EU population would peak just below 507 million in 2018, and then decline gradually to 479 million in 2050 (pale green line). This means that compared to the “zero migration” scenario, the migrants entering the EU between 2015 and 2050 and the children born to these migrants during this period are expected to “boost” the EU population by 61 million.

Whereas the “zero migration” scenario is hypothetical and extremely implausible, the looming secession of the United Kingdom from the EU implies that its earlier rapid population growth through territorial enlargements is likely to reverse soon. If the United Kingdom leaves the EU by the year 2018, the actual EU population is expected to decline by 13 percent from 511 million to 445 million (dotted blue line).

Tempo effect and adjusted indicators of total fertility
2-1 7-2 7-2
Tempo effect and adjusted indicators of total fertility

The period level of fertility is commonly measured by the Total Fertility Rate (TFR). However, the TFR is sensitive to the changes in the age at childbearing, which has been rising in most European countries for several decades. In Italy, Luxembourg, Spain and Switzerland women now have their first child on average after age 30. As births are shifted to later ages, they are both postponed into the future and spread over a longer period of time. This “stretching” of reproduction in turn depresses the period TFR even if the number of children women ultimately have over their life course does not change.

Alternative indicators were proposed to obtain a more accurate measure of the mean number of children per woman in a calendar year. Here we compare two such indicators, the Tempo-adjusted TFR (TFR*), proposed by Bongaarts and Feeney (1998) which is based on birth order-specific total fertility rates and mean ages at birth, and Tempo and Parity-adjusted Total Fertility (TFRp*), elaborated by Bongaarts and Sobotka (2012). The TFRp* offers several improvements over the TFR*. It takes into account the parity composition of women of reproductive age, and thus controls for an additional source of distortion in the conventional TFR. Moreover, it yields considerably more stable results than the TFR*, which is clearly illustrated in the three country graphs shown here. However, the limited availability of detailed data is an obstacle to its use. Wherever possible, we present the results for the TFRp* for 2012, which were computed for 17 European countries, the United States and Japan. For the countries lacking the required data, the data sheet features the TFR* or its estimate, averaged over the 3-year period of 2011–2013 (indicated by asterisk). For the EU countries, the adjusted fertility rate was 1.72 in 2012, by about 10% higher than the conventional TFR of 1.57.

Figures 1-3 illustrate trends in the conventional TFR and its alternatives in 1980–2014 in three countries with different fertility patterns. The graphs illustrate the differences between the two tempo-adjusted indicators, TFR* and TFRp*, and they also show the long-term course of fertility postponement as measured by the rise in the mean age at first birth and, in the Czech Republic and Spain, temporary reversals of TFR trends after the onset of the economic recession in 2008.

In the Czech Republic the intensive shift to later childbearing after 1990 resulted in a dramatic fall of the period TFR to 1.14 in 1999, followed by its subsequent recovery to around 1.5. In contrast, the TFRp* declined gradually to about 1.8 in the late 1990s and further to 1.66 in 2014. The massive gap between the conventional TFR and the adjusted TFRp* in the late 1990s shows how much the TFR can be depressed when women postpone childbearing to later ages.

In Austria, postponement of childbearing started earlier but progressed more gradually. The TFR and the TFRp* have shown relatively stable values since the mid-1980s, hovering around 1.4 and 1.6–1.7, respectively.

Spain shows yet another pattern: conventional and adjusted total fertility both fell in tandem during the 1980s and 1990s. The decline in the period TFR bottomed out at 1.15 in 1998 and modestly recovered until 2008, whereas the TFRp* continued falling until 2007 and briefly converged with the TFR level before rising sharply in the subsequent two years. Recently, fertility was affected by the economic recession, bringing an acceleration of the shift towards later first births and a renewed decline in the period TFR interrupted only in 2014. In contrast, the TFRp* showed a short-term upswing after 2008, which was even more pronounced in the TFR*. This increase was probably caused by a rapid change in the variance of fertility schedule, which can temporarily distort the adjusted measures of fertility, in particular TFR* (Zeng and Land 2001).


Bongaarts, J. and G. Feeney 1998. On the quantum and tempo of fertility. Population and Development Review 24(2): 271-291.

Bongaarts, J. and T. Sobotka 2012. A demographic explanation for the recent rise in European fertility. Population and Development Review 38(1): 83-120.

Zeng, Yi and K. C. Land. 2001. A sensitivity analysis of the Bongaarts-Feeney method for adjusting bias in observed period total fertility rates. Demography 38(1): 17-28.

An intergenerationally equitable normal pension age
3-1 3-2
An intergenerationally equitable normal pension age

As life expectancy rises, how should normal pension ages change in a way that is fair for each generation? Fairness is a core democratic value, but intergenerational fairness is rarely mentioned in policy discussions of pension ages. Up to now there has been no clear, analytical definition of what an intergenerationally equitable normal pension age (IENPA) would be.

A recently developed approach to measuring population ageing (Sanderson and Scherbov 2013) provides the methodological background necessary to conceptualise and calculate IENPAs. The basic idea is that many characteristics relevant to the study of population ageing, such as health, life expectancy, and disability, differ over space and time. The new approach takes these differences into account while measures based solely on chronological age do not.

The IENPA is based on three criteria: (1) members of each cohort receive as much in pension payouts as they pay into the pension plan; (2) the generosity of the pension system, measured as the ratio of average pension receipt to the incomes of people who pay into the pension system is the same for all cohorts; and (3) the pension tax is the same for all cohorts (Sanderson and Scherbov 2015a, 2015b).

The figures show the IENPA for Germany and the UK and plans for the normal pension age of men that have already been legislated. Over the entire period from 2015 to the late 2020s, the IENPA rises close to the same amount as the legislated changes. Using the IENPA has two important advantages. First, although the IENPA and the planned changes reach nearly the same age by the late 2020s, the IENPA rises more continuously, eliminating the irregularities seen in the planned changes. These irregularities are inequitable and can breed discontent. Second, the IENPA provides normal pension ages for the period after the late 2020s that are consistent with the previous legislated changes. The use of IENPA eliminates the need for periodic renegotiations of pension ages.

National pension plans are complex and difficult to compare across countries. Up to now, most of the comparisons of national pension plans have been in terms of static features, such as the replacement rate and the number of years of contribution required for a full pension. There has been virtually no attention paid to the dynamic equity of the plans because there was no way to assess this. With IENPAs, we have an analytic way of determining how fast an intergenerationally fair normal pension age would rise and we can compare this to potential policies that countries might wish to enact.

Life expectancy at older ages can rise, but it can also fall. The IENPA reflects this and can rise or fall along with life expectancy. The use of the IENPA would result in pension systems that are more sustainable and more resilient to demographic shocks.

German legislated pension ages for men
UK legislated pension ages for men

Sanderson, W.C. and S. Scherbov 2013. The characteristics approach to the measurement of population aging. Population and Development Review 39(4): 673-685.
Sanderson, W.C. and S. Scherbov 2015a. Are we overly dependent on conventional dependency ratios? Population and Development Review 41(4): 687-708.
Sanderson, W.C. and S. Scherbov 2015b. An easily understood and intergenerationally equitable normal pension age. In B. Marin (Ed.), The Future of Welfare in a Global Europe (pp. 193–220). Farnham, UK: Ashgate Publishing, Ltd.

New measures of population aging can be found at:

Change in number of live births, 2015-2050
Change in number of live births, 2015-2050
Relative population change, 2015-2050
Relative population change, 2015-2050

Migration plays a key role for the future population change in many European countries as it can reverse, soften or amplify the population trends driven by the expected changes in fertility and mortality.

In this graphic we show how the projected change in population size between 2015 and 2050 is driven by two distinct forces in the selected countries: natural population change (i.e., the difference between the number of births and deaths) and migration, including future births and deaths attributable to migrants arriving since 2015.

The overall picture depicts a divide between north-western and south-eastern Europe. Population is projected to increase in all countries of western and northern Europe, as well as in most of southern Europe. In many fast-growing countries, including Sweden, the UK and France, this increase is driven bya combination of natural population increase and expected immigration. In other countries, including Germany and Italy, immigration is expected to compensate for population losses that would occur as a result of low fertility and shrinking number of births. In contrast, migration will amplify population losses in many countries of central and south-central Europe, with the extreme case of Moldova that may lose as much as 40% of its population. Throughout the region, as well as in Greece, expected migration trends in combination with low fertility will lead to population losses. The Czech Republic stands out as a rare exception, where continuing positive migration balance is projected to sustain population increase in the coming decades.

Total fertility rate in European regions and the USA
Total fertility rate in European regions and the USA

Period fertility trends in different parts of Europe often follow distinct trajectories. This picture of changes in total fertility rate (TFR) from 1980 to 2014 shows that after fertility fell to very low levels in Southern Europe in the 1980s and Central, Eastern, and South-Eastern Europe in the 1990s, two contrasting groups of regions emerged. Western and Northern Europe had relatively higher total fertility rate at or above 1.75, whereas all the other parts of Europe, including the German-speaking countries, had a very low total fertility below 1.50. This division remained stable from 2000 to 2008, when the TFR was gradually increasing in most European countries. During the recent economic recession, fertility rates fell in most countries in Europe between 2008 and 2013, especially in Southern Europe and in the Nordic countries. Outside Europe, even a stronger fall in the TFR was registered in the United States, which now has a similar fertility level as Western European countries. Eastern European countries, especially Russia, bucked this trend, and experienced strong increases in their TFRs after 2008. More recently, South-eastern Europe and German-speaking countries experienced rising fertility as well. These ups and downs in period TFR might be partly driven by changes in the age at motherhood rather than by genuine changes in family size ( see TEMPO box ).

As a result of these contrasting regional trends, the “bifurcation” in the TFR across Europe has diminished recently. Yet, the gap between the “higher-fertility” Western Europe (TFR of 1.86 in 2014) and the “lowest-fertility” Southern Europe (TFR of 1.33 in 2014) remains wide, surpassing half a child per woman.

For more fertility indicators see European fertility datasheet 2015

Life expectancy at birth in selected countries
6 7-2 7-3
Life expectancy at birth in selected countries

In most parts of Europe life expectancy at birth has increased steadily during the last four decades. Among women, it has surpassed the threshold of 85 years in several countries of western and southern Europe including France, Italy, Spain and Switzerland. In Spain, it has surpassed 86 years in 2014, up from 78 in 1980 with average annual increase of 2.7 months. An equally impressive rise has been recorded among men, whose life expectancy still lags behind that of women everywhere in Europe. In some countries of southern and northern Europe as well as in the Netherlands and Switzerland, male life expectancy at birth has surpassed 80 years, up from 72–73 years in 1980. Men in France and Germany witnessed the highest increase on average of more than 3.2 months per year.

While most European regions experience a seemingly unstoppable increase in longevity, parts of central and eastern Europe have taken a different trajectory. Following a period of generally stagnating or even slightly declining life expectancy at birth in the 1980s, life expectancy in some countries of central and south-central Europe saw improving life expectancy since the early stage of their post-socialist transition (see Poland and Slovenia depicted in the chart). However, in some eastern European countries, including Russia, it had plummeted in the early 1990s during the period of turbulent transformation after the collapse of state socialism. The downward trend was particularly pronounced among men and has not only led to increasing gender disparities in life expectancy but also to a much more pronounced gap in comparison to other European regions. These disparities have only partly diminished after 2000. Despite substantial improvement during the last decade, Russian men in 2014 had a life expectancy as low as 65 years, 13 years below the EU average and 16 years lower than men in Switzerland—a top-ranking country within Europe.

Wide differences persist across Europe in the gap between male and female life expectancy. In most countries the male mortality disadvantage has gradually narrowed over time. For instance, the male–female gap in life expectancy in Sweden declined from over six years in the early 1980s to less than four years in the early 2010s. In France, this gap declined from 8 to 6 years between 1980 and 2014. The Baltic states and the eastern Europe countries still display huge life expectancy gaps: if the recent mortality rates were to remain stable, Russian men would die 11 years earlier on average than Russian women.

Regional overview
8-1 8-2-new 8-2-new
Regional overview

Definition of regions in the regional overview takes into account geopolitical criteria as well as similarity in demographic trends in countries they cover. Countries are split into regions as follows: Southern Europe (Cyprus, Greece, Italy, Malta, Portugal, Spain); Western Europe (Belgium, France, Ireland, Luxembourg, Netherlands, United Kingdom); Nordic countries (Denmark, Finland, Iceland, Norway, Sweden); Central-Eastern Europe (Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, Slovenia); South-Eastern Europe (Albania, Bulgaria, Macedonia, Montenegro, Romania, Serbia); Eastern Europe (Belarus, Moldova, Russia, Ukraine); Caucasus (Armenia, Azerbaijan, Georgia).


Sergei Scherbov (responsible for content) – scientific content, population projections, ageing indicators
Marija Mamolo - scientific content, data collection and computation, population projections
Michaela Potančoková – coordination, scientific content, data collection and computation, population projection scenarios, cartography
Tomáš Sobotka – scientific content, fertility indicators, population projection scenarios
Kryštof Zeman – scientific content, data collection and computation, fertility indicators
Assistance: Stefanie Andruchowitz, Jakob Eder, Warren Sanderson


Vienna Institute of Demography (VID) and International Institute for Applied Systems Analysis (IIASA). 2016. European Demographic Datasheet 2016. Wittgenstein Centre (IIASA, VID/OEAW, WU), Vienna.


Preparation and publication of the European Demographic Datasheet was partly supported by the European Research Council under the European Union’s Seventh Framework Programme, projects ERC-2011-StG 284238 (EURREP) and ERC-2012-AdG 323947 (Re-Ageing).


Andrzej Dziekoński, BOOST IT, Michał Kurcewicz, BITAPP