OeNB Bulletin Q2/2025

The picture changes significantly when we focus exclusively on the migrant population (depicted by the green line in chart 2). Among migrants in the euro area, there are still approximately five people of working age for every person aged 65 or older, and this ratio has remained relatively stable since 1990. The notable exception occurred between 2015 and 2020, when a significant increase was observed. Given that migrants account for roughly 15% of the total population, they help to reduce the overall OADR in the euro area by approximately 2.5 percentage points.

When examining Austria in isolation, the rejuvenating effect of migration becomes even more pronounced (chart 3). In Austria, the average age of migrants has remained relatively low, including during the most recent period, and is comparable to the euro area average. However, migration reduces Austria’s OADR by 3.8 percentage points — approximately 50% more than the reduction seen in the euro area. This stronger impact is largely due to the foreign-born population in Austria accounting for nearly 20% of the total population, compared to only about 15% in the euro area (2020 data). Note that these shares only take into account first-generation (i.e. foreign-born) migrants. Given that “imported” fertility rates among migrants at least initially tend to be higher than those of the native population, the effect would likely have been larger had second-generation migrants also been taken into account.

Chart 3

This discussion highlights two important points. First, as shown by Marois et al. (2023), the data for both the euro area and Austria emphasize that even substantial levels of immigration are not able to reverse overall population aging trends. Second, migration can nonetheless significantly slow the population aging process and ease the policy adjustments that become necessary due to population aging – provided that labor market integration is successful.

3.2 Estimating bilateral migration contributions

The next question we want to address is whether it would make a difference to the observed effects if we were to “shut off” a certain origin country completely, say, due to some kind of fragmentation process yielding effectively closed borders. Put differently, are there origin countries that contribute more strongly to reducing the OADR than others, and to which parts of the WAP do they contribute? Unfortunately, there are no age-specific bilateral migration flow and stock data available for all European countries. ⁴ However, the United Nations provide, in addition to data on bilateral migrant stocks, also age-specific migrant stocks (that have no bilateral dimension).

This means that while we know the age structure of all immigrants to a country j , we do not know the age structure of the immigrants of a specific country i to country j . In the following section, we will introduce a technique that allows us to come up with estimates for all countries under a number of assumptions. Let us start by defining the following notations:

M _cj – age cohort c of migrants in destination j (origin unknown)

IM _ij –bilateral stocks of migrants from i in j (age unknown)

C _Ci – relative age c cohort at origin i (i.e. absolute number of people in cohort c at i /population at i )

MI _cij – initial approximation of migrants from i to j in age cohort c

ME _cij – (final) estimate of the migrants from i to j in age cohort c

In the first step, let us come up with an initial approximation of migrants from i to j .

\({MI}_{cij} = {IM}_{ij}*C_{ci}\) (1)

That is, by multiplying IM _ij by C _ci we make an initial approximation of the stock of migrants from i in that age cohort in destination j . We take the absolute value of all migrants from i and multiply it by the proportion of people that are in age cohort c in the origin. The implicit assumption here is that the demographic structure of migrants from i to j is identical to the demographic structure of the population at i . This surely is a strong assumption and will almost certainly be incorrect when considering that emigration is likely to be predominantly concentrated among the younger parts of sending countries’ populations (Gruber and Schorn, 2019). However, this becomes problematic only when comparing situations where the age distribution of migrants from certain origin countries differs more strongly from the age distribution in their home countries than in other countries. ⁵ This could arise if the motives for migration differ significantly among migrants from different countries. While demographics likely will differ less for asylum seekers, the people who come for work reasons will be more likely tilted toward younger segments of the population. However, as long as we can reasonably assume that the migrant population is dominated by one of the two types of migrants, we can consider this as an acceptable workaround.

Next, we calculate the factor x in:

\(M_{cj} = x_{cj}\sum_{i}^{n}{{IM}_{ij}*C_{ci}}\) (2)

In words: the sum of our estimates of bilateral migrant stocks over all origins i in an age cohort deviates from the actual (total) migrant population in this very age cohort by the factor x . This is a consequence of our strong assumptions. However, since we know \(x_{cj}\) we can “force” the distribution of our estimates (\(\sum_{i}^{n}{{IM}_{ij}*c_{ci}}\)) to follow the distribution of \(M_{cj}\) thereby pushing the aggregate error to 0. To do so, let us reformulate x as:

\(x_{cj} = \frac{M_{cj}}{\sum_{i}^{n}{{IM}_{ij}*c_{ci}}}\) (3)

All variables on the right-hand side are known. Then we calibrate the sum of our estimates of the bilateral migration age cohorts with \(x_{cj}\) to make them consistent with the sum of all migrants in that age cohort (\(M_{cj})\): First, we calculate the deviation factor \(x_{cj}\) using the sum of the bilateral estimates (\(\sum_{i}^{n}{{IM}_{ij}*c_{ci}})\) to quantify \(x_{cj}\), which we then use as a scaling factor on our initial estimate in equation (1). This gives us the final estimate of migrants from i in age cohort c :

\({ME}_{cij} = x_{cj}*{IM}_{ij}*c_{ci}\) (4)

with the sum of these estimates over origins corresponding to the actual age distribution of cohorts c :

\(M_{cj} = \sum_{i}^{n}{ME}_{cij}\) (5)

The idea here is simply to inflate or shrink the bilateral estimate for each origin country i in each cohort c by the same factor x . Note that x does not include an origin country index – only an index for the destination j – implying that the source countries for a given destination and cohort would be scaled equally) so that the sum of the bilateral estimates for each cohort corresponds to the observed total number of migrants in this cohort.

Again, the only assumption we need is that there are no systematic deviations in the age structure differences between migrants and their respective home populations across origin countries. This is admittedly a strong assumption but, in our view, a reasonably practical one to gain some preliminary insights into the age distribution of the migrant population by origin – information that is simply not observable in the data.

3.3 Some cautious findings on bilateral migration contributions

Let us note that first, perhaps not all that surprising, migrants to the euro area appear to be concentrated in the working-age population. While migrants constitute only approximately 14% of the total population, they account for slightly less than 17% of the older working-age population and for slightly more than 18% of the younger working-age population.

Table 1

When we assess the countries of origin within these groups, the following picture emerges: First, migration within the euro area is overwhelmingly dominated by movements originating within Europe. In fact, more than 36% of all migrants in the euro area come from other EU countries, and a total of 56% were born in a European country (either within or outside the EU). Note that if the EU were not an economic area but a country, the 36% of intra-EU migration would be considered as internal migration.

Against this backdrop, it is noteworthy that slightly more than 56% of migrants in the older WAP are likely to be of European origin, which makes this group the oldest among the migrant groups while it still remains younger than the native population on average.

Chart 4

Unsurprisingly, migration from Europe is therefore unlikely to substantially alter the demographic structure within the euro area. This is not to say that this group could not mitigate labor market bottlenecks, but in terms of age structure, it does not have a significant impact.

The second remarkable group of migrants (apart from migrants from European countries) comes from the region broadly summarized as MENA (Middle East and North Africa) ⁶ . This group likely also reflects historical migration routes from former colonies, migration from Türkiye and more recent waves of immigration from countries like Syria and Afghanistan. On a European level, migration in this group is significantly tilted toward younger ages, with a higher share of migrants in the younger segment of the working-age population than in the older segment.

Chart 5

When examining Austria, we observe that some trends evident in the euro area are more pronounced here and that there are also some distinct characteristics. In Austria, people born abroad make up a significant part of the working-age population, roughly 25% of the total WAP. Almost half of these migrants come from other EU countries, three-quarters were born in Europe. For historical reasons, a notable migrant group in Austria comes from the Western Balkans, accounting for 5.5% of the WAP. Their age distribution somewhat mirrors that of other European migrants. Their age profile is skewed toward older ages, according to our calculations, but still younger than that of the native population. Migrants from other EU countries and non-EU countries tend to be younger than those from the Western Balkans.

Meanwhile, migrants from the MENA region have achieved comparable significance within the working-age population, representing 4.4% of the WAP. This group shows a marked skew toward the younger segment of the working-age population in Austria, even more strongly than in the euro area aggregate.

As a result, at the current juncture, migrants from European countries are likely to be relatively more concentrated in the older portion of the age distribution, i.e. the group that is more strongly associated with decreasing wage pressures, but are also already strongly concentrated in the older dependent segment. Like the in the EU, migrants from other European countries tend to be concentrated in the older segment of the WAP in Austria as compared to other migrant groups. However, their share in this group is smaller than the EU average. As in the broader EU context, migration from the MENA region significantly contributes to reducing the OADR in Austria but these migrants are concentrated in younger parts of the WAP.

Following the discussion above, it would appear that intra-European migration is indeed most strongly concentrated in older segments of the WAP while migration from MENA – the other significant migrant group – is relatively more strongly represented in the younger segment of the labor market. If these groups are successfully integrated into the labor market, it is more likely that in the short run, migrants from Europe – on average – will be in age groups in which household savings tend to exceed expenditures; for migrants from MENA, however, this is likely to only happen in the medium run.

3.4 Cross-checking the Austrian approximations with administrative data

As stated above, we do not have data on the age distribution of bilateral migrant stocks for all countries at the European level. However, for Austria, data on the bilateral stock of migrants and their respective ages are available, making it possible to get an idea about the error introduced by our approximation approach. These data are taken from the so-called register census (Registerzählung from Statistics Austria), which is a complete survey of the entire Austrian population and its characteristics. These statistical characteristics are not collected from the people themselves via questionnaires but are taken exclusively from administrative registers.

Let us start by looking at our initial estimate \({MI}_{cij}\) of migrants from origin i in age cohort c . In fact, the fit of this (“unscaled”) approximation is not too bad, with a correlation coefficient of 0.94 with the actual value derived from administrative data as can be seen in the chart below.

Chart 6

However, let us look at the fit after distributing the error x equally over age cohorts in order to evaluate whether our suggested correction method improves the fit. Indeed, it does, the fit of our approximation improves remarkably. Chart 7 below clearly shows a high correlation of our estimate with the register-based Austrian data. This results in a correlation coefficient of roughly 0.99, which leads to a far better fit, as can be seen from comparing chart 7 with chart 6.

Chart 7

This implies that our approximation method described above very closely matches the data in Austria’s official registers. When comparing the pyramid above with the one in the administrative data, we can only see minor mismatches. For instance, we overestimate the share of migrants from the EU in the older WAP while at the same time underestimating the share of young WAP migrants from the EU. The opposite is true for the MENA region. The reason for this might be that the proportion of migration for work or study reasons is larger in the EU than in the MENA region, where political aspects may also play a role.

Chart 8

Overall, the very small remaining error and the comparison with Austrian register data underscore that the proposed method might be useful and reliable in estimating the bilateral age distribution of migration flows. This could also be helpful for further analyses undertaken at the bilateral level in the future.

4 Conclusions

We have shown that the impact of migration on population aging is evident, particularly when comparing old-age dependency ratios (OADRs). Migration substantially reduces the OADR, with a notably stronger effect in Austria than in the euro area as a whole, due to much higher migration to Austria. However, while migration has been slowing the population aging process in the past few decades, it has not reversed the overall trend, despite quite significant immigration rates. Migration somewhat delays the effects of population aging, but it does not seem to be able to alter the picture in the longer run.

Assessing the demographic impact of different origin countries/regions of migrants requires several strong assumptions, therefore our results should be interpreted with caution. Even so, the aggregate picture is clear: The most significant origin of migrants in both Austria and the euro area are other EU countries. However, this group of EU migrants is composed of relatively older cohorts, with an age distribution similar to that of the native population, hence offering limited relief in addressing demographic challenges. The primary drivers of OADR reduction, both in terms of age structure and migrant numbers, are immigrants from MENA countries. This group contributes much younger age profiles than migrants from other countries, and it is much larger than migrant groups from more remote origins.

Two key conclusions emerge from this analysis. First, migration has the potential to mitigate the effects of population aging, albeit temporarily, and to help economies to adjust to changing demographic realities. However, again, this reasoning is linked to a framework that necessitates the successful integration of migrants into labor markets. This applies not only to future migration flows but also to the existing migrant population in the euro area. Additionally, the demographic potential of migration is shaped by the age structure of migrants’ countries of origin, further influencing its overall impact.

5 References

Albuquerque, C. A. M., J. Caiado and A. Pereira. 2020.  Population Aging and Inflation: Evidence from Panel Cointegration. In: Applied Economics 23(1). 1.

Alfano, M., R. McKenzie and G. Roy. 2023.  Cross-occupational Effects of Immigration on Native Wages in the UK. In: Applied Economic Letters 23(1). 1.

Altonji, J. and D. Card. 1991.  The Effects of Immigration on the Labor Market Outcomes of Natives. NBER Working Papers 3123. http://www.nber.org/papers/w3123.pdf

Bratsberg, B., R. Oddbjorn, M. Roed and P. Schone. 2014.  Immigration Wage Effects by Origin. In: The Scandinavian Journal of Economics 116(2). 2.

Bobeica, E., E. Lis, C. Nickel and Y. Sun. 2017.  Demographics and Inflation. ECB Working Paper 2006.

Bodnar, K. and N. Nerlich. 2022.  The Macroeconomic and Fiscal Impact of Population Aging. ECB Occasional Paper 296.

Broniatowska, P. 2019.  Population Ageing and Inflation. In: Journal of Population Ageing 12.

Card, D. 1990.  The Impact of the Mariel Boatlift on the Miami Labor Market. In: Industrial and Labor Relations Review 43(2). 2.

Caiumi, A. and G. Peri. 2024.  Immigration’s Effects on US Wages and Employment Redux. NBER Working Paper 32389. http://www.nber.org/papers/w32389.pdf

Cho, D. and K. Lee. 2022.  Population aging and fiscal sustainability: Nonlinear evidence from Europe. In: Journal of International Money and Finance 126(C).

Cohen, E. J. 2024.  Rising Immigration Has Helped Cool an Overheated Labor Market. In: kcFED Economic Bulletin.

Dolado, J., M. Jansen and J. F. Jimeno. 2008.  On the Job Search in a Matching Model with Heterogeneous Jobs and Workers. Banco de Espana Working Paper No. 0813.

Edelberg, W. and T. Watson. 2024.  New Immigration Estimates Help Make Sense of the Pace of Employment. The Hamilton Project, Brookings Massachusetts.

Fitter, J., A. Raggl and P. Ramskogler. 2024.  The (de)globalization of migration: has the polycrisis period changed the patterns of global migration? OeNB Bulletin Q3/2024.

Engler, P., M. MacDonald, R. Piazza and G. Sher. 2023.  The Macroeconomic Effects of Large Immigration Waves. IMF Working Paper 23/259.

Gryzwaczewski, T. 2021.  Russia and Belarus Are Using Migrants as a Weapon Against the EU. In: Foreign Policy September 18.

Gruber, E. and M. Schorn. 2019.  Outmigrating youth: A threat to European peripheries? Österreichische Gesellschaft für Europapolitik - Policy Briefs, July 23.

Goodhart, C. and M. Pradhan. 2020.  The Great Demographic Reversal – Aging Societies, Waning Inequality and an Inflation Revival. Palgrave McMillan: Basingstoke.

IMF. 2022.  World Economic Outlook – Chapter 2.

Juselius, M. and E. Takáts. 2015.  Can Demography Affect Inflation and Monetary Policy. BIS Working Papers No. 485.

Juselius, M. and E. Takáts. 2018.  The Enduring Link Between Demography and Inflation. BIS Working Papers No. 722.

Juselius, M. and E. Takáts. 2021.  Inflation and Demographics through Time. In: Journal of Economic Dynamics and Control 128.

Klinger, S., A. Musayev, J.-M. Natal and E. Weber. 2019.  Immigration and Wage Dynamics in Germany. IMF Working Paper 19/301.

Leahy, J. and A. Thapar. 2019.  Demographic Effects of the Impact of Monetary Policy. NBER Working Paper 26324.

Lis, E., C. Nickel and A. Papetti. 2020.  Demographics and Inflation in the Euro Area: A Two-Sector New Keynesian Perspective. ECB Working Paper 2382.

Lopez, M., J. Passel and M. Rohal. 2015. Modern Immigration Wave Brings 59 Million to U.S., Driving Population Growth and Change Through 2065. Pew Research Centre Report. September.

Marois, G., M. Potanckova and M. Gonzalez-Leonardo. 2023.  Demographic and Labor Force Impacts of Future Immigration Flows into Europe: Does an Immigrant’s Region of Origin Matter. In: Humanities and Social Sciences Communications 10.

Mojon, B. and X. Ragot. 2019.  Can an Aging Workforce Explain Low Inflation. BIS Working Papers 776.

Nickell, S. and J. Saleheen. 2015.  The Impact of Migration on Occupational Wages: Evidence from Britain. Bank of England Working Paper No. 574.

OECD. 2017.  Pensions at a Glance. OECD: Paris.

Peri, G. 2020.  Immigrant Swan Song. In: Finance & Development March 2020.

Philipov, D. and J. Schuster. 2010.  Effect of Migration Size and Age Composition in Europe. Akademie der Wissenschaften: Wien.

Poutvaara, P. 2021.  Population Aging and Migration. IZA Discussion Paper Series 14389.

Rai, K. and B. Garg. 2024.  Demographic Transition and Inflation. In: Economic Systems.

Salisu, A., R. Muhammad and M. Saliu. 2024.  Migration and Inflation Nexus under High and Low Interest Rate Environments: Some Panel Data Evidence. In: International Migration Volume 62. Issue 6. 59–77.

United Nations. 2017.  Migration and Population Change – Drivers and Impacts. United Nations Population Division, 2017/8.

United Nations. 2020. International Migrant Stock 2020. Department of Economic and Social Affairs, Population Division.

United Nations. 2024. World Population Prospects (2024 Revision). Department of Economic and Social Affairs, Population Division.

Zaiceva, A. and K. Zimmermann. 2014.  Migration and the Demographic Shift. IZA DP No. 8743.