## dhicaeeea ## Source details **Canonical URL:** [dhicaeeea](https://www.imf.org/-/media/files/publications/dp/2019/english/dhicaeeea.pdf) ## Other formats - [Markdown version](/-/media/files/publications/dp/2019/english/dhicaeeea.pdf.md) - [Structured JSON version](/-/media/files/publications/dp/2019/english/dhicaeeea.pdf.json) --- ### Executive summary — Main findings - Populations of Central, Eastern, and Southeastern European (CESEE) countries—with the exception of Turkey—are expected to decrease significantly over the next 30 years, driven by low or negative net birth rates and outward migration. - The labor force in CESEE countries is projected to shrink over the next three decades. - Aging populations will increase demands on health care and pension resources. - Some public pension systems currently appear sustainable, but spending would increase substantially if pensions were more in line with pre-retirement earnings. - Workforce aging itself could cause aggregate productivity to deteriorate; for some CESEE countries, this effect could be significant. - Combining labor, fiscal, and productivity channels shows that, without mitigating policies, growth and convergence to Western European living standards would slow considerably. - Fiscal pressures can feed back to growth via tax increases, spending compression, or accommodation through increased public debt and higher borrowing costs, potentially displacing public or private investment. - Trade linkages to other CESEE and Western European countries experiencing demographic pressures worsen the effects through spillovers. - There is considerable heterogeneity across countries; no single “one-size-fits-all” policy applies across the region. ### Demographic projections and drivers - Total populations of CESEE (except Turkey) are expected to decline significantly over the next 30 years; Turkey is the only CESEE country expected to have positive population growth. - The largest rates of decline are projected for Baltic and SEE EU states; Eastern Europe (EE), owing to population size, contributes most to the CESEE total population decline. - Compared with Western Europe, CESEE starts from a younger population overall; birth rates between now and 2050 are expected to be roughly the same, but higher mortality rates and net outward migration drive more rapid declines. - Fertility rate (live births per woman) in CESEE-excluding-Turkey varies between 1.2 and 1.8—below replacement rate of 2.1. - United Nations life-expectancy projections for males at birth: - CESEE average: 72.3 years in 2015–2020, increasing to 77.9 years by 2050. - Western Europe average: 79.9 years in 2015–2020, increasing to 84.7 years by 2050. - About half of CESEE countries are projected to experience both negative natural population growth rates and outward migration over the next 30 years. - Nearly half of CESEE countries are projected to experience net total population losses of 5 percent between now and 2030, and 15 percent by 2050. - Latvia and Bulgaria are projected to lose more than a fifth of their populations by 2050. ### Labor supply implications and projections - Baseline labor force projections (accounting for legislated retirement-age increases): - CESEE region (excluding Turkey) labor force shrinking by 10 percent by 2030 and by 26 percent by 2050. - Bulgaria, Latvia, Poland, and Ukraine projected to experience labor force declines of more than 30 percent by 2050. - Even in more benign cases (e.g., Slovakia and Serbia), the labor force would decline by 20 percent by 2050. - Participation levels and gaps: - Average labor force participation: 68 percent in CESEE versus 74 percent in Western Europe. - Average participation rate for women aged 55 to 64 years: 42 percent in CESEE, 49 percent in Western Europe, 76 percent in Sweden. - Average participation rate for men aged 55 to 64 years: 58 percent in CESEE, 65 percent in Western Europe, 82 percent in Sweden. - Policy levers and limits: - Fertility policies have limited short- to medium-term potential; even immediate shifts to the UN high fertility scenario would not increase the working-age population for at least 15 years. - Inward migration has been used episodically; most CESEE countries lack long-term strategies for inward migration. - Significant scope exists to raise participation, especially of women and older workers. ### Labor-force reform scenarios and headline effects - Two illustrative scenarios (see Annex 2): - Moderate reform scenario: - Moderately paced increases in female labor force participation to the highest Western European rates. - Retirement age increases in line with life expectancy, but not higher than 67. - Ambitious reform scenario: - Rapid increases in female labor force participation to the highest Western European rates. - Raising the retirement age above 70 by 2050. - Aggregate projected impacts: - Moderate reform scenario: - Average labor force decline limited to 3 percent by 2030, but 16 percent by 2050. - Reforms in Albania, Belarus, Bosnia and Herzegovina, Republic of North Macedonia, Moldova, and Serbia would bring net decline in 2050 to below 10 percent. - Bulgaria, Estonia, Latvia, Lithuania, and Poland would face labor force declines of more than 20 percent despite participation increases. - Ambitious reform scenario: - Labor force declines in many countries would stall or be temporarily reversed; in several countries (Belarus, Moldova, and Western Balkans) ambitious reforms could potentially halt long-term decline. - Some countries (including Bulgaria and Latvia) would still see declines of more than 23 percent even with ambitious reforms. - Timing of effects: - Boosting female participation: sizable short-term effects. - Increasing participation of older workers: medium- to long-term effects. - Raising retirement age: largest long-run impact; could reduce median decline in the labor force by nearly 10 percentage points by 2050. ### Age-related public spending: current levels and projections - Current spending levels (region vs euro area peers): - CESEE: 9 percent of GDP on public pensions and 5 percent of GDP on public health care each year. - Euro area peers: 12 percent of GDP on public pensions and 7 percent of GDP on public health care. - Projected increases on current policies (2015–2050): - Costs related to pensions and health care in CESEE expected to increase by nearly 4 percentage points of GDP, on average, between 2015 and 2050. - Almost half of the increase explained by rapidly rising health care spending: - Fiscal cost of public health care systems projected to increase by 2¾ percentage points of GDP. - Projected fiscal cost of public pension systems on current policies appears to rise by just over 1 percentage point of GDP. - Replacement rates and sustainability: - Many CESEE pension systems are defined-contribution and appear fiscally sustainable but have relatively low replacement rates. - On average, a pensioner in 2050 will receive about one-third of their preretirement earnings. - Except for Ukraine, the replacement rate is projected to drop in CESEE countries; notable projected drops include: - Republic of North Macedonia: drop by 28 percentage points. - Moldova: drop by 30 percentage points. - Poland: drop by 25 percentage points. - Bringing replacement rates to 40 percent (ILO recommendation) would: - Increase estimated fiscal costs of pensions by 4 percentage points of GDP, on average. - Increase total pension and health care costs in CESEE by nearly 7 percentage points of GDP, on average, between 2015 and 2050. - Age-related spending is projected to reach 20 percent of GDP by 2050 on average, with some countries (Bosnia and Herzegovina, Croatia, Poland, Slovenia, and Ukraine) spending as much as 25 percent of GDP or more. ### Fiscal implications and reform costs - Labor-market reform fiscal cost estimates to boost female participation through 2050: - Moderate reform package: projected fiscal cost about 0.7 percent of GDP. - Ambitious reform package: projected fiscal cost close to 1 percent of GDP. - Country variation: - Fiscal cost of ambitious reforms close to 1½ percent of GDP in Croatia, the Czech Republic, Poland, and Ukraine. - Fiscal cost of ambitious reforms close to ½ percent of GDP in the Slovak Republic and Romania. - Fiscal-space constraints: - The upfront cost is challenging for countries with limited fiscal space. - The estimated gap between the debt-stabilizing primary balance and the observed primary balance is positive in Belarus, Moldova, Romania, Russia, and Turkey, suggesting fiscal consolidation may be required to stabilize public debt. - Net effect of reforms: - Ambitious labor market reforms would reduce projected pension costs by nearly 5 percentage points of GDP on average; moderate reforms reduce costs by about 2 percentage points of GDP on average. - Labor market reforms can be fiscally costly (training, childcare, subsidies) and may increase health care demand with higher shares of older workers. ### Aging and productivity: empirical estimates and projections - Empirical findings (IMF estimates): - A one-percentage-point increase in the share of older workers (ages 55+) in the total workforce is associated with a decrease in growth of output per worker of 0.7–0.8 percentage points. - Most of this effect stems from lower TFP growth: decreases by about 0.6 percentage points for every percentage point increase in the older-worker share. - Workforce aging is associated with a statistically significant increase in the capital-output ratio (some capital intensification). - Workforce shares and projections: - In 2015, share of older workers (55+) in the CESEE workforce was 15¾ percent on average. - Projected to increase to about 22 percent by 2050 in CESEE. - TFP impact calculation (using coefficient 0.6): - Workforce aging could lead to a 0.09 percentage point decrease in annual TFP growth for CESEE countries on average over 2015–20. - About a 0.08 percentage point reduction over 2040–50 for CESEE. - On average over 2020–50, the average annual decrease of TFP growth is 0.38 in CESEE and 0.34 in Western Europe (values preserved as presented). ### Automation, structural change, and industry composition - Aging raises incentives to automate as prime-age labor becomes scarce and relatively expensive; automation can offset composition effects if machines become sufficiently productive and cheap. - Net impact of automation on employment and wages is ambiguous based on cross-country evidence. - CESEE has a higher share of employed persons—especially male—in occupations associated with larger productivity declines of older workers than the EU-15 average; industrial structure may therefore reinforce aggregate productivity declines from aging. ### General-equilibrium macro simulations (EEUMOD) — aggregate effects - Production-function and EEUMOD prologue results for CESEE (excluding Turkey), PPP-weighted averages: - Real GDP growth rates would be lower by about 1.4 percentage points each year over the whole horizon (2020–50) compared with a "no-demographics" scenario (prologue estimate). - EEUMOD simulation: demographic shock alone reduces level of output by about 27 percent by 2050: - 22 percentage points due to labor supply effect. - 5 percentage points due to the TFP effect. - Including fiscal effect (age-related fiscal costs, deficit-financed) adds 4 percentage points, bringing total decline to 31 percent by 2050. - Representative quantified impacts (CESEE ex. Turkey, 2020–50): - Average yearly impact on GDP growth from 2020–50: decline of 1.2 percentage points, cumulating to a total decline in the output level of 31 percent by 2050. - In per capita terms, average yearly growth from 2020 to 2050 would be lower by about 0.6 percentage points, cumulating to a total decline in the output level of 17 percent. - Contributions to the 1.2 percentage point annual decline: - Labor direct contribution: about 0.4 percentage points. - Productivity direct contribution: about 0.3 percentage points. - Investment response: accounts for 0.5 percentage points. - Private investment projected lower by 26 percent by 2050 compared to the no-demographics scenario. - Fiscal and debt effects under accommodating-deficits baseline: - Average fiscal deficit in CESEE (excluding Turkey) would increase by about 9 percentage points of GDP by 2050. - Public debt would increase by about 75 percentage points of GDP by 2050 compared to the no-demographics scenario. - Current account balances would, on average, be higher by 1.5 percentage points of GDP compared to the no-demographics scenario. ### Spillovers and external effects - External spillovers make GDP 3 percentage points worse by 2050 in a typical-country simulation; about one-tenth of the overall effect on the level of GDP. - Spillover contributions larger for investment (closer to 20 percent) and consumption (nearly a quarter). - Simulations assume goods and capital completely mobile; labor not mobile (no endogenous migration response modeled). - EEUMOD abstracts from remittances; demographic shocks applied only to CESEE and Western European countries/blocs. ### Labor market reform simulation effects (2020–50) - For CESEE (excluding Turkey) overall: - Moderate labor market reforms: improve GDP growth by about 0.2 percentage points (average yearly). - Ambitious labor market reforms: improve GDP growth by about 0.4 percentage points (average yearly). - Convergence impacts: - Moderate package: could lower the per capita income gap between CESEE (excluding Turkey) and Western Europe by about 3 percentage points of GDP. - Ambitious package: could narrow the gap by about 4 percentage points. - Fiscal impacts of reform scenarios: - Moderate reforms: offset about 50 percent of projected increase in aging-related government spending and debt; average government deficit increases by about 4 percentage points of GDP by 2050 (compared with 9 percentage points in baseline); public debt rises by 33 percentage points of GDP compared to no-demographics scenario. - Ambitious reforms: reduce average projected government deficit by about 3 percentage points of GDP (compared with an increase of 9 percentage points in baseline); result in public debt lower by about 28 percentage points by 2050 compared to the no-demographics baseline. - Note: fiscal costs of labor market reforms are not modeled in these simulations. ### Alternative fiscal reactions: higher taxes versus accommodating deficits - Two fiscal reaction assumptions compared: - Accommodating deficits (baseline): age-related fiscal costs accommodated by increased deficits. - Higher taxes scenario: distortionary labor and consumption taxes raised to cover half the deficit each year. - Aggregate CESEE results (average yearly impact over 2020–50): - Real GDP: - Accommodating deficits: annual GDP growth lower by about 1.2 percentage points relative to no-demographics. - Higher taxes: annual GDP growth lower by about 1.0 percentage point relative to no-demographics. - Real GDP per capita: - Higher taxes: per capita GDP growth lower by about 0.5 percentage points. - By 2050, level of per capita GDP is 14.7 percentage points lower under the higher taxes reaction relative to the no-demographics scenario—slightly smaller than the 17 percent fall under accommodating deficits. - Debt/GDP in 2050: - Higher taxes: debt-to-GDP ratio is 12 percentage points higher than the no-demographics scenario by 2050. - Accommodating deficits: public debt increases by about 76 percentage points by 2050 relative to no-demographics. - Conclusion: under the higher taxes reaction, the positive impact of lower debt can outweigh the distortionary cost of taxes in aggregate CESEE results. ### Policy implications and priorities - Overarching objective: more workers working more productively through increasing effective labor inputs, capital deepening, and raising productivity. - No single remedy; priority areas: - Raise labor force participation (especially women and older workers). - Raise retirement ages in line with life expectancy. - Retain and better use existing workforce via education and adult (re)training and lifelong learning. - Invest in automation and labor-saving capital where feasible. - Consider liberalizing immigration regimes for skilled workers and improve institutions/policies to retain local workers and attract foreign workers. - Preserve fiscal space to support infrastructure, research and development, and active labor market policies. - Contain health care spending growth through macro-level controls, micro-level reforms, and demand-side measures while protecting access and equity. - Strengthen governance and institutional quality to slow outward migration and improve productivity outcomes. - Tailoring by country: - Retirement-age reforms particularly beneficial in Belarus, Moldova, Russia, Turkey, and Ukraine. - Latvia and Lithuania: focus on migration policy and productivity due to already high participation rates and sharply decreasing labor supply. - Bosnia and Herzegovina, Moldova, Romania, and Ukraine: substantial scope to raise participation rates and improve output and fiscal sustainability. *Italic: Source — Excerpts and data from the PDF chapter "DEMOGRAPHIC HEADWINDS IN CENTRAL AND EASTERN EUROPE" (dhicaeeea) as provided in the supplied content.* ### Executive Summary ������������������������������������������������������������������������������������������������������ ### Executive Summary ### Main findings - The populations of Central, Eastern, and Southeastern European (CESEE) countries—with the exception of Turkey—are expected to decrease significantly over the next 30 years, driven by low or negative net birth rates and outward migration. - The labor force in CESEE countries is projected to shrink over the next three decades. - Aging populations will increase demands on health care and pension resources. - Some public pension systems currently appear sustainable, but spending would increase substantially if pensions were more in line with pre-retirement earnings. - Aging of the labor force itself could cause aggregate productivity to deteriorate; for some CESEE countries, this effect could be significant. - Assessing labor, fiscal, and productivity channels together shows that, without mitigating policies, growth and convergence to Western European living standards would slow considerably. - Fiscal pressures can feed back to growth via increases in taxation and spending compression, or via accommodation through increased public debt and higher borrowing costs, potentially displacing public or private investment. - Trade linkages to other CESEE and Western European countries experiencing demographic pressures worsen the effects through spillovers. - There is considerable heterogeneity across countries: Turkey is young with a sizable young population, whereas other CESEE countries are older; country rankings vary by expected change in total and working-age populations, share of older workers, speed of aging, and increase in old-age dependency ratios. - No single “one-size-fits-all” policy applies across the region. ### Demographic projections and drivers - Total populations of CESEE (except Turkey) are expected to decline significantly over the next 30 years; Turkey is the only CESEE country expected to have positive population growth. - The largest rates of decline are projected for Baltic and SEE EU states; Eastern Europe (EE), owing to population size, contributes most to the CESEE total population decline. - Compared with Western Europe, CESEE starts from a younger population overall, and birth rates between now and 2050 are expected to be roughly the same, but higher mortality rates and net outward migration drive more rapid declines in population than in Western Europe. - Emigration has exacerbated the problem of contracting labor supply and has accelerated aging, as emigrants have tended to be younger and more educated than those left behind. ### Implications for labor supply - Most countries will face significant declines in their labor force by 2050. - Some countries might temporarily stall or reverse labor force declines by increasing labor force participation rates and retirement ages, but none can ultimately offset the underlying population dynamics fully. - Given limited scope for boosting labor supply in many CESEE countries, policy emphasis will be on retaining and better using the existing workforce (education, adult (re)training, automation), and on considering further liberalization of immigration regimes, especially for skilled workers. ### Implications for age-related public spending - Rising health care and pension expenditures as populations age will challenge public finances. - Under current policies for contributions, payments, and retirement ages, public pensions are fiscally sustainable in most CESEE countries, but costs would increase dramatically if pension payments were to be set closer to pre-retirement earnings. - Policymakers will face tough choices to contain age-related spending, such as the extent of health care coverage and the level of pensions, while preserving fiscal space for infrastructure, research and development, and active labor market policies. ### Implications for productivity - Workforce aging is associated with declines in aggregate labor productivity and total factor productivity growth. - For some countries, given the age composition of their populations, the productivity effect could have very strong adverse effects on growth unless offset by increased labor supply and/or investment in labor-saving machines (automation). - Support for education and adult (re)training and automation may help increase productivity. ### Macroeconomic implications, spillovers, and general equilibrium considerations - Combining the channels of shrinking labor supply, rising age-related public spending, and declining productivity implies substantial headwinds to growth and convergence. - Spillovers across countries—via trade and financial linkages—mean that demographic pressures in one country affect others in the CESEE region and Western Europe. - Feedbacks from fiscal pressures to growth (through taxation, spending compression, or higher debt-servicing costs) are economically important and can amplify adverse effects. - The study emphasizes general equilibrium effects, including spillovers from trade, feedback from fiscal constraints and public debt accumulation, and second-round effects of policy initiatives on growth. ### Policy implications and priorities - There is no single or easy remedy to address demographic pressures. - Policies to increase fertility rates are unlikely to offset the decline in the labor force. - Measures with mitigating potential include: - Increasing labor force participation rates (especially of women and older workers). - Raising retirement ages. - Retaining and better using the existing workforce via education and adult (re)training. - Investing in automation and labor-saving capital (to the extent feasible). - Liberalizing immigration regimes, especially for skilled workers, and improving institutions and policies to retain local workers and attract foreign workers. - Policymakers must balance containing age-related spending with preserving fiscal space for infrastructure, research and development, and active labor market policies. - The appropriate policy package must be tailored for each country; policy choices will be difficult and consequential. ### Scope, contribution, and structure of the study - The study examines demographic pressures across CESEE countries (covering central European states, Baltic states, the Commonwealth of Independent States, western and eastern Balkan states, and Turkey), drawing out macroeconomic implications for public spending, potential output, income convergence, and evaluating policy options. - The study emphasizes general equilibrium effects and spillovers and aims to provide a framework to analyze policy options and inform decisions. - Structure: Chapter 2 presents demographic projections; Chapters 3–5 analyze channels (labor supply, public spending, productivity); Chapter 6 evaluates effects on growth and convergence with attention to spillovers and saving/investment flows; Chapter 7 summarizes and presents policy priorities. ### Regional definitions used in the study - Central, Eastern, and Southeastern Europe (CESEE): Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Republic of North Macedonia, Moldova, Montenegro, Poland, Romania, Russia, Serbia, Slovakia, Slovenia, Turkey, Ukraine. - Baltic states: Estonia, Latvia, Lithuania. - Central European New Member States (CE): Czech Republic, Hungary, Poland, Slovakia, Slovenia. - Eastern Europe (EE): Belarus, Moldova, Russian Federation, Ukraine. - Southeastern European EU Member States (SEE EU): Bulgaria, Croatia, Romania. - Southeastern European Non-EU Member States (SEE non-EU or Western Balkans): Albania, Bosnia and Herzegovina, Republic of North Macedonia, Montenegro, Serbia. - Western Europe (WE): Austria, Belgium, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, Malta, Netherlands, Norway, Portugal, San Marino, Spain, Sweden, Switzerland, United Kingdom. - Note: No data are available for Kosovo. *Executive Summary — DEMOGRAPHIC HEADWINDS IN CENTRAL AND EASTERN EUROPE (from the supplied content)* ### 2. Contributions to Total Population Growth ### 2. Contributions to Total Population Growth ### Birth Rates - Projected CESEE birth rates—the numbers born per head of population—are on average lower than in most other regions. - CESEE birth rates are mostly in line with those in Western Europe and projected to plateau only a little bit below current rates; hence, the projected decline in population relative to Western Europe is mostly not driven by a worsening in birth rates. - Only two CESEE countries—Albania and Turkey—are projected to have significantly positive “natural” (that is, births minus deaths) population growth rates. - Note: The fertility rate (live births per woman) in CESEE-excluding-Turkey varies between 1.2 and 1.8—well below the replacement rate of 2.1. ### Mortality Rates and Life Expectancy - Age-standardized death rates are higher in CESEE countries, particularly for EE countries, than is typical in Western European countries. - Causes of relatively high mortality include cardiovascular diseases, unintentional and intentional injuries, maternal and neonatal conditions, and digestive diseases; excess mortality due to noncommunicable diseases occurs among adults aged 30 years and older. - The high mortality rates imply low expected lifespans—the 10 countries with the lowest expected lifespan of the United Nation’s “more developed regions” are all from the CESEE region. - Probabilities of death at all ages—particularly for older cohorts—are expected to fall for CESEE as a whole, reflecting advances in health care and lower risks from illness and accident. - United Nations life-expectancy projections for males at birth: - CESEE average: 72.3 years in 2015–2020, increasing to 77.9 years by 2050. - Western Europe average: 79.9 years in 2015–2020, increasing to 84.7 years by 2050. - Improvements in life expectancy are uneven and slow; for example, Russia and Moldova are not projected to reach the current CESEE average even by 2050. - Despite assumed improvements, mortality rates are projected to remain significantly higher than those of Western Europe; mortality rates increase over time in all countries because of population aging. ### Migration - Most CESEE countries have experienced significant outward migration since the 1990s; emigrants have tended to be younger, exacerbating demographic pressures and accelerating aging. - The United Nations notes international migration is the component of population change that is most difficult to project. - Migration projections start from the assumption that recent levels, if stable, would continue until 2050, with adjustments for migration policies in each country. - About half of CESEE countries are projected to experience both negative natural population growth rates and outward migration over the next 30 years. - Figure (2017–2030): Contributions to total population growth by births and deaths and by migration show notable negative contributions from migration for many CESEE countries. ### Population Projections and Broad Implications - These demographic factors imply dramatic declines in populations in most CESEE countries over the next three decades. - Nearly half of CESEE countries are projected to experience net total population losses of 5 percent between now and 2030, and 15 percent by 2050. - Latvia and Bulgaria are projected to lose more than a fifth of their populations by 2050. - Only Turkey’s population is expected to increase significantly. - Consequences include shrinking and aging workforces and greater numbers of the aged relative to working age, with likely macroeconomic effects on potential growth, public service costs, and labor productivity. ### Implications for Labor Supply - A falling total population is likely to result in a shrinking working-age population; decline is generally more severe in CESEE than in Western Europe and particularly pronounced for the Baltic states, CE, and SEE EU. - The age structure ("lumpy" cohorts) implies a more rapid short-term decline in working-age population in CESEE than in Western Europe, decelerating by 2035 then accelerating thereafter. - The share of older workers is projected to increase in all CESEE countries. - The ratio of those over 65 years to the working-age population (old-age dependency) is currently lower in CESEE than Western Europe on average, but: - Some CESEE countries (Bulgaria, Croatia, Baltic states) already have higher dependency ratios. - All CESEE countries, even Turkey, will rapidly reach old-age dependency ratios higher than considered problematic in comparatively rich Western European countries. - In some cases (Czech Republic, Poland, Slovenia), the old-age dependency ratio is projected to be higher than the Western European average. - Labor force projections under the baseline scenario (accounting for legislated retirement-age increases) show: - CESEE region (excluding Turkey) labor force shrinking by 10 percent by 2030 and by 26 percent by 2050. - Bulgaria, Latvia, Poland, and Ukraine projected to experience labor force declines of more than 30 percent by 2050. - Even in more benign cases (e.g., Slovakia and Serbia), the labor force would decline by 20 percent by 2050. ### Policy Options to Address Labor-Force Decline - Fertility policies: - Many CESEE countries have used childbirth grants, cash child benefits, tax deductions, housing allowance, and social assistance; family and child benefits are most generous (in terms of share of GDP) in Central European countries. - There is little evidence that direct financial incentives effectively boost fertility. - Even if fertility rates could be raised successfully, a tangible impact on labor supply would only materialize 20 years hence. - Inward migration: - Some countries have used inward-migration policies for seasonal labor shortages and simplified procedures for short-term foreign workers (for example, Czech Republic, Poland, Slovakia for workers from Ukraine). - Most CESEE countries lack long-term strategies for inward migration. - Return-migration measures (for example, Poland) have so far produced limited return migration. - Labor force participation: - Significant scope exists to increase participation in CESEE countries. - Average labor force participation: 68 percent in CESEE versus 74 percent in Western Europe. - Substantial room for improvement exists compared with the highest-performing European countries (Sweden: 82 percent). - Particular scope to raise participation among older workers and women: - Average participation rate for women aged 55 to 64 years: 42 percent in CESEE, 49 percent in Western Europe, 76 percent in Sweden. - Average participation rate for men aged 55 to 64 years: 58 percent in CESEE, 65 percent in Western Europe, 82 percent in Sweden. - Participation of women aged 25 to 45 years is below Western European averages in many CESEE countries, with largest gaps in Turkey and the Balkan countries. - Policy experience: - Some CESEE countries have sought to stem outward migration and overhauled pension systems; a number have legislated increases in statutory retirement ages. - Under current labor market policies, projected declines in the labor force remain drastic despite legislated retirement-age increases (which are expected to raise average participation for men 55–64 from 58 to 61 percent and for women 55–64 from 42 to 46 percent). *Italic: Source — United Nations; World Health Organization; International Labour Organization; ILOSTAT; and IMF staff calculations as presented in the PDF chapter "2. Contributions to Total Population Growth."* ### 1. Total Population (16–64)2. Young Women (25–45) ### 1. Total Population (16–64)2. Young Women (25–45) ### Implications for Labor Supply — scenarios and headline effects - Two illustrative reform scenarios (see Annex 2): - Moderate reform scenario: - Assumes moderately paced increases in female labor force participation rates to the highest Western European rates. - Assumes retirement age increases in line with life expectancy, but not higher than 67. - Ambitious reform scenario: - Assumes rapid increases in female labor force participation rates to the highest Western European rates. - Assumes raising the retirement age above 70 by 2050. - Key comparative effects: - Measures to increase female labor force participation could increase the available labor force compared to baseline, especially before 2030. - Boosting participation of older workers without lifting retirement ages would have a larger impact over time as the share of older cohorts grows. - Raising statutory retirement ages would have the largest long-term impact — could reduce the median decline in the labor force by nearly 10 percentage points by 2050. - Aggregate projected impacts: - Under the moderate reform scenario: - The average labor force decline would be limited to 3 percent by 2030, but would nonetheless be 16 percent by 2050. - Reforms in Albania, Belarus, Bosnia and Herzegovina, Republic of North Macedonia, Moldova, and Serbia would bring the net decline in 2050 to below 10 percent. - Bulgaria, Estonia, Latvia, Lithuania, and Poland would face labor force declines of more than 20 percent despite increases in labor force participation. - Under the ambitious reform scenario: - Labor force declines in many countries would stall or even be reversed, but only temporarily in most cases. - In several countries, including Belarus, Moldova, and the Western Balkan countries, ambitious reforms could potentially halt the decline in the labor force in the long term. - In some countries (including Bulgaria and Latvia) the labor force is projected to decline by more than 23 percent even with ambitious reform measures. - Cross-country heterogeneity in policy gains: - Slovenia and Lithuania (already high female labor force participation rates): would benefit most from increasing the retirement age. - Bosnia and Herzegovina and Moldova: could significantly boost labor supply by increasing female labor force participation. - Belarus: would benefit significantly from increasing participation of older workers, even without changing the retirement age. - Illustration of timing: - Boosting female labor force participation: sizable short-term effects. - Increasing participation of older workers: medium- to long-term increases. - Raising retirement age: largest effect in the long run and helps offset projected labor force decline in several countries. ### Fertility policies and their limited potential to offset labor force decline - Context: - Fertility rates in the CESEE region are among the lowest in the world and, with the exception of Turkey, often well below replacement rates. - Empirical evidence on fertility-supporting policies: - Direct payments on birth, child allowances, and reduced tax rates for larger families have been used in some countries. - The empirical evidence is not encouraging: - The fall in fertility rates has been associated with an increase in the median age of women having their first child, driven in part by increasing costs of raising children. - Where financial incentives affected behavior, the effect was mainly to bring births forward, with little increase in fertility rates. - “The success of policies to influence fertility in low-fertility countries has not been impressive” (United Nations 2013; see also Kalwij 2010). - Policies that reduce the opportunity costs of having children appear more effective (Hoorens and others 2011), although evidence is mixed. - Policies that make it easier for women with children to reenter the workforce would also be consistent with facilitating increased female labor participation. - Simulation results (Box Figure 1.1 summary): - Even if fertility immediately increased to the UN high fertility scenario (2017 vintage), effects would be limited: - The working-age population would still not increase for at least 15 years. - Over longer periods the increased population would still be insufficient to offset the reduction in the labor force; in all countries the effects by 2050 are limited. ### Implications for age-related public spending — pensions and health care - Current spending levels (region vs euro area peers): - CESEE countries currently spend 9 percent of GDP on public pensions and 5 percent of GDP on public health care each year. - Euro area peers spend 12 percent of GDP on public pensions and 7 percent of GDP on public health care. - Projected increases on current policies (2015–2050): - Costs related to pensions and health care in CESEE are expected to increase by nearly 4 percentage points of GDP, on average, between 2015 and 2050 (see Annex 3). - Almost half of the increase is explained by rapidly rising health care spending: - The fiscal cost of public health care systems is projected to increase by 2¾ percentage points of GDP. - The projected fiscal cost of public pension systems on current policies appears to rise by just over 1 percentage point of GDP. - Country variation: - Pension costs are projected to remain unchanged or fall in several countries, whereas they are projected to increase between about 4 to 7 percentage points of GDP in several Western Balkan and Eastern European countries. - Differences are mostly explained by country-specific pension systems (retirement age, entitlement, etc.) and coverage of public health care systems. - Replacement rates and sustainability: - Many CESEE pension systems are based on defined contributions and appear fiscally sustainable but have relatively low replacement rates. - On average, a pensioner in 2050 will receive about one-third of their preretirement earnings. - Except for Ukraine, the replacement rate is projected to drop in CESEE countries; notable projected drops include: - Republic of North Macedonia: drop by 28 percentage points. - Moldova: drop by 30 percentage points. - Poland: drop by 25 percentage points. - Potential political pressure to raise replacement rates: - Bringing replacement rates to 40 percent (the level recommended by the International Labour Organization) would: - Increase estimated fiscal costs of pensions by 4 percentage points of GDP, on average. - Increase total pension and health care costs in the CESEE region by nearly 7 percentage points of GDP, on average, between 2015 and 2050. - Age-related spending is projected to reach 20 percent of GDP by 2050 on average, with some countries (Bosnia and Herzegovina, Croatia, Poland, Slovenia, and Ukraine) spending as much as 25 percent of GDP or more on pensions and health care. - Role of labor market reforms: - Ambitious reforms to increase labor force participation rates and retirement ages would help to offset projected increases in public pension spending. - Moderate and ambitious labor market reform packages (increases in women’s and older workers’ labor force participation, and increases in retirement age) are considered in the fiscal projections; moderate reforms would help offset some projected pension cost increases, while ambitious reforms would offset more but might still be overwhelmed by demographic changes in many countries. - Fiscal policy implication: - Fiscal space needs to be preserved for measures to increase labor participation. *Source: IMF staff calculations; data and figures as presented in the chapter excerpt provided.* ### 3. Age-related Spending in 2050 ### 3. Age-related Spending in 2050 ### Pension spending projections and reform impacts - Increasing replacement rates to the ILO’s 40 percent replacement rate imply an increase in pension outlays; the text analyzes implications for age-related public spending. - Reforms could reduce projected pension costs over the period from 2015 to 2050: - Reforms would reduce costs by about 2 percentage points of GDP on average. - Ambitious reforms would generate average savings of nearly 5 percentage points of GDP. - Moderate labor market reforms are not sufficient to offset long-term costs; ambitious reforms would fully offset projected increases in pension costs in most countries. - Cross-country variation in reform impact is driven by: - The projected increase in the share of population reaching retirement age. - The policy objectives (the gap between current labor force participation rates and the highest Western European rates). - The speeds at which retirement ages need to be increased. - Country examples and patterns: - Substantial scope to increase labor force participation: Hungary, Moldova, and Serbia. - Aligning retirement ages with life expectancy implies large increases from current retirement ages (especially among women) in some countries. - Large shares of the population reaching retirement age in coming decades: Bosnia and Herzegovina, and Slovenia. - After accounting for the fiscal cost of policies, pension fiscal cost is projected to increase in some countries (examples): Albania, the Baltic states, Croatia, and the Czech Republic — but remain well below the projected increase in the absence of reforms. - In other countries (example: Romania, and to some extent the Slovak Republic), the cost of implementing labor market reforms would need careful assessment because projected fiscal savings are more limited. ### Fiscal costs of labor market reforms and net pension outcomes - Labor market reforms have direct fiscal costs that offset some pension savings: - Reducing the tax wedge could lower labor costs and boost job creation but would lead to revenue losses. - Active labor market policies (training, job creation programs, placement services, extended child daycare) have direct budgetary effects. - Ambitious labor market reforms would reduce projected fiscal cost of pensions in most CESEE countries, but could generate additional demand for health care services; evidence suggests sick leave and disability payments tend to increase with a higher share of older workers. - Net change in pension costs (2015–50) is assessed under baseline versus ambitious labor market reform scenario; outcomes vary across countries (see figure references in source). ### Health care spending trends and containment policies - Historical and projected trends: - Public health costs have increased rapidly in advanced economies in the past two decades. - Spending rose by about 1½ percentage points of GDP since the early 2000s in OECD countries (about 1¼ percent in Western Europe). - CESEE countries have, on average, kept government health spending roughly constant in percent of GDP to date, but cost pressures are likely to increase substantially. - Broad reform categories to contain health care spending (Clements and others 2012): - Macro-level Controls: - Budget caps on entire health care budgets or selected sectors. - Supply constraints: “input” controls (types of subsidized prescription drugs, rationing of expensive equipment) and “output” controls (delisting treatments from insured lists). - Price controls: setting salaries of health care professionals and/or prices of drugs and medical treatments; external reference pricing (ERP) has reduced prescription medicine prices by 4 to 20 percent in Bulgaria, Greece, Moldova, the Netherlands, and the Slovak Republic. - Example: Austria imposed financial targets to align health care spending with annual average GDP growth. - Micro-level Reforms: - Public management and coordination: streamlining administration, reviewing fiscal decentralization, consolidating regional authorities, improving public procurement with modern information technology (examples include Bulgaria, the Czech Republic, Estonia, and the United Kingdom). - Contracting and provider payment reforms: shifting from fee-for-service to case-based payments (diagnosis-related groups) and adopting explicit contracts targeting cost control, efficiency, and quality. - Market mechanisms: greater competition among insurers and private provision (examples: Estonia, Germany, the Netherlands, Switzerland) with careful attention to transparency, access, equity, and competition rules. - Demand-side Reforms: - Increasing the share of health care costs borne by patients to avoid excessive consumption of specific services; common mechanisms include patient cost-sharing and changes in private health insurance tax treatment. - Delivery and technology: - Improve hospital management and provider remuneration incentives. - Support primary and preventive care to reduce need for costly later treatments. - Improve provider payment systems to control costs. - Use health information technology (e.g., to collect, store, and share patient data) to potentially reduce costs while protecting outcomes. ### Active labor market policies: costs, variation, and fiscal space - Definition: Active labor market policies assist the unemployed or inactive to find employment (training programs, job placement assistance, employer subsidies, public works). - Estimated fiscal costs to boost female labor force participation through 2050: - Moderate reform package: projected fiscal cost about 0.7 percent of GDP. - Ambitious reform package: projected fiscal cost close to 1 percent of GDP. - Country variation: - Fiscal cost of ambitious reforms close to 1½ percent of GDP in Croatia, the Czech Republic, Poland, and Ukraine. - Fiscal cost of ambitious reforms close to ½ percent of GDP in the Slovak Republic and Romania. - Variation explained by differences in sensitivity of female labor force participation to government spending on active labor market policies and the gap to the highest Western European female participation rate. - Fiscal space constraints: - The upfront cost is especially challenging for countries with limited fiscal space. - The estimated gap between the debt-stabilizing primary balance and the observed primary balance is positive in Belarus, Moldova, Romania, Russia, and Turkey, suggesting fiscal consolidation may be required to stabilize public debt. ### Trade-offs and broader fiscal implications - Reforms can help contain pension and health care costs, but: - Labor participation reforms impose fiscal costs and social choices (for example, the extent of state-provided health care). - The fiscal costs considered understate full implications for public finances because a full assessment would account for effects of demographic changes on growth, revenues, and debt sustainability (addressed in the next chapter). *Italic source: IMF staff calculations; content excerpted from "3. Age-related Spending in 2050" (DEMOGRAPHIC HEADWINDS IN CENTRAL AND EASTERN EUROPE).* ### Box 3. Costs of Active Labor Market Policies ### Box 3. Costs of Active Labor Market Policies ### Aging and productivity: overview - Between 2015 and 2030, the share of older workers (ages 55+) in the CESEE workforce will increase by 2.5 percentage points; by 2050, it will expand by another 3 percentage points. - The chapter assesses how these demographic changes might affect productivity, complementing Chapter 3’s examination of direct effects on labor supply. - Theory points to multiple channels: - Positive effects: greater experience among older workers; incentives to automate as skilled labor becomes scarce (references cited). - Negative effects: age-related deterioration in physical and mental capabilities, depreciation of knowledge, and slower adoption of new technologies. - Magnification/externality effects: innovation and idea-generation peak in mid-career cohorts, so shifts in age structure can affect aggregate innovation beyond individual productivity changes. ### Literature and empirical evidence (key results preserved) - Macro-based studies generally find a hump-shaped relationship: productivity highest for workers in their 40s; increases in shares of older workers (above ages 50–55) coincide with reductions in aggregate productivity. - Selected quantitative findings from prior studies (as reported): - A 5 percentage point shift from the cohort aged 40–49 to the cohort aged 50–59 is associated with over a 10% decrease in TFP (Feyrer 2007). - A 1 percentage point increase in the share of workers aged 55–64 leads to a decline in the growth of output per worker of 0.25–0.7 percentage points (Aiyar et al. 2016). - A 1 percentage point increase in the share of workers aged 55–64 is associated with a reduction in TFP growth of 0.74 percentage points (IMF 2017). ### IMF empirical estimates (from Table 4 and text) - Data: panel of up to 167 countries, 1990–2015; instrumentation with 10-year lagged population cohorts; controls and robustness checks described in Annex 4. - Main estimated effects (first specification, columns 1–4): - A one-percentage-point increase in the share of older workers (ages 55+) in the total workforce is associated with a decrease in growth of output per worker of 0.7–0.8 percentage points (column 1). - Most of this effect stems from lower TFP growth, which decreases by about 0.6 percentage points for every percentage point increase in the older-worker share (column 2). - No evidence that human capital is an important transmission channel (column 3). - Workforce aging is associated with a statistically significant (albeit smaller) increase in the capital-output ratio, indicating some capital intensification (column 4). - Alternative specification (columns 5–8): - Productivity may be highest around age 50; a rising share of workers in the 45–54 cohort can mitigate the negative impact of increases in the 55+ share (column 6). ### Projections for CESEE and Western Europe - Method: UN World Population Prospects medium fertility scenario multiplied by baseline labor force participation rates from Chapter 3 to derive workforce shares by five-year age cohorts. - Baseline shares and projections: - In 2015, the share of older workers (ages 55+ years) in the total workforce was 15¾ percent on average in CESEE economies. - This share is projected to increase to about 22 percent by 2050 in CESEE. - The share of the most productive cohort (ages 45–54) is projected to increase in CESEE until the mid-2040s. - In Western Europe, the share of older workers is about 16½ percent in 2015, with most further increases occurring before 2030. - TFP impact calculation: projected differences in the share of workers ages 55+ for each 5-year period from 2015 to 2050 multiplied by 0.6 (the estimated effect on TFP growth). - Workforce aging could lead to a 0.09 percentage point decrease in annual TFP growth for CESEE countries on average over 2015–20. - About a 0.08 percentage point reduction over 2040–50 for CESEE. - For Western Europe, a 0.27 percentage point reduction in 2015–20 and about a 0.02 percentage point reduction in the last decade of the projection horizon. - Temporal pattern: Western Europe sees the strongest TFP growth impact early in the horizon, decreasing thereafter; CESEE sees a more muted early impact, hovering around 0.15 percentage points between 2025 and 2045, and slightly positive in 2045–50 (regional averages). - Cross-country heterogeneity (averages noted): - On average over 2020–50, the average annual decrease of TFP growth is 0.38 in CESEE and 0.34 in Western Europe. - Country examples: Belarus — relatively small increase in older-worker share and negligible average impact on TFP growth (2020–50); Moldova — share of older workers almost constant until 2040 then picks up rapidly 2040–50, contributing to an average decline of productivity growth of 0.23 percentage points. ### Uncertainty, inference limits, and time variation - Statistical uncertainty: estimated coefficient for aging on aggregate TFP growth is –0.608 with a 95 percent confidence interval from –0.942 to –0.274 (standard error 0.171). - Inference caveat: historical sample older-worker shares range about 2½ percent to 29 percent; projected shares range about 10 percent to 31 percent. Overlap exists but historical dynamics may not capture future dynamics fully. - Behavioral relationship may change over time: - Longer healthy life expectancy could push out the threshold age of “older workers.” - Evidence that average age of inventors has been rising. - Technological adoption triggered by labor scarcity could neutralize or offset negative productivity effects. ### Occupational and industrial composition effects - Differences in industrial structure and occupation composition alter aging impacts: - Productivity of manual workers might decline faster with age; productivity of service or professional occupations might increase with age. - CESEE has a higher share of employed persons—especially male—in professions associated with declining productivity of older workers than the EU-15 average; thus industrial structure may reinforce aggregate productivity declines from aging (Box Figure 4.1 categories preserved as described). *Sources: International Labour Organization (ILO); PWT 9.0; UN World Population Prospects; and IMF staff calculations.* ### Box 4. Industrial Structure and Productivity ### Box 4. Industrial Structure and Productivity ### Automation and Productivity: conceptual findings - Aging workforce implies a composition effect: "an increase in the share of older and less productive workers per se reduces aggregate labor productivity." - Scarcity of prime-age workers raises their wages, creating incentives for firms to install machines to boost output of scarce workers. - Machines can automate tasks but workers remain needed to operate robots; automation can generate new roles for workers. - Net effects of aging on aggregate productivity depend on the balance of the composition effect and incentives to automate; when machines become "sufficiently relatively productive and cheap, then they could be installed in such numbers as to offset the productivity effect." - Possible dynamic: "initial effect of aging would be a decrease in aggregate productivity, then an increase as machines become cheaper and more capable." - Job disruption is likely during transition; machines can produce "re-instatement effects" through new jobs but transition may increase labor-market mismatching. - Empirical evidence on automation effects is mixed: - Acemoglu and Restrepo (2018b) find that increases in automation in the United States were accompanied by decreases in employment and real wages. - Schneider, Hong, and Le (2018) applying the same method to Japanese data find the opposite. - Cross-study note: Acemoglu and Restrepo (2017) find a positive relationship between change in (log) GDP per capita and overall population aging (ratio of those above 50 to those between 20 and 49); these results differ from other specifications in the chapter. ### Prologue: first-pass implications for growth and convergence (production-function exercise) - Method: projected labor supply and productivity changes fed into a standard Cobb-Douglas production function (see Annex 5). - Key aggregate impacts for CESEE (excluding Turkey), PPP-weighted averages: - Real GDP growth rates would be lower by about 1.4 percentage points each year over the whole horizon (2020–50) compared with a "no-demographics" scenario. - Level of output in 2050 would typically be about 35 percent lower than in the no-demographics scenario. - The fall in labor supply typically explains about 60 percent of the total decline. - Notable country-level effects: - Labor-channel effects notably strong in Bulgaria, Latvia, and Poland. - Workforce-aging effect on TFP growth significant, with large GDP per capita falls in Moldova and Slovakia. - Countries vary: Poland and Slovakia much more negatively affected than Albania and Turkey. - Convergence implications: - Simple average real GDP per capita as a share of the Western European average projected to increase from 52 percent in 2020 to 60 percent in 2050, when adjusting 2023 potential growth rates by aging impact. - In some cases—Belarus, Croatia, Poland, and notably the Czech Republic and Russia—convergence may reverse. - Speed of convergence slows: across CESEE, GDP per capita relative to Western Europe increases by 8 percentage points on average by 2050 under aging-adjusted 2023 potential growth, but would have increased by 22 percentage points with no demographic effects. - Caveats: these estimates are indicative; they do not model endogenous responses of investment, adjustment of relative prices, or spillovers from global demographic change. ### Channels of economic adjustment (EEUMOD multicountry model setup) - Model: Europe-oriented version (EEUMOD) of the IMF’s Flexible System of Global Models (see Annex 6). Models production, spending decisions, and trade/financial interactions. - Demographic effects modeled through three channels (treated as transitory except population shocks, fading after 2050): 1. Lower aggregate population and employment: - Decreased output, lower financial and human wealth vs. no-demographics. - Households lower consumption and increase savings. - Firms reduce investment; capital stock permanently lower. - Labor becomes scarcer and more expensive relative to capital → increase in capital-labor ratio. - Demand for imports declines; net exports and current account improve (household saving effect dominates fiscal deficit increase). 2. Lower productivity: - Reinforces above effects and decreases per capita consumption (real wages fall further). - Leads to additional decline in investment. 3. Fiscal consequences (age-related fiscal costs from Chapter 4): - Health care and pension expenditures increase as a percentage of GDP. - Higher public spending supports private consumption (partially offsetting falls in real wages and wealth). - Increasing public debt raises sovereign risk premium and crowds out private investment. - Fiscal and simulation assumptions: - Increase in age-related fiscal costs assumed accommodated by increased deficits (and therefore debt stock) for baseline simulations, acknowledging alternatives (higher taxes or lower spending) would also slow growth. - EEUMOD assumes sovereign risk premium increases by three basis points for each percentage point of GDP increase in the level of public debt. - Goods and capital assumed completely mobile; labor not mobile (no endogenous migration response). - Simulations abstract from remittances; demographic shocks applied only to CESEE and Western European countries/blocs. ### General-equilibrium outcomes from EEUMOD simulations (representative CESEE excluding Turkey) - Output and channels (2050 impacts): - Demographic shock alone reduces level of output by about 27 percent by 2050: - 22 percentage points due to the labor supply effect. - 5 percentage points due to the TFP effect. - Including fiscal effect (age-related fiscal costs, deficit-financed) adds 4 percentage points, bringing total decline to 31 percent by 2050. - The EEUMOD estimate (31 percent total decline) is similar to the prologue estimate (35 percent) though somewhat smaller because higher relative wages raise the capital-labor ratio, mitigating decline. - Other macro responses over time: - Real consumption, real investment, real exports, and real imports decline steadily with shrinking labor supply and workforce aging. - Fiscal deficit increases with rising cost of aging (assumed wholly deficit financed). - Net exports and current account balances improve. - Real exchange rate initially depreciates to support net trade adjustment, then appreciates over the long term as real exports decline with output. - Aggregate growth impact summary: - "Demographic shifts would lower growth rates in all CESEE countries (excluding Turkey). Compared to the 'no-demographics' path, the average annual real GDP growth over the 2020–50 period would be lower by about" [text truncated in source here]. *Source: dhicaeeea - Box 4. Industrial Structure and Productivity (IMF staff content).* ### 1.2 percentage points, cumulating to a total decline in the output level of ### dhicaeeea - 1.2 percentage points, cumulating to a total decline in the output level of ### Effects on GDP and GDP per capita - Average yearly impact on GDP growth from 2020–50: decline of 1.2 percentage points in the CESEE region (excluding Turkey), cumulating to a total decline in the output level of 31 percent by 2050 (Figure 22, panel 1). - In per capita terms, average yearly growth in CESEE from 2020 to 2050 would be lower by about 0.6 percentage points, cumulating to a total decline in the output level of 17 percent (Figure 22, panel 2). - Country highlights: - Bulgaria: greatest impact on GDP, driven mainly by the very poor outlook for labor supply (see Figure 8). - Ukraine: greatest impact on GDP, driven by both labor supply and TFP, given the rapid increase in the share of older workers (see Table 1). - Southeastern Europe and Serbia: least affected, as outlooks for labor supply, TFP, and fiscal balances are relatively better. ### Convergence and regional patterns - Convergence continues in most countries, albeit more slowly. - By 2050, the gap to Western European income levels would narrow for the majority of CESEE countries; except in Croatia, Russia, and other Eastern European countries, per capita income levels would be closer to Western European levels by 2050. - Euro area Central Europe would surpass average Western European income levels by 2050 (Figure 22, panel 3). - Income levels are aggregated using projected purchasing power parity GDP weights. ### Spillovers - Simulations incorporate mutually reinforcing spillover effects: own demographic changes (labor supply, productivity, fiscal stresses) plus spillovers from demographics in neighboring countries on foreign demand (Box 6). - External spillovers make GDP 3 percentage points worse by 2050 in a typical-country simulation; this is about one-tenth of the overall effect on the level of GDP. - Spillover contributions are larger for other variables: - Investment: contribution closer to 20 percent. - Consumption: contribution nearly a quarter. - When all countries face demographic shocks, weaker demand for exports and import compression reduce consumption and investment, worsening household welfare. ### Private investment and capital dynamics - Model assumption: capital and labor are complements—investment declines with employment. - Capital-to-labor ratio increases as labor becomes scarce and the appreciation of the real exchange rate makes capital cheaper; phenomenon could encourage automation (see Box 5). - Contributions to the 1.2 percentage point average yearly decline in GDP (CESEE ex. Turkey, 2020–50): - Labor direct contribution: about 0.4 percentage points. - Productivity direct contribution: about 0.3 percentage points. - Investment response: accounts for 0.5 percentage points. - Private investment projection: lower by 26 percent by 2050 compared to the no-demographics scenario (Figure 22, panel 4). - Variation across countries in investment projections largely aligns with GDP projections and debt accumulation. ### Public finances, debt, and current account implications - Under the accommodating-deficits fiscal reaction (baseline simulation), average fiscal deficit in CESEE (excluding Turkey) would increase by about 9 percentage points of GDP (Figure 22, panel 5). - Age-related spending increases (health care and pensions), negative output impact, and rising interest payments (higher sovereign premium) magnify fiscal deterioration. - Public debt would increase by about 75 percentage points of GDP by 2050 compared to the no-demographics scenario. - Current account balances (2020–50 average impact): - CESEE current account balances would, on average, be higher by 1.5 percentage points of GDP compared to the no-demographics scenario (Figure 22, panel 6). - Countries with rapid population declines (e.g., Ukraine) would experience the largest increase. - Southeastern Europe would have a substantially smaller increase. - Central European euro-area countries would see external balances decrease somewhat compared with the no-demographic scenario. - Overall, current account flows from demographic adjustment would likely not materially affect world saving-investment balances. - Cross-check with IMF’s External Balance Assessment Model: region likely to see an increase in net savings over the next 20 years before declining as it enters late-stage demographic change; scale of increase somewhat less than EEUMOD due to different general equilibrium and saving behavior assumptions. - Note on investment drivers: demographic forces (labor supply and decline in productivity) responsible for about two-thirds of the fall in private investment; remaining one-third mostly attributed to increase in sovereign risk premium and associated increase in user cost of capital. ### Impact of labor market reforms - Labor market reforms operate via three channels: changes in the size of the labor force, workforce aging effects on TFP, and fiscal feedback effects from changes to health care and pension costs and sovereign risk (Figure 23). - For CESEE (excluding Turkey) overall, simulated effects from 2020–50: - Moderate labor market reforms: improve GDP growth by about 0.2 percentage points (average yearly). - Ambitious labor market reforms: improve GDP growth by about 0.4 percentage points (average yearly). - Per capita effects: - Ambitious reforms can make some states better off over the next decade than in a no-demographic scenario; total GDP effects are less pronounced over longer periods due to population decline. - Convergence impacts: - Moderate package: could lower the per capita income gap between CESEE (excluding Turkey) and Western Europe by about 3 percentage points of GDP. - Ambitious package: could narrow the gap by about 4 percentage points. - Eastern European states: ambitious reforms could add about 11 percentage points of GDP per capita to narrowing the income gap by 2050. - Central European states: ambitious scenario impacts more modest—about 1½ percentage points of GDP per capita for Baltic states, Slovakia, Slovenia; 4 percentage points for Czech Republic, Hungary, Poland. - Fiscal impacts of reform scenarios (period 2020–50): - Moderate reform scenario: - Offsets about 50 percent of the projected increase in aging-related government spending and debt. - Average government deficit projected to increase by about 4 percentage points of GDP by 2050 (compared with 9 percentage points in the baseline). - Public debt rises by 33 percentage points of GDP compared to a no-demographics scenario (compared with 75 percent in the baseline). - Ambitious reform scenario: - More than offsets baseline results. - Reduces average projected government deficit by about 3 percentage points of GDP (compared with an increase of 9 percentage points in the baseline). - Results in public debt lower by about 28 percentage points by 2050 compared to the no-demographics baseline. - Note: fiscal costs of labor market reforms are not modeled in these simulations. ### Alternative fiscal reactions: higher taxes versus accommodating deficits - Two fiscal reaction assumptions compared: - Accommodating deficits (baseline): increase in age-related fiscal costs accommodated by increased deficits. - Higher taxes scenario: distortionary labor and consumption taxes raise revenue equal to half of the deficit under the accommodating-deficits scenario each year. - Rationale: higher taxes contain debt (lower sovereign premium and cost of capital) but also lower demand and reduce employment incentives; net effect depends on initial indebtedness. - Aggregate CESEE results (average yearly impact over 2020–50): - Real GDP: - Accommodating deficits: annual GDP growth lower by about 1.2 percentage points relative to no-demographics. - Higher taxes: annual GDP growth lower by about 1.0 percentage point relative to no-demographics. - Real GDP per capita: - Higher taxes: per capita GDP growth lower by about 0.5 percentage points. - By 2050, level of per capita GDP is 14.7 percentage points lower under the higher taxes reaction relative to the no-demographics scenario—slightly smaller than the 17 percent fall under the accommodating deficits reaction. - Debt/GDP in 2050: - Higher taxes: debt-to-GDP ratio is 12 percentage points higher than the no-demographics scenario by 2050. - Accommodating deficits: public debt increases by about 76 percentage points by 2050 relative to no-demographics. - Conclusion: under the higher taxes reaction, the positive impact of lower debt outweighs the negative impact of distortionary taxes in the CESEE aggregate case. ### Mechanisms considered and limitations - Simulations using EEUMOD incorporate a narrow range of demographic effects: total population size, labor force size, and age composition effect on productivity. - Empirical evidence indicates additional mechanisms may matter for external balances (expected lifespan, retirement age, pension support for saving rates of workers). - Cross-check with IMF External Balance Assessment (which includes four demographic variables expressed as deviations from the world average: population growth, old-age dependency ratio, share of prime-age savers, life expectancy of prime-age savers) corroborates main findings on current account trends and drivers. *Source: IMF staff calculations and EEUMOD simulations as presented in the supplied content.* ### Box 7. Another Perspective on External Balances: The IMF’s External Balance ### Box 7. Another Perspective on External Balances: The IMF’s External Balance Assessment Model ### Demographic effects on external balances and interest rates - Simulations mimic a profile of increasing then declining current account balances as countries progress through demographic change: earlier in demographic transitions national savings are high; later, savings decline and the current account deteriorates as the elderly run down their assets in retirement. - Model evidence and empirical work typically find population aging accompanied by falling real interest rates: - The scale of the change to the “world” rate in simulations is small: only 0.1 percentage points. - The fall is greater for Central, Eastern, and Southeastern European countries: 0.6 percentage points for the region. - Countervailing forces can push rates up: an increase in the share of retirees (assumed to consume only and have low propensity to save) would put upward pressure on real interest rates; increases in age-related public spending can raise risk premiums. - Cautions in interpretation: - Projecting world rates requires demographic projections for countries that account for the bulk of world savings (outside the scope here). - The model abstracts from richer demographic dynamics. - In EEUMOD simulations, the fall in equilibrium interest rates is not sufficient to offset increases in sovereign rates arising from risk premiums; there is no “hidden dividend” for debt sustainability. - Countries may face more frequent encounters with the zero nominal bound. ### Key demographic trends and aggregate impacts for CESEE - Total population of the CESEE region is expected to decline significantly; only Turkey is expected to have positive population growth. - Drivers of adverse population dynamics: - High mortality rates and persistent net outward migration. - Projected CESEE birth rates are on average lower than in most other regions but mostly in line with Western Europe. - Mortality rates are projected to remain significantly higher than Western Europe. - About half of CESEE countries are projected to experience both negative natural population growth rates and outward migration over the next 30 years. - Population and aging projections: - Nearly half of CESEE countries are projected to experience net total population losses of 5 percent between now and 2030, and 15 percent by 2050. - CESEE countries are projected to age more quickly than the Western European average; younger and poorer CESEE countries generally expected to age more quickly. - Labor force and dependency impacts: - The labor force in CESEE countries is projected to decline by 10 percent by 2030 and by 26 percent by 2050—even after accounting for legislated increases in retirement ages that raise participation rates. - In half of CESEE countries, the ratio of elderly to those aged 20 to 64 years is projected to exceed 50 percent. - The share of workers aged 55 years or older in the workforce is expected to increase to over 20 percent. - Fiscal and growth consequences: - The aging population will increase demand for public spending on pensions and health care for the elderly in all CESEE countries, in most cases by over 5 percentage points of GDP and in some by 10 or more. - Projections (over 2020–50) indicate average yearly growth—in per capita terms—would be lower by 0.6 percentage points than without these demographic developments (a cumulative loss of 17 percent). - The average fiscal deficit across CESEE (excluding Turkey) would increase by 9 percentage points of GDP from 2020 to 2050 absent offsetting policies. - The cost of letting debt accumulate includes higher debt servicing and sovereign risk, deterring investment; raising taxes to meet expenditure pressures also has significant output costs. ### Quantified model results (selected tables and figures as presented) - Table 6. Effects of Demographic Shocks on GDP Growth Rates (Average yearly impact over 2020–2050, percentage points) - Production Function (Section VI.A) / EEUMOD (Section VI.C) - Real GDP 21.4 21.2 - Real GDP per-capita 21.0 20.6 - Table 7. Impact of Labor Market Reforms (Percentage points) - Baseline Moderate labor reforms Ambitious labor reforms - GDP growth 21.1621.0020.88 - GDP per capita growth 20.6020.4620.43 - GDP 230.9227.0224.2 - GDP per capita 216.9213.1212.2 - Figure and model notes: - Current account norms by country group 2015–50 (Percent of GDP) and contribution of demographic variables to Current Account Norm 2030 (Percent of GDP) are shown in model outputs (EBA model results, EEUMOD). ### Implications for policy: broad strategy - Overarching objective: more workers working more productively through increasing effective labor inputs, capital deepening, and raising productivity. - Recognized limits: policies to boost fertility have limited evidence of effectiveness; limited political appetite for substantially higher inward migration in many cases; even ambitious reforms may be overwhelmed by population changes in most countries. ### Policies to increase effective labor inputs - Boosting working-age population: - Consider worker import arrangements and permanent immigration as ways to address shrinking labor supply; national debates on migration focus on socially agreeable pace. - Boosting labor intensity (participation and employment rates): - Significant room exists to increase participation, especially of women and older workers. - Raising participation rates could fully offset decline in working-age population in Albania and Bosnia and Herzegovina, and more than offset it in Belarus, Republic of North Macedonia, and Moldova; significant contributions in Poland, Romania, and Slovakia. - Typical policy levers supported by cross-country evidence: - Lower labor tax wedges, lower unemployment replacement rates, active labor market policies oriented toward training schemes (Card, Kluve, and Weber 2010). - Product market liberalization (Égert and Gal 2016; IMF 2016d). - Targeted policies: - Raising female labor employment: more spending on “in kind” family benefits (such as childcare) empirically increases female employment rates (Égert and Gal 2016). - Raising employment of older workers: reduce generosity of unemployment benefits that most negatively affect employment of older and less skilled workers; promote workplace flexibility and part-time options; complement reforms by raising legal retirement ages. - Hours worked: in theory can increase effective labor inputs, but hours in CESEE are already substantially higher than in Western Europe and can be expected to decrease as incomes rise. - Mixed evidence exists on labor market liberalization and wage bargaining system reforms; what matters most is internalizing macroeconomic considerations in bargaining arrangements (Organisation for Economic Co-operation and Development 2018). - Policies to retain and attract skilled workers: - CESEE experienced large outward migration of well-educated young workers, reducing labor force and productivity and slowing convergence. - Key migration determinants: differences in per capita income, quality of institutions, and employment prospects (Atoyan and others 2016). - Policy implications: active labor market policies to mitigate emigration impacts; strengthen institutions and economic environment to encourage potential emigrants to stay; consider liberalizing immigration regimes for skilled workers. - Boosting health and life expectancy: - Reducing mortality (notably from heart diseases) could yield substantial life expectancy gains (United Nations estimate: a 5.9-year gain in life expectancy at birth for the region by reducing heart disease mortality to top-population levels), raising participation rates and extending viable working lifespans. ### Policies to support capital deepening - Private investment responses are central to growth and convergence; returns on investment depend on infrastructure, human capital, and institutions. - Subsidizing private capital investment risks inefficient allocation unless clear distortions exist; financial sector reforms can encourage efficient allocation. *Source: Box 7. Another Perspective on External Balances: The IMF’s External Balance Assessment Model (excerpt).* ### references therein). For the most part, CESEE countries have open capital ### Summary and Implications for Policies ### Policies to Boost Productivity - Allocation of labor and capital: - Differences in total factor productivity can be explained by differences in the efficiency with which capital and labor are allocated to higher-productivity firms (Restuccia and Rogerson 2017; Hsieh and Klenow 2016 and references therein). - Product market reforms are associated with higher capital stock (Égert and Gal 2016) and help boost total factor productivity growth, especially if they boost competition (IMF 2015b and references therein). - Other reforms that facilitate more efficient resource allocation would also help increase productivity (Hsieh and Klenow 2016). Examples include strengthening state-owned enterprise governance or privatizing state-owned enterprises, reductions in red tape, and reducing the size of informal sector. - Human capital: - Growth differences across countries over long periods of time are primarily driven by human capital (C. Jones 2005). - Two policies to boost human capital in the context of demographic change: - Preschool, primary, secondary, and tertiary/vocational education: There is a strong relationship between educational attainment and growth rates. Education becomes all the more important when dependency ratios increase. Public spending and educational attainment show little connection (Hanushek and Woessmann 2017), but measures such as regular assessments (Woessmann 2016) would not require substantial increases in public spending. - Lifelong learning: Lifelong learning will likely increase in importance with policies to encourage greater participation of older workers. The Organisation for Economic Co-operation and Development’s Survey of Adult Skills shows that Nordic countries—with high levels of participation and human capital—spend proportionately more on lifelong learning than other countries. - Inclusive growth dimension: - Exclusion of underrepresented groups from labor force by not equipping them with qualifications can seriously dampen growth (Hsieh and others 2018). - Technological change displaces unskilled workers and exacerbates inequality (Acemoglu 2002). - Governments should preserve spending on education and training—even though the numbers of the young will decrease, the need to increase educational attainment motivates increased spending. Attention will need to be paid to retraining older workers; the region lags Western Europe in the numbers of adults engaged in retraining and lifelong learning programs (Organisation for Economic Co-operation and Development). - Governance: - Improving the quality of institutions, including government efficiency, helps slow outward migration and improve productivity growth. - High levels of “social infrastructure” are associated with both higher average skill levels and total factor productivity (Hall and Jones 1995). - For CESEE countries, the quality of institutions has been found to be a key constraining factor to the efficient use of technologies and resources (IMF 2016b). The effect will likely be larger for CESEE countries that typically have relatively high levels of educational attainment compared with other countries of similar incomes (European Bank for Reconstruction and Development 2019). ### Policies to Ensure Sustainability of the Public Finances - Projected fiscal pressures: - On current fiscal policies, costs related to pensions and health care in CESEE are expected to increase by nearly 4 percentage points of GDP, on average, between 2015 and 2050. - Almost half of the increase is explained by rapidly rising health care spending. - Bringing replacement rates to the recommended minimum level of 40 percent would increase estimated fiscal costs of pensions by 4 percentage points of GDP, on average. - Total pension and health care costs in the CESEE region would therefore increase by nearly 7 percentage points of GDP, on average, between 2015 and 2050. - Reform scenarios: - The moderate reform scenario helps offset about half of the projected increase in aging-related government spending and debt for the period 2020 to 2050. - The ambitious reforms would fully offset projected increases in pension costs in most countries; when all factors are accounted for, including the reduction in debt pressures, the ambitious labor reform scenario could more than offset the baseline results. - Raising retirement ages with improvements in life expectancy would help directly by reducing the number of pensioners, and indirectly by helping complement efforts to boost the labor force participation of older workers. - Higher participation of older workers would require higher spending on public health (European Bank for Reconstruction and Development 2019). - Policy implication: - Fiscal space needs to be preserved for measures to increase labor participation and raise skills, motivating a broader examination of tax systems and more efficient public expenditure. ### Policy Priorities (heterogeneity across CESEE countries) - Labor supply: - All countries except Turkey face the prospect of sharply decreasing workforces, based on projections for working-age population and participation rates. - The issue is particularly pressing for Bulgaria, Latvia, Poland, and Ukraine. - Participation rate priorities differ across countries: - Participation of younger women is noticeably low in Moldova and Turkey. - Participation of older women is low in Bosnia and Herzegovina, Croatia, Republic of North Macedonia, Romania, Turkey, and Ukraine. - Participation of older men is particularly low in Bosnia and Herzegovina, Romania, Slovenia, Turkey, and Ukraine. - Retirement age reforms: - Reform of retirement ages is likely necessary across most CESEE countries, but would be especially beneficial in Belarus, Moldova, Russia, Turkey, and Ukraine, both from the point of view of labor supply and fiscal sustainability. - Workforce aging and old-age dependency: - Workforce aging is mixed; Moldova and Slovakia are noted as aging rapidly. - Old-age dependency is expected to increase substantially in general but will not substantially affect Moldova and Turkey. - Old-age dependency is a more pressing issue in Central European countries, notably Poland and Slovenia. - All CESEE countries will experience increased fiscal pressures from age-related spending, especially Albania, Bosnia and Herzegovina, Croatia, Lithuania, Moldova, Poland, Russia, Slovenia, and Ukraine. - Policy tailoring: - Latvia and Lithuania face both sharply decreasing labor supply and already have relatively high participation rates, indicating policies should be directed at migration policies and conditions to boost productivity. - In countries such as Bosnia and Herzegovina, Moldova, Romania, and Ukraine, labor supply is also expected to decrease sharply, but there are opportunities to substantially improve output and fiscal sustainability with policies directed at raising participation rates. ### Data Sources and Projection Reliability (Annex findings) - Data sources and projection coverage: - Projections used are published by the UN Population Division, specifically the 2017 vintage of the World Population Prospects. - Population historical estimates and projections cover the period from 1950 to 2015 and from 2015 to 2100, respectively, at five-year intervals. - The World Population Prospects contains data and projections for population by country (total, by age cohort, male and female), young- and old-age dependency ratios, fertility, mortality, life expectancy, and net migration. - Projections are made for several scenarios: a “medium variant”; a “high variant” scenario that assumes higher fertility; a “low variant” scenario with lower fertility; and constant fertility and instant-replacement scenarios. - The World Population Prospects also includes 80 and 95 percent prediction intervals around the medium trajectory. - Forecast accuracy and migration: - Previous forecast errors were not noticeably larger for CESEE economies compared with those for Western Europe; there is little evidence that the projections were biased. - Forecast errors have been smaller on average for CESEE countries than other European countries; the median error of the CESEE projections is notably smaller than average error, in fact close to zero. - The UN tended to overestimate total populations in Southeastern Europe and underestimate those in Central Europe, with exceptions. - Incorporating migration projections has mostly improved forecast accuracy, with notable exceptions of Bosnia and Herzegovina, Bulgaria, and Estonia. - Migration is difficult to predict, but migration projections aided overall forecast performance rather than worsened it. - Uncertainty: - The United Nations publishes ex ante uncertainty estimates quantified as 80 and 95 percent confidence intervals around the medium scenario. - There is more uncertainty about fertility than mortality rates; this implies quite low levels of uncertainty about working-age population over the next 15 years. - For the total population of the CESEE region as a whole, the United Nations estimates only a 2½ percent chance that growth rates will not be negative in 20 years’ time. *Source: IMF staff analysis and annex material in the provided content.* ### Annex Figure 1.2. Forecast Errors ### Annex Figure 1.2. Forecast Errors ### UN projection forecast errors and confidence intervals - UN World Population Prospects 2000 revision: projection of total population in 2010 against actual, percent of actual (figure shows country-level forecast errors for CESEE economies; data labels use ISO codes). - Key qualitative findings: - For all CESEE subregions, negative population growth is evaluated to be a near certainty within 10 years at most. - By country, uncertainty around the medium path is greater for Balkan economies and Turkey, consistent with greater fertility uncertainty and younger populations. - Uncertainty is less for older economies with conventional “Western” demographic profiles. - UN confidence-interval properties: - Confidence intervals for total population growth paths shown (Annex Figure 1.3) and for working-age population growth (Annex Figure 1.4 and Annex Figure 1.5). - Two points emphasized for working-age population: - Even with Turkey included in the CESEE aggregate, population growth is always negative, even at 95 percent confidence limits. - There is hardly any uncertainty estimated for working-age population for the next 15 years—because those who will be of working age during the next 15 years have already been born; fertility uncertainty dominates mortality uncertainty. - Evaluation of UN uncertainty estimates: - UN estimates of uncertainty around fertility and mortality are found to be very good (Azose, Sevcikova, and Raferty 2016). - Not including uncertainty around migration paths is potentially problematic: prediction intervals for population paths for North America and Europe are substantially underestimated when migration uncertainty is not included. - Adding uncertainty about migration would tilt error bands downwards; therefore, lack of migration uncertainty should not be interpreted as bias toward being too pessimistic about population prospects in Europe. ### Baseline labor force projection methodology - Labor force decomposition formula used: - Laborforce = ∑_j Population_j * Participationrate_j, where j is five-year cohort of men or women from 16 to 80 years old. - Population projections source: - UN population projections (2017 vintage, medium fertility scenario). - Participation rates: - Current cohort participation rates from International Labour Organization data used for future participation rates. - Where planned increases in statutory retirement ages exist, participation rates of older cohorts (those aged 55 to 70 years) are increased to match rates in countries with similar statutory retirement ages. - Working-age focus: - Much of subsequent analysis focuses on working-age rather than total population. ### Policy scenarios for labor force participation and retirement - Scenarios change participation rates for older workers and women; population projections remain the same. - Female labor force participation (FLFP) scenarios (ages 25–45): - FLFP assumed to increase to the highest EU levels of the corresponding age-gender cohort. - Moderate FLFP increase: average annual FLFP increase of 0.5 percentage points (corresponding to average annual increase in the EU15 countries over 1995 to 2016). - Ambitious FLFP increase: average annual FLFP increase of 1.3 percentage points (highest average FLFP increase achieved in an EU country—Spain—over 1995 to 2016). - Increasing participation of older workers (aged 55 years or older) without changing retirement age: - Participation rates of older men and women assumed to rise to the maximum achieved in EU countries, gradually in the moderate scenario and rapidly in the ambitious scenario. - Participation rates start increasing from the first projection period; target participation rate changes with projected increases in the retirement age. - Raising statutory retirement age: - Moderate scenario: retirement age linked to changes in life expectancy until reaching a ceiling of 67 for both men and women; participation rates at the average of EU countries with similar statutory retirement ages. - Ambitious scenario: retirement age increases to 67 for both men and women by 2030 and links further increases to increases in country- and gender-specific life expectancy in 2030–2050. - Reform packages: - Moderate reform scenario = moderate FLFP increase + retirement age linked to life expectancy (ceiling 67) + participation of older workers improving to maximum achieved in EU countries with corresponding statutory retirement ages. - Ambitious reform scenario = rapid FLFP increase + retirement age increases to 67 by 2030 (and subsequently linked to life expectancy) + participation of older workers at the maximum achieved in EU countries with corresponding statutory retirement ages. ### Participation rate baselines and policy scenario tables (selected features) - Baseline participation rates for workers aged 55–64 are tabulated for men and women across 2015, 2020, 2025, 2030, 2035, 2040, 2045, 2050 (Annex Table 2.1). - Moderate policy scenario: Annex Table 2.2 provides men and women 55–64 participation rates for 2015–2050 reflecting gradual improvements and alignment with EU maxima. - Ambitious policy scenario: Annex Table 2.3 provides men and women 55–64 participation rates for 2015–2050 reflecting more rapid improvements (examples include higher rates by 2050). - Statutory retirement ages (Annex Table 2.4) list country values; example entries: - Albania: Retirement age men 65.0, women 63.0 - Belarus: Retirement age men 60.0, women 55.0 - Average WE: 65.6, 65.2 ### Baseline pension and health care projection framework - Pension expenditure to GDP decomposition (equation 1): - PE / GDP = (PE / pensioners) × (pensioners / pop65+) × (pop65+ / pop15−64) × (pop15−64 / workers) - Intuition: pension spending grows in line with the old-age dependency ratio; assuming constant benefit and coverage ratios, an increase in population ages 65+ due to unexpected longevity increases pension spending for the same level of GDP. - Health care expenditure to GDP decomposition (equation 2): - HE_GDP = (HE_0−64 / pop0−64) * (GDP_workers / pop0−64_workers) × (1 + α × (pop65+ / pop0−64)), where α = (HE_65+ / Popul65+) / (HE_0−64 / Popul0−64) - Intuition: an exogenous increase in the number of elderly increases health care spending because older population receives relatively high per capita health benefits. ### Policy effects on pension spending (operationalized) - Replacement rate policy: - Replacement rate set to take a value of at least 40 (minimum recommended replacement rate). Pension spending increases in countries and years where the replacement rate was originally below 40. - Labor force participation effects: - Increases in participation of young women and older workers reduce pension spending via the old-age dependency ratio. - Retirement age effects: - Increasing statutory retirement age reduces pension spending indirectly via the coverage ratio (reducing the share of pensioners in total population above 65). ### Production function and link to productivity - Cobb-Douglas production function used: - y_it = (K/Y_it)^α_(1−α) * A_it * h_it (standard Cobb-Douglas with constant returns to scale; y_it is output per worker, K/Y is capital per worker, A_it is TFP, h_it is human capital per worker, α is capital share). - Workforce age-structure effect on productivity (growth specification): - ∆ log y_it = f(∑_s β_s w_sit), where w_sit is share of each age cohort s in the workforce and β_s is contribution of cohort s to productivity growth. - Channels through which workforce age structure may affect labor productivity: - Capital intensity, human capital, and TFP may act as transmission channels from workforce aging to labor productivity. - Empirical approach: - Analysis follows Feyrer (2007); Aiyar, Ebeke, and Shao (2016); and Adler and others (2017) with a regression specification examining the impact of workforce aging on productivity growth and its main drivers. *Source: United Nations; and IMF staff calculations.* ### Annex 4. Estimating the Effects ### Annex 4. Estimating the Effects of Aging on Productivity ### Empirical specification and identification - Baseline estimating equation: - ∆ logYL_it = α_i + γ_t + ∑_s β_s w_sit + δ yadr_it + φ oadr_it + ε_it, - where YL_it denotes output per worker, α_i is a country fixed effect, γ_t are time dummies for several periods (1990–95, 1998–99, and 2008–09), oadr_it is the old-age dependency ratio, and yadr_it is the young-age dependency ratio. - Decomposition: four separate regressions estimated for components of labor productivity—(physical) capital, human capital, and TFP—to assess whether workforce aging operates through factor accumulation or TFP. - Instrumentation to address endogeneity: - Worker age shares instrumented with the 10-year lagged shares of the respective population cohorts. - Youth and old-age dependency ratios instrumented with the 10-year lag of the population share under the age of 4 and the population share of those aged 55 to 59 years, respectively. ### Data and sample - Panel data set with 167 countries over 1990–2014. - Population data: UN World Population Prospects (2017), available from 1950 onwards. - Workforce data: ILOSTAT. - Young and old-age dependency ratios: share of population aged 0 to 14 and 65+ to the active population aged 15 to 64, respectively. - Output per worker, capital stock, human capital, and TFP data: Penn World Table 9.0 (data end in 2014). ### Robustness checks and alternative specifications - Specification including all age cohorts except the 45- to 54-year reference group: - All coefficients except those for the older workers were insignificant. - The value for the significant coefficient was about –0.8. - Sample restrictions and sensitivity: - Excluding small countries with populations of less than 1 million, excluding oil and commodity exporting countries, and limiting the sample to European countries yielded a coefficient with an unchanged sign and broadly similar magnitude while maintaining statistical significance at the 1 percent level. - Regional interactions: - Interacting the old-age worker share with sub–CESEE regional dummies produced results either broadly consistent with main findings or insignificant. - Nonlinearities and additional controls: - Adding a quadratic term of the old-age worker share produced implausible results, possibly due to instrumentation difficulties. - Adding the uninstrumented share of manufacturing in total value added yields a slightly lower coefficient on the older worker share. - Occupational and education interactions: - Interacting the old-age worker share with occupational shares or using educational attainment shares within the older worker cohort did not yield additional insights—likely related to instrumenting difficulties for interacted regressors. ### Simplified production-function benchmark (Cobb-Douglas, balanced growth) - Balanced-growth assumption: capital-to-output ratio constant implies: - gY = gK = gA/α + gL, - where gY is output growth, gK is capital growth, gA is TFP growth, gL is labor growth, and α is the labor income share. - Aging impact expressed as difference between “aging” and “no-change” scenarios: - gY_aging – gY_no-change = (gA_aging – gA_no-change)/α + (gL_aging – gL_no-change), - “no-change” scenario: size of labor force constant and rate of TFP growth equal to the initial year (here, 2023), implying gL_no-change = 0 and gA_no-change = gA_2023. - Aging scenario inputs: - gL_aging: projected growth rate of labor supply. - gA_aging – gA_no-change: change in TFP growth derived by multiplying the change in the share of older workers by the estimated coefficient. - Formula reported in source for TFP channel: - gA_aging – gA_no-change = –0.608 * (Share55 + aging – Share55 + 2023) ### Inputs required for productivity projections (as summarized) - Labor force projections: IMF staff estimates (Section III). - Projected share of older workers (55+) and initial value in 2023: UN demographic database. - Regression coefficient of the share of older workers on total factor productivity growth: IMF staff estimates (Section V). - Labor income share: Penn World Tables version 9.1. *Source: Annex 4. Estimating the Effects of Aging on Productivity* ### References ### References ### Demographics, Aging, and Labor Market - Disney, Richard. 1996. Can We Afford to Grow Older? A Perspective on the Economics of Aging. Cambridge, MA: MIT Press. - Dixon, Sylvia. 2003. “Implications of population ageing for the labour market.” Labour Market Trends 111(2): 67–76. - Eggertsson, Gauti B., Manuel Lancastre, and Lawrence H. Summers. 2018. “Aging, Output per Capita and Secular Stagnation.” NBER Working Paper 24902, National Bureau of Economic Research, Cambridge, MA. - European Commission. 2015. The Aging Report: Economic and budgetary projections for the 28 EU Member States (2013–2060). Brussels: European Commission. - European Commission. 2018. The 2018 Ageing Report: Economic and Budgetary Projections for the 28 EU Member States (2016–2070). Brussels: European Commission. - Ferrero, Giuseppe, Marco Gross, and Stefano Neri. 2017. “On Secular Stagnation and Low Interest Rates: Demography Matters.” ECB Working Paper 2088, European Central Bank, Frankfurt. - Hopenhayn, Hugo, Julian Neira, and Rish Singhania. 2018. “From Population Growth to Firm Demographics: Implications for Concentration, Entrepreneurship and the Labor Share.” NBER Working Papers 25382, National Bureau of Economic Research, Cambridge, MA. - Jaimovich, Nir, and Henry Siu. 2009. “The Young, the Old, and the Restless: Demographics and Business Cycle Volatility.” American Economic Review 99 (3): 804–26. - Kalwij, Adriaan. 2010. “The Impact of Family Policy Expenditure on Fertility in Western Europe.” Demography 47 (2): 503–19. - Kreuger, Dirk, and Alexander Ludwig. 2007. “On the Consequences of Demographic Change for Rates of Returns to Capital, and the Distribution of Wealth and Welfare.” Journal of Monetary Economics 54: 49–87. - Liang, James, Hui Wang, and Edward P. Lazear. 2018. “Demographics and Entrepreneurship.” Journal of Political Economy 216: S140–96. - Lindh, Thomas, and Bo Malmberg. 1999. “Age Structure Effects and Growth in the OECD 1950–1990.” Journal of Population Economics 12: 431–49. - Maestas, Nicole, Kathleen J. Mullen, and David Powell. 2016. “The Effect of Population Aging on Economic Growth, the Labor Force and Productivity.” NBER Working Paper 22452, National Bureau of Economic Research, Cambridge, MA. - Organisation for Economic Cooperation and Development. 1998. “Work-force ageing.” OECD Employment Outlook 1998. Paris: OECD Publishing, pp. 123–51. - United Nations. 2012. Changing Levels and Trends in Mortality. New York: United Nations. - United Nations. 2013. World Fertility Report: Fertility at the Extremes. New York: United Nations. - United Nations. 2015. World Mortality Report 2015. New York: United Nations. - United Nations. 2017a. World Population Prospects: Methodology of the United Nations Population Estimates and Projections, 2017 Revision. New York: United Nations. - United Nations. 2017b. World Population Prospects: Volume II: Demographic Profiles, 2017 Revision. New York: United Nations. - Wattenberg, Ben J. 1987. The Birth Dearth. New York: Pharos Books. - Veen, Stephan. 2008. Demographischer Wandel, alternde Belegschaften und Betriebsproduktivität. Munich: Rainer Hampp Verlag. - Werding, Martin. 2008. “Ageing and Productivity Growth: Are there Macro-level Cohort Effects of Human Capital?” CESifo Working Paper Series 2207, CESifo Group Munich. ### Productivity, Innovation, and Human Capital - Feyrer, James. 2007. “Demographics and Productivity.” Review of Economics and Statistics 89 (1): 100–09. - Feyrer, James. 2008. “Aggregate Evidence on the Link Between Age Structure and Productivity.” Population and Development Review 34: 78–99. - Galenson, David W., and Bruce Weinberg. 2000. “Age and the Quality of Work: Modern American Painters.” Journal of Political Economy 108: 761–77. - Galenson, David W., and Bruce Weinberg. 2005. “Creative Careers: The Life Cycles of Nobel Laureates in Economics.” NBER Working Paper 11799, National Bureau of Economic Research, Cambridge, MA. - Göbel, Christian, and Thomas Zwick. 2012. “Age and Productivity: Sector Differences.” De Economist 160 (1): 35–57. - Hall, Robert E., and Charles I. Jones. 1995. “Why Do Some Countries Produce So Much More Output per Worker than Others?” Quarterly Journal of Economics 114: 83–116. - Hsieh, Chang-Tai, Erik Hurst, Charles I. Jones, and Peter J. Klenow. 2018. “The Allocation of Talent and U.S. Economic Growth.” Manuscript. - Hsieh, Chang-Tai, and Peter J. Klenow. 2016. “Productivity and Misallocation.” NBER Reporter 2016 Number I: Research Summary, National Bureau of Economic Research, Cambridge, MA. - Jones, Benjamin F. 2005. “Age and Great Invention.” NBER Working Paper 11359, National Bureau of Economic Research, Cambridge, MA. - Jones, Charles I. 2005. “Growth and Ideas.” In Handbook of Economic Growth, edited by P. Aghion and S. Durlauf, volume 1B, 1063–11. New York: Elsevier. - Lutz, Wolfgang, Jesus Crespo Cuaresma, and Warren Sanderson. 2008. “The Demography of Educational Attainment and Economic Growth.” Science 319: 1047–48. - Romer, Paul M. 1990. “Capital, Labor, and Productivity.” Brookings Papers on Economic Activity: Microeconomics 337–67. - Restuccia, Diego, and Richard Rogerson. 2017. “The Causes and Costs of Misallocation.” Journal of Economic Perspectives 31 (3): 151–74. - Woessmann, Ludger. 2016. “The Importance of School Systems: Evidence from International Differences in Student Achievement.” Journal of Economic Perspectives 30 (3): 3–31. - Hanushek, Eric A., and Ludger Woessmann. 2010. “Education and Economic Growth.” In International Encyclopedia of Education, edited by Penelope Peterson, Eva Baker, and Barry McGaw, volume 2, 245–52. Oxford: Elsevier. - Hanushek, Eric A., and Ludger Woessmann. 2012. “Do Better Schools Lead to More Growth? Cognitive Skills, Economic Outcomes, and Causation.” Journal of Economic Growth 17: 267–321. - Hanushek, Eric A., and Ludger Woessmann. 2017. “School Resources and Student Achievement: A Review of Cross-Country Economic Research.” In Cognitive Abilities and Educational Outcomes: A Festschrift in Honour of Jan-Eric Gustafsson, edited by Monica Rosén, Kajsa Yang Hansen, and Ulrika Wolff, Chapter 8. New York: Springer. - Maestas, Nicole, Kathleen J. Mullen, and David Powell. 2016. “The Effect of Population Aging on Economic Growth, the Labor Force and Productivity.” NBER Working Paper 22452, National Bureau of Economic Research, Cambridge, MA. ### Structural Reforms, Macroeconomics, and Policy Methodologies - Égert, Balázs, and Peter Gal. 2016. “The Quantification of Structural Reforms in OECD Countries: A New Framework.” OECD Economics Department Working Papers 1354, Organisation for Economic Co-operation and Development, Paris. - Gagnon, Etienne, Benjamin K. Johannsen, and David Lopez-Salido. 2016. “Understanding the New Normal: The Role of Demographics.” Finance and Economics Discussion Series 2016–080, Board of Governors of the Federal Reserve System, Washington, DC. - International Monetary Fund (IMF). 2014. Fiscal Monitor: Back to Work: How Fiscal Policy Can Help. Washington, October. - International Monetary Fund (IMF). 2015a. Technical Background Note: 2015 Refinements to the External Balance Assessment (EBA) Methodology. Washington: International Monetary Fund. - International Monetary Fund (IMF). 2015b. “Structural Reforms and Macroeconomic Performance: Initial Considerations for the Fund.” International Monetary Fund, Washington, DC. - International Monetary Fund (IMF). 2016a. “Methodological Note on EBA-Lite.” IMF Policy Paper, International Monetary Fund, Washington, DC. - International Monetary Fund (IMF). 2016b. Regional Economic Issues: Central, Eastern and Southeastern Europe—Effective Government for Stronger Growth. Washington, November. - International Monetary Fund (IMF). 2016c. “Republic of Latvia: Selected Issues Paper.” IMF Country Report 16/172, International Monetary Fund, Washington, DC. - International Monetary Fund (IMF). 2016d. “Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies.” In World Economic Outlook. Washington, April. - International Monetary Fund (IMF). 2017. “Asia: At Risk of Growing Old before Becoming Rich?” In Regional Economic Outlook: Asia and Pacific. Washington, April. - International Monetary Fund (IMF). 2018. External Sector Report—Refinements to the External Balance Assessment Methodology—Technical Supplement. Washington: International Monetary Fund. - Phillips, Steven, Luis Catão, Luca Ricci, Rudolfs Bems, Mitali Das, Julian Di Giovanni, D. Filiz Unsal, Marola Castillo, Jungjin Lee, Jair Rodriguez, and Mauricio Vargas. 2013. “The External Balance Assessment (EBA) Methodology.” IMF Working Paper 13/272, International Monetary Fund, Washington, DC. - OECD. 2017. “Economic Policy Reforms 2017: Going for Growth.” Paris: OECD Publishing. - OECD. 2018. Employment Outlook 2018. Paris: OECD Publishing. - European Bank for Reconstruction and Development. 2019. Transition Report 2018–2019. London. - Schwierz, Christoph. 2016 “Cost-Containment Policies in Hospital Expenditure in the European Union.” European Commission Discussion Paper 037, Brussels. ### Health, Pharmaceuticals, and Sectoral Studies - Organisation for Economic Cooperation and Development/European Observatory on Health Systems and Policies. 2017. Austria: Country Health Profile 2017, State of Health in the EU. Paris: OECD Publishing. - Kanavos, Panos, Anna-Maria Fontrier, Jennifer Gill, Olina Efthymiadou, and Nicola Boekstein. 2017. The Impact of External Reference Pricing Within and Across Countries. London: London School of Economics. - Schneider, Todd, Gee Hee Hong, and Anh Van Le. 2018. “Land of the Rising Robots.” Finance and Development 28. ### Technology, History, and Miscellaneous - Habakkuk, John. 1962. American and British Technology in the Nineteenth Century. Cambridge: Cambridge University Press. - Simon, Julian L. 1981. The Ultimate Resource. Princeton: Princeton University Press. *References (content unit: dhicaeeea - References). * --- _Source: https://www.imf.org/-/media/files/publications/dp/2019/english/dhicaeeea.pdf_