Aging, Immigration and the Welfare State in Austria

Transcription

1 Aging, Immigration and the Welfare State in Austria Thomas Davoine with Helmut Hofer, Christian Keuschnigg and Philip Schuster Institute for Advanced Studies (IHS, Vienna) Vienna, -3 September 03 / 3

2 Content The challenge The approach 3 Results / 3

3 Population aging in Austria Life Expectancy at birth (Women) Newborns (/000) 37 8 Aged 5-64 / Aged Source: projections from Statistik Austria (0) At constant policy, each worker is to support about x more retired persons over the next 40 years 3 / 3

4 Options Family life and work policies Pension reforms 3 Immigration effects 4 / 3

5 European Union Family life and work policies Euro Area Source: Commission services, Eurostat, EUROPOP00. Graph. - Projection Fertilityof rates total by fertility country rates in Europe in EUROPOP00 (number of births per woman) LV PT HU DE RO AT ES PL SK IT MT CZ CY EL SI LT BG EU7 LU EA EE NL DK BE FI NO SE UK FR IE Source: Commission services, Eurostat, EUROPOP00. Source: Eurostat in Ageing Working Group (0) Note: A slight reduction is projected for IE, FR, UK, SE and NO by 060. Literature (Gauthier, 007, review): mixed evidence on the effectiveness of family-friendly policies (public childcare, flexible work arrangements,...) 46 5 / 3

6 Pension reforms Different reforms: increase the retirement age increase social security contributions (or taxes) reduce pension payments Pension reforms are not popular 6 / 3

7 Immigration effects: illustration of potential for Austria Migrants are young: less than 5% are 40 or older (00, Austria) DELSA/ELSA/WD/SEM(0) Figure 6. Evolution and composition of the foreign population in Austria, absolute numbers and share in the total population, Other foreigners Foreigners from Turkey Foreigners from the former Yugoslavia All foreigners Share of foreigners in total population Share of foreigners in total population Source: Data provided by Statistics Austria. Source: OECD (0) The composition of recent immigration flows is shown in Figure 7. As can be seen, free movement is by far the most important category of entry for new permanent-type migration. Migrants from the enlarged European Union (EU-7) accounted for almost 60% of inflows in 009. Family migration remains the most important category for migrants from outside the EU, followed by humanitarian 7 / 3

8 ...but, is it here to last? Will it matter? Austria, 00 Austria, 070 foreigners 85+ natives foreigners 85+ natives employed unemployed non-participating employed unemployed non-participating 8 / 3

9 Scientific tools at hand and contribution Börsch-Supan, Ludwig and Winter (006): OLG and population aging analysis with endogenous labor supply: agg. savings rate with exogenous labor supply: agg. savings rate Existing tools and contribution: Analysis OLG Aging Migration End. labor supply Hours Retirement Storesletten (000) X X X Börsch-Supan & all (006) X X X Jaag & all (00) X X X X This paper X X X X X 9 / 3

10 Content The challenge The approach 3 Results 0 / 3

11 How OLG-CGE tools help economic analysis Incentive and general equilibrium effects Incentives: an increase in taxes or social security contributions discourages household labor supply GE: variations in labor supply trigger changes in wages CGE: take both into account to assess increases in government revenue Realistic life-cycle In OLGs: people are born, (study), work, retire, die Flexible framework Easy to add and remove institutional features in OLGs If rich enough, useful for policy reform analysis Here, quantitative answer to the question is the migration effect here to last? / 3

12 The price to pay (/3): wage differentials assumptions 00 wage differentials by skill class and origin natives foreigners low medium high low medium high 5 to 9 years to 4 years to 39 years to 54 years to 69 years We do not explain the gap now; we have to assume it will not change / 3

13 The price to pay (/3): migration motivation Barriers to migration are very different... hard to pin down in a model! Is there a maximum tolerance by destination countries? If yes: what is the number?; will it change over time? Migrants take it into account, weighing (econonomic and non-economic) pro s and con s Simplification: we ignore the questions and rely on someone else s crystal ball (exogeneous migration flows in and out) 3 / 3

14 The price to pay (3/3): migration outside options How much better (wages, pension rights) is it for migrants leaving and how much worse was it for migrants coming? Simplification: it s the same, if you leave the country (we don t know the cost of migration barriers anyway) For Austria: acceptable (009 flows to Germany:%; to ex-yugoslavia:%) Simplification: it s the same, if you arrive in the country For Austria: challengeable (009, stock from Germany:5%; from ex-yugoslavia:8%) We could avoid this simplification, but we lack the data (assets of immigrants at arrival) Not to mention households only differ by age (or have little differences in skill), in particular have same gender 4 / 3

15 Features of the OLG model OLG basis: max C y,c o,l y u(c y ϕ(l y )) + βu(c o ) s.t. A o = ( + r) (( τ)wl y C y ) C o A o With more heterogeneity: more than stages in life-cycle with different wages, unknown death 3 skill classes 8 generations Additional households decisions Labor market: education, participation, effort to find a job, hours, retirement Goods market: consumption versus saving Additional institutions Taxes: on labor income, on capital income, on consumption Social security contributions: worker side, firm side Social insurances: UI, PAYG pensions, welfare benefits,... Subsidies: to investment, to training,... 5 / 3

16 Content The challenge The approach 3 Results 6 / 3

17 Simulation results Social Pension Empl. GDP / Pension Social security per (hours/ capita system security contrib. benef. worker/ deficit deficit /year) ( %) ( %)* ( %)* (%GDP 0) (%GDP 0) Initial state (00) Demographic scenarios (050) Main High migration (+5%) Selected policy reforms (050) pension benefits (0.6% GDP) contributions (.3% GDP) retirement age ( year) * deviation from growth trend 7 / 3

18 Simulation results - methodological bias Assuming exogenous (constant) retirement age underestimates the benefits of pension reforms, such as decreases in pension payments Ret. Pension Empl. GDP / Pension Social age per (hours/ capita system security benef. worker/ deficit deficit /year) ( %) ( %)* (%GDP 0) (%GDP 0) Initial state (00) Selected policy reforms (050) Endo ret. age: pension benefits (3% GDP) Const ret. age: pension benefits (3% GDP) * deviation from growth trend 8 / 3

19 Summary key findings and policy implications Immigration helps a little: +5% migration = + year retirement age Reason: immigrants are young when they arrive, but they also age Health insurance is important: from 00 to 050, increases of at least... social security deficit: 6.8% to 4.7 % of GDP pension deficit: 4.5% to 9.7% of GDP Assuming constant retirement age moderately bias the results s.s. deficit if endogenous retirement decisions & cuts : 4.5% to 5.6% of GDP s.s. deficit if exogenous retirement decisions & cuts : 4.5% to 6.% of GDP Moderate to strong reforms are needed: to maintain social security deficit years increase in retirement age, -5% pensions, +5% SSC; or +8 years increase retirement age (increase in life expectancy: 7 years) Retirement age increase is less damaging to economic growth than pension cuts, SSC increases 9 / 3

20 Thank you for your attention 0 / 3

21 Appendix A. Pension financing is not the only challenge Age-related public health expenditures per capita (EUR, 007, Austria) 7,000 6,000 5,000 4,000 3,000,000, At constant policy, technology and use, mechanical increase in public expenditures due to population ageing / 3

22 Appendix B. Labor market elasticities in OLG Taxlab low skill medium skill high skill elasticities of - hours worked retirement participation / 3

23 Appendix C. Main demographic scenario - details absolute numbers Population (5+) Share of foreigners Dependency ratio Pensioners (in % of population) Effective retirement age Unemployment rate Employment (yearly hours per worker) Effective employment (yearly hours per capita) increase from basis in % Labor costs Net wages Pension payment per beneficiary increase from basis in % GDP/capita Capital/capita Consumption/capita in % of basis GDP Health expenditure Pension expenditure Social security deficit Social security deficit (constant population) Pension deficit Pension deficit (constant population) / 3