Skill Composition of Migration and the Generosity of the Welfare State: Free vs. Policy-Restricted Migration

1. Skill Composition of Migration and the Generosity of the Welfare State:Free vs. Policy-Restricted Migration Assaf Razin September 2008 Based on: Cohen and Razin (2008) Preliminary and Incomplete Draft

2. Research Question: • The paper analyzes the effect of the generosity of the welfare state on the skill composition of immigrants. • Our analysis is divided into two extreme cases: • Free migration: equilibrium is supply side determined; • Policy-Restricted Migration: equilibrium is “Policy” side determined.

3. Research Issue: • In free migration theory, welfare-state generosity adversely affects the skill composition of immigrants. • In the political-economy theory, If migration is policy-restricted, welfare-state generosity positively affects the skill composition of immigrants. • The reason, in both cases, is the two sides redistributive nature of the welfare state—the tax finance and the benefit sides. • If migration is free, a generous welfare-state attracts unskilled immigrants but deters the skilled immigrants; because, the latter are net contributors and the former net consumer of the system. • If migration is restricted by host country policy, skilled immigrants are more desirable by host countries, as they are net contributors to the welfare system, helping the tax financing of the system; whereas, unskilled immigrants are less desirable by host countries, as they are net contributors to the welfare system, shouldering the tax financing of the system on the native born voter.

4. Motivation for a New Approach : • Empirical literature as to the effect of the generosity of the system on the skill mix of immigrants is inconclusive: • One strand of the literature brings out evidence that welfare state serves as a magnet for immigrants: • Southwick (1981); Gramlich and Laren (1984); Blank (1988); Borjas (1999); Gelbach (2000); McKinnish (2005, 2007) • But, there is also evidence finding that welfare benefits has no significant effect over migration: • Walker (1994); Levine and Zimmerman (1999)

5. Literature (continue) • On immigration into European countries: • Peridy (2006) finds that welfare-state benefits have positive effect on migration. • De Giorgi and Pellizzari (2006) find that this positive effect is coherent across different levels of education among the immigrants. • Docquier, Lohest and Marfouk (2006) find that while all education groups are indeed attracted to welfare-benefits, the unskilled are motivated by social expenditure much more than the skilled. • These studies, however, do not differentiate between free and policy-restricted migration.

6. Parsimonious Model’s Predictions and Empirical Strategy • Our parsimonious model predicts a differential effect on migration and its skill composition, depending on whether migration is free or policy-controlled. to obtain unbiased estimates of the generosity of the welfare state on migration (and on its skill composition), one must control for the migration regime.

7. States within the US • Studies of migration between states of the US (such as Borjas (1999), for example), are evidently confined to a single migration regime (namely, free migration), and can obtain unbiased estimates.

8. Migration-regime and Empirics • studies that employ samples that are confined to the policy-controlled migration regime, but at the same time employ a model of the migrants' choice, whether to migrate and to which country, are inconsistent. Therefore, in this case the estimates convey little information on the migrant choices (or on the welfare state as a magnet), but rather on the migration policy choices by the host country.

9. Migration-regime mix • Studies that refer to both migration regimes, without controlling for them, also convey little information on either migration policy by the host country or on the individual migrant's migration choices

10. The Effect of Generosity in Our Sample With no differentiation between free vs. policy-restricted migration

11. Main Drawback of “old” Approach • This pooled regression does not separate source-host countries pairs into free and restricted migration groups. • The welfare-state benefits has a negative and significant effect on the skill composition of immigrants. But we argue that this is the result of the existence of “free migration” group which “dominates” the sample. • Our key hypothesis is that if migration is restricted by policy, the effect of the welfare state should be positive!

12. Theory: A Policy-Restricted Migration Model • Production: • Labor: • Government Budget: • Individuals:

13. Theory: Policy-Restricted Migration Stylized Model • Indirect Utility function: • Median Voter:

14. Policy-Restricted Migration: Key Effects • The share of skilled immigrants, σ, has two effects on the indirect utility. • an increase in σ increases aggregate labor productivity, and thereby the aggregate level of benefits per capita, b. • an increase in σ depresses skilled labor wages and raises unskilled labor wages. • If the median voter is unskilled, both effects are positive. Thus, the unskilled median voter is unambiguously for skilled immigration-- σu=1. • If the domestic median voter is skilled, however, the two effects are conflicting-- σs < 1. • Therefore: σs <σu=1.

15. Theory: Migration Policy in the Stylized Model • Comparative statics: • If the median voter is an unskilled individual, then an exogenous increase in the tax rate, , does not change the σu (since it is already at the corner). • If, however, the median voter is a skilled individual, an exogenous increase in the tax rate, , will change the migration policy, σs, in the direction of larger share of skilled immigrants (using the envelope theorem). • The reason is: if the tax rate increases, the redistribution fiscal burden upon a skilled median voter rises; therefore, additional skilled immigrants ease the burden.

16. First prediction of the parsimonious model • First, an increase in the generosity of the welfare state (τ) adversely affects the skill composition of migrants, in the free-migration regime

17. Second prediction of the parsimonious model • An increase in the generosity of the welfare state (τ) has a more pronounced effect on the share of skilled migrants of total migrants when the migration-regime is policy-controlled

18. A Free Migration Model • We make two simplifying assumptions: • (1) the source country is relatively small. Therefore, its emigrants cannot affect labor supply, productivity and wages in the host countries. • (2) individuals in the source country are heterogenous with respect for the reservation utility, for each skill level • (reasons for the heterogeneity: migration costs due to housing, family size, portablity of savings, etc.).

19. Free Migration Stylized Model • Denote skilled and unskilled migration: Equilibrium w.r.t the cut-off migrant: • The two equations implicitly determine the equilibrium quantities of immigrants, ms and mu:

20. Predictions of Free Migration Theory • Exogenous increase of the tax rate, , deters skilled immigrants and attracts unskilled immigrants. • The reason is simple: • If migration has no effect on the markets in the host countries, then only the direct effect of the tax rate exists: more benefits on the one hand, and less disposable income on the other hand. • Consequently, those who are net contributors to the welfare state (skilled workers) suffer and those who are net earners of the welfare state (the unskilled workers) benefit.

21. Discussion • Predictions hinge on the redistributive nature of the welfare state: • Under free migration, equilibrium reflects (among other) the interests of the immigrants. • Thus, a generous welfare state inflicts a fiscal burden on skilled immigrants. This serves as a deterrent. • In the policy-restricted migration regime, however, political-economy considerations of the native born in the host country are prominently at play. • Therefore, the fiscal burden of a generous welfare state falling on skilled native born, induces this group to endorse higher rates of skilled immigration • (The unskilled native born is in favor of maximum level anyhow).

22. Empirical Evidence: Countries • We confront the prediction of the model with a cross section data of source-host developed country pairs (similar notions of “skills”). • We decompose the data into two groups: • The first group contains only source-host pairs of countries which enable free mobility of labor among EU member states. • (Europe begun with labor mobility treaties in1951. EU prohibits any kind of discrimination between native born and immigrants, regarding labor market accessibility and welfare-state benefits eligibility.)

23. Empirical Evidence: Countries • These are 16 European countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Portugal, Spain, Sweden, U.K., Norway and Switzerland. • The second group includes only source-host pairs of countries, within which the source country residents cannot necessarily move into either of the host country without any restriction. • The host countries are the same 16 countries. • The source countries are 10 developed countries: U.S., Canada, Japan, Australia, New Zealand, Israel, Taiwan, Hong Kong, Korea and Singapore.

24. Identification • This decomposition to Free Migration and Policy –Restricted groups is the identification variable. • We assume that migration is free within the first group, and is effectively restricted by policy within the second group. • Thus, even if migration supply from the countries in the second group is different than that of the first group, still, their skill composition is determined solely by the migration policy of the host countries. • We further assume, plausibly, that the categorizing of both groups is exogenous to the skill composition of immigrants. • Hence we can identify the differential effect of the generosity of the welfare state on the skill composition of immigrants across the two groups.

25. Data • Immigration Data (Docquier and Marfouk (2006)): • The data contains bilateral stock of immigrants, based on census and register data, for the years 1990 and 2000. • Immigrants are at working age (25+), defined as foreign born, subdivided into three classes of education level: low-skilled (0-8 schooling years), medium-skilled (9-12 schooling years) and high-skilled (13+ schooling years). • Stock variables are more reliable than flow data, as flow data disregards return migration movements, which may distort the estimated effects. • Moreover, endogenous dynamic of equilibrium rates, are better captured by stocks than flows.

26. Data (continue) • Data for welfare-state benefits per capita is based on OECD's Analytical Database. • We use lagged average rates between 1974-1990. • Social benefits encompass all kinds of social public expenditures, in cash or in kind, including, for instance, old age transfers, incapacity related benefits, health care, unemployment compensations and so on. • The data is PPP-converted to 1990 U.S. dollars.

27. Econometric Approach: Difference in Differences • Assume that emigration stock rates, in 2000, are explained by the following equation: • We subtract the low-skilled equation from the high-skilled equation:

28. Discussion • The dependent variable, Δm, can be considered as a measure for the skill composition of immigrants. • The higher Δm is, the more upgraded is the skill composition the immigrants. • Hence, a positive estimation of a certain coefficient indicates a positive effect on the skill composition measure of the immigrants, and vice versa. • Note that the skill-differences model eliminates part of the error term, θs,h. • Any variable whose impact over immigration is skill invariant, is canceled out. • Additionally, including past migration stocks (in 1990) accounts for all time invariant effects.

29. Findings

30. Table 1 • Table 1 presents the main estimation results. Columns 1 and 2 report OLS regressions results; columns 3 and 4 report IV regression results, using the legal origin of the host countries as the instrumental variable. The difference between columns 1 and 3, on one hand, and columns 2 and 4, on the other hand, is in the variables of the vector X_{s,h}. Columns 1 and 3 contain only the migration stocks, skilled and unskilled, in the year 1990. Columns 2 and 4 include also the shares of the skilled and the unskilled labor (both native-born and migrants) in the host country in the year 1990.

31. First null hypothesis • The first null hypothesis, is that Δβ₂<0. Indeed, the coefficient is negative and significant. That is, the generosity of the welfare state adversely affects the skill composition of migrants in the free migration regime.

32. First null • The magnitude of the coefficient is even higher in the IV regressions than the OLS regressions. Whether we include the full set of the variables in X in the regressions (columns 2 and 4) or not (columns 1 and 3) does not seem to have much of an effect on on the magnitude of the coefficient.

33. The second null hypothesis • The second null hypothesis is that Δβ₃>0. Indeed, it is positive and significant in all four regressions. That is, the effect of the generosity of the welfare state on the skill composition of migrants is more pronounced in the policy-controlled migration regime. The magnitude of the coefficient is even higher in the IV regressions than the OLS regressions.

34. 2nd null • Again, whether we include the full set of the variables in X_{s,h} in the regressions (columns 2 and 4) or not (columns 1 and 3) does not seem to have much of an effect on on the magnitude of the coefficient.

35. Inertia? • The effect of skilled composition of migration in 1990 on the skill-composition of migration in 2000 is negative and significant in all four regressions, in the free migration regime. and the effect of high-skilled migration stock in 1990 on the skill-mix of migration in 2000 is positive. An interpretation of this result is that in the free-migration regime there is an inertia over time for each skilled group of migrants, in the free migration regime More unskilled migrants bring about further waves of unskilled migrants; and similarly, more skilled migrants bring about further waves of skilled migrants

36. Wage depression effect • We also find in the free migration regime that the host country share of skilled workers in 1990 has a negative and significant effect on the skill composition of migrants in 2000. An interpretation of this result is that high share of skilled labor in the host country depresses wages of skilled labor and deters skilled migrants.

37. Skilled labor in policy-controlled regime • Note that the effect of any control variable on the dependent variable in the policy-controlled regime is given by the sum of the coefficient of this control variable and the corresponding interaction term. With this in mind, we can see that the effect of the1990 stock of unskilled migrants on the skill composition of migrants in 2000 is positive in the policy-controlled regime, in all four regressions.

38. Fiscal burden effect? • The interpretation of this result , consistent with our model, is that having initially (e.g., in 1990) a large stock of unskilled migrants, which poses a fiscal burden on the welfare state, induces its voters to opt for more skilled migrants in order to ease the burden. The1990 stock of skilled migrants has a positive effect on the skill composition of migrants in 2000 in the policy-controlled regime in all four regressions.

39. Discussion of Results • The variable of interest is the interaction between the first group dummy variable and the benefits per capita. • Our hypothesis is that Δβ3>0, where Δβ3 is the coefficient of the variable of interest (benefits pc 1974-1990 (H) X R). • As can be seen, the coefficient is positive and significant at 5% in all of the regressions. • Specifically, it is different from the benefits' coefficient (benefits pc 1974-1990 (H)), which captures, as predicted, the adverse effect of the welfare state generosity on the skill composition of immigrants, in the "free migration" group.

40. ?Time Series Analysis • Over the past decade Spain has been the biggest importer of labor in the EU. Waves of migrants from outside of the EU: North Africa, South Africa, South America; and from Eastern Europe, from the inside of EU. By 2007, there were nearly 4.5 foreigners living legally in Spain. Legal migrants alone, including those who gained Spanish nationality, account for 12 percent of the population, up from 3 percent in 1998.

41. Appendix

42. Appendix

43. Appendix

44. Appendix