Immigrant Assimilation and Welfare Participation - CiteSeerX

1 downloads 0 Views 1MB Size Report
Lennart Flood, Bjorn Gustaffson, Dan-Olof Rooth, participants at the .... the new country (Borjas and Trejo 1991 and 1993; Borjas and Hilton 1996; Baker.
Immigrant Assimilation and Welfare Participation Do Immigrants Assimilate Into or Out of Welfare? Jorgen Hansen Magnus Lofstrom abstract This paper analyzes differences in welfare utilization between immigrants and natives in Sweden using a large panel data set for the years 1990 to 1996. We find that immigrants use welfare to a greater extent than natives and that differences cannot be explained by observable characteristics. Welfare participation decreases with time spent in Sweden. Refugees assimilate out of welfare at a faster rate than nonrefugee immigrants, but neither group is predicted to reach parity with natives. Increases in unemployment and immigration, as well as the change in the composition of immigrants, contributed to the increase in welfare utilization in Sweden.

I. Introduction There has been a dramatic increase in the expenditure on social assistance (SA) in Sweden since the early 1980s.1 According to the National Board of Health and Welfare, total real expenditures between 1983 and 1996 increased form 4.4 billion Swedish kronor (SEK) to 11.9 billion SEK. As we will show, immigration Jorgen Hansen is an assistant professor of economics at Concordia University. Magnus Lofstrom is an assistant professor of economics and political economy at the University of Texas at Dallas. The authors would like to thank two anonymous referees, Thomas Bauer, Anders Bjo¨;rklund, Don DeVoretz, Lennart Flood, Bjorn Gustaffson, Dan-Olof Rooth, participants at the Canadian Economic Association’s annual meeting 2000, the CEPR/TSER workshop at Bar-Ilan University, the Canadian International Labour Network’s conference 2000, the European Economic Association’s annual meeting 2000, and seminar participants at Gothenburg University, Lund University, SOFI, Simon Fraser University for helpful comments. Financial support from the European Commission (grant SOE2-CT97-03052) and the Swedish Council for Social Research is gratefully acknowledged. The data used in this paper are provided by Statistics Sweden. For information about accessing these data and user restrictions, contact Statistics Sweden at ⬍[email protected]⬎. [Submitted February 2000; accepted August 2001] ISSN 022-166X  2003 by the Board of Regents of the University of Wisconsin System 1. The term social assistance is used synonymously with public assistance and welfare in this paper. THE JOURNAL OF HUMAN RESOURCES

• XXXVIII • 1

Hansen and Lofstrom is central to the increase in welfare costs. For example, by the mid-1990s, expenditures on social assistance for immigrants equaled expenditures for natives, even though immigrants represented only 10–11 percent of the total population. Immigrants are also greatly overrepresented in the welfare population in the United States and Germany (see, for example, Bean, Van Hook, and Glick 1997; Borjas and Trejo 1991; Riphahn 1998). It is quite clear that the concern about immigrant welfare usage is not specific to Sweden, but is also central to the immigration debates in other countries. For example, concerns about the rising welfare costs in the United States led the Congress to pass The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, which denies noncitizens who arrived after 1996 the right to receive most types of public assistance. The concerns about immigrant overutilization of public assistance are obvious in Germany as well. Immigrants without permanent residency in Germany may lose the right to stay in the country or may be denied residency extensions if they rely on social assistance. In Sweden, the main reason for the growth in welfare expenditure is, not surprisingly, an increase in the utilization of SA. In 1983, 7.8 percent of all households received SA. This figure had increased to 10.7 percent by 1996. This represents a quite remarkable increase of more than 37 percent. The rise is especially strong among the immigrant population. In particular, households from refugee countries are overrepresented among the households that receive SA. During the same period, the proportion of immigrants in the Swedish population increased significantly, from 7.6 percent in 1983 to 10.8 percent in 1996. The overrepresentation of immigrants in the welfare population in combination with the increase in immigration can also explain part of the rise in welfare costs. One important reason for the increase in welfare participation, and the consequent growth in expenditures, is the growth in the unemployment rate in the 1990s in Sweden, which grew from 1.7 percent in 1990 to slightly more than 8 percent in 1996. For immigrants, the labor market deteriorated even more. In 1990, approximately 4 percent of the immigrant population was unemployed. This had increased to 23 percent by 1996. The increase in welfare expenditures in Sweden in the 1990s can partly be explained by the large inflow of immigrants who arrived during this period who were not eligible for unemployment insurance and therefore had to rely on social assistance for their subsistence. In this paper we try to answer two questions central to the debate of immigrant welfare utilization by using a unique large Swedish panel data set, Longitudinal Individual Data (LINDA). The first question we ask is whether the overrepresentation of immigrants among public assistance receiving households is due to differences in observable characteristics, such as age, family composition and the level of education, or if it is due to unobservable heterogeneity.2 For example, if the higher immigrant welfare-participation rates are partially caused by differences in educational attainment, then policies directed towards increasing the educational level among immigrants may reduce the fiscal burden of social assistance in the future. However, 2. Examples of unobserved heterogeneity leading to differences in welfare-participation rates between immigrants and natives include behavioral differences (for example, dissimilarity in the reservation wage) and differences in the labor markets faced by the two groups (possibly due to discrimination).

75

76

The Journal of Human Resources if the observed differences in SA utilization depend on differences in preferences, then these types of policies would have a very limited effect, if any effect at all. The second question deals with the important long term effects of immigration on welfare expenditures. Are immigrants likely to assimilate ‘‘into’’ or ‘‘out of’’ welfare dependency, that is, do immigrants increase or decrease their participation in social assistance with time spent in the host country? If immigrants’ participation rates will change with time spent in the new country, the initial, upon arrival, costs of welfare should not be used to infer lifetime welfare costs of immigrants. For example, if immigrants assimilate out of welfare, initial welfare costs will overstate the long-run social assistance expenditures incurred by the immigrant population. Previous studies of immigrant welfare dependency in the economics literature generally find that immigrants are on average more likely to receive welfare than native-born individuals. However, differences in observable socioeconomic characteristics explain greater participation rates among immigrants than natives in the United States, Australia, and Germany (Blau 1984; Maani 1993; and Riphahn 1998, respectively). Previous studies have also found that time spent in the host country affect participation rates of immigrants. For example, in the United States, Canada, and Germany immigrants appear to increase welfare utilization with time spent in the new country (Borjas and Trejo 1991 and 1993; Borjas and Hilton 1996; Baker and Benjamin 1995; Riphahn 1998). In other words, the existing literature suggests that immigrants assimilate into welfare dependency.3 The longitudinal data used in this paper provide a clear advantage over the data used in previous studies. Most of the above-mentioned papers utilized cross-sectional data, except for two, Borjas and Hilton (1996) and Riphahn (1998), who used panel data for the United States and Germany, respectively. Borjas and Hilton estimate linear probability models with fixed effects using a relatively short panel based on survey information. Riphahn controls for both unobserved heterogeneity and attrition. However, the data are not a representative sample of immigrants in Germany. Unfortunately, it only includes a small sample of guest workers and no refugees. In this paper we take advantage of a recently collected large representative panel data set containing information on more than 300,000 individuals annually for the period 1990–96. The data are collected from administrative records implying essentially no attrition and less measurement error than what would be expected in survey data. Another significant advantage is that the longitudinal data set allows us to use methods controlling for unobserved heterogeneity. It is essential to control for unobserved effects since many of the factors determining whether a household receives welfare or not, including the reservation wage and stigma effects from participating in welfare programs, are unobserved by the econometrician. The key findings in this paper are that immigrants are more likely to participate in the social assistance program than natives even when controlling for observable characteristics, and that immigrants assimilate out of welfare with time spent in the new country. The former of these findings contradicts what has generally been found 3. Regarding Baker and Benjamin’s (1995) finding that immigrants appear to assimilate into welfare in Canada, Crossley, McDonald, and Worswick (2001) show that these results are sensitive to years included in the sample. As reported below, we do not find the same sensitivity of the results, with respect to survey years, in the Swedish data used here.

Hansen and Lofstrom previously in the literature. The self-selection of immigrants coming to a relatively generous welfare state is likely to be one of the reasons for this result. We also find that immigrants reduce welfare-participation rates with time spent in the new country. Although refugees display substantially higher participation rates upon arrival compared to nonrefugee immigrants, they assimilate out of welfare much more rapidly than their nonrefugee counterparts. We also find that roughly 50 percent of the observed increase in welfare utilization in Sweden in the 1990s can be attributed to the increases in both unemployment and immigration. The result, that immigrants assimilate out of welfare, appears to contradict previous findings in regards to the assimilation of immigrants’ welfare utilization. However, even after 20 years in the host country we find that both refugee and nonrefugee immigrants show significantly higher social-assistance-participation rates than statistically similar natives—by between 8 and 10 percentage points. These numbers are quite close to the findings of Borjas and Hilton (1996) and Baker and Benjamin (1995). These results suggest that immigrants in a relatively generous welfare state, like Sweden, display similar welfare participation behavior as immigrants in less generous welfare states, like the United States, relative to natives after having spent some time in the new host country. The paper is organized in the following way. In Sections II and III we give background information about immigration into Sweden and the social assistance program. Section IV describes the data and variables while Section V depicts trends and differences, between immigrants and natives, in welfare participation. In Section VI we test whether differences in welfare utilization can be explained by differences in socioeconomic characteristics. Assimilation issues are also analyzed in this section. Finally, we conclude in Section VII.

II. Historical Background—Immigration into Sweden The inflow of immigrants to Sweden has undergone a number of changes during the last six decades. Figure 1 shows annual immigration to Sweden from 1940 to 1998, both in terms of actual immigrant inflow and inflow expressed as a proportion of the population in the corresponding year. Overall, annual immigration has amounted to about 0.4 percent of the population, but it was notably higher during 1990s. The proportional inflow is slightly higher than the U.S. experience of the 1990s, but quite similar to the experiences of Canada and Australia. Naturally, the large inflow of immigrants has also changed the composition of the population in Sweden. In 1991, almost 10 percent of the Swedish population was born outside Sweden. This is a relatively high figure compared to the United States (7.9 percent), but is lower than the proportion of foreign-born in Canada (16.1 percent) and Australia (22.3 percent) (SOPEMI 1998). The reasons people immigrate to Sweden have changed substantially during the post-war period. In principle, we can distinguish between three categories of immigrants, based on the reasons for immigration: economic migrants (due to the recruitment of labor, for example), tied movers (based on family ties) and refugees. In the late 1940s, a large fraction of the immigrants arrived in Sweden as refugees. In the period from 1950 to 1970, however, most immigrants were recruited by the Swedish

77

78

The Journal of Human Resources

Figure 1 Immigration into Sweden, Annual Inflow and Proportion of Population, 1940–98 Source: Statistics Sweden, Historical Population Development Table, 1999.

industry or arrived because of family ties. Since 1970, the proportion of immigrants arriving as refugees has increased significantly, from less than 10 percent of the immigrant inflow in 1970 to about 70 percent in the early 1990s. In 1994, this proportion dropped from 70 percent to about 50 percent, mostly due to improved conditions in the Balkan countries.

III. Social Assistance in Sweden The Swedish welfare system is well known internationally for the degree of income security that it provides for its citizens. Recently, this generous system has been the target of a number of reforms, mainly due to the recession that hit Sweden, and many other countries, in the early 1990s. As an ultimate safety net, people in Sweden are covered by social assistance (SA). In order to be eligible for SA, all other welfare programs, such as unemployment compensation, housing allowance (bostadsbidrag), child allowance (barnbidrag), maintenance allowance (underha˚llsbidrag), and various pensions, must be exhausted first. The benefit levels vary, both across family types and regions, but are intended to cover expenses essential for a ‘‘decent’’ living.4 To be eligible for SA benefits, a family must have income and assets below certain specified benefits levels (known as norms). Unit 1998, the norms were determined in each of the 288 municipalities 4. For example, in 1994, for a single individual, the norm varied between 1,613 SEK and 4,107 SEK per month across the municipalities. The benefits are meant to cover expenses for so-called necessary consumption, such as food, basic clothing, leisure, health, newspapers, telephone, and fees for TV.

Hansen and Lofstrom

Figure 2 Real Expenditures on Social Assistance in Sweden, 1983–98, in 1996 SEK Source: National Board of Health and Welfare, Social Assistance, Table 1, 1998.

in Sweden and served as guidelines for the social worker, who decided the actual size of the benefits. SA benefits were paid according to a schedule that set a guaranteed amount for a family of a given size. These benefits were reduced at a 100 percent reduction rate as the family’s income rose. Expenditures on social assistance have increased quite substantially in Sweden during the last 15 years. In Figure 2, we show total expenditures on SA from 1983 to 1998. During the 1980s, total expenditures on welfare increased by more than 30 percent. However, the most rapid increase in expenditures took place in the 1990s. Between 1990 and 1996, welfare costs increased by roughly 100 percent. In 1996, 11.9 billion SEK was spent on social assistance, or approximately 2 percent of all government expenditures. Figure 2 also divides welfare expenditures separately for native Swedes and immigrants. Clearly, welfare expenditures increased at a much faster pace for immigrants than for native-born Swedes between 1983 and 1998. Throughout the 1990s, immigrants and natives accounted for about the same amount of welfare expenditures. This is quite remarkable given that immigrants represent a 10 percent minority of the population in Sweden.

IV. Data A. Description of the Data and Sampling Procedures The data used in this paper are taken from a recently created Swedish longitudinal data set, Longitudinal Individual Data (LINDA). LINDA is a register-based data set

79

80

The Journal of Human Resources and it consists of a large panel of individuals and their household members, which are representative for the population from 1960 to 1996. LINDA is a joint endeavor between the Department of Economics at Uppsala University, The National Social Insurance Board (RFV), Statistics Sweden, and the Ministries of Finance and Labor. The main administrator of the data set is Statistics Sweden. For a more detailed description of the data used here, including the sampling structure, see Edin and Fredriksson (2000). LINDA contains a 3 percent representative random sample of the Swedish population, corresponding to approximately 300,000 individuals for the period studied here. The sampled population consists of all individuals, including children and elderly persons, who lived in Sweden during a particular year. The sampling procedure used in constructing the panel data set ensures that each cross-section is representative for the population in each year. Attached to LINDA is a nonoverlapping representative random sample of immigrants containing the same variables, and created in the same fashion, as the general sample. The immigrant sample consists of 20 percent of all individuals born abroad. We merged this sample with the general population sample. This generated a sample of the Swedish population in which immigrants are overrepresented. The overrepresentation can be adjusted for by using appropriate methods. The sample used in this study consists of information from LINDA for the year 1990–96.5 We excluded all households in which the sampled person is younger than 18 years or older than 65 years. Because welfare participation in Sweden is based on household characteristics and household income, the appropriate unit of observation is the household. A common approach in the literature is to let the household be represented by the household head, meaning that the characteristics of the household coincide with those of the household head. However, we are not able to identify the head of the household in LINDA and instead we let the household be represented by the sampled individual. This means that the value of different observable characteristics, such as age and education, in the subsequent analysis refer to the person in the household that was originally sampled. Furthermore, a household is defined as an immigrant household if the sampled person was born abroad, and as a refugee household if he was born in a refugee country, as defined by the Swedish Immigration Board, or in a sub-Saharan country.6 If the person representing the household is an immigrant or a refugee, we have information about the year of arrival in Sweden.7 We also estimated models with a slightly different definition of an immigrant household in which the household was defined to be an immigrant household if any household member was an immigrant. This appears to have no impact on the analysis presented below. 5. We lack information about welfare use prior to 1990. 6. The countries defined by the Swedish Immigration Board as refugee countries: Ethiopia, Afghanistan, Bulgaria, Bangladesh, Bosnia, Chile, Sri Lanka, Cuba, Iraq, Iran, India, Yugoslavia, China, Croatia, Lebanon, Moldavia, Peru, Pakistan, Poland, Russia, Soviet Union, Romania, Somalia, Syria, Togo, Turkey, Ukraine, Uganda, and Vietnam. 7. All immigrant households included in LINDA, whether defined as refugees or not, have obtained residence permits. This means, for instance, that asylum seekers who have not yet obtained a residence permit are not included in our sample. Furthermore, the data do not allow us to identify the exact year of arrival for immigrants who arrived in 1968 or earlier.

40.46 N/A 31.34% 58.22% 10.43% 51.98% 0.57 5,979 6.01% 1,009,780

39.91 12.52 32.40% 57.42% 10.19% 53.34% 0.59 5,961 6.04% 1,647,390

Nonwelfare Recipient

Source: Longitudinal Individual Data for Sweden (LINDA), 1990–96.

Age Years since migration Education Elementary school High school College Single Number of children SA norm Unemployment rate Sample size

Total

Natives

43.54% 54.65% 1.81% 89.47% 0.52 4,859 6.40% 44,372

32.31 N/A

Welfare Recipient

36.29% 51.99% 11.73% 51.74% 0.71 6,137 6.04% 304,239

39.31 15.87

47.31% 48.56% 4.13% 80.04% 0.70 5,341 6.22% 33,125

35.08 13.34

Welfare Recipient

38.61% 47.93% 13.46% 41.79% 0.99 6,742 6.24% 183,162

35.91 9.63

45.97% 44.57% 9.46% 55.67% 1.01 6,353 6.81% 72,712

34.54 5.37

Welfare Recipient

Refugee Country Nonwelfare Recipient

Immigrants Nonrefugee Country Nonwelfare Recipient

Table 1 Mean Observable Characteristics by Immigrant Status and Welfare Receipt

Hansen and Lofstrom 81

82

The Journal of Human Resources B. Variable Definition and Descriptive Statistics To answer the questions regarding immigrant assimilation and welfare participation raised in the introduction, we estimate models where the dependent variable is a binary variable equal to one if the household received SA for at least one month during the year, and equal to zero otherwise. The Swedish municipalities provided data on social assistance benefit levels. We were able to assign a social assistance norm, which determines the benefit level, to each household in the sample in 1994 and in 1996. The municipality in which the household resides, as well as the family composition, such as marital status, ages, and number of children, determines the norms. Unfortunately, we have not been able to obtain similar information for the other years. For that reason, we assign the 1994 norms to all years prior to 1995 and the 1996 norms to the years 1995 and 1996.8 By including the norms into the models, we can obtain estimates of the effects of higher benefit levels on public assistance utilization. To control for local variation in unemployment rates, we include county unemployment rates, obtained from Statistics Sweden’s labor force surveys. These are assigned to each household in each year based on the household’s region of residence. In Table 1, we present average characteristics for the household by welfare recipiency for the period 1990 to 1996. In general, we observe that households on welfare are younger, less educated and, to a larger degree, single, as compared to households not on public assistance.9 For immigrants, we observe that SA recipients have on average been in the country for a shorter period than those households off SA. Interestingly, refugee households have on average higher post-secondary education compared to native Swedish and nonrefugee immigrant households. Moreover, the fraction of college-educated households receiving SA is substantially larger among refugees than among the other two groups.

V. Trends and Differences in Welfare Participation As Figure 2 shows, real expenditures on welfare increased substantially in Sweden during the 1990s. During this period, there was a substantial increase in the number of households receiving SA. The National Board of Health and Welfare reports that 7.9 percent of all households in Sweden in 1990 received social assistance. By 1996, the participation rate had increased to 10.7 percent. This represents an increase by 35 percent. Furthermore, the average monthly amount received did not change much and increased by slightly less than 5 percent. This suggests that the increase in expenditures is not due to an increase in the generosity of the welfare system. 8. To ensure that the findings reported below are not sensitive to the limited availability of municipal welfare benefits rules, we reestimated the models shown in Tables 3 and 4 using a sample restricted to the two years the norm is available for, 1994 and 1996. The results from the restricted sample, available upon request from the authors, are very similar to the ones obtained utilizing the full sample and the conclusions reported in this paper remain the same. 9. Since participation in SA is based on household characteristics, the entries in Table 1 refer to those of the household representative.

Hansen and Lofstrom Table 2 Welfare Participation by Immigrant Status and Arrival Cohort, 1990 and 1996 Sample Size

Natives Immigrants Nonrefugee country All cohorts Arrival cohort 1968–75 1976–80 1981–85 1986–90 1991–96 Refugee country All cohorts Arrival cohort 1968–75 1976–80 1981–85 1986–90 1991–96

Welfare Participation Rates Difference 1990–96

1990

1996

1990

1996

147,319

151,096

3.18%

4.68%

1.50

30,419

53,648

8.29%

10.15%

1.86

19,011 6,200 3,405 1,803

22,557 10,595 6,005 8,422 6,069

7.01% 8.90% 11.34% 13.92%

6.64% 10.01% 10.99% 13.58% 17.85%

⫺0.38 1.10 ⫺0.35 ⫺0.34

13,055

53,095

16.49%

31.99%

15.50

2,536 4,121 4,094 2,304

3,330 6,458 6,818 16,009 20,480

7.85% 11.65% 16.17% 35.24%

9.10% 11.66% 16.46% 30.06% 48.80%

1.25 0.01 0.29 ⫺5.18

Source: Longitudinal Individual Data for Sweden (LINDA), 1990 and 1996.

Welfare-participation rates have been shown to be different for immigrants and natives in many countries (see, for example, Borjas and Trejo 1991; Maani 1993; Riphahn 1998). Table 2 shows that this is true for Sweden as well.10 Immigrants from both refugee and nonrefugee countries are more likely to receive social assistance than native-born Swedes. Furthermore, refugees participate to a greater extent in the social assistance program than nonrefugees. The table also shows that the increase in the welfare participation rate over the period 1990–96 was greater for immigrants than it was for natives. Table 2 also shows substantial differences in welfare participation rates across arrival cohorts. The table suggests that immigrants reduce their social assistance reliance with time spent in Sweden. However, the observed assimilation out of welfare could be due to a decline in the skill level of later cohorts, so-called negative cohort effects. In a series of articles, Borjas (1985 and 1994, for example) has shown 10. The immigrant sample used in Table 2 corresponds to 20 percent of the immigrant population for each year. This combined with the dramatic increase in immigration into Sweden during the period analyzed explains the large difference in number of immigrant observations in 1990 and 1996.

83

84

The Journal of Human Resources that such cohort effects can cause overestimated assimilation rates of immigrants. Notwithstanding, Table 2 shows that the welfare utilization rate decreases relative to natives, for any given arrival cohort, over the period studied here. The native welfare participation rate increased by 1.5 percentage points from 1990 to 1996. No immigrant arrival cohort displays a greater increase in social assistance utilization rate than 1.25 percentage points over the same period. This indicates that negative cohort effects are not the source of the observed assimilation pattern. Furthermore, this suggests that the increase in immigrant welfare participation does not stem from an increase in immigrants’ propensities to participate in the social assistance program, but is instead, at least partially, due to the substantial increase in immigration that Sweden experienced in the 1990s. We will explore the effect of the rise in immigration on the observed increase in welfare participation below. Immigrant assimilation out of welfare becomes quite clear when the difference in participation rates between immigrants and natives are shown by years since migration, as in Figure 3. Refugees in particular seem to assimilate quickly. Their initial welfare-participation rates are between 40 and 50 percentage points higher than natives. After 10 or 11 years in Sweden, the difference drops to about 10 percentage points. Nonrefugee immigrants also appear to assimilate out of welfare. It should be noted that these comparisons are flawed in several ways. For example, the average age of natives is held roughly constant while the average age of the immigrant population increases with years since migration. Controls for cohort effects are also important to incorporate. Differences between immigrants and natives, or changes in

Figure 3 Observed Differences in Welfare Participation, Native-Born Swedes and Immigrants, by Years Since Migration Source: Longitudinal Individual Data for Sweden (LINDA), 1990 to 1996.

Hansen and Lofstrom differences over time, in socioeconomic and geographic characteristics may also partly explain the pattern. To accurately analyze assimilation rates, an empirical model needs to be estimated. We now turn our attention to such a model.

VI. Empirical Specification and Results To analyze welfare utilization we need insight into why households participate in social assistance. Clearly, the labor market conditions household members face will affect the probability that the household will end up on welfare. In other words, the factors we believe will affect employment probabilities also need to be incorporated into our welfare utilization models. It is also quite plausible that immigrants and natives do not face the same labor market conditions. For example, human capital obtained abroad may be viewed differently than human capital acquired in the new country. Indication of this was found in Betts and Lofstrom (2000) for the United States. Another possibility is discrimination against immigrants. It is therefore important to allow the employment factors to affect welfare participation probabilities differently for immigrants and natives. Individuals’ preferences and tastes for leisure will also affect the probability of being on welfare through differences in the reservation wages. The so-called stigma effect of receiving welfare payments also depends on individuals’ preferences. The individual differences in reservation wages and the potential stigma of being on welfare are inherently unobservable but can be controlled for by using estimation methods that account for differences across individuals, including random or fixed effects. A. Do Differences in Observable Characteristics Explain Differences in Welfare Participation? In the introduction we asked whether the overrepresentation of immigrants among welfare-recipient households is due to differences in observable characteristics, to behavioral differences, or to differences in labor market conditions. To answer this, we formulate random-effects probit models of welfare participation. To be specific, the estimated models can be described as follows. Let (1) y*it ⫽ X itβ ⫹ ε it

∀ i ⫽ 1,2, . . . , n and t ⫽ 1,2, . . . , Ti

where (2) ε it ⫽ µ i ⫹ vit y*it is a latent variable representing preferences for welfare utilization of household i at time t. X is a vector of socioeconomic and geographic characteristics, including age, educational attainment, marital status, and the number of children. In addition, it contains information about the municipal social assistance norm and the county unemployment rate. The unobserved household specific effect, assumed to be time invariant, is represented by µ i and vit and is a white-noise error term. We assume that these unobserved stochastic terms have the following properties: (3) µ i , vit ⬃ N(0, Ω)

85

86

The Journal of Human Resources with (4) Ω ⫽

冤 0 1冥 σ 2µ

0

This implies that (5) Var(εit ) ⫽ σµ2 ⫹ 1 and (6) Corr(εit , εis ) ⫽ ρ ⫽

σ 2µ σ 2µ ⫹ 1

We do not observe y*, but we assume that we can observe the sign of it, and based on that we can formulate the following decision rule: (7) yit ⫽ 1 if y*it ⬎ 0, and 0 otherwise where yit ⫽ 1 represents utilization of SA. The set of control variables is very similar to what has been used in previous studies for other countries. The vector X also includes three not mutually exclusive nativity indicator variables. The following three immigrant dummies are defined: Immigrant (equal to one for all immigrants), Nordic Immigrant (equal to one for all immigrants from the Nordic countries) and Refugee Immigrant (equal to one for all immigrants from the previously defined refugee countries). This implies that the coefficient on the Immigrant dummy captures differences between natives and all immigrants while the coefficient on the Refugee Immigrant dummy captures differences between nonrefugee immigrants and immigrants from refugee countries. Given this definition of the nativity variables, the disparity between immigrants from refugee countries and natives is the sum of the two estimated coefficients for immigrants and refugees. Similarly, the difference in probability of welfare participation between Nordic immigrants and natives is the sum of the Immigrant and Nordic Immigrant estimated coefficients. If higher welfare-participation rates among immigrants are simply due to differences in observable characteristics between natives and immigrants, the estimated relevant coefficients on the nativity variables should not be significantly different from zero when these controls are included in the model. The results and marginal effects calculated at the mean of the observables are presented in Table 3. All immigrants appear to be more likely to participate in the social assistance program even after observable characteristics are controlled for. Model 1 shows the differences in the probability of receiving welfare between the three immigrant categories and natives, adjusting for the national trend over the period. The overall difference between nonrefugee immigrants and natives is 6.9 percentage points. The probability that a Nordic immigrant household participates in the social assistance program appears to be slightly less than a nonrefugee immigrant’s household. A refugee household is substantially more likely to be on welfare than a native household; the difference is about 18.6 percentage points. The results for Model 2 indicate that differences in observable characteristics ex-

Hansen and Lofstrom plain very little of the differences in welfare participation between natives and immigrants. Immigrants from nonrefugee countries are about 6.6 percentage points more likely to receive welfare compared to statistically similar natives. Only 5 percent of the total observed difference is due to differences in observable characteristics. The estimated difference for a refugee household drops by 1.6 percentage points when the observable socioeconomic variables are included in the model. A very small proportion, slightly less than 9 percent, of the higher observed-participation rates of refugees can be explained by differences in age, education, household composition, and geographic location. Table 2 shows that it is important to allow for differences in welfare utilization between arrival cohorts. To ensure that changes in the composition of immigrants over the period do not explain the large differences even after observables have been controlled for, we reestimated Model 2 in Table 3 with arrival cohort dummies. The results are shown in Table 3 as Model 3. The differences between immigrants and natives remain for all arrival cohorts. It is quite clear that differences in welfare participation between immigrants and natives are not due to differences in socioeconomic characteristics. Our findings that differences in observable characteristics between immigrants and natives do not explain the higher welfare-participation rates of immigrants differ from the findings for the United States, Canada, Australia, and Germany. The analysis performed here does not tell us whether the differences between immigrants and natives are due to behavioral differences or differences in the labor market opportunities between the two groups. Two possible reasons may explain the differences in findings across countries. First, the Swedish labor market may view immigrants and their observable characteristics differently from the previously mentioned countries’ labor markets. Another possible reason is that immigrants do not randomly choose the destination country. Instead, they may self-select according to preferences, relative expected earnings, and the generosity of the welfare system in the new host country. If so, immigrants may select Sweden partially due to its fairly generous welfare system. It should be noted that our analysis does not allow us to determine the extent to which each possible reason contributes to the discrepancy in findings across countries. B. Assimilation Into or Out of Welfare? Our results indicate differences in the welfare utilization between immigrants and natives, even after controlling for observable characteristics. Our next step is to analyze how immigrants’ welfare participation behavior changes with time spent in the host country. In doing so, it is important to control for unobserved effects. We do this by estimating a random effects probit model, similar to the one discussed above. To allow for different effects of observable variables for native Swedish households and immigrant households (both refugee and nonrefugee), we specify an interacted model, where all the observable variables are interacted with indicator variables for immigrants and refugees. However, we impose equal year effects across the three groups in order to identify the assimilation effects. This model is not fully interacted in the sense that we do not interact all the variables with the Nordic immigrant variable. The reason for this is that estimated models including these interactions

87

Arrival 1968–75

Arrival 1976–80

Arrival 1981–5

Arrival 1986–90

Refugee immigrant

Nordic immigrant

Immigrant

Variable 0.9734 (0.0327) ⫺0.1059 (0.0395) 1.6376 (0.0341)

Estimated Coefficient

Model 1

0.1165

⫺0.0075

0.0692

Marginal Effect

Table 3 Random Effects Probit Models of Welfare Participation

0.9758 (0.0328) ⫺0.1390 (0.0404) 1.5519 (0.0344)

Estimated Coefficient

Model 2

0.1044

⫺0.0093

0.0656

Marginal Effect

1.5913 (0.0644) ⫺0.0500 (0.0395) 1.9987 (0.0687) ⫺0.4497 (0.0743) ⫺0.6690 (0.0808) ⫺0.7504 (0.0736) ⫺1.0144 (0.0691)

Estimated Coefficient

Model 3

⫺0.0664

⫺0.0491

⫺0.0438

⫺0.0294

0.1308

⫺0.0033

0.1042

Marginal Effect

88 The Journal of Human Resources

Yes No No No 0.7723 405,426 ⫺77,411

Yes Yes Yes Yes 0.7473 405,426 ⫺74,133

Yes Yes Yes Yes 0.7266 405,426 ⫺72,570

⫺0.6586 (0.0841) ⫺1.3864 (0.0991) ⫺1.6859 (0.0937) ⫺1.8215 (0.1104) ⫺0.1192

⫺0.1104

⫺0.0907

⫺0.0431

Note: Standard errors appear in parentheses. Data used are from LINDA, 1990–96. Estimates are based on a 25 percent random subsample of the full LINDA sample. P-value of likelihood ratio test for no within-group correlation is less than 0.0001 for all models.

Included controls Time fixed effects Individual characteristics Regional characteristics County fixed effects Within-group correlation Number of observations Log likelihood

Arrival 1968–75*refugee

Arrival 1976–80*refugee

Arrival 1981–5*refugee

Arrival 1986–90*refugee

Hansen and Lofstrom 89

90

The Journal of Human Resources indicated no differences between Nordic immigrants and other nonrefugee immigrants. Hence, we limit interactions to the nativity variables for immigrants and refugee immigrants. However, we allow for different mean effects among Nordic immigrants and other immigrants by including the Nordic immigrant dummy variable. We incorporate information about years since migration to assess whether immigrants assimilate into or out of welfare dependency. We also know from Section II that the composition of immigrants has changed substantially over the last 30 years, and it is therefore important to control for differences in arrival cohorts. The results from the model described above can be found in Table 4. In the first two columns, we present the results for native Swedes, while Columns 3 and 4 present the differences between immigrants from nonrefugee countries and natives. Finally, the last two columns show the estimated differences between refugees and nonrefugees. Note that to obtain marginal effects for nonrefugee immigrants, it is necessary to add the estimated marginal effects for natives, shown in Column 2, to the effect specific to immigrants, as shown in Column 4. Similarly, to get the marginal effects for immigrants from refugee countries, it is necessary to add the shown effects in Columns 2, 4, and 6.11 We focus our discussion on the main parameters affecting assimilation into or out of welfare dependency. In passing, we note that the effects of education and family composition have the expected signs. An interesting result is that the effect of the social assistance norm is positive and significant, implying that, everything else held constant, welfare-participation rates are higher in municipalities where the norm is relatively high. Immigrants are less sensitive to variation in the benefit levels than natives. In fact, it appears that the social assistance norm has no significant effect on the welfare utilization of both refugee and nonrefugee immigrants. Furthermore, our results indicate that welfare participation is weakly, in terms of statistical significance, positively associated with changes in the local unemployment rate for all three groups. One of the main purposes of this paper is to study whether immigrants assimilate into or out of welfare dependency. The estimated coefficients for the age and years since migration variables can be used to trace out differences in the probability of welfare recipiency between immigrants and natives by years since migration. It is also necessary to choose values for the other characteristics included in the matrix X. We have chosen 1996 as our baseline year and a household located in one of the major cities. The sample person is a high school graduate who is married with one child. These choices represent the mode for these variables. The predicted differences in welfare participation between immigrants and natives, calculated for the representative household described above, are shown in Figures 4, 5, and 6. The figures indicate that immigrants assimilate out of welfare with time spent in Sweden. Figure 4 shows that both nonrefugee and refugee households reduce their utilization of welfare, relative to natives with years since migration. In particular, the assimilation process is substantially faster for refugee households. Despite much higher initial participation rates, after 14–15 years refugee households display very similar welfare-participation rates as nonrefugee households. However, it should be 11. The reason for defining the immigrant status variables in this way is so that we can easily test for differences across immigrant groups.

Hansen and Lofstrom noted that neither group appears to reach the participation rates of native households within a 20-year period in Sweden.12 In Section II we showed that the composition of immigrants has changed substantially over the last 30 years. It is therefore important to control for arrival cohorts. Figures 5 and 6 show differences in welfare participation between natives and immigrants by arrival cohorts. The trends in Figure 5 indicate quite small differences in the predicted assimilation patterns among arrival cohorts for immigrants from nonrefugee countries. Figure 6 shows a different pattern for refugee households. Our results indicate that the latest arrival cohorts have higher participation rates than all the earlier arrival cohorts of refugee immigrants. There is a concern that the findings above may be due to nonrandom selective return migration. That is, if those immigrants who are less likely to participate in welfare programs are also the ones who are most likely to stay, our estimates would overstate the decline in welfare participation with time spent in Sweden. However, Edin, LaLonde, and Aslund (2000) also use data extracted from LINDA and find that the immigrants who stay in Sweden are more likely to receive welfare than the ones who leave. Furthermore, they find that return migration among refugees is low and that earnings assimilation of this group is not sensitive to emigration.13 Based on this, we do not think that our findings are driven by selective return migration. To ensure that our results are not specific to the random effects assumptions, we also estimated a fixed effects logit model. We were concerned about the assumption made in the random effects model that the household specific error terms are uncorrelated with the observed independent variables. The assimilation results derived from the fixed effects logit model are remarkably similar to the random effects probit results. For that reason and because the effects of the time invariant variables are of interest in a study like this, we only report the above-discussed random effects probit results. To summarize our results concerning immigrant assimilation in welfare utilization, we find strong support for the hypothesis that immigrant households in Sweden tend to assimilate out of rather than into welfare dependency. This result differs from the results found in the existing literature on the welfare assimilation of immigrants. However, even after 20 years in the host country we find that both refugee and nonrefugee immigrants show significantly higher social-assistance-participation rates than statistically similar natives by between 8 and 10 percentage points. Remarkably, our findings are very similar to those of Borjas and Hilton (1996), using U.S. data, and of Baker and Benjamin (1995), using Canadian data, whose estimates 12. The figures show differences in welfare-participation rates, assuming that age at migration is equal to 18. To assess the impact of age at migration, we recreated Figure 4, setting age at migration to 30. Our findings suggest that age at migration has a very minor impact on nonrefugee immigrants’ participation behavior. Refugee immigrants, however, display slightly slower assimilation rates the older the immigrant is upon arrival in Sweden. We also estimated models including explicit controls for age at migration and found no differences to the findings reported in the paper. 13. Edin, LaLonde, and Aslund (2000) also find that if an immigrant is to leave Sweden, departure is most likely to take place within the first few years after arrival. Motivated by this finding and the policy in Sweden of giving most refugee immigrants welfare upon arrival in Sweden, we reestimated the model in Table 4 excluding all immigrants who had been in Sweden for fewer than three years. The results show that assimilation patterns are not sensitive to this sample selection.

91

Number of children

Single

College

High school

Age squared/100

Age

Constant

Variable ⫺4.3198 (0.2416) 0.0320 (0.0071) ⫺0.0833 (0.0093) ⫺0.2287 (0.0256) ⫺1.2294 (0.0696) 1.2829 (0.0655) ⫺0.0360 (0.0519)

Estimated Coefficient

⫺0.0024

0.0863

⫺0.0827

⫺0.0154

⫺0.0056

0.0022

Marginal Effect

Table 4 Random Effects Probit Model of Welfare Participation

2.2770 (0.3268) ⫺0.0032 (0.0128) 0.0196 (0.0165) 0.0296 (0.0440) 0.1661 (0.1006) ⫺0.4816 (0.0973) 0.1651 (0.0742)

Estimated Coefficient

0.0111

⫺0.0324

0.0112

0.0020

0.0013

⫺0.0002

0.1533

Marginal Effect

Immigrant Dummy

1.3447 (0.3419) ⫺0.0098 (0.0120) 0.0637 (0.0155) 0.2528 (0.0411) 0.7707 (0.0891) 0.0173 (0.0949) 0.0269 (0.0646)

Estimated Coefficient

0.0018

0.0012

0.0519

0.0170

0.0043

⫺0.0007

0.0905

Marginal Effect

Refugee Immigrant Dummy

Variables Interacted with

92 The Journal of Human Resources

0.2245 (0.0260) 0.0581 (0.0233) 0.0161 (0.0105) 0.0011

0.0039

0.0151

0.0134 (0.0393) ⫺0.1035 (0.0337) ⫺0.0083 (0.0075) ⫺0.0286 (0.0397) ⫺0.0374 (0.0137) ⫺0.0216 (0.0339) ⫺0.2718 (0.0852) ⫺0.2835 (0.1260) ⫺0.1508 (0.1571) ⫺0.1254 (0.1971) 0.7240 ⫺72,019 405,426 ⫺0.0084

⫺0.0102

⫺0.0191

⫺0.0183

⫺0.0015

⫺0.0025

⫺0.0019

⫺0.0006

⫺0.0070

0.0009

⫺0.1678 (0.0112) 0.4016 (0.0428) ⫺0.2441 (0.0983) ⫺0.6840 (0.1518) ⫺0.9724 (0.1938) ⫺1.2386 (0.2588)

⫺0.0254 (0.0380) ⫺0.0168 (0.0297) 0.0023 (0.0085)

⫺0.0834

⫺0.0654

⫺0.0460

⫺0.0164

0.0270

⫺0.0113

0.0002

⫺0.0011

⫺0.0017

Note: Standard errors appear in parentheses. Data used are from LINDA, 1990–96. Estimates are based on a 25 percent random subsample of the full LINDA sample. Model includes both county and year fixed effects. P-value of likelihood ratio test for no within-group correlation is less than 0.0001.

Within-group correlation Log likelihood Sample size

Arrival 1968–75

Arrival 1976–80

Arrival 1981–5

Arrival 1986–90

Years since migration2 /100

Years since migration

Nordic immigrant

Local unemployment rate

SA norm/1000

Single*number of children

Hansen and Lofstrom 93

94

The Journal of Human Resources

Figure 4 Predicted Differences in Welfare Participation between Native-Born Swedes and Immigrants, by Years Since Migration, Age at Migration ⫽ 18, 1996 Baseline Note: Based on estimates presented in Table 4.

Figure 5 Predicted Differences in Welfare Participation between Native-Born Swedes and Immigrants from Nonrefugee Countries, by Arrival Cohort and Years Since Migration, Age at Migration ⫽ 18, 1996 Baseline Note: Based on estimates presented in Table 4.

Hansen and Lofstrom

Figure 6 Predicted Differences in Welfare Participation between Native-Born Swedes and Immigrants from Refugee Countries, by Arrival Cohort and Years Since Migration, Age at Migration ⫽ 18, 1996 Baseline Note: Based on estimates presented in Table 4.

imply differences of around 12 and 8 percentage points respectively.14 This finding suggests that immigrants in a comparatively generous welfare state, like Sweden, display welfare-participation behavior similar to that of immigrants in less generous welfare states, like the United States, relative to natives after having spent some time in the new host country. The model presented in Table 4 can be used to simulate effects of changes in immigration and unemployment on welfare-participation rates. The findings in this paper suggest that the increase in welfare participation is at least partially due to the increases in both immigration and unemployment, which Sweden experienced in the 1990s. To investigate how much these two factors contributed to the increase in social assistance utilization we predicted four time series of welfare participation. These are shown in Figure 7. Series 1 is derived based on the observed sample means of the included variables and the estimated parameters. Series 2 is derived in the same way, but we hold county unemployment rates constant at the observed 1990 levels. As expected, the deteriorating Swedish labor market is one of the reasons welfare-participation rates increased. However, changes in unemployment rates ex14. We used the estimated coefficients Borjas and Hilton (1996) report in Table 7 and the ones Baker and Benjamin (1995) report in Table 2 to calculate the predicted differences in participation rates. The predictions were calculated by assuming 20 years since migration for the most recent cohort and age at migration was set equal to 18 years. The models estimated by Benjamin and Baker (1995) do not control for age at migration.

95

96

The Journal of Human Resources

Figure 7 Simulated Welfare-Participation Series, with and without Changes in the County Unemployment Rates, the Proportion and the Composition of Immigrants Note: Based on estimates presented in Table 4.

plain only about 20 percent of the increase between 1990 and 1996. As discussed above, the increase in immigration may also have contributed to the rise in welfare utilization. In Series 3, we hold constant both the unemployment rates and the proportion of immigrants in the population to the 1990 proportions. The result from this exercise suggests that 40 percent of the increase in welfare utilization rates between 1990 and 1996 stems from these two sources. In other words, the increase in immigration contributed as much to the upward trend as did the higher unemployment rates. In the last series, Series 4, we also hold constant the composition of immigrants. In other words, the proportion of immigrants who are Nordic from nonrefugee and refugee countries is held constant to the 1990 shares. This, too, appears to have had an impact on social assistance utilization. Altogether, the worsened labor market coupled with the increase, and the change, in composition in immigration explains about 50 percent of the higher welfare-participation rates observed in 1996, compared to 1990.15 15. Our results indicate that changes in unemployment and immigration cannot explain all of the observed increase in welfare participation, but only roughly half of the increase. An obvious question then is, what are some of the other factors that caused the social-assistance-participation rates to increase in the 1990s? Because we control for the generosity of the welfare system by including the welfare norm, in addition to the observation that benefit levels only increased slightly, changes in the social assistance rules are not a likely reason for the unexplained portion of the increase. Cost savings measures implemented due to budget concerns, however, affected other government transfer programs. For example, the unemployment benefits ratio for the largest unemployment insurance program in Sweden (arbetslo¨shetskassa) declined from 90 percent to 75 percent during the period studied here. For a given unemployment rate, the probability

Hansen and Lofstrom

VII. Conclusion and Summary This paper analyzes differences in the welfare utilization between immigrants and natives in Sweden using a large panel data set, LINDA, for the years 1990 and 1996. Welfare expenditure in Sweden, as with many western countries, increased substantially in the 1990s. Closely linked to the increase in welfare costs in Sweden is an increase in immigration. In this paper we find that immigrants use welfare to a greater extent than natives use it. Furthermore, nonrefugee immigrants utilize social assistance less than refugee immigrants. Differences in welfare participation between immigrants and natives cannot be explained by observable socioeconomic characteristics. Immigrants appear to reduce their welfare-participation rates with time spent in Sweden in both an absolute sense, as well as relative to natives. Although refugees display substantially higher public-assistance-participation rates upon arrival in Sweden compared with nonrefugee immigrants, they assimilate out of welfare at a faster rate than nonrefugee immigrants. However, it should be noted that neither group is predicted to reach the participation rates of native households within a 20-year period in Sweden. We also find that approximately 50 percent of the observed increase in welfare utilization in Sweden in the 1990s can be attributed to the increases in both unemployment and immigration, as well as to the increase in the proportion of refugees in the immigrant population. Immigration has been at the heart of welfare debates in many countries in the 1990s. Due to the continued increase in immigration in many countries such as the United States, Germany, and Sweden, this is unlikely to change anytime soon. Given the rapid decrease in welfare utilization of immigrants, particularly refugees, with time spent in the new country, the welfare cost upon arrival should not be used to make long-term prediction of welfare costs caused by immigration. However, immigrants are overrepresented among the welfare population even after observable characteristics are controlled for. This suggests that in future research it is important to analyze why this is the case. One question that arises is: Are immigrants’ observable skills not recognized to the same extent as natives’ skills in the Swedish labor market or do immigrants have higher reservation wages?

References Baker, Michael, and Dwayne Benjamin. 1995. ‘‘The Receipt of Transfer Payments by Immigrants to Canada.’’ Journal of Human Resources 30(4):650–76. Bean, Frank D., Jennifer V.W. Van Hook, and Jennifer E. Glick. 1997. ‘‘Country of Origin, Type of Public Assistance, and Patterns of Welfare Recipiency among U.S. Immigrants and Natives.’’ Social Science Quarterly 78(2):432–51. Betts, Julian R., and Magnus Lofstrom. 2000. ‘‘The Educational Attainment of Immigrants: Trends and Implications.’’ In Issues in the Economics of Immigration, ed. George J. Borjas, 51–115. Chicago: University of Chicago Press.

a household will apply and receive welfare is higher in 1996 than it was in 1990. Hence, the less generous unemployment compensation program is likely to have increased participation in social assistance.

97

98

The Journal of Human Resources Blau, Francine D. 1984. ‘‘The Use of Transfer Payments by Immigrants.’’ Industrial and Labor Relations Review 37(2):222–39. Borjas, George J. 1985. ‘‘Assimilation Changes in Cohort Quality and the Earnings of Immigrants.’’ Journal of Labor Economics 3(4):463–89. ———. 1994.’’The Economics of Immigration.’’ Journal of Economic Literature (32)4: 1667–717. Borjas, George J., and Stephen J. Trejo. 1991. ‘‘Immigrant Participation in the Welfare System.’’ Industrial and Labor Relations Review 44(2):195–211. ———. 1993. ‘‘National Origin and Immigrant Welfare Recipiency.’’ Journal of Public Economics 50(3):325–44. Borjas, George J., and Lynette Hilton. 1996. ‘‘Immigration and the Welfare State: Immigrant Participation in Means-Tested Entitlement Programs.’’ Quarterly Journal of Economics 111(2):575–604. Crossley, Thomas F., James Ted McDonald, and Christopher Worswick. 2001. ‘‘Immigrant Benefit Receipt Revisited: Sensitivity of the Choice of Survey Years and Model Specification.’’ Journal of Human Resources 36(2):379–97. Edin, Per-Anders, and Peter Fredriksson. 2000. ‘‘LINDA—Longitudinal Individual Data for Sweden.’’ Working Paper Number 19, Department of Economics. Uppsala, Sweden: Uppsala University. Edin, Per-Anders, Robert J. LaLonde, and Olof Aslund. 2000. ‘‘Emigration of Immigrants and Measures of Immigrants Assimilation: Evidence from Sweden.’’ Swedish Economic Policy Review (7)2:163–204. Maani, Sholeh A. 1993. ‘‘Immigrants and the Use of Government Transfer Payments.’’ Australian Economic Review 0(104):65–76. Riphanh, Regina T. 1998. ‘‘Immigrant Participation in Social Assistance Programs: Evidence from German Guestworkers.’’ IZA Discussion Paper Series No. 15. Bonn, Germany. SOPEMI. 1998. Trends in International Migration. Paris: Organization for Economic Cooperation and Development.