The Impact of Welfare Policy On Immigrants' Social Trust

Purpose: The incorporation of immigrants and the social cohesion of heterogeneous societies are salient issues worldwide. Some scholars argue that immigration threatens social cohesion because it fosters distrust towards people different from ourselves which, in turn, erodes reciprocity and collaboration (Alesina and Ferrara, 2000, 2002; Putnam 2007). The main research question of this paper asks whether comprehensive welfare states can promote social cohesion by reducing the social trust differences between natives and immigrants. To answer the question we test two competing hypotheses: the cultural hypothesis which claims that social trust stems from one's culture and remains stable over one's life course (Uslaner, 2008) against the institution-centered hypothesis which argues that generous welfare states promote social trust and therefore foster social cohesion (Kumlin, 2004; Kumlin and Rothstein, 2005, 2008). Thus far, no study has examined this process in an international comparative setting with immigrant populations.

Method: To determine whether the social trust levels of non-western immigrants changes to resemble that of natives and whether comprehensive social policy contributes to this process we relied on two sources of data. At the individual-level we used the second, third and four waves of European Social Survey (ESS). These were collected between 2004 and 2008 across 24 European countries and 102,229 native-born individuals, 5,956 first generation immigrants and 8, 251 second generation immigrants. At the country-level data came from the European System of Social Protection Statistics (ESSPROS) which provides annual comparable statistics across European nations on social protection benefits. To test the impact of comprehensive social protection benefits on immigrants' social trust we distinguished between universal and means-tested benefits. We utilized Bayesian inference with Markov chain Monte Carlo (MCMC) algorithms to fit the multilevel models. Part of the modeling strategy was to adjust for both country and individual-level confounders as fixed effects while allowing for heterogeneity between individuals by migratory status (natives versus immigrants) and between countries. The random effect on the intercept at the country level (μ_0j) was assumed to be normally distributed with a mean of zero and variance (σ²_μ0) (Blakely and Subramanian, 2006).

Results: Our findings supported the institution-centered hypothesis as countries' universal social spending increased the social trust of immigrants (β₁ = .30, p < .01) whereas social spending on means-tested benefits did not exert an influence on immigrants' level of trust (β₂ = .10, p > .10). Moreover, the claim that social trust is not stable over one's life course was further supported by the fact that the level of trust of second generation immigrants was more similar to the level of trust of natives than to the level of trust of the first generation immigrants.

Implications: These findings suggest that egalitarian social policies promote the incorporation of immigrants and stress the importance of interventions that foster equal access to the systems of social protection in receiving societies.