124 Chapter 6 compensation hypothesis into its two component parts: (1) that increased mobility increases individual job insecurity and (2) increased job insecurity in turn increases preferences for redistribution, I find that increased mobility increases individual job insecurity and job insecurity increases the likelihood of supporting redistribution, reinforcing the idea that personal economic vulnerability drives greater support for redistributive policies (Hacker et al., 2013; Häusermann et al., 2015; Margalit, 2013; Meltzer & Richard, 1981; Rehm, 2009; Walter, 2010, 2017). Taking all the chapters together, this dissertation contributes to the theoretical debate around how immigration can be expected to alter the welfare state through showing empirically that immigrant origins matter for welfare state responses and by challenging assumptions that migration weakens the welfare state or undermines welfare support. This dissertation interprets associations with increased spending or replacement rates not as evidence of strain, but potentially as policy responses to compensation mechanisms especially as there is no association found between increased immigration and the generosity index. Furthermore, this dissertation demonstrates the importance of not only disaggregating the dependent variable – welfare state effort – into its component programmes, as done in the previous literature (e.g., Gaston & Rajaguru, 2013; Lipsmeyer & Zhu, 2011; Römer, 2023; Soroka et al., 2016; Taschwer, 2021) but also disaggregating the independent variable – immigration – into the various subcategories that make up the whole. Moreover, this dissertation supports a more conditional perspective, such as that economic conditions or inclusive institutions may help sustain welfare generosity despite rising diversity. However, as with all observational research, this dissertation has certain limitations. First, there are inherent challenges in establishing any kind of causal inference. While the analyses identify robust associations between immigration and welfare generosity, these are associations and not causality. In addition, the potential for endogeneity through reverse causality or omitted variable bias – remains a key limitation. I attempt to account for these factors by using econometric methods such as an instrumental variable approach, lagged independent variables, or by including a wide-range of controls, but there is always the possibility that it is not enough to be considered fully exogenous. 6.2 Policy Implications For policymakers, these results should help shed light on a polarised debate frequently influenced by implicit biases and prejudices related to race and difference. A key takeaway is that immigration does not necessarily pose a threat to welfare states in Europe. Instead, it is possible for welfare states to maintain or even enhance their levels of social protection despite, or
RkJQdWJsaXNoZXIy MjY0ODMw