Intra-EU Labour Migration and Attitudes Towards the Welfare State 109 increases, the log-odds of choosing a higher category (greater support for redistribution) also increases. Model 1 shows the results for a full range of micro-level factors that are considered important influences on an individual’s attitude towards redistribution. I find a number of results for the control variables that are consistent with previous literature (Burgoon et al., 2012). For example, having experienced a spell of unemployment lasting three months or more means an individual is more likely to agree with the statement above, while having a right-leaning ideology means that one is more likely to disagree with the statement. Additionally, if you are female, you are more likely to agree, while feeling financially well-off means that you are more likely to disagree. Importantly, the analysis finds a significant and positive association between job insecurity and preferences for redistribution. Consequently, when an individual expresses job insecurity, then the log-odds of being in a higher category/more supportive of government redistribution increase versus those who feel more job secure. The effects of the control variables stay consistent across all the Models in Table 5.2. In Model 2, I add three macro-level controls to test what contextual level variables play a role in affecting attitudes towards redistribution. The analysis shows that GDP per capita and the unemployment rate are key influences on attitudes towards redistribution and that they have opposite effects. GDP per capita is statistically significant and negative, suggesting that on average individuals in richer countries are more likely to disagree with the statement, aligning with earlier literature which finds that demand for redistribution reduces as economic prosperity increases (Dallinger, 2010). In addition, individuals in countries with higher levels of unemployment are more likely to agree with the statement. This result continues to provide evidence towards the idea that job insecurity can influence individual attitudes towards redistribution. On average, a poorer country with higher unemployment, which is suggestive of more uncertainty in the labour market, will be more supportive of the concept of redistribution. In addition, social spending is not associated with attitudes on redistribution, despite previous researchers hypothesising that a country with higher levels of spending or generosity garners more support for redistribution. As a robustness test to check whether the problem is the operationalisation of ‘generosity’, I utilise the unemployment replacement rate (Model 3) and find no change. The coefficient for the welfare state remains insignificant and there are no changes to any of the other independent variables. Moreover, the significant result for job insecurity remains after including the contextual level controls. Moving forward, religiosity and generosity are dropped from the specifications as they are consistently insignificant and do not affect the other results when excluded.
RkJQdWJsaXNoZXIy MjY0ODMw