The Structure of Inequality, Social Trust and Redistribution: A Cross-Country Analysis in a Panel of OECD Countries

Sharp income disparities, once a feature of a distant pre-War past, have been increasingly realised during the last three decades within the majority of OECD countries. This trajectory may be attributable to a wide array of factors, including among others the integration of the world economy, demographic changes, labour market reforms, the level of access to good quality public education and the structure and efficiency of the redistributive mechanism. The latter appears as crucial for alleviating income inequalities produced by market forces, i.e., gross inequalities. These are recorded generally as steeper in contrast to net ones, i.e., after the interference of the of the redistributive mechanism. However, the degree of redistribution seems to be rather heterogeneous among the OECD countries with some countries equalising generally incomes more radically than others. To which factors can we attribute this variation in the intensity of the redistributive effort?

The standard political economy approach to what determines the level of redistribution is expressed by the medium voter theorem, which predicts that redistribution will be higher the further the median voter’s gross income lies from the mean of the gross income distribution. This view accounts only for pecuniary incentives from the side of the median voter, yet, empirically such prediction has not fared well and most of the relevant studies disregard the role of gross income inequalities in influencing (demand for) redistribution.

Nevertheless, a recent wave of research has attempted a more thorough examination of the relationship between gross inequality and redistribution by arguing that what matters eventually is not aggregate measures, as expressed by GINI coefficients, but rather structural features of such distribution. Moving away from the medium voter theorem that relies on pecuniary incentives alone, these arguments stress social norms induced by social identification among income classes, that is participation of individuals to social networks influenced by the income structure. Two competing theories, this of social affinity and social rivalry fall within this category and argue that relative income positions may drive (demand for) redistribution either for reasons of altruism or antagonism respectively. Namely, the social affinity hypothesis asserts that closer economic proximity between income classes should foster social ties between those classes and thus formulate accordingly alliances towards increased redistribution. Alternatively, the social rivalry hypothesis suggests that proximity of higher to lower socio-economic classes would invoke sentiments of antagonism, at least from the side of the members of higher socio-economic classes and thus bring weaker (demand for) redistribution.

The social identification arguments and their implications for both the demand side of redistribution and the actual levels observed, provide deeper insights in addition to pecuniary incentives arguments alone. But how participation of individuals to specific social groups or networks based on income is translated eventually to specific redistributive attitudes and moreover to the extensiveness of actual redistribution?

A probable linkage between structure of gross inequality and redistribution may be that of social or generalized trust, that is the confidence that others will not behave in opportunistic manner and they will look after our interests, if they have the option to do so. Social trust should be considered as a factor determining stronger (demand for) redistribution, for reasons of social identification as identified above. In particular, based on the social affinity hypothesis, closer economic proximity among income classes, captured by relative income positions, should nurture stronger trusting attitudes for reasons of familiarity and closer social ties and further stronger (demand for) redistribution. In case that the social rivalry hypothesis is relevant, the opposite effect should be expected for reasons of sustained antagonism among income classes and thus weaker trusting attitudes and (demand for) redistribution.

The paper will contribute towards these issues by firstly focusing on the effects that the structure of gross inequality exerts on the levels of social trust; and secondly, by examining how social trust potentially affects redistribution. Empirically, the relationship will be rigorously tested in a panel dataset of 25 OECD countries by examining the impact of gross earnings positions of income classes, (i.e. percentiles ratios of gross earnings, between the poor, middle and rich classes) on (their respective) trusting attitudes, and testing the relevance of the social affinity and social rivalry hypotheses. At a later stage, I will investigate how these trusting attitudes possibly promote redistributive components of public social expenditure (at macro-level) and also attitudes to redistribution (at individual level). Instrumental variables analysis will be pursued at macro-level analysis, so to check for an independent effect of social trust to the levels of public social expenditure.