Skip to content


Work Package 10: Unemployment and Earnings Inequalities

Social class notoriously influences several aspects of individual and group life chances. In contemporary societies the employment status of a person and the amount of earnings he/she gets from the occupation he/she carries out play a central role in determining his/her level of living. Therefore studying how class affects individual risks of unemployment can be considered as a crucial test to validate the ESeC. The same, of course, holds in the case of earnings.


Our hypotheses regarding the above topics can be summarised as follows. Leaving aside employers and the self-employed – because, by definition, they do not run any risk of being fired – we expect, first, that moving from higher to lower classes monotonically increases the risks of being unemployed. Second, we would expect that people belonging to lower classes are more likely to carry out insecure and precarious jobs that lead them to experience much more frequent and longer unemployment spells during their work history than their counterparts belonging to middle and higher classes. Third, we also hypothesise that, at least in southern European countries, incumbents of occupations classified within lower classes should be observed more frequently in the informal sector of economy.

Regarding income from employment, we believe that, despite a lot of within class variability, people belonging to higher classes will receive, on average, higher salaries and enjoy more fringe benefits and allowances than do members of middle and, even more so, lower classes. Moreover we are inclined to hypothesise that earnings differentials between occupational classes have increased over the last ten years in most EU countries.


Obviously, apart from class, several micro variables – such as gender, birth cohort, education – and macro variables – such as national labour markets regulations, institutional arrangements of individual societies, type of welfare system – can affect the risks of unemployment and the level of individual earnings. Therefore, in order to validate the ESeC and to estimate the net effect accounted for by this class schema, we shall carry out multivariate comparative analyses. Moreover, since we hypothesise that between class disparities in the risks of unemployment extend over the whole working life of individuals and earnings inequalities change over time, we shall carry out cross-cohort event history analyses. Multilevel models applied to longitudinal micro data could be the best statistical approach in order to validate ESeC.


In principle, ECHP would be the best data source to perform a comparative, dynamic analysis of class disparities in contemporary EU societies. However, it is unclear whether the occupational classification used in the ECHP is as detailed as required for a criterion validity test. If this is the case, we will use data from national panel studies, such as the British Household Panel Study, the German Socio-economic Panel, the Swedish Level of Living Survey, the Italian Household Longitudinal Study and the likes. In the latter case, we will attempt to increase the degree of contextual variability. Tentatively such a result could be reached by treating central and northern Italian regions, on the one hand, and southern ones, on the other, as different countries.