A RIF regression approach to evaluate wage changes: a focus on Italy

This paper investigates how microeconomic characteristics affect wage levels and wage inequality in Italy, before and during the economic crisis. We use EU-SILC (European Union Survey on Income and Living Conditions) data at individual level (the unit of analysis are employees aged 16-64) for 2005 and 2013. After analysing how the structure of employment has changed between 2005 and 2013 in Italy, we perform the Recentered Influence Function (RIF) regression of Gini index, variance, median and two extreme deciles (q10 and q90) on (log of) gross individual wage. The RIF regression allows us to estimate the impact of changes on covariates on the whole unconditional distribution of the measures of interest. Thus, the changes in wage inequality are decomposed into two components: the composition effect, which captures the impact due to individuals’ endowments, and the wage structure that depends on the labour market characteristics of the Country. Finally, the composition effect and the wage structure are computed for each covariate, highlighting the factors that contribute the more to the inequality over time. The five statistics confirm that the greater weight is associated with the return effect. In particular, the analysis reveals that the increase in wage inequality might be almost fully explained by the low efficiency of the Italian labour market structure in contrasting it with adequate labour policies and support measures.