IZA DP No. 15534: Using Distribution Regression Difference-in-Differences to Evaluate the Effects of a Minimum Wage Introduction on the Distribution of Hourly Wages and Hours Worked
This paper evaluates the effects of the newly introduced German minimum wage on the distribution of hourly wages and hours worked. The study is based on the German Structure of Earnings Survey (GSES), the only large scale data set for Germany that includes information on hourly wages and hours worked. We provide a full distributional analysis based on counterfactual distributions that would have prevailed, had the minimum wage not been introduced. Our results suggest that its introduction almost eliminated wage rates below its threshold and, depending on the specification considered, led to spill-over effects up to 20 percent above it. We show that inequality in hourly wages fell between 2014 and 2018, but that the long-term trend of rising inequality would already have been stopped after 2014 without the minimum wage. We demonstrate that the existence of pre-trends leads to an upward bias for the estimation of the minimum wage effect. We do not find any significant shifts in the distribution of weekly working hours. As a methodological contribution, we provide a transparent treatment of distribution regression difference-in-differences (DR DiD) based on bite measures.