Are real entry wages rigid over the business cycle? Empirical evidence for Germany from 1977 to 2009

So far little empirical evidence exists on how real wages of newly hired workers react to business cycle conditions. This paper aims at filling this gap for Germany by analyzing the cyclical behavior of real wages of newly hired workers while controlling for 'cyclical upgrading' and 'cyclical downgrading' in employee/employer matches over the cycle. The analysis is undertaken for the 1977 to 2009 period using administrative longitudinal matched employer-employee wage data. I find that an increase in the unemployment rate of one percentage point decreases the real wages of job entries within given firm-jobs by about 1.27 percent. In light of the magnitude of the entry-wage cyclicality it seems that introducing wage rigidity in the Mortensen-Pissarides model in order to amplify realistic volatility of unemployment is not supported by the data. Further I show that the procyclicality of the employment/population ratio is identical to the procyclicality of real entry wages. This counters the view of many macroeconomists that wages are much less cyclical than employment and unemployment.