Online Appendix for the article: Higher and higher? Performance pay and wage inequality in Germany

End-point correction " Performance pay jobs " are defined as jobs that have paid for performance at least once in the past. Thus, job matches that are observed over a longer period are more likely to be observed as PP jobs. For this reason, job matches that are observed at the beginning of the observation period in 1984 may be misclassified as non-PP jobs if they paid for performance before 1984. In order to correct for this, an end-point correction is applied in analogy to Lemieux et al. (2009). It proceeds in three steps: First, PP jobs are estimated as a function of calender year and the number of years an individual job-match is observed in the sample. Second, the distribution of years that the job-matches are observed in the sample is held constant at a time in the middle of the observation period. Third, the share of PP jobs is predicted based on this hypothetical distribution of observation years. These shares deviate from the uncorrected shares at the beginning of the observation period. The corrected shares are depicted in Tab. 1 and Fig. 1 in the main article.