1. Low rank = unskilled or semi-skilled work
2. Middle rank = skilled work
3. Upper rank = highly skilled work
4. Executive rank = executive work
Table 2 presents average wage growth associated with transitions from one rank to the next
between two periods for all workers and by occupation. The diagonal shows the average wage
growth of workers who did not change rank. The last column computes the average wage (in
level) of all workers in a given rank at time t. Note that average wage growth associated with
a promotion is slightly higher for promotion from the middle to the upper rank and upper
to executive rank than it is for the low to middle rank. On the other hand, there remains
individual variations in wage changes within each rank. Comparing the difference in wage
level between rank L and M, there is a difference of 483 marks, which corresponds to a 30%
difference in average wage between the two ranks. It is 57% between rank M and U and 24%
between U and EX.
Concerning the evidence on serial correlation in wage increases and promotions, the analysis
is limited by the fact that the sample period is not long enough to observe several episodes of
mobility per worker. On average, the number of years workers stay in the sample is about 8
years. Over that period, the average number of time a worker experience a change is 1.1. As
a result, the sample size when considering the workers who experience a change twice or more
is very small. This limits the possibility of making reliable inferences on serial correlation in
promotions and wage increases.
Summarizing the findings, average wage growth at promotion is higher than without pro-
motion but is lower than the average difference in wage growth between two consecutive ranks.
There is also some evidence that previous wage growth predicts promotion as previous pe-
riod average wage growth for those experiencing a promotion or a job change the following
period is higher than for those experiencing no change in job within the firm. These findings
suggest that promotion to a higher rank plays an important role in the wage determination
process in German firms. The last two findings suggest that individual-specific variations in
wage changes are also important. Overall, these findings show evidence that individual and
job characteristics are both important factors in determining wage outcomes within German
firms.
on the occupational rank variables.
16