Wage drift was negative
Under collective wage agreements, employee compensation usually increases slightly each year. Many employees see their gross wage go up by more than the agreed contractual wage growth, for example because they are given an individual wage increase, a promotion-related pay rise or a bonus. In the macro statistics, however, gross wage growth lagged behind contractual wage growth in recent years. Between 2014 and 2018, annual gross wage growth was 0.5 percentage points below contractual wage growth on average. This is remarkable because gross wage increases tended to outstrip contractual wage rises during earlier periods of increasing labour market tightness. The difference between gross and contractual wage growth is the contribution of the incidental wage components, or wage drift. Wage drift can be broken down into two components: (i) additional and/or performance-based individual remuneration at the micro level and (ii) composition effects at the macro level. CPB and DNB have studied recent trends in wage drift based on a new dataset and using various decomposition analyses.
Sharp growth in youth employment and flex jobs
Changes in the composition of employment at the macro level are reflected in wage drift. This is typically a countercyclical effect, as mostly lower paid employees (young people, flex workers and lower educated workers) lose their jobs in a recession, while more expensive employees remain in employment. As the proportion of low-paid workers falls, the average wage level goes up. This results in a positive contribution of the composition effect to wage growth and wage drift.
The CPB/DNB study calculates this composition effect in two different ways. Firstly, in considering the characteristics of employees and jobs, we found an estimated average annual negative composition effect of -0.2 percentage points between 2014 and 2018. In particular, the influx of young employees and flex workers slowed down wage growth in recent years (see Figure 1). Another finding that stands out is that the positive composition effect of the level of education decreased in spite of the steady increase in the average level of education and higher wages of highly educated employees. One obvious explanation is that in times of economic boom the influx of lower educated employees, with low average hourly wages, exceeds that of highly educated employees. Secondly, our other decomposition analysis, which explicitly considered the development of new and existing jobs, also found an estimated average annual negative composition effect, of -0.3 percentage points.
Figure 1. Composition effect broken down by background characteristics