Workforce characteristics and productivity in the DRC

My last post looked at the wide dispersion of worker productivity across companies in the DRC.  Increasing the amount of capital per worker in a company was associated with increased productivity, but there were a lot of productivity differences left unexplained.

Another explanation for the productivity differences is that in some companies the workforce is more highly skilled, so they can produce more for every hour worked.  The graph below shows productivity plotted against the percentage of production workers who are skilled rather than unskilled.  There is almost no relation between the two variables, contrary to expectations.

Source: World Bank Enterprise Surveys

If the skill-productivity link isn't convincing, then how about a link between the number of managers and productivity?  Managers (and other non-production staff generally) may help production staff to reach their best performance.  On the other hand, non-production staff are not directly producing goods so they may lower overall productivity.  The next graph shows that there is almost no connection either way.

These graphs are a challenge to explain.  It looks like workforce productivity doesn't depend much on the workforce.  One possibility is that my approach is too blunt to examine the relations in sufficient detail.  There will be other possibilities too.