Sunday, 5 April 2015

Productivity per worker in the DRC varies hugely by company

Productivity of workers is an issue of concern to companies, workers, and governments.  It partially determines how much profit companies can get from hiring workers, and how high workers' salaries are, and what the output of a whole economy is.

One way of measuring worker productivity is by sales per worker.  In the DRC, there are a very wide range of values for the quantity.  The graph shows the distribution across companies of the natural logarithm of sales per employee.  The dispersion is so wide that we have to take logarithms of sales per employee in order to show it on a graph.  For example, the company with the highest sales per employee sells a billion times more per employee than the company with the lowest sales per employee, and we could find similar, slightly less extreme dispersal for other companies too.

 Source: World Bank Enterprise Surveys

Part of the explanation is likely to be that some companies have more equipment than others.  The equipment does some of the work that an employee would do, so that for every employee, they are doing their own work and there's some extra work being done by the equipment.  Here's a graph of sales per employee plotted against capital (in the form of machinery, vehicles, and equipment) per employee:

There seems to be a reasonable positive relation between capital and sales, which explains part of the difference in productivity across workers.  The line is fitted by the least squares method, with the fit having an R-squared of 0.16 meaning that 16 percent of the difference in sales per worker can be explained by difference in capital per worker.  Equipment is part of the explanation for the variation in productivity per worker, but there is a lot left to explain.

No comments: