Is there any empirical evidence on the value added by workers?
That is, their "productivity" minus the wage payments, social security benefits etc. that are paid by the firm?
I'm eventually looking for cross-sector or time variation, but anything would be a good starting point right now.
Normally when we think about value added, we think of marginal revenue product of labor: how much more revenue does adding another worker add to the firm?
If you assume that the market is competitive, then the MRPL should be exactly equal to what you said: wage, benefits, etc. Otherwise another worker would be willing to work for an epsilon less.
In other words, the "value added" by your definition (their "productivity" minus payments) should be zero. From my experience, researchers measure worker's productivity by wages.
If you want to find where this does not hold, then I guess you'll need more detailed data. Perhaps comparing workers with same job description but different compensation.
