Rent sharing, wage floors, and development

Faced with more favourable demand conditions, many firms raise wages. However, we show that firms with labour market power, lower productivity, and binding wage floors will absorb these positive revenue productivity shocks as excess profits instead of increasing wages or employment. Our prediction follows from a simple but novel theoretical insight under a standard framework of monopsonistic competition, and we empirically test this theory in South Africa using firm-level administrative data. We first explain how firm wage-setting behaviour changes at a productivity threshold directly related to the wage floor and then show how the predicted wage, employment, and profit patterns are evident in the cross-section of firms covered by collective bargaining agreements. We then replicate and extend a leading method of identifying rent-sharing elasticities, but estimated separately by firm revenue productivity bins. As predicted by the theory, we find that firms below the threshold increase wages and employment less, and profits more, in response to revenue productivity shocks, and that there is a break at the threshold where wage floors bind. The study complicates the conclusions emerging from the literature on firm rentsharing, and forms part of an explanation for 'stalled' development and 'jobless growth'.