Non-linear taxation with monopsony power

How does monopsony power affect optimal income taxation and welfare? I study this question in a Mirrleesian framework where firms observe workers’ abilities while the government does not. Monopsony power does not generate efficiency losses, but determines what share of the labor market surplus is translated into profits. Monopsony power increases the tax incidence borne by firms and mitigates (exacerbates) inequality in labor (capital) income. I derive intuitive expressions for the optimal marginal tax rate on labor income and the welfare effect of raising monopsony power. Monopsony power has an ambiguous effect on optimal tax rates and welfare. On the one hand, monopsony power enables the government to use firms as a screening device in order to extract information on hidden ability. This lowers optimal tax rates and raises welfare. On the other hand, monopsony power generates a distributional conflict over profits. This raises optimal tax rates and lowers welfare provided profits flow back to individuals with below-average welfare weights. I calibrate the model to the US economy and find that monopsony power raises optimal tax rates at low earnings levels and lowers optimal tax rates for middleand high-income earners. Moreover, the welfare effect of eliminating monopsony power is very sizable and ranges between –3.4% and +6.8% of GDP depending on the government’s redistributive preferences.