**You Sponsor Mine, I Procure Yours: Pharmaceutical Sponsorships And Procurement in Public Hospitals**(Job Market Paper)

Using unique data linking hospital procurement contracts with sponsorships from pharmaceutical firms, I investigate possible instances of influence peddling and conflict of interest in Romanian hospitals. Procurement contracts related to sponsorships are 11% higher than procurement contracts not related to sponsorships. Sponsorships increase the probability of receiving a procurement contract by 5 percentage points. Sponsoring a doctor in hospital management has a slightly larger effect than sponsoring a regular doctor: the difference is economically significant only for direct contracts, which are the least transparent. Contracts linked to sponsorships are associated with shorter procurement times: being linked to a management sponsorship is associated with a decrease of 4 days in the time between announcement and signing, but there is no such association with sponsorships to regular doctors.

**Heterogeneous returns to capital and their implications for optimal tax policy**(with Aart Gerritsen, Bas Jacobs and Kevin Spiritus)

There is increasing evidence that people differ in the rates of return on their savings even after controlling for risk. We determine the implications of this stylized fact for the optimal taxation of labor and capital income. We allow for two distinct reasons why returns are heterogeneous: because individuals differ in their innate ability to invest their savings, or because those with higher wealth have more incentives to acquire the necessary financial knowledge and advice that allows them to achieve high returns. In either case, a strictly positive tax on capital income is part of a Pareto-efficient redistributive tax system. We write optimal nonlinear capital- and labor-income taxes in terms of sufficient statistics. Numerical simulations show that small degrees of return heterogeneity may justify significant capital-income taxes.

**Why is the Long-Run Tax on Capital Income Zero? Reinterpreting the Chamley-Judd Result**(with Bas Jacobs)

Why is it optimal not to tax capital income in the long run in Chamley (1986) and Judd (1985)? This paper demonstrates that the answer follows from standard intuitions from the optimal commodity-tax literature. We show that the steady state assumption is critical for the Chamley-Judd result: in the steady-state, Engel curves for consumption become linear in labor earnings and consumption demands become equally complementary to leisure over time. From the optimal tax literature, we conclude that consumption should be taxed uniformly, which means that the optimal capital income tax is zero. We show that the intuition that capital income should not be taxed because the consumption distortions become infinite only applies when restrictions are imposed on the utility function. These restrictions ensure that consumption demands are equally complementary to leisure in the long run, thereby confirming standard optimal-tax intuitions. We also demonstrate that the optimal capital-income tax is zero irrespective of whether factor prices are determined in partial or general equilibrium. This result contradicts the intuition that optimal taxes on capital income are zero because the entire burden of capital income taxes is shifted to labor through general-equilibrium effects on factor prices.