“Signaling Export Quality under Firm Heterogeneity” (Job Market Paper)
Abstract: What are the economic consequences of dishonesty in international trade? Dishonest firms misrepresent products and reduce consumer beliefs about export quality, but honest firms can signal to mitigate the information asymmetry. I develop a quantifiable export model in which firms with heterogeneous productivity signal unobserved quality choices through prices. As consumers might have difficulty distinguishing whether a low price is due to dishonesty or high productivity, prices become less effective at signaling product quality. The model delivers three main predictions: (1) honest and dishonest firms overlap in productivity, (2) there is an inverted U-shaped relationship between the measure of dishonest firms and the cost advantage of misrepresented products, and (3) trade barriers are ineffective in reducing the share of firms choosing dishonesty. These predictions match patterns found using the FDA Import Refusal Report. I exploit the relative export volume and export price growth between high and low productivity firms to structurally estimate the model for Chinese exporters. I find that the potential for dishonesty leads to 22% fewer exports and a 28% reduction in export profits.
“Fiscal Policy in a Production Network” (with Gernot Muller, Ernesto Pasten, Raphael Schoenle, and Michael Weber)
Abstract: Heterogeneity across sectors raises challenges for stabilization policy because one common monetary policy stance does not fit all sectors. In this paper, we study optimal stabilization policy in a multi-sector New Keynesian model. We show that the optimal policy implies a simple rule for government spending at the sectoral level. It calls for an adjustment of spending in response to sectoral inflation and output gaps. In a second step, we use government procurement data aggregated to the sectoral level and quarterly frequencies for 2001-2018 to estimate the response of government spending at the sectoral level. We find evidence that government spending is indeed adjusting to sectoral output gaps and inflation.
Abstract: When human capital investments are imperfectly observed by employers, distinct groups of agents with identical investment costs may achieve different equilibrium outcomes. Coate and Loury (1993) showed that affirmative action may still lead to unequal investments between groups. Is there a policy that can close the investment gap due to statistical discrimination without fail? This paper introduces a novel policy: wage bill equality, which commits firms to equate the total wages paid to each group relative to its population size. Under a competitive economy with endogenous human capital formation, I show that wage bill equality is the unique policy that always achieves equality by eliminating all discriminatory equilibria. The underlying mechanism is that wage bill equality corrects incentive incompatibilities by only rewarding agents who invest, and disproportionately so for the discriminated group. Compared to affirmative action, wage bill equality leads to higher wages for every worker in the discriminated group.
“Discrimination as a Poisonous Fog” (with Roland Fryer)
Abstract: We introduce discrimination in multiple sectors to study how sectoral spillovers may lead to racial differences. Agents make a general human capital investment that is imperfectly observed by institutions of two sectors. Taste-based discrimination in the prejudiced sector lowers agents' incentives to invest, thereby spilling over and reducing fair institutions' beliefs about investment. If the size of the prejudiced sector exceeds a critical threshold, agents are trapped in the equilibrium with the lowest investment. To effectively address spillovers in discrimination, the policymaker must target a sufficient mass of prejudiced institutions. We document evidence that black children have lower returns to schooling in health, social outcomes, and police interactions than white children, even though returns to schooling in wages are similar. The analysis suggests that spillovers may be an institutional reason for the racial education gap.
Abstract: How effectively can agents coordinate on a network when they have limited depth of reasoning? I map this problem to a complete information game that combines the network, agents' beliefs, and each depth of reasoning. The complete information game fully characterizes agents' equilibrium actions, which I interpret using a directed acyclic graph with absorbing states. Using these tools, I present conditions to agents' beliefs such that as their depths go to infinity, their asymptotic actions exist and are equivalent to rational actions. Then, I develop an algorithm that computes equilibrium actions for an environment with both rational agents and agents constrained by limited depth of reasoning. The analysis also begets two technical innovations in studying nonstandard networks.