Essays in Public Economics
This dissertation consists of three chapters and analyzes individuals' and firms' response to tax and government spending policies.
The first chapter focuses on the economic incidence of a large value added tax (VAT) cut in the restaurant industry in France. In particular it estimates the share of the tax cut falling on workers, firm owners and consumers by analyzing the VAT cut applied to French sit-down restaurants -- a drop from 19.6 percent to 5.5 percent -- in July 2009. Theoretically, we develop a standard partial equilibrium model of consumption tax incidence that includes consumption substitutability between the taxed good and an untaxed good, and markets for production inputs, which we allow to vary with the tax. Empirically, we use firm-level data and a difference-in-differences strategy to show that the reform increased restaurant profits and the cost of employees, and aggregate price data to estimate the decrease in prices produced by the reform. We compare sit-down restaurants to (a) non-restaurant market services and (b) non-restaurant small firms, and find that prices decreased by around 2 percent, the cost per employee increased by 3.9 percent and the return to capital increased by around 10 percent. Using these reduced-form estimates we conduct a welfare analysis and find that (1) the effect on consumers was limited, (2) employees shared around 20 percent of the benefit eighteen months after the reform, and 29 percent thirty months after the reform, and (3) the reform mostly benefited capital owners, who received around 50 percent of the tax cut, both in the short-run and in the long-run.
The second chapter derives a more general result on the price effect of VAT changes. It shows that prices respond asymmetrically to increases and decreases in VAT. We combine monthly commodity price data with information on value added tax (VAT) rates across several European countries for the period 1996-2015 and show that prices respond more to VAT increases than to VAT decreases. We explain this asymmetric pass-through to prices with a simple fairness argument developed in the behavioral economics literature. Our finding cautions against using incidence estimates derived in previous studies without accounting for the direction of the tax change and questions the effectiveness of reductions in VAT to achieve redistribution or stimulate economic growth.
The third chapter of the dissertation focuses on a major government spending program in the United States: unemployment insurance (UI). Specifically, I use data on geographic mobility from the Survey of Income and Program Participation (SIPP) and state-year variation in unemployment insurance (UI) laws to evaluate the link between unemployment benefit generosity and mobility decisions of unemployed workers in the United States in the period 2001-2012. My empirical strategy uses a proportional hazard model to study whether the probability of moving during the unemployment spell depends on UI benefit generosity. I find that (1) higher UI weekly benefit amounts increase unemployed workers' geographic mobility, (2) the effect is stronger for more liquidity constrained unemployed workers (3) UI weekly benefit amounts are a stronger determinant of unemployed workers' geographic mobility than UI duration. Theoretically, I develop a spatial equilibrium model with risk-averse and liquidity constrained unemployed workers, and explain the empirical findings with the presence of high monetary moving costs.