Inferring Capital-Labor Substitution from Firm Size Distortions (Job Market Paper)

I propose a novel identification strategy to estimate the elasticity of substitution between capital and labor, taking advantage of a set of size-dependent policies. These policies create a break in the size distribution of French firms. To the extent that firms around the threshold distort their capital-labor ratio, the elasticity of substitution can be recovered from firm-level distorted input choices and the amount of bunching. On the other hand, because firms have an incentive to under-report their size, part of the apparent distortions is only due to evasion. I extend the canonical span of control model with a labor friction and a misreporting margin, and develop formulas to link the elasticity to observed distortions. I merge two comprehensive administrative datasets and use the discrepancies between a self-reported employment count and a more accurate measure that the administration calculates from employees' payroll data. I provide direct evidence for evasion, show that under-reporting surges around the threshold, and accounts for a large fraction of the distortions. I find that firm-level elasticities are close to zero and that the aggregate elasticity in the economy, is around 0.3. Additional channels on the extensive margin are important to reach higher elasticities of substitution. (Paper)

A Unified Law of Mortality: Implications for the Long Run Effects of Early Conditions

(with Adriana Lleras-Muney)

How do social and economic conditions experienced early in life affect the evolution of health and mortality rates over the lifetime? To answer this question, we build and estimate a simple dynamic model of health. A key insight of the approach is that if mortality depends on health, then the evolution of mortality rates by age places constraints on the evolution of the underlying distribution of (unobserved) health. Therefore mortality rates can be used to infer how health has evolved over time and across countries. We estimate our model using high quality cohort life tables from the Human Mortality Database. We use the model and the estimated parameters to understand how unexpected shocks, like wars and infectious disease epidemics experienced early in life, affect the age-profile of health and mortality. We also investigate implications for SES gradients and optimal health care expenditures. (Paper)

The Uncertainty Cost of Wrongful Discharge Laws

(with Yanos Zylberberg)

This article estimates the cost of uncertainty for firms and workers related to wrongful-discharge regulations. We collect unique French data on 90,000 litigations after which either the employee or the employer appealed the labor court decision. We then use textual analysis on appelate court decisions to extract case characteristics and qualify the motives behind both verdicts. To isolate exogenous and unexpected variation in the final judgment, we exploit quasi-random judge assignment within the same regional jurisdiction. Combining this unique information with high-quality employer-employee data, we then estimate the effect of these decisions — financial penalties or reinstatement — on employment and firm structure in the short- and longer-run. Decisions for which motives behind second-instance verdicts strongly contradict the first-instance litigations are used to distinguish an ex-post “cash-flow” effect from a forward-looking “uncertainty” effect.

Work in Progress

The Macroeconomic Effects of Anti-Competitive Practices (with Ludovic Panon)

This paper develops a quantitative framework to assess the macroeconomic effects of antitrust policies. We first construct a new measure of “competition shocks” at the firm and sector level from an automatic textual analysis of about 3,000 antitrust decisions taken by the French competition authority over the last 20 years. We then combine this measure with an administrative data set covering the universe of French firms and analyze the impact of these competition shocks on market structure. Our model with heterogeneous firms and endogenous markups extends the framework of Atkeson and Burstein (2008) to allow for collusion, mergers and input-output spillovers.

Dynamic Hierarchies, Job Ladders, and Misallocation