Essays in Empirical Financial Economics

Abstract

This collection of essays investigates empirical relationships in corporate finance, labor economics, and asset pricing.

The first chapter utilizes matched employer-employee data to study the effects of stronger corporate governance on the earnings and composition of the firm's workforce. I build a new dataset that links over 2,000 public companies to their employees in Texas. Using the passage of shareholder-sponsored proposals to declassify the board of directors, I show that vote passage lowers a firm's average employee earnings by 11\%, directionally consistent with the previous literature. This has often been interpreted as wage decreases for individual workers. However, I show that this decrease is driven in a large part by the changing composition of the workforce.

The second chapter, coauthored with Harrison Hong and Yen-Cheng Chang, cleanly identifies the price effects of stock market indexing. We take advantage of a discontinuity in the assignment of firms to the Russell 1000 and Russell 2000 indexes to estimate the causal effects of index addition and deletion. We find that additions to the Russell 2000 result in a persistent 5\% price increases and deletions result in an equal and opposite price decline. We then show that these indexing effects have decreased over time and identify the types of funds that provide liquidity to indexers.

The third chapter, coauthored with Harrison Hong, shows evidence for a channel through which corporate social responsibility benefits firms. We focus on prosecutors, who are only susceptible to the halo effect, not product quality signalling or delegated giving. Using the enforcement of the Foreign Corrupt Practices Act, we find that social responsibility is associated with 2 million dollars less in fines, though it is uncorrelated with bribe characteristics and cooperation, which should entirely determine optimal fines. We use data on donations and text-mine related press releases to show that this bias is likely a halo effect and not prosecutorial conflict of interest.