We investigate variation in information production by sell‐side analysts and its potential role in hedge fund activist intervention, an important external corporate governance mechanism that creates shareholder value. Using textual analysis to derive an activism dictionary from intervention objectives and tactics, we find substantially more activism content in pre‐intervention analyst reports of target firms than propensity score matched control firms. Activism content is associated with more detailed reports containing more quantitative information. Target firm intervention‐date stock returns are significantly higher when activist intention (13D) filings are supported by reports with more general and objective‐specific activism content. Of activists' public letters to stakeholders, 31.9% directly mention sell‐side analysis, amplifying the association between target returns and analyst report information. The relationship between analyst information and activism returns is robust to using brokerage closures as an exogenous shock and is consistent with analyst incentives. Activist funds with no prior disclosed position in target firms and more experienced funds capture higher returns from sell‐side information. Overall, our results suggest sell‐side analysts play a significant informational role in supporting hedge fund activism.

We study the New York City taxi market to examine whether an excise tax hike on cigarettes corresponds to smoker taxi drivers more frequently cheating their customers. Increased cheating could be motivated by both financial pressures and as a reaction to unfair treatment (as surveyed smokers view cigarette tax hikes as quite unfair). We examine this question using detailed ride-level data where we can identify a rare but fraudulent overcharging technique (cheating) and a subsample of taxi drivers who smoke (affected taxpayers, identified via tickets for smoking in a cab). In difference-in-differences regressions we find that following a cigarette tax hike, taxi drivers who smoke are approximately 1.5 times more likely to cheat customers than other drivers. Our findings are strongest in the subsample of smokers with consistently low earnings.