Working Papers

This dissertation examines how political agency affects climate policy and asset prices. Using the surprise election of Donald Trump as a natural experiment, I study how changes in political leadership influence environmental regulations and financial market outcomes.

We develop a novel measure of partisan corporate speech using techniques from natural language processing. Using the entire corpus of tweets from companies listed on the S&P 500, we first establish a large increase in the amount of partisan corporate speech between 2011 and 2022. This increase in partisan speech is disproportionately driven by corporates using speech commonly associated with Democratic politicians; in particular, statements related to climate change as well as diversity, equity, and inclusion. We also explore how intraday stock returns respond to partisan corporate tweets.

Politically divergent firms are considerably less likely to merge. This effect concentrates in years with high political polarization, and when firms plan to integrate their operations. The results hold after including industry-by-year and deal fixed effects, and controlling for geographical distance, product similarity, and nonpolitical differences in corporate culture or ESG policies. Following politically divergent mergers, employee turnover is higher, particularly of employees whose views misalign with the merger partner. Overall, the rise in political polarization changed the landscape of the real asset market in the U.S., with fewer mergers between politically divergent firms or firms from politically divergent states.

When people exchange ideas, both truths and falsehoods can proliferate. We study the role of explanations for the spread of truths and falsehoods in 15 financial decision tasks. Participants record the reasoning behind each of their answers with incentives for accuracy of their listeners responses, providing over 6,900 unique verbal explanations in total. A separate group of participants either only observe one orator choice or additionally listen to the corresponding explanation before making their own choice. Listening to explanations strongly improves aggregate accuracy. This effect is asymmetric: explanations enable the spread of truths, but do not curb the contagion of falsehoods. To study mechanisms, we extract every single argument provided in the explanations alongside a large collection of speech features, revealing the nature of financial reasoning on each topic. Explanations for truths exhibit a significantly richer message space and higher argument quality than explanations for falsehoods. These content differences in the supply of explanations for truths versus falsehoods account for 60% of their asymmetric benefit, whereas orator and receiver characteristics play a minor role.