Seeds of Societal Progress: Essays on Economic Inequality, Criminal Justice, and Climate Change

Abstract

Significant economic and social problems threaten 21st-century America, including economic inequality, historic levels of incarceration, and climate change. This dissertation integrates economic theory, econometric techniques, experimental methods, and machine learning to explore how the general public—via public opinion and social networks—can influence, mitigate, or even foster such societal problems. The first chapter develops a social network employment model and shows that minorities suffer disadvantages in economic and social opportunities, all else equal, simply because their social group is smaller—even given the absence of discriminatory intent. I term this phenomenon social network discrimination. The second chapter exploits variation in the timing of district attorney (DA) elections during the steepest rise in U.S. incarceration (roughly 1986–2006) to show that, on average, being in a DA election year increased per capita admissions and months sentenced to state prisons. Interacting results with county characteristics suggests sentencing outcomes respond to public opinion—including anti-Black/pro-White racial bias and preferences regarding the harshness of courts. The third chapter designs and runs a nationally representative survey on 2,430 respondents across the U.S. both to study what factors matter most for U.S. public engagement in climate action, as well as to experimentally test divergent approaches to mobilizing climate action. I find that either communicating a vision for climate change mitigation or communicating its projected damages (or doing both) increases various measures of climate-friendly attitudes. I employ machine learning techniques to identify which groups of Americans are best targeted by such climate-friendly interventions.