Saggi di economia applicata

In recent years, the relationship between political actors and the economic system has received growing attention. This thesis focuses on corporations and aims to answer two key questions: Who spends money on politics, and why do they do so? We investigate two primary forms of non-market strategy—electoral campaign financing and lobbying—to understand how firms respond to political uncertainty through corporate political expenditures. The first two chapters explore these behaviors in-depth, examining how and why firms engage with political actors and the strategic considerations that drive their political choices. The final chapter shifts focus toward new methodologies, using synthetic data to address privacy concerns and validate economic analyses. These three chapters offer new insights into corporate political activity and the innovative tools that can enhance research in applied economics. In the first chapter, we perform an exploratory analysis to identify which types of firms contribute to electoral campaigns and how they do so. Our findings reveal that most firms adopt a bipartisan strategy, with a preference for the Republican Party. This preference is associated with relatively stable firms that invest less in research and operate in heavy industries such as chemicals, mining, and metals. Beyond these internal characteristics, we also assess external influences on the extensive (whether to contribute) and intensive (how much to contribute) margins. We show that perceived political risk increases the likelihood of campaign financing, particularly when tied to economic issues. Furthermore, rising political polarization in the U.S. Congress correlates with a higher probability of contributing to the electoral campaign and larger total contributions, although the share going specifically to the Republican Party is not significantly affected. Finally, a heterogeneity analysis demonstrates that firms in less research-intensive sectors donate more. Based on these findings, the second chapter explores why firms adopt a bipartisan contribution strategy by focusing on the interplay between campaign contributions and lobbying. I propose an event-study framework to test whether bipartisan contributions are a form of “insurance” against political surprises. Using the unexpected outcome of the 2016 presidential election as a case study, I observe that firms that did not contribute in 2016 significantly increased their lobbying expenditures after Donald Trump’s victory—by up to 46.9% (intensive margin) and 3.9% (extensive margin). To contextualize these findings, I compare the 2016 election with the 2012 election of Barack Obama, where I do not find significant results. I also test whether campaign contributions boost firm profits or returns but find no meaningful evidence in several specifications. Additional robustness checks using the 2008 and 2000 elections further enforce the role of political uncertainty. The 2008 election demonstrates that the insurance-effect results are not linked to first-time presidencies, while the 2000 election of George W. Bush highlights how similar uncertainties can prompt non-contributing firms to lobby in response. In the last part of Chapter 2, I add data from the Federal Procurement Agency to show that this mitigation effect is more pronounced for firms working in sectors that hold more government contracts. The third chapter turns toward a new frontier in artificial intelligence: the generation of synthetic data and its validation for economic analysis to address privacy constraints. To validate this new methodology, we replicate three empirical exercises using synthetic datasets. Our results underscore both the limitations and the advantages of using synthetic data for empirical analysis.