This thesis focuses on how selection into opportunities and self-selection affect labor market outcomes, particularly focusing on early career and political outcomes. While work and voting decisions are motivated by different sets of determinants, they are strongly affected by institutions that are fundamentally present in modern and democratic economies. Here, I present three instances in which individual decisions – affected by preferences and opportunities – determine collective outcomes.

In the first part of the thesis, we investigate a mechanism to rationalize the empirical evidence of the market failure of occupational licensing. Entry in many occupations is regulated to screen out the least able producers but the available evidence suggests that this objective is rarely achieved. Using microdata covering the universe of Italian law school graduates (2007-2013), we show that this result is due to the strong intergenerational transmission of regulated professions. We find that having relatives already active in the profession substantially increases the probability of passing the entry exam (and earnings), especially so for those who performed poorly in law school. We do not find evidence of intergenerational transmission of occupation-specific human capital. Counterfactual simulations show that positive selection emerges if family connections are assumed away.

In the second part of the thesis, I investigate how knowledge taught at university beyond degrees affects the labor market outcomes of graduates. Using novel data covering the universe of Italian graduates, I find that returns to combinations of bachelor’s and master’s degrees vary substantially even for combinations with the same undergraduate degree, suggesting that both types of programs require consideration. Multidisciplinary university careers relate positively to economic outcomes, while combinations in the same field perform worse. Quantitative courses alone do not explain higher returns.

In the third part of the thesis, we investigate how uncertainty in access to credit affects political outcomes in the US. There is a tight connection between credit access and voting. We show that uncertainty in access to credit pushes voters toward more conservative candidates in US elections. Using a 1% sample of the US population with valid credit reports, we relate access to credit to voting outcomes in all county-by-congressional districts over the period 2004-2016. Specifically, we construct exogenous measures of uncertainty to credit access, i.e. credit score values around which individual total credit amount jumps the most (e.g. around which uncertainty on access to credit is the highest). We then show that a 10pp increase in the share of marginal voters located just around these thresholds increases Republican votes by 2.7pp and reduces that of Democrats by 2.6pp. Furthermore, winning candidates in more uncertain constituencies tend to follow a more conservative rhetoric.