Using a new dataset of Italian publicly traded companies between 1994 and 2008, this paper aims to quantify the value of different types of political connections. Conducting an event study on stock returns, we examine how the stocks of connected companies perform before and after the politicians to whom they are connected are either elected to Parliament, or appointed as government ministers. First, we check whether political connections in general lead to abnormally positive returns. Then, we ask whether political connections have a different effect on stock returns, depending on whether companies have politicians themselves (direct connection) or some of their relatives (indirect connection) among their administrators. Finally, we test the hypothesis that connections are effective only when the politicians are members of the governing coalition. Contrary to most studies, political connections are not always associated with positive stock returns. Taking stock performance as a proxy for the benefits of connections, we conclude that only certain political connections are in fact valuable to companies. Being connected with the (future) governing coalition has the predicted positive effect, whereas gaining or maintaining a connection with the opposition coalition has no effect, or even a negative effect. Also, only indirect connections are found to increase the company’s value, while direct connections are not. However, given the size and composition of our sample, we cannot confidently conclude that indirect connections do differ from direct ones.

This dissertation uses Italian data to investigate why parties nominate for reelection incumbents accused of corruption. In some elections, Italian parties renominated incumbent parliament members (MPs) regardless of whether they were accused of corruption by the judiciary. Once they were renominated, accused MPs were reelected to office. In other elections, parties refrained from renominating accused MPs. Non-renominated MPs were then unable to reenter parliament.

To explain this variation, I argue that media coverage of corruption decreases the chances of allegedly corrupt incumbents being renominated. Based on two dimensions of media coverage, I develop two hypotheses. By giving prominence to the issue of corruption, the media increases its public salience, discouraging party leaders from renominating legislators accused of corruption. Prominence of corruption in the media lowers the renomination chances of allegedly corrupt legislators. Then, by reporting on corruption accusations against specific legislators, the media enables voters to identify them and punish their party. When corruption is salient to voters, media mentions of corruption allegations against specific legislators decrease their chances of being renominated.

To test the first hypothesis, I analyze how the Italian media covered corruption before each election. I find a negative correlation between media prominence of corruption and the renomination rates of MPs accused of corruption. For the second hypothesis, I conduct a statistical analysis of MPs' renomination probabilities in two elections. Controlling for relevant confounders, newspaper mentions of corruption allegations against MPs appear to decrease their renomination chances. These findings indicate that media scrutiny promotes electoral accountability by influencing candidate selection.