Incumbency and the role of parties in re-election

Incumbent MPs very frequently seek and get re-election by relying on the personal vote and consensus they have built and strengthened during the years of their mandate. Parties have no incentive not to re-nominate them (assuming they have the power to do so) to balance their nominations based on gender because it would jeopardize, at least in part, their stable block of consensus (Darcy and Choike 1986). For these reasons, MPs turnover is usually low, and this creates obstacles to entry for those who, like women, present themselves to the electorate for the first time (Darcy 1988; Studlar and McAllister 1991; Matland and Brown 1992; Welch and Studlar 1996). This disparity is stronger where the electoral system is based on the personal vote (such as in single-member districts or open-PR). On the contrary, in closed-list PR, nominations are often the prerogative of parties’ leaders who can include a greater number of women on their lists without excluding powerful incumbents (Andersen and Thorson 1984; Katz 1986). Nevertheless, in PR the position on the list seems to be influenced by the incumbency effect: the first positions are firmly occupied by outgoing parliamentarians, more often men than women, perpetuating an old inequality that struggles to be overcome (Meserve, Pemstein, and Bernhard 2020). The article proposed an innovative research design based on the differential probabilities of election of male and female incumbents. Using an original dataset of the universe of candidates to the Italian national elections from 1948 to 2018, the article advances our knowledge on the incumbency effect: is it a “gendered” phenomenon? Italy is a typical case in Seawright and Gerring’s (2008) typology, but it is also a crucial case because it shows an extremely high internal variability having changed substantially its electoral system four times from 1948 to 2018. The diachronic study of the Italian case allows us to parameterize important variables and to clarify the causal mechanisms identified by comparative studies (George and Bennett 2005). By focusing on a single country, many categories of omitted variables are eliminated because they do not vary within a single country. Furthermore, I can deploy in-depth country knowledge to build and analyse the model and the causal mechanisms more carefully than could be conceivable in a larger cross-country analysis. Results show that being an outgoing parliamentarian significantly increases the chances of being elected, but there are significant differences between the two genders: male incumbents have an additional advantage over female incumbents. The causal mechanism runs mainly through party nomination. Women do not suffer from specific disadvantages compared to men at the time of (re)nomination, but they are placed in lower list-position in case of closed-list PR and, in case of personal-vote systems, they are moved to a different district (thus losing their “incumbent capital”) more often than men. In case of open PR, voters do not overly reward outgoing men compared to outgoing women, and, in SMD, the balance of power is reversed: outgoing women get, on average, more votes than outgoing men.