One of the most widely-accepted statements in political science is that personal vote does not exist in closed-list proportional representation systems (CLPR). Under this sort of electoral rules, parties present ranked ballots and voters simply select one list over another. The electoral risk associated with nominating lesser known candidates is weakened by the fact that citizens tend to vote for the party list, not for the individuals on it (Jones et al. 2002). According to this conventional wisdom, it would make no difference whether the candidates are nominated randomly, or if they are simply appointed by parties after the elections, once they know how many seats they need to fill. In addition, there would be no point for candidates in CLPR systems to cultivate a personal vote. In this kind of electoral systems, legislators´ behavior would be heavily constrained by the intra-party selection process but essentially unconstrained by the inter-party competition (Strom 1997). In fact, most seats in CLPR systems are “safe”, and “selection is equal to election”. Taking advantage of a quasi-experimental research design through propensity score matching, we estimate the causal effect of personal vote-earning attributes on vote shares of parties in two Spanish general elections. Specifically, we hypothesize that the presence of local-born politicians and candidates with previous political experience in the party lists increases parties’ vote share in a given municipality through three mechanisms: increased visibility in the media, higher influence in the policy-making process, and successful promotion of pork-barrel legislation. To test this argument, we use a new dataset that contains municipal-level observations of the vote shares of the two major Spanish parties, PSOE and PP, for the legislative elections of 2004 and 2008.
