In flexible list systems, two parallel competitions take place: parties compete for the greatest share of seats while candidates run against their co-partisans to obtain their own seat. This paper investigates the intraparty competition pattern by building a four-category typology of electoral candidates according to both their electoral success and their popularity in terms of vote share. The aim is to identify the profile of the “unlucky winners”, i.e. the candidates who gathered more preference votes than some of their elected co-partisans but who did not obtain a seat because of a too low list position. These candidates reveal a tricky feature of the representative democracy and of flexible list systems where (decentralized) personalization is encouraged but does not really pay off. The paper studies Sweden and Belgium, two countries whose electoral systems provide that the electoral popularity does not automatically lead to the election. On the basis of a database of Swedish and Belgian candidates running for legislative elections and on these candidates’ answers to a candidate survey (the CCS surveys), I quantitatively assess whether the profile and the attitudes of these candidates swimming against the tide and who would benefit from a more personalized system differ from their list counterparts. The paper questions the effects of personalized electoral systems through the eyes of the key actors of personalized politics, i.e. the politicians.
