In Chapter 4, we began to consider the conditionality of electoral system effects through an analysis of the number of parties in mixed-member systems. We saw that FPTP rules produce their expected effects in established democracies, most notably by constraining the number of parties. At the same time, we saw a very different result under FPTP rules in new democracies, which tended to produce many more parties. We argued that these different outcomes were due to far greater strategic defection in established democracies by candidates and their supporters who faced little chance of success in SMDs, but this argument was merely conjecture; we did not demonstrate the behavior that underlay the different results. To get at these mechanisms, we need to delve more deeply into whether voters and elites in different contexts do, in fact, behave in this way. We need to address the question: Is there, in fact, more strategic defection in established democracies than in new ones? In Chapters 5 and 6, we address this issue of strategic defection under different contexts.

In this chapter, we study votes cast under mixed-member systems to learn more about the factors that shape voting behavior. Our principal aim is to explore the conditions under which voters will be likely to cast strategic ballots, whereby they withdraw support from their preferred candidate in order to affect the race, but our strongest findings in this chapter relate to the personal vote. In the study of mixed-member electoral systems, work on strategic voting is particularly well established. Ticket splitting in which a greater number of votes are cast for large parties in the SMD tier than in the PR tier (and a smaller number of votes are cast for minor parties in SMDs) has been put forward as evidence that voters react strategically to restrictive electoral rules – such as FPTP – that tend to deny representation to minor parties. On this topic, the literature gives substantial attention to strategic voting, but does not account sufficiently for another factor that can drive ticket splitting: the personal vote – additional SMD votes cast for a candidate due to the candidate's personal appeal to voters.