Party System Change in Romania: Institutional versus Structural Explanations.

For about one decade and a half (1990-2004), party competition in Romania was mostly a conflict between two major blocs, left and right. Things changed radically in the last few years; what we see now is a three-party competition between left (PSD), right (PNL) and center (PDL). In this paper we seek for an explanation for this change. One promising candidate is the electoral system, which was changed prior to the 2008 parliamentary election, from party list PR to 'personalized' (Single Member District) PR. The goal of this change was to give voters the possibility to choose from among candidates, rather than parties. If voters did change their behavior as a result of this reform, 'ideological' voting should decrease, so aggregate support for a party in the previous election (2004), held under party list PR rules, should be a less reliable predictor of support for the same party in the last election (2008) than 2000 results were for the 2004 results. Indeed, if we correlate the locality-level party vote in 2000 and 2004, and then in 2004 and 2008, the former correlations are substantially higher than the latter, which seems to indicate a shift from ideological to personalized voting. However, if we compute the correlations cross-sectionally rather than longitudinally, i.e., we correlate the Chamber and the Senate vote for the same party in 2000, 2004 and 2008, we see that the correlation in 2008 is as high as it was in previous elections, a result that disconfirms the hypothesis of an impact of the electoral system change. Instead, we advance an alternative hypothesis, one that is consistent with both results, namely, that realignment took place in the last few years (post-2004). If that is the case, aggregate support for a party in the previous election should become a less reliable predictor of its support in the last election (2008). Yet on the other hand, voters who switched party allegiance will vote for their new preferred party in both Chamber and Senate elections, so the 2008 correlation will be as high as it was in 2004, a result inconsistent with an institutional explanation (one that looks at the effects of electoral system change), but consistent with a structural (i.e., realignment) account. To show that this is indeed the case, we present a theory of party system change applied to Romania, one that accounts for the change as the joint effect of structural, external, political and institutional factors. We are confident that institutional factors play an important role; however, we argue that the increased importance of the presidency and the staggering of presidential and parliamentary elections play a more important role compared to the change of the electoral system. In the last part we analyze survey and exit poll data to show how the change in the party system is reflected in the changing profile of the constituencies of major parties, particularly the PD-L. [ABSTRACT FROM AUTHOR]