What kind of Stability? The Survival of Parties and the Programmatic Structure of Party Competition in four Eastern European Countries

The scholarly literature generally considers party system stability a necessary condition of electoral accountability. In case the menu of choices radically changes from one election to another, voters are not able to reward or punish parties which entered government. However, the current paper shows that stability – often understood as the survival of parties – is not sufficient to ensure accountability. The stability of the programmatic structure seems as important since political parties can dramatically shift their appeal over time. Thus, the current paper suggests that we need to systematically take into account the distinction between the stability of party labels and the stability of the programmatic structure of party competition. The highly institutionalized party systems of Western democracies characterized by historic cleavages and high rates of party survival hide this distinction, and they have led scholars to think of stability as a unidimensional concept. Party systems in less advanced democracies provide a more fertile ground to separate the two phenomena. Eastern European party systems provide examples of all four possible combinations defined by the stability of labels and the stability of the programmatic structure. Therefore, the paper sets out to examine four of these countries (Hungary, Latvia, Poland, Romania) and maps the evolution of their party systems over time, from 2004 to 2016. The paper relies on an original novel dataset of parties’ issue positions and emphasis collected by the POLCON ERC project. The dataset covers party competition as reported by two national newspapers during a two-months long window of observation before national elections. The dataset relies on core sentence coding, it maps the salience and positions on the major issues as well as the relationship between competing actors in each campaign. The paper relies on weighted multidimensional scaling to model the programmatic structure of the national political spaces and their over time evolution. In addition, I use party age weighted by the parties’ vote share to measure the stability of party labels over time. The paper confirms the stability of party labels and the stability of the programmatic structure as two distinct dimensions. As the results show, Romania is a country where party labels survive in an unstable programmatic environment. By contrast, Latvia is the opposite case, with a low survival rate despite a stable programmatic structure. Hungary serves to illustrate the case of high stability of both, while Poland shows low levels on both dimensions. Since the period of observation overlaps with the deepest economic crisis in recent history, the paper uses punishment for the crisis to assess how different forms of stability influence electoral accountability. None of the four countries managed to evade the effects of the crisis, therefore, a comparison of a pre- and post-2008 dynamics shows the way they were affected by patterns of electoral punishment. Depending on the form of stability, parties either collapsed or radically altered the programmatic structure of competition. The four types of stability are tentatively explained with the forms of linkage mechanisms which evolved during the transition process.