Party Entry, Exit and Candidate Turnover

We analyse party entry and exit through the lens of candidate turnover using a dataset on 200,000 candidates in 61 Central and Eastern European (CEE) elections. We study candidate turnover in entering and exiting parties as defined by widely used datasets on volatility and new party entry (ParlGov database - Döring and Manow 2016; MARPOR – Volkens et al 2017; Mainwaring, Gervasoni & España-Najera 2017; Powell & Tucker 2014; Tavits 2008). Genuinely entering parties should rely on new candidates and genuinely exiting parties should take most of their candidates with them. Generally, candidate novelty is high amongst new parties, but all of the datasets include prominent cases with low novelty. Several significant parties have intermediate levels of novelty – such partially new parties defy classification as new or continuing. Full party exit is rare as most exiting parties leave behind many important candidates. We complement the quantitative analysis of candidate turnover with in depth discussion of the most problematic cases. The contentious cases of party entry and exit significantly affect volatility indices – particularly those related to party entry and exit (Type A in Powell & Tucker 2014 and extra-system volatility in Mainwaring, Gervasoni & España-Najera 2017). The impossibility of coding partially new parties “correctly” as new or old challenges the dichotomous notion of party novelty. The problem is more common in Central and Eastern Europe than in long-standing democracies, but significant instances of partially new parties are found everywhere (Kadima, Danish People’s Party, En Marche; see Sikk & Köker 2017). We conclude with suggestions on improved ways to measure party system change.