Tavits: The development of stable party support

Tavits. 2005. The development of stable party support: Electoral dynamics in post-communist Europe. AJPS.

In Brief

Like Ferree (2005), Tavits studies 'volatility' (Y). Tavits's purpose is not so much to test her own theory as to pit three approaches against one another. The three approaches:

1. Institutional. When institutions make it easy for new parties to enter the system, there is volatility. Thus, three hypotheses:
   • More ENPP means more volatility.
   • Older parties (on average) means less volatility.
   • Increased ideological distance between the parties means less volatility.
2. Economic voting. If voters make decisions based on recent economic performance, than volatility is highest when inflation is high.
3. Cleavage structures. Unclear what she means by this, but "electoral volatility will be lower in societies with well-structured and salient social cleavages."

Comments and Criticisms

Conceptual Problems

Tavits lacks Ferree's (2005) conceptual clarity. She doesn't seem to know what sort of cleavage structure will reduce volatility. Based on her measurement, she seems to think that increasing ethnic heterogeneity will increase volatility, though it's not clear why. She also seems to think that a close urban-rural split will lead to less volatility than a bigger split. (This seems backwards; if the groups are of equal size, then you would expect more contestation, since both groups can hope to win by mobilizing.)

Compare this with Ferree, who argued that if societal cleavages produce one group that can form a winning coalition on its own, then there will be stability; if there are zero or more than one, there will be volatility. Ferree looked at only one cleavage structure: African ethnicity. I'm not sure how to map her argument onto a setting with multiple cleavages (ethnic vs Russian, rural vs urban, labor vs newly rich, etc), but I don't think Tavits has come close to doing a good job.

In her regression, Tavits claims that the cleavage variables aren't significant. First, she probably should have squared the ethnic heterogeneity variable, at the very least (in light of Ferree's argument). More significantly, she needs some in-depth analysis: How many potential winning coalitions are there in each country? Is there an anti-Russian majority? Is there an (orthogonal) rural coalition? That's two cleavages. What she really needs to do is see whether each country's cleavages overlap (producing one winning coalition) or not.

Empirical Problems

Her study includes several post-communist states, including Russia (see Figure 1). But Russia is a major outlier when it comes to ethnic heterogeneity--and probably on other variables too, given how much larger it is than the other countries (and how much more of an international player it is).