Grigorii V. Golosov, "The Structure of Party Alternatives and Voter Choice in Russia: Evidence from the 2003-2004 Regional Legislative Elections," Party Politics, 12 (November, 2006), 707-725.

First Paragraph:
Throughout the post-communist period, the persistent underdevelopment of political parties has been viewed as a characteristic of Russia's transitional politics (Fish, 1995; Golosov, 1998; McFaul, 2001). Each of the elections to the lower chamber of the country's legislature, the State Duma, produced results that were vastly different from those of the previous elections. Hence the extremely high degrees of party system volatility, which can be conventionally explained with reference to the lack of stable party identifications in the electorate (White et al., 1997). Contrary to such explanations, recent survey-based studies have produced some evidence that anti-party sentiment within Russia's electorate is gradually decreasing over time (Pammett and DeBardeleben, 2000), and that party identifications tend to consolidate rapidly (Brader and Tucker, 2001; Miller et al., 2000; Miller and Klobucar, 2000). However, the data from the most recent Duma elections, held in December 2003, are scarcely in correspondence with these findings. The level of volatility remains very high. It seems that even if party identifications consolidate, voters fail to translate these identifications into actual electoral behaviour.

Figures and Tables:
Table 1. Russian political parties in the regional legislative elections, December 2003 to March 2004
Table 2. Average volatility between national and regional elections in Russia, by time of regional elections
Table 3. Weighted least squares regression of voting in regional legislative elections

First Paragraph of Conclusion:
The working hypotheses advanced in this study have been tested and found valid, which justifies my expectation that voter choice in the regional legislative elections has been strongly affected by the structures of party alternatives. While there are reasons for believing that voter preferences throughout the period under observation remained relatively stable, the results of regional legislative elections differed from the results of the national ones in the same regions. The scope of these differences was systematically dependent on two factors: the number of parties in regional elections and the presence or absence of electoral blocs that are the principal means of expressing local concerns of the electorate. The source of volatility between national and regional elections therefore resided not with the voter but rather with the political parties themselves. Using a market metaphor employed by some of the students of national electoral politics in Russia (Rose et al., 2001), the 'supply side' of the party system is more important than its 'demand side'.