A Cross-Regional Analysis of Economic Voting in Post-Communist Russia, 1999-2003

Abstract

Since the 1993 State Duma elections, there had been no ruling party in Russia. Instead many parties, including the Communist Party of the Russian Federation, the Liberal Democratic Party of Russia, the “Yabloko, ” liberals, nationalists, and “parties of power, ” had shared seats in the State Duma. But one of the so-called parties of power, United Russia, won an outright victory in the December 2003 elections. Some existing studies attribute the cause of this party's success to the economic performance of President Vladimir Putin's administration. I think we need to examine the accuracy of this view.To this end, in this article, I use statistical methods and analyze cross-regional variations in election results. It would appear that variations in voting patterns of the Russian electorate are clearly manifested by statistical analysis of the relationship between some economic indicators at the regional level and electoral results at the district level. In this article I use multiple regressions. These time dependent variables consist of PR vote shares of the parties of power at the district level in 1999 and 2003, and independent variables consist of inflation, unemployment and growth rates in each region. But there is a problem with the method of analysis. These regressions concern relationships between 89 regions, federal subjects, and 225 electoral districts. In sum, regions and districts can be conceptualized as a hierarchical structure, with districts nested within regions. To solve the problem, I use multilevel modeling, which allows effective study of the relationships between variables observed at different levels in the hierarchical structure.Judging from the results, the analysis, unlike those of previous studies, demonstrates that in 1999 Russian voters evaluated an anti-inflation policy by their government, but in 2003 they expected a future unemployment policy as well as evaluated a promotion of economic growth. These results may oblige reconsideration of the previous studies.