The work began with the idea that all candidates are important to the electoral process – even those whose candidacies are futile. The reasoning emerges from the assumption that candidates' strategic decisions are interactive, that at each point in the campaign process, decisions are made in light of the actions taken by other contestants. A fundamental goal of the work was to capture the effects of these interactions. To achieve this goal within the limits of comparative research, the survey did not randomly sample national candidates, but rather sampled as extensively as possible within case study regions. The regions included in the initial survey were Kostroma, Saratov, Yaroslav'l, and Chelyabinsk. In 1999, the sample regions were expanded to include five ethnically defined republics: Bashkir, Chuvash, Komi, Tartarstan, and Udmurt. The empirical work in this book concentrates on the 1999 data to provide the most extensive comparison among candidates, including a comparison of behavior in Russian regions and the ethnic republics.

The regions included in the study were first identified in 1994. The ethnic republics were chosen to mirror conditions within the first set of Russian regions, but also to provide some variation in the level of titular citizens in each region. These initial regions were selected based on variation on key indicators that were identified as crucial determinants of electoral behavior in Russia and elsewhere. These included region size, population, rural population, percent of pensioners, employment in industry, unemployment, the size of regional budgets, and titular population.