Cross-efficiency of using economic resources: application of DEA and network methodology in regional studies
Elizaveta KorneevaIn this paper I provide the methodology for evaluating effectiveness of using economic resources in the State Duma elections though Data Envelopment Analysis (DEA), which I use for generating the network matrix for further analysis. DEA is a nonparametric method for estimation the effectiveness of Decision-Making Units (DMU). DMU in this case is regions of Russia, which are trying to provide a certain level of support in federal elections. DEA is widely used in different scientific fields, however, in political science still has not seen widespread application. Therefore, DEA in this research measures effectiveness of using different economic resources of the regions to get share of voters in the elections. The aim of regional authorities in this case is to use the minimum amount of economic resources to ensure maximum results on the elections.
The most important advantage of this technique is that it can handle multiple economic variables even those not generally applicable together. I use results of federal elections in Russia to the State Duma that took place in 2003, 2007, 2011 and 2016 in regional dimensions. Therefore, the database consists of many cases of electoral results in the regions that allow to compare a dynamic of effectiveness of economic voting in the same regions in cross-time period. In this case the research of the regions of Russia is relevant because Russia includes 85 subjects of federation.
The output of DEA is United Russia’s share of the vote in State Duma elections in regional dimension. The input variables are a product of the broad electoral behavior literature and anticipated results. As input variables of DEA model, I use the average rate of pensions, the average income per capita, the proportion of the population with incomes below the subsistence level, gross regional product per capita, investment in fixed assets per capita, unemployment rate.
In order to construct network of regions I use DEA cross-efficiency outputs. Cross-efficiency outputs show effectiveness on the region if it gets economic resources of another. I proposed combination of DEA and network methodology which allows to understand why the use of the same economic resources may be more or less effective when it used in other regions.