CC BY
60Effect of Income on Political Preferences of Russian Voters*
Rustem NUREEV, Doctor of economics, Professor
Head of Macroeconomics Department, Financial University; Professor at National Research University - Higher School
of Economics, Moscow
nureev50@gmail.com
Sergey SHULGIN, Ph.D. in economics
Russian Academy for National Economy and Public Administration (RANEPA), Moscow sergey@shulgin.ru
Abstract. There was direct correlation between the voters' income and electoral support for incumbent in Russia during the 1990-s and early 2000-s. The results of election to the State Duma (the parliament) in 2011 and Russia's presidential elections in 2012 show the opposite. Using income data and electoral results in the Russian regions for each candidate (G. Zyuganov, S. Mironov, V. Zhirinovsky, M. Prokhorov, V. Putin) we defined the level of electoral support in different income groups. Results show the substantial changes in last 8 years in voting behavior. There is the effect of Putin's inversed threshold and the greatest proportion of votes negatively correlated (-1.58), with a group of people with incomes of 14,250 to 21,250 rub/month. Such inverse correlation may be due to a protest voting. Putin's electoral support has a positive correlation in low-income group. In this paper we analyze the determinants of voting behavior and show how the income distribution affects the voters' political preferences (based on the results of the presidential elections in 2012). For each candidate we defined the level of electoral support in different income groups. Also we analyzed income distribution of absent voters.
Аннотация. В статье анализируются детерминанты электорального поведения и показывается, каким образом распределение населения по доходам влияет на политические предпочтения избирателей (по итогам федеральных выборов 2012 г.) Определен пороговый уровень доходов, при достижении которого люди начинают проявлять социальную активность и заинтересованность в участии в электоральной системе (демократии). Показано влияние распределения доходов в российских регионах на политические предпочтения избирателей.
Key words: Political preferences, regional studies, electoral behavior, income distribution.
1. review
S. Kuznets was the one of the first who showed the importance of the distribution of income inequality for economic growth and social and economic progress (Kuznets, 1955; 1971; 1979). However, the focus of his research was not the problem of electoral behavior of voters. Almost ten years later, it becomes the subject of a special study (Lewis-Beck, 1988). Analyzing the Western democracies, the author suggested indirect, but reliable way to assess the economic factors on the electoral process. The book summarizes and complements the classical set of economic factors to explain the behavior of voters. M. Lewis-Beck believes that the majority of voters rarely appeal to the main macroeco-
nomic indicators in assessing the economic situation and prospects of the economy.
According to M. Lewis-Beck, changes in voters' disposable income also have minor effects on electoral behavior. This paradox can be partly explained by the weak faith of the population in the government's ability to influence the personal financial situation (by arguments like: "The economic policy of the government matters, but does not affect me"). According to sociological researches in 1980-s, influence of government policy on personal well-being was felt by only 45% of population in UK, 44% in France, 40% in Germany, 34% in Italy, 49% in Spain, and only 20% in USA. Lewis-Beck uses his survey to show that in almost all the developed capitalist countries, the economic
reasons are the most important in the vote. The motives of party self-identification (right/left in Europe, the Republicans/Democrats in the U.S.), appear much stronger than the motives of social or religious identity. The following factors strengthen the economic value of vote: the openness of the national economy, economic growth or its expectation, the presence of the ruling coalition or single party government. Among the developed countries studied by Lewis-Beck, the economic motives of the electoral behavior are mostly significant in the United States. In any case, the economic motive affects voting through the personal assessment of the economic development of the country's voters.
For assessment Lewis-Beck suggests three components: "Retrospective" (evaluation of the past compared to the present), "Prospective" (assessment of the future) and "Affective" (unexplained irritation, etc.). According to his study, the most important is "Prospective" evaluation of public policies, the second — "affective" component and the last — "retrospective" component of assessment.
Respondents were asked to rate the influence of the government on unemployment, inflation, personal well-being, balance of trade, economic growth, public debt and a number of other parameters. Unemployment was the most important parameter in all countries (UK, France, Germany, Italy, Spain) and inflation was on the second place. Other parameters (such as personal well-being, balance of trade, economic growth, public debt, etc.) were significantly less important than unemployment and inflation. We can see from Lewis-Beck that voters in Western democracies assess their economic situation and current trends primarily through their assessment of the future.
According to R. Kiewiet and D. Rivers (1984) voters are not inclined to attach great importance to the current macroeconomic situation. Authors believed that voters were rather farsighted than myopic, and votes do not tend to react with enthusiasm for the short-time economic improvement. Voters do not live by one day and are able to assess the dynamics of the economic situation. The authors in their studies used "Eurobarometer" data by George Gallup Institute. The authors suggested that economic motives of voting were particularly strong in the case of deterioration of the situation. Growth of economic indicators, as it turned out, did not lead to a significant increase of electoral support for incumbent. Economic growth matters only in case of a sharp change of direction in economic development (the typical example — Ronald Reagan's victory in the presidential election in 1984).
A. Sobyanin and B. Suhovolskiy (1995) studied the electoral process in Russia and demonstrated numerous examples of electoral frauds using electoral sta-
tistics. According to A. Lavrov (1997) social structure affects voters' political preferences. Lavrov argued that the higher share of urban population and share of population employed by the government (in public administration and state industry) and the share of people with tertiary education lead to the stronger electoral support for centrist and democratic candidates. And vice versa, support for the left politicians in 1990-s increased with a higher share of rural and agrarian population and with higher share of pensioners.
L. Smirnayagin (1999) studied the stability of political preferences and proposed a "degradation index", to explain the shifts in voters' political preferences. He estimated degradation index for Russia in 1990-s as 0.54. This means that 54% of the voters were ready to change their political preferences in the next election. This high percentage of voters who were ready to switch their preference means that formation of civil society in Russia is uncompleted.
V. Mau, O. Kochetkova, K. Yanovsky, S. Zhavoronkov, Yu. Lomakina (2000) studied the impact of different economics indications on electoral behavior. They argued that in late 90-s (1995-2000) the most important for electoral behavior were income and wages, tax payments, share of urban population. At that period the higher was the voters' income (wages etc.) the higher was support for ruling party. The similar findings were in later studies by O. Kochetkova (2004), according to which the support for incumbent politicians positively correlated with incomes and negatively correlated with unemployment and wage arrears.
U. Seresova (2005) agued that economic indicators were significant for the electoral process but were not the most important ones. She suggested that electoral behavior was better explained by the level of regional modernization and the role of traditional culture.
However, most of studies analyzed the situation of the electoral behavior of the 1990-s and early twenty-first century. In this paper we deal with a new political reality. In this article we further develop the approach suggested by S. Shulgin (2005) who examined how income distribution in different countries affected democratic institutions. Author used income distribution to analyze the levels of freedom of press measured by Freedom House.
2. DATA
In this paper we use official Russia's electoral statistics for presidential election 2012. All our findings consequently contain errors associated with reliability of official electoral statistics. There is an extensive literature that indicated the frauds during Russian elections. We discussed this problem in several articles (Economic Sub-
jects, 2010, etc.). The article (Enikolopov et al., 2013) discusses the results of the parliamentary elections in 2011. Authors compared election results in Moscow precincts attended by independent observers, with the election results in precinct where observers were not allowed.
The second part of our data describes income levels and income distributions in Russian regions. This statistics come from Russian Statistical Agency (RusStat). RusStat estimates income distribution based on data from the Household Budget Survey (HBS). Household Budget Survey was carried out by state statistics on a regular basis in all regions of the Russian Federation. The unit of observation in this survey is the household and its members.
3. DATA ANALYSIS
Using statistics on income level and income distribution, for each region we construct income distribution function. Income distribution function for given level of income evaluates the share of people within region who have such level of income.
RusStat's statistical yearbook "Regions of Russia" (Regiony Rossii, 2011) in Table 5.9 gives the distribution of population by per capita income (as a share of regional population). Table 5.8 (from yearbook) gives the share of total income by 20 per cent population groups (from the poorest 20% of population to the richest 20% of population).
We use data on income distribution and per capita income level to construct cumulative function that shows how many people has income below a certain value. For example, in the Belgorod region 4.2% of people have incomes of up to 3,500 rubles, 6.3% from 3 500 to 5 000 rubles, 10.6% from 5 000 to 7 000 rubles, 16.2% from 7 000 to 10000, 21.4% from 10 000 to 15 000, 23.0% from 15 000 to 25 000, 9.5% from 25 000 to 35 000 and 8.8% of disposable income — over 35 000 rubles per month (see Table 1).
Then, in the Belgorod region cumulative function of income shows that 4.2% of people have incomes of up to 3 500 rubles, 10.5% to 5 000 rubles, 21.1% up to 7 000 rubles, 37.3% to 10 000, 58.7% to 15 000, 81.7% to 25 000, 91.2% to 35 000, and the remaining 8.8% — of disposable income over 35 000 rubles a month.
For each region we build a linear approximation of the distribution function of per capita income (see the example in Figure 1). To determine how many people in the Belgorod region have income less than 6 000 rubles, we find average on intervals of distribution function 5 000 (10.5%) and 7.000 (21.1%), and the resulting 15.8%.
Income distribution data exist for 82 Russia's regions (for all 83 regions in Russia, with exception of Republic of Chechnya).
We use income distribution functions for each region to construct the variable "share of population with incomes below X". Figure 2 shows the distribution of Russia's regions by this variable ("share of population with incomes below X") on 4 different X (Fig. 2a X=5000 rubles per month, Fig. 2b X=10000 rubles per month, Fig. 2c X=20000 rubles per month, Fig. 2d X=30000 rubles per month).
We use electoral statistics to construct electoral variable "the share of votes for candidate N". We estimated "the share of votes for candidate N" as a share of voters participated in president election. We constructed electoral variables for all five candidates (Zhirinovsky, Zyuganov, Mironov, Prokhorov, Putin). Also we constructed electoral variable for non-voters — as a share of voters who were registered but did not participate.
Next, we looked for correlations between income variables "share of population with incomes below X" and electoral variables "the share of votes for candidate N".
On Figure 3 presented scatterplots for electoral variable "share of votes for Zhirinovsky" and income variable "share of population with incomes below X" for different X (Fig. 3a X=5000 rubles per month, Fig. 3b X=10000 rubles per month, Fig. 3c X=20000 rubles per month, Fig. 3d X=30 000 rubles per month). On each scatterplot on Figure 3, vertical axis represents the same electoral variable ("share of votes for Zhirinovsky") and horizontal axis — income variable "share of population with incomes below X" for different income levels (5000, 10000, 20000, 30000 rubles per month)
Figures 4, 5, 6, 7, 8 present scatterplots ("income variable" vs "electoral variable") for other candidates (Fig. 4: for Zyuganov, Fig. 5: for Mironov, Fig. 6: for Prokhorov, Fig. 7 for Putin, Fig. 8: for non-voters.)
Table 1. Distribution of population by per capita income (as a percentage of the total for the Belgorod Region, 2010).
Per capita income, rub. per month
to 3500,0 from 3500,1 to 5000,0 from 5000,1 to 7000,0 from 7000,1 to 10000,0 from 10000,1 to 15000,0 from 15000,1 to 25000,0 from 25000,1 to 35000,0 more 35000,0
Belgorod region 4,2 6,3 10,6 16,2 21,4 23,0 9,5 8,8
Figure 1. Example of cumulative distribution function approximation of average monthly income (for the Belgorod region, 2010).
a)
b)
c)
d)
Figure 3. The share of votes for Zhirinovsky (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
a)
Zhirinovskiy share of income less than 5000
CD oo
8o„ o ° ° OSJ 0Cd O
o O 0 0 " 0 o" o
0
tot_share[, 6]
b)
c)
d)
4. MODEL: ELECTORAL BEHAVIOR AND INCOME DISTRIBUTION
Previously we defined electoral variables as "the share of votes for candidate N" and income variables as "share of population with incomes below X". In our analysis we are looking for correlations between electoral variables and income variables. We analyze such correlations on all possible income levels (up to 100000 rubles per month).
To analyze correlation between electoral and income variables we used model of simple pair regression (1):
Share of votes = a + a x
- - 01 (1) x Share _of _population_with_income_less_than_X + e.
The advantage of this approach is simplicity (since we use a large number of such pairs of simple regression to assess the most relevant interval). At the same time, simple regression leaves many possible interpretations in addition to correlation between the independent and the dependent variables. For example, we can expect that income depends on other variables, which also affect the electoral preferences (level of urbanization,
education level, gender, age, etc.). Realizing that this approach can be criticized, we nonetheless underscore its advantage. It reveals the link between income and electoral support for the candidate. Many other important variables (education, urbanization, gender, age) are correlated with income, but we are interested in correlation between electoral behaviors of different income groups.
Figure 9 shows the distribution parameter estimation of the set of regressions where the dependent variable is the share of the vote for Zhirinovsky, and the explanatory variable is the "share of population with incomes below X". Figure 9a shows the distribution of F-statistics, and Figure 9b — the distribution of t-statistics of the coefficient of the explanatory variable.
In simple regression F-statistics coincides with the absolute value of t-statistics, we use t-statistics when the sign is important. Sign in the t-statistics is the sign of correlation between dependent and independent variables. The negative sign indicates the negative correlation between the share of votes for a candidate and a share of people with certain level of income.
In Figure 9b points 1-4 correspond to the results of the regression estimates, based on data that are dis-
a)
Zuganov share of Income less than 5000
O
o
o il
O^J 0
trcrf
tp
0
o
o o O
&
20
40
60
SO
100
tot_share[, 5]
b)
c)
d)
Figure 4. The share of votes for Zyuganov (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
played in Fig. 3a — 3d. Point 1 in Fig. 9b corresponds to the t-statistic (-4.60) for the coefficient of the explanatory variable b (-0.155) regression, based on data in Fig. 3a (for the income share of less than 5 thousand rubles). Point 2 in Fig. 9b corresponds to the t-statistic (-3.75) for the coefficient of the explanatory variable b (-0.0687) regression, constructed from data in Figure 3b (for revenue share is less than 10 thousand rubles.). Point 3 in Fig. 9b corresponds to the t-statistic (-2.818) for the coefficient of the explanatory variable b (-0.045) regression, based on data in Fig. 3c (for revenue share is less than 20
thousand rubles.). Point 4 in Fig. 9b corresponds to the t-statistic (-2.198) for the coefficient of the explanatory variable b (-0.045) regression, based on data in Fig. 3c (for revenue share is less than 30 thousand rubles.)
In addition to the four points (1-4), for which we have provided examples of the distribution of votes and the percentage of people with a certain level of income (in Fig. 3a-3d), the graph 9b contains coefficients of t-statistics for the income groups built around a set of distributed income from 0 to 100 thousand rubles. Five percent significance level t-statistics (for 82 observations)
a)
Mironov share of income less than 5000
0
O Go
o Ô o o Vr
* Is*
Ô
o G ù ° O o
<3
20
40 60
tot_share[, 5]
SO
100
b)
c)
E V
d)
Mironov
share of income less than 20000
o
0 O ö
ÖÖ
0 O
Q 0 0 ûo
% 00 o
û o O
çp
O o
D o
o <3 O O
o
0 Û
o 0
o
s
20
40 GO
tat_share[. 21]
30
100
Figure 5. The share of votes for Mironov (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
corresponds to the level of 1.99 (5%), which on the Fig. 9b reaches a level of income 35 thousand rub. The coefficient of the variable "proportion of people with incomes below the X" is no longer statistically significant when x is greater than 35 thousand rubles per month, in regressions explaining the share of votes cast for Zhirinovsky.
Similarly graphs 9a, 9b present the results of regressions explaining the share of votes for Zhirinovsky's presidential election in 2012, if the schedule 9a contains the results of regression in which the share of votes for Zhirinovsky explained by the percentage of people with
incomes from 0 to X, and a deferred variable on the horizontal axis, then the graph 9c shows the results that explain the voting share for Zhirinovsky in the proportion of people with income from Y to X. The curves shown in the graph 9a, a special case of the reduced dependence in graph 9c (at Y = 0).
In the graph 9c we consider all possible income groups, for example, not only income group from 0 to 5000 (point 1 on the chart 9a and Figure 3a), but also of income from 1000 to 5000, from 2000 to 5000, from 3000 to 5000, from 4000 to 5000, not only income group
a)
b)
c)
d)
Figure 6. The share of votes for Prokhorov (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
0 to 20,000 (as a point on the graph 3 9a and Figure 3c), but also of income from 5000 to 20000, 10000 to 20000, 15000 to 20000.
Zhirinovsky remains relevant in high-income areas, which suggests that a certain number of supporters of Zhirinovsky are present among middle-income voters, and among the richest of the voters.
Fig. 10 shows examples of the distribution of population groups with income from Y to X for the Belgorod region (Fig. 10a) and in Moscow (Figure 10b). Each point on this graph represents the percentage of people (axis
Z) in the region with an income in the range from Y to X. For example, a group of people with incomes between 20 and 60 thousand rubles (X = 60000, Y = 20000), in the Belgorod region corresponds to the value of Z = 24.38 (i.e. the number of 24.38% of the total population), while in Moscow Z = 34.13 (i.e. 34.13% of the total Moscow's population has income of 20 to 60 million).
Thus, the graph 9c coordinate Z (height above the plane XY) has the value of F-score statistics regression, in which the share of votes for Zhirinovsky explained by the proportion of people with income from Y to X.
a)
b)
c)
d)
Putin
share of income less than 20000
o «g,
0 o Û
ú o o
□ Û O
0 ® □ « L J®*® o
Ö o°a °%>
4
20
40
60
SO
100
tat_share[, 21]
Figure 7. The share of votes for Putin (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
Analysis of the results shown in the graph 9c shows that there is a hump with a lower limit of 7000 rubles. Grey-black color on chart 9d shows the coefficient of the independent variable positive (b = 0.15). Simplified interpretation of this threshold can be illustrated by a hypothetical example. If the region of 1000 voters passed a group of people with incomes up to 7000 rubles a month in a group of people with incomes above 7,000 rubles, then 150 of them will vote for Zhirinovsky.
Similar to Fig. 9, we construct the graphs for the other candidates for president of Russia. Figure 11 shows the
distribution parameter estimation of the set of regressions in which the share of the vote for Zyuganov, due to "percentage of the population living below the X rubles."
The analysis of results (see Figure 11a-11d) estimates regressions on the entire set of groups and their ability to explain votes received by Zyuganov in the presidential election of 2012. Graphs 11a and 11b show that the groups with the boundaries from 0 to 28000 are insignificant (at the 5% level). The importance of communication with the vote for Zyuganov groups having certain income shows an increase in the lower limit of the group
Figure 8. The share of non-votes (vertical axis) vs. "share of people with incomes less then": a) 5000 rubles per month, b) 10 000 rubles per month, c) 20 000 rubles per month, d) 30 000 rubles per month.
a)
b)
c)
d)
and reaches a maximum for the group with income from 9250 to 21,750, an increase in the upper limits of income, the importance of a vote for Zyuganov again disappears (height "hill" decreases with increasing upper limit). Grey-black (see Fig. 11d) shows that, for this group of the population with income from 9250 to 21,750, the coefficient of the independent variable is positive (0.35). This means an increase in the group for the 1000 population, increases the vote for Zyuganov at 350. For Zyuganov, we can also select a group of high-income (ranging from 40
thousand to 60 thousand), the size of which is negatively related to the share of Zyuganov votes (the white dots area in the graph 9d). Apparently the presence of this group explained the significance of the positive impact of falling after reaching a maximum in the range of 9250 to 21,750 rubles per month.
Fig. 12 shows parameters distribution of estimation of set of regressions in which the share of votes for Mironov is determined by "percentage of the population living below the X rubles."
Iitll lull IhM ¡hII lull litis
OhII lull lull lull Iitll litis
a) F-statistics
b) t-statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistics (grey-black indicates the areas of the positive correlation; white dots - negative correlation; light grey - not statistically significant at the 5% level)
a) An example of the distribution of income in the Belgorod region
b) An example of the distribution of income in Moscow
Figure 10. Examples of the distribution of population groups with income from Y to X.
Analysis of the influence of the size of the set of different income groups for votes received by Mironov in the presidential election shows a general pattern similar to one, which can be observed in the data for Zhirinovsky. While with the increase of the lower limit of the size of a significant association of the wealthy with the number of votes for Mironov disappears, unlike for Zhirinovsky. This suggests a narrower area in which the electorate is concentrated. The maximum is reached in the range of 13,250 to 24,250, and with the growth of the upper boundary, the relationship to the number of votes for Mironov is falling faster than for Zhirinovsky, i.e. within this population (with incomes greater than 13,000 rubles a month) voters with increasingly higher incomes are less likely to support Mironov.
Fig. 13 shows the parameters distribution of estimation of set of regressions in which the share of votes for Prokhorov is determined by "percentage of the population living below the X rubles." Analysis of the results of the evaluation of the set of regressions explaining the vote received Prokhorov shows that he has a certain threshold value (15000 rubles a month), above which the voters are beginning to support Prokhorov. However, a more important feature of the Prokhorov's electorate is that it belongs to the highest income group. Increase of the lower limit does not reduce the significance, and a group with a very high lower bound (more than 40 thousand rubles a month) also has a positive effect on the share of the vote for Prokhorov.
Fig. 14 shows the parameters distribution of estimation of set of regressions in which the share of votes for
Putin is determined by "percentage of the population living below the X rubles." Analysis of the results of the evaluation of the vote for Putin showed no significant relationship for almost the entire set. The observed significant dependence is similar to patterns seen for Mironov and Zhirinovsky, but the coefficients of the explanatory variables b is minimal (b = 0.01), although statistically significant. For Putin, as well as for Zhirinovsky and Mironov, there is an income threshold (14,250) less obvious than for Zyuganov, the upper limit of the population (at 21,250). However, the white dots area on the chart 14d shows that the correlation coefficient of this group with the votes cast for Putin is negative (-1.58), the area of positive correlation vote for Putin is in the area with low incomes. Thus, there is an opposite threshold effect — after reaching the income threshold (14,000) voters reduce support for Putin.
Fig. 15 shows the parameters distribution of estimation of set of regressions in which the share of voters who do not come to the polls is determined by "percentage of the population living below the X rubles."
The analysis of the proportion of voters who did not participate in the presidential elections shows that, unlike shares cast for candidates, there are no incomes negatively affecting the proportion of people who took part in the vote. We see that the greatest proportion of the electorate who voted is not associated with a group of people with incomes of 13,100 to 22,400, with the upper boundary of the growth, this relationship becomes less significant, but another local maximum is achieved for the group from 4000 to 90000.
a) F- statisticcs
b) t- statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistics (grey-black indicates the areas of the positive correlation; white dots - negative correlation; light gray - not statistically significant at the 5% level)
a) F- statistics
b) t- statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistic (grey-black indicates the areas of the positive correlation; white dots - negative correlation; light gray - not statistically significant at the 5% level)
a) F- statistics
b) t- statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistic (grey-black indicates the areas of the positive correlation; white dots - negative correlation; light gray - not statistically significant at the 5% level)
a) F- statistics
b) t- statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistic (grey-black indicates the areas of the positive correlation; white dots - negative correlation; light gray - not statistically significant at the 5% level)
a) F- statistics
b) t- statistics
c) F-statistics (for regressions explaining the share of votes for Zhirinovsky by the share of population with income more than Y and less then X)
d) F-statistics (grey-black indicates the areas of the positive correlation; light gray - not statistically significant at the 5% level)
A possible explanation for this is the presence of a closer relationship between the share of non-voters and the population with low or high incomes, while at the intermediate level this relationship is somewhat weaker.
5. PRELIMINARY RESuLTS
According to the previous research (Mau, Kochetkova, Yanovsky, Zhavoronkov, Lomakina, 2000; Kochetkova, 2004; others) there was direct correlation between voters' income and electoral support for incumbent in Russia during the 1990-s and early 2000-s. The results of Russian presidential elections in 2012 show the opposite trend. For each candidate we defined the level of electoral support in different income groups.
Firstly, the effect of the income threshold of votes for certain candidates (Zhirinovsky: 7000 rubles per month, Mironov: 13,000 rubles per month): people change their behavior when it reaches the threshold. At the same time Mironov's electorate concentrated in a narrower range of income, while Zhirinovsky has a significant proportion of voters among the citizens with a high level of income.
Secondly, a special case represents the electorate of Zyuganov, whose electorate is formed by a group of people with "average" income, for which the lower and upper limits are defined (from 9250 to 21 750 rubles per month). Thirdly, the high-income groups of population (with incomes of 40 thousand rubles a month) are mostly associated with the electorate of Prokhorov. This suggests that the growth of income potentially increases the electoral support of the candidates of this type.
Fourthly, there is the effect of Putin's return threshold and the greatest proportion of his votes negatively correlated (-1.58) with a group of people with incomes of 14,250 to 21,250. Inverse correlation may be due to a protest vote against the representative of the party in power. The zone of positive correlation of votes for Putin is in the low-income area. In the future, we plan to look more closely at regional differentiation factors that more accurately compare data from different regions to refine the preliminary results of our research.
REFERENCES
Eggertsson T. (1990). Economic Behavior and Institutions: Principles of Neoinstitutional Economics. Cambridge: Cambridge University Press.
Enikolopov R., Korovkin V., Petrova M., Sonin K., Zakharov A. (2013). Field experiment estimate of electoral fraud in Russian parliamentary elections. Proceedings of the National Academy of Sciences of the United States of America, 110, 2, 448-452. Furubotn E.G., Richter R. (1997). Institutions and Economic Theory. The Contribution of the New Institutional Economics. Ann Arbor, MI: University of Michigan Press.
Goldman M. (2003). The Piratization of Russia: Russian Reform Goes Awry. L., N.Y.: Routledge.
Gregory P. (1998): Russian and Soviet Economic Performance and Structure. P. Gregory, R. Stuart (eds.). N.Y.: Addison-Wesley.
Herschenkron A. (1962): The Approach to European Industrialization: a Postscript. Economic Backwardness in Historical Perspective: A Book of Essays. Cambridge: Harvard University Press.
Kahkonen S., Olson M. (eds.) 2001. A New Institutional Approach to Economic Development. New Delhi: Vistaar Publications.
Kiewiet D.R., Rivers R.D. (1984). A Retrospective on Retrospective Voting. Political Behavior, 6, 1-26.
Kochetkova O. (2004) Economic factors in voting behavior. Ph.D. thesis, Moscow.
Kuznets S. (1955). Economic Growth and Income Inequality. American Economic Review, 45, 1, 1-28.
Kuznets S. (1971). Lecture to the Memory of Alfred Nobel, December 11.
Kuznets S. (1979). Growth, Population, and Income Distribution: Selected Essays. N.Y.: Norton.
Laranjaa M., Uyarrab E., Flanagan K. (2008). Policies for Science, Technology and Innovation: Translating Rationales into Regional Policies in a Multi-Level Setting. Research Policy, 37, 823-835.
Lewis-Beck, M.S. (1988). Economics and Elections: The Major Western Democracies. Ann Arbor, MI: University of Michigan Press.
Long-Term Social and Economic Development of Russian Federation (2008). M.: Ministry for economic development of the Russian Federation.
Mau V., Kochetkova O., Yanovskiy K., Zhavoronkov S., Lomakina Yu. (2001). Economic Factors of Electoral Behavior and Social Consciousness (the Experience of Russia 1995-2000). M.: IEPP
Mueller D.C. (2003). Public Choice III. A Revisited Edition of Public Choice. Cambridge: Cambridge University Press.
North D. (1991). Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press.
Olson M. (1982). Rise and Decline of Nations. New Haven: Yale University Press.
Olson M. (1993). Why is Economic Performance even Worse after Communism is Abandoned. Virginia political economy lecture series. Fairfax. Virginia. March 17.
Olson M. (2000). Power and Prosperity: Outgrowing Communist and Capitalist Dictatorships. N.Y.: Basic Books.
Ordeshook P.C., Shvetsova O. (1999). If Hamilton and Madison were Merely Lucky? What Hope is there for Russian Federalism? Constitutional Political Economy, 6 (2), 107-127.
Persson T., Tabellini G. (2000). Political Economics. Explaining Economic Policy. Cambridge, MA: MIT Press.
Seresova W. (2005) Social factors of the electoral process in the regions of Russia or Who voted poor. Logos 1 (46).
Shulgin S. (2005). Income Inequality and the Demand for the Independent Mass Media. Public Choice Society Conference paper.
Solow R. (1956). A Contribution to the Theory of Economic Growth. Quarterly Journal of Economics, 70, 65-94.
The Institutional Economics (2005). The Institutional Economics of Russia's Transformation. Burlington: Ashgate.
Tullock G. (1994). The New Federalist. Vancouver: The Fraser Institute.
Appendix 1
Population, by per capita income in 2010
(as a percentage of the total population of the subject of Russian Federation)
Per capita income, rub. per month
to 3500,0 from 3500,1 to 5000,0 from 5000,1 to 7000,0 from 7000,1 to 10000,0 from 10000,1 to 15000,0 from 15000,1 to 25000,0 from 25000,1 to 35000,0 over 35000,0
The Russian Federation 3,9 5,6 9,4 14,7 20,2 23,5 10,8 11,9
Central Federal District
Belgorod region 4,2 6,3 10,6 16,2 21,4 23,0 9,5 8,8
Bryansk region 6,2 8,9 13,8 19,1 22,0 19,4 6,4 4,2
Vladimir region 5,5 9,1 14,8 20,6 23,1 18,7 5,3 2,9
Voronezh region 7,4 9,3 13,6 18,2 20,9 19,0 6,6 5,0
Ivanovo region 7,5 11,2 16,9 21,6 21,8 15,5 3,8 1,7
Kaluga region 4,4 6,9 11,7 17,5 22,3 22,3 8,4 6,5
Kostroma region 5,4 8,8 14,5 20,4 23,1 19,1 5,6 3,1
Kursk region 4,4 7,2 12,2 18,2 22,7 21,9 7,8 5,6
Lipetsk region 3,8 6,4 11,1 17,2 22,5 23,2 8,8 7,0
Moscow region 2,0 3,6 7,0 12,3 19,4 25,9 13,3 16,5
Orel region 7,8 9,7 14,2 18,7 20,9 18,3 6,1 4,3
Ryazan region 4,8 7,9 13,3 19,3 23,1 20,7 6,7 4,2
Smolensk region 4,2 7,0 12,1 18,2 22,9 22,1 7,9 5,6
Tambov region 7,6 9,3 13,6 18,1 20,7 18,9 6,7 5,1
Tver region 3,7 7,0 12,6 19,5 24,2 21,9 7,0 4,1
Tula region 3,6 6,4 11,3 17,7 23,1 23,2 8,5 6,2
Yaroslavl region 4,9 7,6 12,4 18,2 22,4 21,4 7,6 5,5
Moscow 1,0 1,9 3,8 7,1 12,4 20,4 13,9 39,5
North-Western Federal District
Karelia Republic 2,4 5,2 10,3 17,5 24,2 25,0 9,1 6,3
Komi Republic 2,4 3,9 7,1 12,1 18,6 24,8 13,1 18,0
Arkhangelsk Region 1,9 3,9 7,9 14,2 21,7 26,7 12,1 11,6
including the Nenets Autonomous District 0,3 0,6 1,6 3,8 8,5 18,4 15,6 51,2
Vologda Region 4,5 7,6 12,8 19,0 23,2 21,2 7,1 4,6
Kaliningrad Region 3,4 6,0 10,9 17,3 23,0 23,7 8,9 6,8
Leningrad Region 4,4 7,2 12,2 18,2 22,8 21,9 7,8 5,5
Murmansk region 1,1 2,4 5,4 10,7 18,7 27,4 15,0 19,3
Novgorod region 5,1 7,3 11,7 17,1 21,4 21,7 8,5 7,2
Pskov region 6,3 9,3 14,5 19,8 22,3 18,7 5,7 3,4
St. Petersburg 2,7 4,0 7,0 11,6 17,6 23,9 13,1 20,1
Southern Federal District
Republic of Adygea 7,9 10,3 15,0 19,6 21,2 17,4 5,3 3,3
Republic of Kalmykia 21,7 18,0 19,5 18,3 13,6 7,1 1,3 0,5
Krasnodar Territory 5,0 6,9 11,0 16,1 20,8 22,1 9,2 8,9
Astrakhan region 5,9 8,1 12,5 17,7 21,4 20,7 7,6 6,1
Volgograd region 4,0 7,2 12,6 19,0 23,6 21,8 7,2 4,6
Rostov region 5,8 8,1 12,7 17,9 21,6 20,6 7,5 5,8
Per capita income, rub. per month
to 3500,0 from 3500,1 to 5000,0 from 5000,1 to 7000,0 from 7000,1 to 10000,0 from 10000,1 to 15000,0 from 15000,1 to 25000,0 from 25000,1 to 35000,0 over 35000,0
North-Caucasian Federal District
Dagestan Republic 4,9 7,3 11,9 17,5 21,9 21,7 8,2 6,6
Ingush Republic 11,2 13,9 18,7 21,4 19,2 12,0 2,6 1,0
Kabardino-Balkar Republic 8,9 11,4 16,3 20,4 20,8 15,6 4,3 2,3
Karachay-Cherkessia Republic 9,5 12,5 17,5 21,1 20,4 14,0 3,4 1,6
Republic of North Ossetia - Alania 5,0 8,3 13,7 19,8 23,2 20,1 6,2 3,7
Chechen Republic
Stavropol Territory 7,0 9,5 14,3 19,1 21,5 18,6 6,0 4,0
Per capita income, rub. per month
to 3500,0 from 3500,1 to 5000,0 from 5000,1 to 7000,0 from 7000,1 to 10000,0 from 10000,1 to 15000,0 from 15000,1 to 25000,0 from 25000,1 to 35000,0 over 35000,0
Volga Federal District
Bashkortostan Republic 5,2 6,7 10,5 15,3 20,0 22,1 9,7 10,5
Mari El Republic 12,6 13,4 17,2 19,7 18,7 13,1 3,5 1,8
Mordovia Republic 9,0 11,6 16,6 20,5 20,8 15,3 4,1 2,1
Tatarstan Republic 4,0 5,9 9,8 15,2 20,6 23,3 10,4 10,8
Udmurt Republic 6,2 9,4 14,8 20,2 22,5 18,4 5,4 3,1
Chuvashia Republic 8,7 11,7 16,9 21,0 20,9 15,0 3,9 1,9
Perm Territory 4,2 5,7 9,4 14,3 19,6 23,1 10,9 12,8
Kirov region 4,8 8,1 13,5 19,6 23,3 20,4 6,4 3,9
Nizhny Novgorod region 3,9 6,3 10,8 16,6 22,0 23,2 9,3 7,9
Orenburg region 6,0 8,8 13,7 19,0 22,1 19,6 6,5 4,3
Penza region 6,6 9,4 14,5 19,6 22,0 18,5 5,8 3,6
Samara Region 4,8 5,9 9,3 13,8 18,7 22,3 10,9 14,3
Saratov region 7,7 10,4 15,4 20,0 21,5 17,1 5,0 2,9
Ulyanovsk region 7,6 9,7 14,3 18,9 21,1 18,3 6,0 4,1
urals Federal District
Kurgan region 7,3 9,2 13,6 18,2 20,9 19,1 6,7 5,0
Sverdlovsk region 3,0 4,5 7,8 12,7 18,8 24,2 12,4 16,6
Tyumen Region 1,9 3,1 5,7 10,1 16,3 24,0 14,2 24,7
including:
Khanty-Mansi Autonomous Area - Yugra 0,9 1,8 4,0 8,1 14,9 24,8 16,0 29,5
Yamal-Nenets Autonomous District 0,4 0,9 2,1 5,0 10,7 21,4 16,6 42,9
Chelyabinsk region 4,1 6,3 10,6 16,3 21,5 23,1 9,5 8,6
Siberian Federal District
Altai Republic 5,2 8,3 13,6 19,4 22,9 20,1 6,5 4,0
Buryatia Republic 7,3 9,0 13,3 17,8 20,7 19,3 7,0 5,6
Tuva Republic 11,1 13,3 17,8 20,8 19,5 13,0 3,1 1,4
Khakassia Republic 6,4 9,3 14,4 19,6 22,1 18,8 5,8 3,6
Altai Territory 8,3 11,5 16,9 21,2 21,3 15,2 3,8 1,8
Trans-Baikal Territory 6,4 8,6 13,1 18,1 21,4 19,9 7,1 5,4
Krasnoyarsk Territory 4,8 6,4 10,2 15,1 20,0 22,5 10,1 10,9
Per capita income, rub. per month
to 3500,0 from 3500,1 to 5000,0 from 5000,1 to 7000,0 from 7000,1 to 10000,0 from 10000,1 to 15000,0 from 15000,1 to 25000,0 from 25000,1 to 35000,0 over 35000,0
Irkutsk Region 6,7 8,3 12,4 17,1 20,6 20,2 7,8 6,9
Kemerovo Region 5,3 7,5 11,9 17,1 21,4 21,5 8,3 7,0
Novosibirsk Region 5,1 7,1 11,4 16,6 21,1 21,9 8,8 8,0
Omsk Region 5,9 7,9 12,2 17,3 21,1 20,9 8,0 6,7
Tomsk Region 4,5 7,1 11,9 17,7 22,4 22,1 8,1 6,2
Far Eastern Federal District
Sakha Republic (Yakutia) 1,5 3,1 6,3 11,7 19,2 26,7 14,0 17,5
Kamchatka 0,4 1,3 3,6 8,4 17,0 28,8 17,2 23,3
Primorye 2,9 5,3 9,7 15,9 22,2 24,8 10,2 9,0
Khabarovsk Territory 1,2 2,7 6,0 11,7 19,8 27,7 14,3 16,6
Amur Region 4,0 7,2 12,7 19,1 23,7 21,7 7,1 4,5
Magadan region 0,9 2,0 4,5 9,2 16,8 26,5 15,8 24,3
Sakhalin Region 0,7 1,7 3,8 7,8 14,9 25,3 16,4 29,4
Jewish Autonomous Region 3,8 6,5 11,5 17,8 23,0 22,9 8,4 6,1
Chukotka Autonomous District 0,3 0,8 2,0 5,1 11,6 23,6 17,8 38,8
Source: Regions of Russia. Socio-economic indicators. 2011: Stat. Sat / Rosstat. M., 2011, p. 164-165.
Appendix 2
The results of the presidential elections in 2012
Zhirinovsky Zyuganov Mironov Prokhorov Putin
Belgorod region 59561 6.62% 211079 23.45% 35601 3.96% 49807 5.53% 533716 59.30%
Bryansk region 42974 6.14% 146340 20.91% 23453 3.35% 32141 4.59% 448018 64.02%
Vladimir region 53615 8.40% 132400 20.75% 41895 6.57% 60315 9.45% 341301 53.49%
Voronezh region 81081 6.22% 292379 22.42% 47974 3.68% 69813 5.35% 800024 61.34%
Ivanovo region 37650 7.25% 95005 18.30% 23060 4.44% 37016 7.13% 321170 61.85%
Kaluga region 37634 7.42% 101459 20.01% 21427 4.23% 40911 8.07% 299175 59.02%
Kostroma region 28204 8.09% 90714 26.02% 16094 4.62% 26517 7.61% 183984 52.78%
Kursk region 49744 8.20% 122775 20.24% 23101 3.81% 38002 6.26% 366745 60.45%
Lipetsk region 44697 7.13% 132408 21.13% 24722 3.95% 34778 5.55% 382179 60.99%
Moscow region 236028 6.66% 686449 19.36% 149801 4.23% 396379 11.18% 2015379 56.85%
Orel region 33549 7.45% 130934 29.09% 15066 3.35% 27632 6.14% 237868 52.84%
Ryazan region 47068 7.58% 132981 21.42% 25562 4.12% 37903 6.10% 370945 59.74%
Smolensk region 38246 7.94% 111182 23.07% 20930 4.34% 32516 6.75% 273232 56.69%
Tambov region 28179 4.54% 107797 17.38% 13973 2.25% 19594 3.16% 444978 71.76%
Tver region 49384 7.40% 131591 19.71% 32835 4.92% 59302 8.88% 387308 58.02%
Tula region 50218 5.79% 147019 16.95% 29601 3.41% 43917 5.06% 587952 67.77%
Yaroslavl region 51816 7.72% 133476 19.89% 41212 6.14% 71007 10.58% 365892 54.53%
Zhirinovsky Zyuganov Mironov Prokhorov Putin
Moscow 267418 6.30% 814573 19.18% 214703 5.05% 868736 20.45% 1994310 46.95%
The Republic of Karelia 26579 8.59% 50957 16.47% 18886 6.10% 37798 12.22% 171380 55.38%
Komi Republic 40314 7.67% 70135 13.34% 22738 4.32% 43759 8.32% 341864 65.02%
Arkhangelsk Region 51169 8.90% 91648 15.94% 33223 5.78% 60108 10.45% 333344 57.97%
Nenets Autonomous District 2114 9.04% 4040 17.27% 1239 5.30% 2349 10.04% 13346 57.05%
Vologda Region 49492 8.13% 93417 15.35% 40306 6.62% 57064 9.38% 361720 59.44%
Kaliningrad 35625 7.79% 97570 21.33% 16139 3.53% 62016 13.56% 240421 52.55%
Leningrad Region 54857 6.77% 114951 14.18% 47518 5.86% 80874 9.98% 501893 61.90%
Murmansk region 32933 8.09% 65190 16.00% 20566 5.05% 39291 9.65% 244579 60.05%
Novgorod region 22955 7.41% 54875 17.70% 22066 7.12% 27017 8.72% 179501 57.91%
Pskov region 23760 6.71% 73073 20.64% 16164 4.57% 25824 7.30% 211265 59.69%
St. Petersburg 110979 4.65% 311937 13.06% 157768 6.61% 370799 15.52% 1403753 58.77%
Republic of Adygea 11164 5.06% 45311 20.55% 6637 3.01% 13145 5.96% 141257 64.07%
Republic of Kalmykia 3374 2.54% 23295 17.51% 3562 2.68% 8029 6.04% 93500 70.30%
Krasnodar 176119 6.54% 496909 18.46% 88976 3.31% 181844 6.75% 1715349 63.72%
Astrakhan region 21918 5.07% 67662 15.64% 18595 4.30% 21873 5.06% 297448 68.76%
Volgograd region 87657 6.86% 240998 18.85% 55325 4.33% 71142 5.56% 810598 63.41%
Rostov region 132418 6.27% 423884 20.06% 76633 3.63% 134461 6.36% 1324042 62.66%
Republic of Dagestan 1523 0.11% 84669 5.94% 4163 0.29% 6427 0.45% 1322567 92.84%
Republic of Ingushetia 1944 1.17% 7422 4.45% 1761 1.06% 1934 1.16% 153274 91.91%
Kabardino-Balkaria 11888 3.08% 53261 13.81% 11753 3.05% 8937 2.32% 299529 77.64%
Karachay-Cherkessia 2851 0.98% 16937 5.81% 2162 0.74% 2629 0.90% 266410 91.36%
Republic of North Ossetia - Alania 13063 3.16% 87017 21.05% 12864 3.11% 6848 1.66% 289643 70.06%
The Chechen Republic 140 0.02% 182 0.03% 165 0.03% 129 0.02% 611578 99.76%
Stavropol Territory 83543 6.99% 215600 18.03% 37551 3.14% 75724 6.33% 770874 64.47%
Republic of Bashkortostan 83704 3.64% 326250 14.18% 57329 2.49% 83667 3.64% 1731716 75.28%
Mari El Republic 24895 6.53% 84200 22.09% 15175 3.98% 24282 6.37% 228612 59.98%
Republic of Mordovia 13635 2.34% 42060 7.23% 6448 1.11% 9353 1.61% 506415 87.06%
The Republic of Tatarstan 52994 2.23% 229711 9.66% 41878 1.76% 69708 2.93% 1967291 82.70%
Udmurt Republic 49160 6.27% 116277 14.82% 26803 3.42% 67362 8.59% 515755 65.75%
Republic of Chuvashia 39707 5.65% 144676 20.58% 31201 4.44% 38838 5.52% 438070 62.32%
Perm 53879 4.60% 184639 15.78% 51535 4.40% 127098 10.86% 736496 62.94%
Kirov region 54531 7.90% 127982 18.54% 36005 5.22% 63993 9.27% 399810 57.93%
Nizhny Novgorod region 110808 5.96% 353964 19.05% 63189 3.40% 125432 6.75% 1187194 63.90%
Zhirinovsky Zyuganov Mironov Prokhorov Putin
Orenburg region 74414 7.33% 252947 24.92% 41104 4.05% 58849 5.80% 577411 56.89%
Penza region 48915 6.39% 150786 19.70% 24213 3.16% 39908 5.21% 492031 64.27%
Samara Region 117828 7.56% 320128 20.55% 61361 3.94% 125423 8.05% 912099 58.56%
Saratov region 66985 5.06% 206818 15.63% 43267 3.27% 59006 4.46% 934685 70.64%
Ulyanovsk region 46384 6.96% 160089 24.03% 27783 4.17% 37437 5.62% 387540 58.18%
Kurgan region 41340 8.57% 83955 17.40% 19280 3.99% 27725 5.75% 305777 63.39%
Sverdlovsk region 107819 5.20% 251690 12.14% 113353 5.47% 237780 11.46% 1337781 64.50%
Tyumen Region 59083 7.07% 95398 11.41% 20455 2.45% 43047 5.15% 611281 73.10%
Khanty-Mansi Autonomous Area - Yugra 57400 8.11% 97651 13.80% 23276 3.29% 50526 7.14% 469822 66.41%
Yamal-Nenets Autonomous District 17456 5.21% 18738 5.59% 4979 1.49% 7807 2.33% 283313 84.58%
Chelyabinsk region 97869 5.66% 254542 14.72% 88177 5.10% 138907 8.03% 1124538 65.02%
Altai Republic 5704 5.60% 17229 16.92% 3406 3.34% 6265 6.15% 68110 66.87%
Republic of Buryatia 22211 5.34% 75082 18.04% 13994 3.36% 24430 5.87% 275466 66.20%
Republic of Tyva 2574 1.74% 6370 4.32% 2023 1.37% 2925 1.98% 132828 90.00%
Republic of Khakassia 20991 8.48% 50872 20.56% 8878 3.59% 19400 7.84% 144519 58.40%
Altay 97961 8.33% 261665 22.26% 45883 3.90% 83778 7.13% 674139 57.35%
Trans-Baikal Territory 49612 9.95% 71636 14.37% 15015 3.01% 29466 5.91% 327407 65.69%
Krasnoyarsk Territory 112222 8.61% 235058 18.03% 46123 3.54% 109827 8.42% 784337 60.16%
Irkutsk Region 88419 8.24% 242097 22.57% 41152 3.84% 94008 8.76% 594861 55.45%
Kemerovo Region 112067 6.82% 133705 8.14% 37450 2.28% 75519 4.60% 1267837 77.19%
Novosibirsk Region 104223 7.70% 304761 22.53% 41001 3.03% 124205 9.18% 762126 56.34%
Omsk Region 74857 7.68% 234035 24.01% 39284 4.03% 72540 7.44% 541469 55.55%
Tomsk Region 35139 7.67% 86403 18.85% 16966 3.70% 53028 11.57% 261581 57.07%
The Republic of Sakha (Yakutia) 20010 4.37% 65871 14.39% 20193 4.41% 29712 6.49% 317933 69.46%
Kamchatka 16504 10.54% 25009 15.97% 5430 3.47% 14015 8.95% 93738 59.84%
Primorye 85396 8.63% 201493 20.36% 43168 4.36% 78639 7.95% 567177 57.31%
Khabarovsk Krai 68500 10.47% 115436 17.65% 31944 4.88% 62145 9.50% 367239 56.15%
Amur Region 39717 9.94% 67433 16.87% 13594 3.40% 23070 5.77% 251182 62.84%
Magadan 6399 9.18% 13946 20.01% 2607 3.74% 6769 9.71% 39196 56.25%
Sakhalin Region 20016 8.77% 45730 20.03% 8856 3.88% 22337 9.78% 128565 56.30%
Jewish Autonomous Region 6632 8.35% 14796 18.63% 2763 3.48% 5102 6.42% 48912 61.59%
Chukotka 2106 7.18% 2651 9.04% 633 2.16% 2209 7.53% 21310 72.64%
Source: http://www.cikrf.ru/
