SUPPLEMENTARY INFORMATION
International Trade as a Limiting Factor in Democratization: An Analysis of Sub-National Regions in Post-Communist Russia by Alexander Libman Frankfurt School of Finance & Management and Anastsssia Obydenkova Universitat Pompeu Fabra (Barcelona)
SUPPLEMENTARY INFORMATION
Appendix A: Data Table A1: Description and sources of variables Variable Description Average growth Growth rate of the GDP in %, inflation-corrected, averrate of the GDP age value for seven years Average income Average income per capita in the region, ‘000 RUR, per capita average value for seven years Coefficient of Standard deviation of the share of FSU trade of the variation of the region in 2000-2006 / average share of the FSU trade of FSU trade the region in 2000-2006 Democracy Index of democracy based on expert opinion, higher values represent higher level of democracy Democracy Index of democracy squared squared Distance from Distance between the capital of the region and the City Moscow of Moscow, 0 for the City of Moscow and Moscow Region Dummy border 1 if the region shares a common border with a foreign region state, 0 otherwise Dummy Chukotka 1 if region is Chukotka, 0 otherwise Dummy FSU bor1 of the region shares a common border with a postder Soviet state belonging to the CIS (excluding Latvia, Lithuania and Estonia), 0 otherwise Dummy Ingushet1 if region is Ingushetiya, 0 otherwise iya Dummy Kalmykia 1 if region is Kalmykia, 0 otherwise Dummy Moscow 1 if region is Moscow City, 0 otherwise City Dummy republic 1 if the region is an ethnic republic, 0 otherwise Dummy St. Pe1 if region is St. Petersburg, 0 otherwise tersburg Education Share of the population with university degree
Period 2000-2006
Source Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2004
NA
Moscow Carnegie Center Moscow Carnegie Center Rosstat
NA
Own calculation
NA NA
Own calculation Own calculation
NA
Own calculation
NA
Own calculation
NA NA
Own calculation Own calculation
2002
Russian Census
2000-2004
SUPPLEMENTARY INFORMATION Variable Industrial decline in the 1990s Initial GDP Investments Oil and gas
Openness Share of agricultural goods in the trade turnover Share of agriculture Share of energy in the trade turnover Share of FSU exports Share of FSU imports Share of FSU trade
Share of machinery in the trade turnover Share of metals and mining in the trade turnover Share of Russians
Description Decline of industrial output of the region in 1991-2000 (product of annual industrial output growth rates) Gross regional product per capita in 2000, ‘000 RUR Total investments in fixed capital per capita, RUR / GDP per capita, ‘000 RUR Extraction of oil in the region, mln. tonnes * 1.4 + Extraction of gas in the region, bln. sq. m * 1.2, average value for seven years (Export + Import (mln. USD) / GDP (mln. RUR), average value for seven years Trade turnover (export + import) in agricultural goods / Total trade turnover of the region, average value for 7 years Share of agriculture, fishery and forestry in the regional GDP, %, average for seven years Trade turnover (export + import) in energy / Total trade turnover of the region, average value for 7 years Exports to the post-Soviet countries of the CIS / total exports of the region, average value for 7 years Imports to the post-Soviet countries of the CIS / total imports of the region, average value for 7 years Trade turnover (export + import) with the post-Soviet countries of the CIS (i.e. 11 states, excluding Latvia, Lithuania and Estonia) / Total trade turnover (export + import) of a region, average value for 7 years Trade turnover (export + import) in machinery / Total trade turnover of the region, average value for 7 years
Period 1991-2000
Source Rosstat
2000
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
2000-2006
Rosstat
Trade turnover (export + import) in metals and mining / 2000-2006 Rosstat Total trade turnover of the region, average value for 7 years Share of ethnic Russians in the total population of the 2002 Russian Census region Temperature Long-term average mid-January temperature, Celsius NA Rosstat Notes: (1) Rosstat stands for the Russian Statistical Agency, (2) oil and gas are measured in coal equivalents in order to capture the extraction of both commodities through a single variable; (3) the choice to measure education through the university education is motivated by the fact that during the period of study secondary schooling covered almost 100% of the respective age cohort in Russia. Table A2: Summary statistics Variable
Obs
Mean
Std. dev.
Min
Max
79
6.58
2.37
0.36
14.11
Coefficient of variation of the FSU trade
79 79
4.65 0.42
2.48 0.39
1.57 0.05
18.40 2.20
Democracy
79
29.01
6.28
17.00
45.00
Democracy squared Distance from Moscow
79 79
880.73
382.25
289.00
2025.00
Dummy border region
79
2.37 0.47
2.75 0.50
0.00 0.00
11.88 1.00
Dummy Chukotka Dummy FSU border
79 79
0.01 0.30
0.11 0.46
0.00 0.00
1.00 1.00
Dummy Ingushetiya
79
0.01
0.11
0.00
1.00
Dummy Kalmyjia
79
0.01
0.11
0.00
1.00
Average growth rate of the GDP Average income per capita
SUPPLEMENTARY INFORMATION Dummy Moscow City Dummy republic
79
0.01
0.11
0.00
1.00
Dummy St. Petersburg
79 79
0.25 0.01
0.44 0.11
0.00 0.00
1.00 1.00
Education
79
0.52
0.16
0.13
0.91
Industrial decline in the 1990s Initial GDP
79 79
32.76 219.21
24.41 81.74
6.67 103.89
176.92 600.90
Investments
79
16.62
119.24
0.00
1060.44
Oil and gas Openness
79 79
0.01 0.12
0.01 0.14
0.00 0.00
0.10 0.67
Share of agricultural goods in the trade turnover
79
11.10
6.23
0.00
30.91
Share of agriculture Share of energy in the trade turnover
79 79
0.18 0.19
0.23 0.19
0.00 0.00
0.92 0.86
Share of FSU exports
79
0.25
0.23
0.01
0.88
Share of FSU imports Share of FSU trade
79 79
0.19 0.23
0.18 0.19
0.00 0.02
0.69 1.42
Share of machinery in the trade turnover
79
0.14
0.19
0.00
0.81
Share of metals and mining in the trade turnover Share of Russians
79 79
-11.48
8.45
-35.60
0.00
Temperature
79
0.77 6.58
0.24 2.37
0.01 0.36
0.97 14.11
Appendix B: Carnegie Index of Democracy The Carnegie Index of Democracy is a sum of 10 expert-based indices evaluating the level of democracy in the Russian regions using a 5-point scale, where 5 corresponds to the highest level and 1 to the lowest level. The index was compiled in the 2000s by the Carnegie Center in Moscow, an independent think tank established by the Carnegie Endowment for International Peace, and is therefore unlikely to be biased by the intra-Russian political considerations. The index is also available for the 1990s, yet for this period it was calculated backwards in the 2000s; hence the data can be treated as less reliable. The index we use in this paper was calculated for the period of 2000-2004 and published in the Social Atlas of the Russian Regions by the Independent Institute for Social Policy. The description of the main components of the index is listed below: all components are highly correlated. Two components (corruption and economic liberalization) should be treated with caution, as they could represent not only political, but also economic dimension of the Russian regional systems, yet their exclusion has no influence on the results. The variable has been actively used in the literature as a measure of Russian sub-national democracy (e.g. Lankina and Getachew 2006); hence, our study seems to be consistent with the overall stand of research on this topic. Table B1: Components of the Carnegie Index Component Comments Civil society NGOs, referenda, freedom of public political activity Corruption State capture in a broader sense, i.e. interconnections between political and business elites and their interventions in the political decision-making Economic liberalization Specific directed interventions of regional administration, ignoring property rights of influential players (e.g. potential opposition) Elites Existence of a mechanism of leader changes through elections, existence of multiple political elites Freedom of elections Elections at all levels (national, regional, local) included Freedom of local municipalities Extend of authorities of local municipalities vis-à-vis the control of the regional government Independence of the media Independence of the media from political and other pressures Openness of regional political life Transparency of regional politics and its involvement in the overall national politics
SUPPLEMENTARY INFORMATION Political pluralism
Existence of stable political parties, representation of parties in regional legislatures Regional political organization Real balance of power between the executive and the legislative, elections / appointments of crucial political actors, independence of courts and police, protection of citizen rights Source: Obydenkova and Libman (2012) In addition, Carnegie Center also published a second index of democratization measuring the degree of freedom and competitiveness of regional elections of the 1999-2002 cycle and using more objective criteria (mostly from electoral statistics). This index is not applicable for our study, as it captures the electoral cycle at the very beginning of the rapid economic growth, i.e. before the effects of the limiting factor could have become pronounced. Interestingly, while this index is correlated with the index we use, there are several outliers: most notably, Ivanovo scores as a region with one of the most democratic elections in 2000 and as a region significantly below average in the rating for 2000-2004. This is consistent with the main claim of the paper: in the region with large dependence from the trade with the FSU after the start of economic growth the adverse effects on democracy have been more pronounced.
Appendix C: Additional control variables in the growth regressions We add a number of controls to our main regressions to check for robustness of the results • We re-estimated regressions controlling for two sets out outliers: (a) three regions traditionally exhibiting growth patterns that are different from those of the rest of the country (Ingushetia, Kalmykia and Chukotka) - and (b) capital cities Moscow and St. Petersburg. The first three regions are different primarily because they served as tax havens for individual large corporations (e.g. Sibneft or Yukos) or for a broad set of companies throughout the period. The second set of outliers is special because Moscow City and St. Petersburg enjoy special status due to public support; furthermore, headquarters of many companies are located in these cities, thus their GDP includes elements actually produced in other parts of Russia but counted as the GDP of the regions where headquarters of the companies are located. • We added the share of agriculture, dummy border region and average mid-January temperature to check for the impact of industry structure, location and climatic conditions on the results. • We dropped the openness variable (which may be correlated with the share of trade with the FSU). • We controlled for the commodity trade structure. Because we have no industrial structure of trade with the FSU and with other countries reported separately for individual regions, we attempt to solve the problem by controlling for shares of main trade good classes (energy, machinery, agriculture, metals and mining) in the overall trade turnover. • It is possible that the regions, which were more dependent on the trade with the FSU, also had a stronger decline in the 1990s, and thus, the growth we observe in the 2000s is higher simply because of a very strong drop in output (due to the disorganization effect). The Russian Statistical Agency does not report the regional GDP data before 1995; thus, to check this hypothesis, we took the growth rates of the industrial output and calculated the overall decline of the industrial output in 1991-2000. There are also two modifications of this specification: one drops initial GDP (because it may measure the same as the decline in the 1990s), and one controls for the cities of Moscow and St. Petersburg with two dummies (because, in these cases, a drop in industrial output could be due to the boom in the service sector and not overall economic decline). • We control for direct influence of democracy on economic growth, which may exist besides the indirect influence (change in the level of democracy because of the trade with the FSU). If we merely observe the indirect effect, one might speculate whether trade with the FSU is indeed a limiting factor and the mechanism of this study does exist; if we observe both direct and indirect effect (as we do in our sample), the conclusion regarding the existence of the limiting factor is more probable. We indeed find a direct (non-linear) effect of democracy on growth in line with Libman (2012), yet the effect of trade on growth persists in this case as well. All modifications entirely confirm our main results: the effect of FSU trade on growth remains positive and significant.
SUPPLEMENTARY INFORMATION
Table C1: Effects of trade with the FSU on the economic growth; dep. var.: average GDP growth rate, inflationcorrected, 2000-2006
Initial GDP Oil and gas Education Openness Investments
(C1) OLS -0.033** (0.015) 0.005** (0.002) 25.686*** (5.893) 44.468** (18.172) 0.012*** (0.003)
(C2) OLS -0.050*** (0.017) 0.008*** (0.002) 17.065** (7.550) 44.670** (17.231) 0.013*** (0.003)
Share of agriculture Dummy border region Temperature Democracy Democracy squared
(C3) (C4) (C5) (C6) (C7) (C8) (C9) OLS OLS OLS OLS OLS OLS OLS -0.019 -0.018 -0.014 -0.023 -0.017 -0.026 -0.006 (0.016) (0.015) (0.015) (0.016) (0.016) (0.018) (0.016) 0.003 0.003 0.002 0.003 0.002 0.002 0.004* (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) 23.318*** 26.316*** 23.443*** 25.352*** 22.349*** 25.527*** 20.569*** (5.197) (5.323) (5.164) (5.428) (5.846) (5.558) (4.786) 39.658** 34.400** 41.859** 36.375** 28.913* 41.936** (16.644) (16.821) (17.662) (17.322) (14.733) (17.463) 0.013*** 0.014*** 0.012*** 0.013*** 0.013*** 0.012*** 0.012*** (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) -0.003 (0.050) 0.684 (0.433) -0.024 (0.031) -0.432* (0.248) 0.007* (0.004)
Share of metals and mining in trade turnover
-0.068 (1.367)
Share of machinery in trade turnover
-1.806 (1.907)
Share of energy in trade turnover
2.154* (1.135)
Share of agricultural goods in trade turnover
0.610 (2.331)
Industrial decline in the 1990s Share of FSU trade Dummy Kalmykia Dummy Chukotka Dummy Ingushetia
4.547*** (1.249) -1.658* (0.917) 4.286*** (0.711) -3.718*** (0.755)
Dummy Moscow City Dummy St. Petersburg Constant Observations R2 J.-B. test
-1.013 (1.367) 79 0.485 24.77***
4.025*** (1.208) -2.107** (0.967) 4.563*** (0.724) -4.296*** (0.774) 5.864*** (2.140) 0.697 (1.080) 0.766 (1.508) 79 0.519 23.8***
5.061*** (1.334)
4.387*** (1.184)
4.831*** (1.287)
5.338*** (1.338)
5.269*** (1.249)
5.165*** (1.383)
-2.259 (1.898) 5.256*** (1.302)
-1.122 (1.531) 79 0.415 16.7***
-1.256 (1.276) 79 0.377 10.63***
-1.432 (1.231) 79 0.433 12.99***
-1.856 (1.447) 79 0.42 17.82***
4.961 (4.003) 79 0.442 21.67***
-1.070 (1.363) 78 0.462 9.684***
0.096 (1.433) 79 0.433 13.44***
Note: see Table 1. Dagestan and Arkhangelsk in regression (C1), (C2), (C7), (C9), Dagestan in all other regressions. After exclusion of outliers shares of FSI trade remain significant and positive.
SUPPLEMENTARY INFORMATION Appendix D: Impact of the FSU trade on growth and democracy In order to establish the presence of a simultaneous effect of trade with the FSU on economic growth and democracy going in the opposite directions for both variables, we estimate a system of equations using the three-stage least squares (3SLS). The system includes two equations: a growth regression (described in the main part of the paper) and a regression explaining the level of democracy in the Russian regions. In both equations, we explain the levels of growth and democracy by trade with the FSU and by a set of controls accounting for further explanations of economic and political development. Specifically, we apply the following composition of specifications: • (D1) Growth rate and democracy depend on a set of parameters, but not on each other; democracy, however, depends on income per capita • (D2) Growth rate does not depend on democracy, democracy depends on the growth rate of the GDP in the region, but not on the income per capita • (D3) Growth rate does not depend on democracy, democracy depends on both income per capita and GDP growth rate (i.e. short-term and long-term growth) • (D4) Growth rate depends on democracy in the region (possibly in a non-linear way), democracy does not depend on growth rate • (D5) Growth rate depends on democracy, democracy depends on growth rate, but not on income per capita • (D6) Growth rate depends on democracy, democracy depends on growth rate and income per capita • (D7) – (D9): the same as (D4) – (D6), but accounting for the fact that democracy squared should also be treated as an endogenous variable in the 3SLS We find that the economic growth is positively associated with the FSU trade in all specifications. The effect of trade on democracy is less robust, yet still present in the majority of the estimations and is negative (Obydenkova and Libman 2012), what is consistent with the theoretical predictions made in the paper. The reasons for low robustness of the effect of trade on democracy are most likely linked to the findings reported in the main part of the paper: trade affects democracy only in regions exhibiting low volatility of trade (and hence, where trade is more likely to be a limiting factor).
SUPPLEMENTARY INFORMATION Table D1: Effects of trade with the FSU countries on the economic growth and democracy, system of equations, 2000-2006
Education Income per capita Share of Russians Dummy republic Distance
(D1) 3SLS Economic Democracy growth 12.596 23.573*** (20.104) (6.320) 0.341 (0.338) 14.987*** (4.486) 1.695 (2.470) -0.469* (0.247)
GDP growth rate Oil and gas Openness Initial GDP Investments Share of FSU trade Constant Endogenous variables Chi-sq Observations
0.003 (0.005) 37.229 (49.417)
0.003 (0.003) 39.715** (17.691) -0.020 (0.014) 0.013*** (0.003)
-7.131* 5.006*** (3.889) (1.351) 15.222*** -1.189 (5.597) (1.264) Democracy, economic growth 43.98*** 57.11*** 79 79
(D2) 3SLS Economic Democracy growth 31.666 24.316*** (19.979) (6.289)
0.004 (0.002) 40.161** (17.695) -0.024* (0.013) 0.012*** (0.003)
(D3) 3SLS Economic Democracy growth 29.343 23.608*** (21.911) (6.319) 0.294 (0.363) 12.220** (5.530) 0.794 (2.716) -0.472* (0.253) -0.981 (0.608) 0.004 0.004 (0.006) (0.002) 93.244 41.057** (61.888) (17.683) -0.021 (0.014) 0.012*** (0.003)
-5.877 4.777*** (5.062) (1.341) 15.981** -1.047 (7.624) (1.264) Democracy, economic growth 45.01*** 56.66*** 79 79
-2.657 4.846*** (5.065) (1.346) 19.715*** -0.907 (7.242) (1.257) Democracy, economic growth 47.56*** 54.14*** 79 79
14.634** (5.860) 1.360 (2.823) -0.384 (0.236) -0.451 (0.711) 0.005 (0.005) 65.320 (64.759)
Notes: *** 1% significance level; ** 5% significance level; * 10% significance level; standard errors in parentheses
SUPPLEMENTARY INFORMATION
(Table D1 continued)
Education Income per capita Share of Russians Dummy republic Distance
(D4) 3SLS Economic Democracy growth 9.549 21.074*** (20.125) (6.509) 0.413 (0.339) 13.886*** (4.513) 1.289 (2.485) -0.503** (0.248)
0.002 (0.002) 29.111 (17.990) -0.007 (0.014) 0.011*** (0.003) -1.427*** (0.452)
(D6) 3SLS Economic Democracy growth -0.652 19.590*** (21.083) (6.506) 0.377 (0.351) 17.089*** (5.454) 2.028 (2.685) -0.462* (0.251) 0.770 (0.563) 0.002 0.003 (0.005) (0.003) -7.814 30.061 (58.747) (18.362) -0.010 (0.015) 0.012*** (0.003) -1.380*** (0.458)
0.018** (0.007)
0.025*** (0.007)
0.024*** (0.007)
-7.041* 5.517*** (3.893) (1.390) 16.448*** 14.336** (5.623) (7.005) Democracy, economic growth 42.49*** 59.75*** 79 79
-13.383*** 5.979*** (5.102) (1.350) 9.609 19.202*** (7.701) (6.886) Democracy, economic growth 38.66*** 67.56*** 79 79
-10.682** 5.779*** (4.808) (1.388) 11.873* 18.629*** (7.028) (6.981) Democracy, democracy squared 44.45*** 64.30*** 79 79
GDP growth rate Oil and gas Openness Initial GDP Investments Democracy Democracy squared Share of FSU trade Constant Endogenous variables Chi-sq Observations
0.002 (0.005) 35.883 (49.430)
0.003 (0.003) 34.405* (18.373) -0.011 (0.015) 0.012*** (0.003) -1.061** (0.460)
(D5) 3SLS Economic Democracy growth 6.550 18.327*** (20.200) (6.336)
18.223*** (5.985) 1.818 (2.855) -0.349 (0.241) 1.143 (0.696) 0.003 (0.005) -27.210 (65.278)
SUPPLEMENTARY INFORMATION
(Table D1 continued)
Education Income per capita Share of Russians Dummy republic Distance
(D7) 3SLS Economic Democracy growth 8.138 20.397*** (20.117) (6.969) 0.447 (0.338) 13.457*** (4.501) 0.986 (2.475) -0.514** (0.248)
GDP growth rate Oil and gas Openness Initial GDP Investments Democracy Democracy squared Share of FSUs trade Constant Endogenous variables Chi-sq Observations
0.002 (0.005) 34.898 (49.427)
0.003 (0.003) 30.912 (24.420) -0.012 (0.017) 0.012*** (0.003) -1.257 (0.923) 0.021 (0.017)
-7.001* 5.685*** (3.892) (1.545) 16.977*** 16.829 (5.611) (12.236) Democracy, democracy squared, economic growth 37.87*** 55.66*** 79 79
(D8) 3SLS Economic Democracy growth 31.921 21.417*** (20.000) (6.806)
11.476* (6.096) 0.112 (2.940) -0.423* (0.247) -0.560 (0.714) 0.005 (0.005) 69.386 (64.974)
0.003 (0.003) 32.000 (23.854) -0.016 (0.016) 0.013*** (0.003) -1.025 (0.904) 0.017 (0.017)
-5.598 5.469*** (5.105) (1.508) 19.382** 13.446 (7.835) (11.996) Democracy, democracy squared, economic growth 38.88*** 53.85*** 79 79
(D9) 3SLS Economic Democracy growth 20.176 20.758*** (22.183) (6.970) 0.478 (0.369) 9.708* (5.831) 0.079 (2.862) -0.554** (0.268) -0.881 (0.617) 0.003 0.003 (0.006) (0.003) 85.071 31.565 (62.305) (24.418) -0.013 (0.017) 0.012*** (1.230) -0.923 (0.021) -0.017 (1.230) -2.829 5.584*** (5.114) (1.544) 22.262*** 16.546 (7.526) (12.237) Democracy, democracy squared, economic growth 43.11*** 53.88*** 79 79
SUPPLEMENTARY INFORMATION
Appendix E: FSU trade as share of GDP In the main specification of the paper we look at the marginal increase of the share of FSU trade in the total trade conditional on the overall economic openness. In a further robustness check we calculated the total FSU trade per unit of regional GDP (as a product of our FSU share measure used so far and the openness to trade) and regressed the growth rate on this variable including and excluding the overall openness to trade. The findings do not change. In the same way, the results for the system of equations for democracy and growth remain the same (for the insignificant coefficient of FSU trade the p-value is actually 0.101). Table E1: Effects of trade with the FSU on the economic growth; dep. var.: average GDP growth rate, inflationcorrected, 2000-2006
Initial GDP Oil and gas Education Investments
(E1) OLS -0.032* (0.017) 0.005** (0.002) 27.050*** (5.897) 0.013*** (0.003)
Openness FSU trade per unit of GDP
256.730** (121.223)
(E2) OLS -0.033* (0.017) 0.005** (0.002) 25.957*** (6.227) 0.012*** (0.003) 15.659 (16.224)
235.838* (119.832)
Share of Russians Dummy republic Distance from Moscow
(E3) 3SLS
77.313 (47.906)
67.896 (53.073)
-0.025* (0.014) 0.004 (0.003) 22.591*** (6.690) 0.012*** (0.003) 2.494 (19.287)
-482.226 (293.729)
277.300** (109.799)
-665.880** (334.000)
287.351*** (107.491)
0.002 (0.005) 6.413 (20.141)
-0.244 (1.197) 79 0.352
Note: see Tables 1 and D1
0.005 (0.005) 14.658 (20.231)
-0.022 (0.014) 0.004 (0.003) 21.387*** (6.587) 0.011*** (0.003)
-398.979 (293.869)
287.190*** (106.510)
-682.151* (348.499)
297.335*** (104.202)
14.592*** (4.606)
18.873*** (6.292)
1.807 (2.470)
2.018 (2.943)
1.518 (2.505)
1.999 (2.993)
-0.460* (0.243)
-0.268 (0.240)
-0.477* (0.247)
-0.240 (0.244)
-1.109** (0.487)
-1.221** (0.482)
-1.133** (0.485)
-1.410*** (0.478)
0.018** (0.008)
0.021*** (0.008)
0.019** (0.008)
0.024*** (0.008)
0.507 (0.332)
-0.444 (1.201) 79 0.346
0.002 (0.005) 9.678 (20.359)
-0.026* (0.015) 0.005* (0.003) 24.295*** (6.755) 0.012*** (0.003)
18.011*** (6.202)
0.580* (0.335)
GDP growth
Observations R-squared Endogenous variables Chi-squared
0.005 (0.005) 15.144 (19.359)
(E6) 3SLS
15.156*** (4.543)
Democracy squared
Constant
(E5) 3SLS
-0.027* (0.015) 0.005* (0.003) 23.910*** (6.837) 0.012*** (0.003) 6.974 (19.706)
Democracy
Income per capita
(E4) 3SLS
14.390*** (5.429) 79
16.011** (7.412) 79
Democracy, GDP growth 41.75*** 46.41***
0.601 -0.669 9.046 (8.140) 79
17.304** (7.333) 79
Democracy, GDP growth 37.83*** 48.67***
14.941*** (5.509) 79
16.229** (7.428) 79
Democracy, GDP growth 37.96*** 46.09***
0.928 (0.651) 7.233 (8.180) 79
19.933*** (7.319) 79
Democracy, GDP growth 33.61*** 51.30***
SUPPLEMENTARY INFORMATION
Appendix F: Alternative time frame of the study In the paper we use 2006 as a cutoff year for our study to represent the fundamental change in the Russian political system. However, as an alternative cutoff one could consider the year 2004. First, gubernatorial elections were abolished in 2004, hence, although elected governors continued running Russian regions, their behavior could have already changed. Second, the Carnegie index was calculated for 2000-2004. In order to check for the robustness of our results, we re-estimated the regressions using only 2000-2004 period; the findings mostly stay the same as they were in the main modification. Table F1: Effects of trade with the FSU on the economic growth and democracy, 2000-2004
Dep. var. Initial GDP Oil and gas Education Openness Investments Share of trade with the FSU
(F1) OLS Growth rate -0.015 (0.023) 0.002 (0.004) 21.493*** (6.679) 33.751** (15.186) 0.015*** (0.006) 4.577*** (1.627)
Share of exports in the FSU
(F2) OLS Growth rate -0.016 (0.024) 0.002 (0.004) 20.024*** (6.985) 35.668** (16.460) 0.015*** (0.005)
(F3) OLS Growth rate -0.023 (0.022) 0.002 (0.004) 25.458*** (7.148) 22.207 (14.761) 0.016*** (0.006)
(F4) OLS Growth rate -0.014 (0.023) 0.001 (0.004) 22.555*** (7.483) 31.079** (15.255) 0.016*** (0.006)
(F5) 3SLS Democracy
0.003 (0.005) 11.978 (20.246) 52.324 (45.189)
-6.703* (3.853) 4.012*** (1.383)
Share of imports in the FSU
3.392** (1.492)
(F6) 3SLS
Growth rate -0.002 (0.020) 0.001 (0.004) 18.619** (8.800) 35.096 (22.478) 0.014*** (0.004) 4.940*** (1.869)
Share of Russians
13.738*** (4.503) 1.116 (2.460) -0.439* (0.244)
Dummy republic Distance from Moscow
-1.289** (0.622) 0.021** (0.010) 14.923** -6.692 1.164 (3.158) -0.365 (0.234)
Growth rate of the regional GDP
0.148 (0.570)
Income per capita
Endogenous variables R-squared Chi-squared
-8.313* (4.379)
-1.433** (0.626) 0.023** (0.010)
Democracy squared
Observations
0.005 (0.005) 21.628 (18.138) 42.448 (52.824)
Growth rate -0.003 (0.020) 0.001 (0.004) 18.654** (8.644) 33.899 (22.073) 0.014*** (0.004) 4.957*** (1.835)
2.577* (1.473) 2.648 (1.596)
Democracy
Constant
Democracy
-1.023 (1.673) 79
-0.467 (1.538) 79
-1.449 (1.790) 79
-1.629 (1.845) 79
0.301
0.289
0.304
0.319
Note: see Tables 1 and D1
0.422 (0.427) 16.128*** 20.750** (5.557) (9.574) 79 79 Democracy, economic growth
14.244* 18.431* (8.244) (9.508) 79 79 Democracy, economic growth
42.45***
40.78***
38.81***
38.57***
SUPPLEMENTARY INFORMATION
Appendix G: Impact of stability of trade with the FSU on growth and democracy In addition to the main approach based on the analysis of interaction terms, we also apply the following procedure to investigate the impact of stability of trade with the FSU on growth and democracy. We divide the sample into two groups: regions with coefficients of variation of the FSU trade above average or below average (Ivanovo, as expected, is in the second group). Then we estimate the growth regression and the democracy regression separately for each of the sub-groups. For the growth regression, we find that both sub-groups of regions grow faster if the FSU trade is larger. Thus, some sort of positive effect due to the lower impact of economic disorganization here seems to be present in both cases. However, for the democracy regression, we find a significant and negative effect of the FSU trade only for regions with small coefficients of variation. If the coefficient of variation is large, the effect is insignificant (and the sign is positive). Thus, the FSU trade seems to simultaneously result in increasing growth and decreasing democracy only in regions with more stable FSU trade, where the dependence of economic growth on this trade is likely to be higher; therefore, the logic described above should be more probable. Because the sub-groups of the sample with high and low coefficients of variation are different in terms of the number of regions, we have also split the sample using the median coefficient of variation, but the results are unambiguously confirmed. Table G1: Impact of trade with the FSUs on democracy and economic growth, conditional on the coefficient of variation of the FSU trade share
Dep. var. Initial GDP Oil and gas Education Openness Investments
(G1) OLS Growth 0.011 (0.018) -0.001 (0.003) 22.279*** (7.131) 32.105** (12.955) 0.012*** (0.004)
(G2) OLS Growth -0.018 (0.024) -0.059 (0.066) -11.735 (11.978) 49.552 (51.803) 0.015*** (0.004)
0.190 (0.161) 17.980 (29.459) 137.108 (93.565)
23.191*** (5.674) 7.788** (3.426) 0.171 (0.281) 10.342 (11.922) 1.531 (7.016) Mean coefficient of variation
5.044*** (1.349) -2.663 (1.577) Median coefficient of variation
High 28 0.617
Low 39 0.536
18.949*** (5.253) 3.130 (2.406) Mean coefficient of variation
Low 51 0.477
High 28 0.677
Low 51 0.364
Distance
Separation of high and low coefficient of variation Coefficient of variation Observations R2
0.001 (0.002) 16.287 (23.290) 17.179 (28.223)
4.224*** (1.135) -1.480 (1.176) Mean coefficient of variation
Dummy republic
Constant
(G4) OLS Democracy
-0.256 (7.283) -8.404*** (2.829) -0.141 (0.702) -10.299** (3.961) 31.044*** (8.875) Mean coefficient of variation
Share of Russians
Share of FSU trade
(G3) OLS Democracy
Note: see Table 1
(G5) OLS Growth 0.005 (0.013) 0.000 (0.002) 23.063*** (5.488) 52.337 (44.197) 0.016*** (0.005)
(G6) OLS Growth -0.021 (0.024) -0.014 (0.065) -0.119 (11.849) 50.047*** (15.082) 0.012*** (0.004)
(G7) OLS Democracy
(G8) OLS Democracy
-0.001 (0.003) 23.348 (26.681) -28.877 (105.803)
0.335** (0.135) 3.115 (29.264) 86.126*** (29.990)
11.348** (5.098) 2.166 (2.208) Median coefficient of variation
-11.769 (9.777) -12.624*** (4.291) -0.086 (0.771) -8.804* (4.601) 40.819*** (11.306) Median coefficient of variation
20.789*** (4.959) 5.648** (2.748) -0.132 (0.252) -3.062 (9.790) 9.383 (6.429) Median coefficient of variation
High 40 0.477
Low 39 0.350
High 40 0.621
SUPPLEMENTARY INFORMATION
Appendix H: Sources of data for the process-tracing The interviews had been collected in the period of August 2008 – December 2009. The interviewees were chosen on the basis of their presence in a region and their involvement in regional politics from 2000 onwards. In addition, the interviews with regional academic elite and journalists provided very different and more complete insight in the dynamics of regional political and economic development during the period under investigation. We also added to the sample of the interviewees some Moscow-based experts, whose main focus has been research on the post-Soviet trade, in order to obtain more systematic insight about the institutions governing the FSU trade and their possible impact on the political development of the region. The media reports have been scrutinized from the database of EastView, which includes a significant number of Russian newspapers – both regional and federal. However, these data have been augmented by reports of the regional newspapers not available in the database. In addition, we also looked at the Internet sources: they are particularly instructive from the point of view of reports of the opposition. We have, however, also specifically included newspapers with well-known political attitudes (like the pro-Communist Sovetskaya Rossiya or the governmental official Rossiyskaya Gazeta) to obtain a significant variation of statements and positions on the regional development. Among the key newspapers and magazines studied, the following should be mentioned: Novye Izvestiya, Kommersant, Nezavisimaya Gazeta, Rossiyskaya Gazeta, Rossiyskie Vesti, Sovetskaya Rossiya, Vremya Novostei, Trud, Izvestiya, and Ekspert (federal); and Tekstil’nyi Krai Rossii, Ivanovo-Press, Rabochiy Krai, Ivanovo-Voznesensk and Professional (regional and local). One should, however, point out that the federal newspapers were a richer source of information in terms of facts and analyses of Ivanovo politics than many regional newspapers, which is consistent with the main results of the paper.
Appendix I: The choice of the timeframe of the study Upper and lower cutoff: The upper cutoff year of 2006 seems to be reasonable given the political changes in the Russian system: Before 2006 most governors in the Russian regions still had received their office through public elections; after 2006 most governors were already appointed by the federal center. However, it is more important to justify why we did not consider the period of 1990s. Partly it is due to the poor quality of statistics (the Russian government started systematically reporting trade with the FSU states only in the second half of the 1990s). However, it is also reasonable if one considers the economic development of the Russian Federation. In the 1990s (until 1999) the country was hit by severe transition recession. Most of the companies were almost insolvent and forced to survive using various barter schemes; debt arrears accumulated enormously; wages were systematically delayed for months or years. Gaddy and Ickes (2002) describe this period as a virtual economy. Under these conditions the ability to control the limiting factor had little impact on regime transition: the revenues generated in the Russian economy during this period were rather linked to the asset-stripping and search for public subsidies than to actual production, thus, there were no profits for businessmen to compete for and no wages for employees to ensure their loyalty to the regime, which could have been linked to the access to the limiting factor of post-Soviet trade. The situation was somewhat different for export of natural resources; however, even in this case the rents generated were much smaller than, say, rents from operation with short-term government bonds (until the crisis of 1998). In 1999-2000 Russia entered a period of rapid economic growth. Many companies were able to recover and to expand. In this environment the access to the limiting factor turned into an important issue. History of Ivanovo in 1991-1999: The choice of the lower cutoff is clearly confirmed by detailed analysis of our case study. In the early 1990s Ivanovo was hit by the transition recession encountered by Russia, resulting in enormous decline of the local textiles industry (unable to compete with cheap imports from Turkey and China and to survive in the environment of shrinking consumer demand). Hence, during this period the control over limiting factor had no real influence on the political development in the region: the macroeconomic environment in Russia (outside of control of any regional governor) made any economic development unsustainable. Most textiles companies had to cut their sales and their production, and therefore did not require the quantities of cotton they used to buy during the Soviet period. While accessing cotton was more and more difficult with every year, the small production volumes ensured that this problem was not really crucial for the survival of the enterprises. Ivanovo was ruled consequently by two governors: Adolf Laptev (1990-1996) and Vladislav Tikhomirov (1996-2000); the latter was also the predecessor of Laptev as the president of the Ivanovo Region Executive Committee in 1987-1990. Both governors started as the members of the Soviet nomenklatura (Tikhomirov, for example, was a former secretary of the Communist Party of the Soviet Union regional committee for agriculture); Laptev had some
SUPPLEMENTARY INFORMATION support of the Communist Party (which, in fact, opposed his resignation), but did not protest in any way when President Yeltsin decided to replace him by Tikhomirov (publicly explaining his resignation by the advanced age of 60 years; a detailed description of this campaign was published by Vera Rodionova in various issues of the Politicheskiy Monitoring IGPI for 1996). After his appointment in 1996, Tikhomirov passed direct gubernatorial elections (the same year) without any strong opposition from other political forces, being able to obtain 50.1% of votes in the first round of elections (Kommersant, 3 December 1996). Thus, the regional political system, on the one hand, did not appear to be very competitive, but, on the other hand, the regional governors seemed to have almost no independent power base without central support.