Do political connections make businesspeople richer? Evidence from Russia,2003–2010

Introduction

An extensive body of literature shows that firms and businesspeople benefit from having political connections (Faccio, 2010; Szakonyi, 2018): being connected allows them to secure procurement contracts and government loans in hard times and crises (Claessens et al., 2008; Faccio et al., 2006; Khwaja and Mian, 2005; Li et al., 2008), brings tax benefits (Wu et al., 2012), and increases firms’ market valuation (Liu et al., 2013; Su and Fung, 2013). Connectedness is found to be especially important where institutions are weak (Gehlbach et al., 2010; Li et al., 2006) but proves beneficial even in well-institutionalized environments (Acemoglu et al., 2016; Goldman et al., 2009) that are generally expected to discourage informality.

Research into the effects of political connectedness, however extensive, does not always focus systematically on the top-level business political engagement, and almost never compares different kinds of political connections and their efficiency (e.g., connections at the parliamentary vs. executive, national vs. regional level). Most studies are also essentially cross-sectional in their design—meaning they do not explore the possibility of change in the way connections work over time in any given jurisdiction, only showing how connected firms and businesspeople differ from the matchable unconnected ones. As a result, the significance of the very political context making these connections meaningful for businesspeople remains unexplored. Here, we seek to fill these gaps by assembling and quantitatively analyzing a unique comprehensive dataset of Russia’s richest businesspeople’s political connections for the years 2003–2010.

Political connections as an asset

Following the mainstream in studying businesspeople’s political engagement, we treat such engagement as a non-market strategy pursued to build competitive advantage (Baron, 1995; Szakonyi, 2020: 6–9, 30–78). In this vein, getting connected could be considered an asset that the businessperson acquires for a benefit that should outweigh the costs. Alternatively, it could be viewed as an investment; any given type of connection could then be considered an investment vehicle.

Performance of political investment vehicles is likely to vary. On the one hand, different types of connection (such as, for instance, legislative or executive) provide for different abilities to influence public policy and could differ in the costs and benefits they entail. Some types of connection, e.g., top-level connections, could be naturally expected to be systematically “stronger,” while also being more “expensive.” If getting connected is strategic though, all of the connectedness actually observed would have to be at least weakly beneficial—those businesspeople who would not gain from getting connected would simply choose not to. And the connected ones would seek the types of connections they consider would work best for themselves given their individual circumstances.

At the same time, neither the costs nor the benefits of connectedness are strictly predetermined at the moment of becoming connected, and may vary over time—something the owners of connections cannot fully appreciate in advance. In particular, performance of different types of connection as “investment vehicles” could change as the political system that the connections are embedded into changes. Whenever that happens, the increasing costs and/or diminishing returns of different kinds of connection would be reflected in the connected businesspeople’s fortunes, relative to other businesspeople in a similar position, thus reflecting the relative potency of different kinds of connection. The return on such connections, if their status changes systematically, would then also change systematically across all individuals having such connection—an effect we seek to capture in this article using evidence from Russia.

The case of Russia

Russia presents a number of advantages for this kind of research (Szakonyi, 2020: 21–25). Not only does Russia have one of the highest numbers of billionaires globally, thus providing the researcher with enough observations for rigorous quantitative analysis, but as a country with a poorer quality of institutions it should also be expected to have more politically connected businesspeople (see Faccio, 2006: 373). Besides, Russia’s federal system allows exploring the issues of regional vs. national connectedness, and its business elites are well studied—both internationally (e.g., Schimpfössl, 2018) and especially by investigative journalists at home, so there are extensive personal level data available about Russia’s richest people and their connections. Russia therefore stands out in terms of the scale of the phenomenon and the quality of data. At the same time, Russia’s socioeconomic development in the 2000s was largely emblematic of any other nascent market economy as it experienced an economic boom caused by high oil prices, accompanied by growing dependence on natural resources and rising social inequality. Therefore, the patterns of business–power relationships it displays over this period should in principle be typical of a country with imperfect political, market, and rule of law institutions.

Another advantage of Russia as a case is that it provides researchers with a number of clear-cut and testable natural expectations about the potency of different kinds of connections grounded in Russia’s political dynamics in the 2000s. First, the way federal relations in the 2000s evolved suggests that over time regional-level connections should have lost their strength due to a series of centralizing political reforms implemented by the federal authorities, the most important of which led to the effective weakening of the heads of the regional executive (Golosov, 2018). Second, as autocracy consolidated in Russia in the 2000s and the federal executive tightened its grip over the parliament (Remington, 2016), the connections at the federal legislature should have weakened. These two processes would ultimately prove an essential element in the broader process of building a personalistic and increasingly deinstitutionalized authoritarian regime in Russia.

And third, against this political background, the processes of etatization and even gradual nationalization unfolded, which brought to life various forms of proxy and informal business ownership for federal-level politicians (Åslund, 2019; Miller, 2018). This process should have been additionally exacerbated in the context of the informality that was already prevalent in Russia at least since Soviet times, and more specifically by the existence of informal networks reportedly permeating Russian politics and economics (Ledeneva, 2006). As certain formal instruments of connectedness weakened, the informal ones would have to become relatively stronger.

All these political processes are well documented. Our ambition is to investigate whether the actual economic return on being connected echoed these changes in Russia. Following our knowledge of the case, our hypothesis is that those businesspeople who were informally connected to the federal executive should by 2010 have seen their wealth grow significantly faster than those with no connections, while those formally or regionally connected should not.

Proving this hypothesis correct will confirm more rigorously the existing understanding of the change in business–state relations in Russia in the 2000s that so far only rests on anecdotal evidence (e.g., Barsukova, 2019; Matveev, 2019). More importantly, it will also serve to validate the very procedure of detecting and mapping the actual political structure (its real effects and dynamics) by inferring it from businesses’ worth depending on their owners’ position within this structure—their connectedness and the type of connections they sport.

This latter use is especially important in political settings where nominal structures and institutions observed cannot be taken at face value. In particular, that would be the case of electoral authoritarianism, an increasingly common type of political regime “that permits certain institutions normally associated with democracy, such as elections and political parties, to exist, while remaining authoritarian in the basic patterns of power distribution and reproduction” (Golosov 2013, 618), and reduces such institutions to a mere façade disguising the actual way the regime works (Schedler, 2006: 3). The procedure used in this article should allow us to peek behind the façade into such regimes’ inner workings.

Data

Our dataset contains the following information about Russia’s richest businesspeople. First, we include their fortunes for each year over the 2003–2010 period, compiled from the rankings published by a Russian business weekly Finans between 2004 and 2011 (see Section 1 in the Online Appendix for the links). On average, each yearly ranking includes 500 businesspeople. The rankings rely on expert assessment of the value of industrial assets and the real estate that the participants own, as well as their other incomes. To account for decreasing marginal utility, we log-transform assets in our analysis. See Section 2 in the Online Appendix for empirical distributions of assets before and after transformation.

Second, we record the businesspeople’s connections with government, coded as a categorical variable with eight distinct values that depend on whether the connection is (a) federal or regional, (b) executive or legislative, and (c) formal or informal. To have a formal connection, they should personally hold a public office. To have an informal connection with an official, they should be reported to have had close relations with them before the official took up the position, or be a close relative. Actual occurrence of each type varies, and the regional legislative informal subtype is completely unobserved, suggesting a lack of interest toward this kind of connection among Russia’s richest people. We also create a specific category for those businesspeople who have held an office at the federal government in the past. Information about political connections has been compiled manually from several open sources. ¹ See Section 3 in the Online Appendix for the description of sources used and Section 4 for the frequencies of different connection types. Third, we use the industry that the businessperson has been most active in (categorized into 13 industries; see Section 5 in the Online Appendix for the full list) as a control variable.

Results

To estimate the impact of political connectedness on assets and their change across time, we use panel GLS regression with standard errors clustered on the level of individual businesspeople. Formally, the estimated model can be presented as follows. Analyzed data consists of observations (person–years) indexed as i = 1 , … , I . There are also two groups of dichotomous variables, power connections indexed k = 1 , … , K and industries indexed h = 1 , … , H . Then, assets can be modeled using these two groups of variables and time:

where β o is the intercept, i.e., predicted assets in year 2003 for person with no connection; β 1 k is the effect of connection type k on assets in year 2003; β 2 is the baseline effect of time on assets, i.e., predicted asset growth from 2003 to 2010 for a person with no connection; β 3 k is the effect of connection type k on asset growth from 2003 to 2010; β 4 h is the effect of industry h on assets; and ε i is the error term. Overall, we estimate the effect of political connections on both asset level and asset growth in one integrated model as, respectively, direct effects and interactive effects with time.

Results are presented in Table 1 as effects of different types of political connections on both asset level and asset growth. Since assets in the analysis are log-transformed, estimates can be translated back into the original scale (millions of dollars) by exponentiating. Table entries should be interpreted as follows. Effects of specific connection types on both assets and asset growth are estimated against the baseline of no connections. Specifically, average estimated assets in 2003 for a businessperson with no political connections are 4.50 on the log scale (US$90 million). To get estimated assets in the same year for a businessperson with a specific connection type, it is necessary to add the respective effect. For instance, for a businessperson with a formal connection to the federal executive average estimated assets in 2003 would be 4.50 + 0.12 = 4.62 on the log scale (US$101 million).

OPEN IN VIEWER

Table 1.

Political connections, assets, and asset growth: estimates from the GLS regression.

	Assets in 2003	Asset growth
No connection (baseline)	4.50***	1.48***
	(0.09)	(0.06)
Federal executive formal	0.12	−0.20
	(0.49)	(0.75)
Federal executive past	0.72*	−0.23
	(0.34)	(0.37)
Federal executive informal	−1.69***	2.63***
	(0.41)	(0.67)
Federal legislature formal	0.40**	−0.03
	(0.14)	(0.22)
Federal legislature informal	0.81	−0.54
	(0.49)	(0.57)
Regional executive formal	0.10	0.22
	(0.45)	(0.46)
Regional executive informal	0.63	−0.64
	(0.41)	(0.48)
Regional legislature formal	0.57	−0.54
	(0.32)	(0.38)

Note: 3,689 observations (person–years) and 1,035 persons. Standard errors in parentheses. Standard errors clustered on the level of persons. Results controlled for businessperson’s industry. See Section 6 in the Online Appendix for full results.

*

p < 0.05, **p < 0.01, ***p < 0.001.

Similar logic applies to estimated asset growth, which for a businessperson with no political connections is estimated as 1.48 on the log scale. It is possible to obtain the growth estimate for a businessperson with a specific connection type by adding the respective effect on growth. For instance, estimated asset growth for a businessperson with a formal connection to the federal executive would be 1.48 − 0.20 = 1.28 on the log scale. Unlike asset levels, growth cannot be directly translated into the original scale as the baseline should be accounted for. Consider the same example with a businessperson who has a formal connection to the federal executive. Recall that they have assets of 4.62 on the log scale in 2003 (US$101 million), yielding 4.62 + 1.28 = 5.90 in 2010 (US$365 million). Corresponding growth, therefore, is the difference that equals US$264 million.

Our analysis reveals that, in the end, only one type of political connection has an effect on asset growth that is significantly different from zero—informal connection to the federal executive. The effect is positive, very large in magnitude, and significant on the 99.9% confidence level. Starting from a significantly lower base in 2003, businesspeople with informal connections to the federal executive increase their assets much faster than those with no political connections (also see Lamberova and Sonin, 2018).

To ensure the reliability of this result, we ran a number of robustness checks. Specifically, we estimated (a) the effects of political connections on asset levels independently in 2004 (too few observations in 2003) and 2010, (b) the effects of political connections on asset growth estimated directly (rather than via panel GLS specification), and (c) the effects of political connections on asset levels and growth via panel GLS specification controlling for the number of connections per businessperson. The results of these additional analyses, presented respectively in Sections 7, 8, and 9 of the Online Appendix, confirm the presence of a positive and significant effect of informal connection to the federal executive on asset growth during the studied period.

Figure 1 presents the effect of informal connections to the federal executive on wealth in graphical form. In 2003, businesspeople with such connections start from a significantly lower base compared to those with no connections. However, their assets grow at a much higher rate. As a result, differences in assets become statistically insignificant by 2006—businesspeople informally connected to the federal executive catch up. By 2008, they become richer than people with no connections and by 2010 this difference becomes significant. The left-hand side of Figure 1 presents this effect in the log scale whereas the right-hand side reconstructs the original scale by exponentiation (standard errors omitted). It highlights major differences in asset growth rates and shows that by 2010 an average businessperson with informal connections to the federal executive is predicted to accumulate a fortune of almost US$1 billion.

OPEN IN VIEWER

Figure 1.

Changes in assets for businesspersons with informal connections at the federal executive compared to those with no power connections.

An essential question concerns causality: Do informal connections to the federal executive lead to asset growth? Or are businesspeople with the fastest growing fortunes more likely to acquire such connections? In our case, causality is easy to establish without any additional identification strategies. Of the 11 businesspeople coded as having informal connections to the federal executive in our sample, all acquired these connections before accumulating enough assets to show up in the data. Therefore, we can safely claim that in Russia in 2003–2010 it was the informal connection to the federal executive that helped businesspeople to accumulate wealth, and not the other way around.

Businesspeople with informal connections to the regional executives, who do not grow faster than those without any connections within the same time frame, make an instructive comparison point. If anything, asset growth for businesspeople with informal connections to the regional executives is slower than for those with no connections—but not significantly so. This effect reflects the general trend toward the weakening of the regional authorities following Vladimir Putin’s centralization. Combined with the gradual strengthening of informal executive connections at the federal level, our findings reveal the economic underpinnings of the political process that incorporated Russia’s regional autocracies into a single nation-wide authoritarian system, which was under way in the mid-2000s (Golosov, 2011)—as the federal center adapted the regions’ autocratic political practices, it also assimilated those of economic cronyism.

Building the “vertical of power” in federal relations was also supplemented by a weakening of the parliament vis-à-vis the executive by the end of the 2000s—even though having formal connections at the federal legislature provides for a significantly stronger base in 2003, it does not lead to a higher growth rate. The “oligarchic” model of business–state relations, still typical of the early 2000s (Hoffman, 2011), thus gave way to a model that is now highly centralized around the federal executive and dominated by informal personal connections.

Conclusion

Our analysis reveals that the economic return on being connected (as observed in the businesspeople’s wealth dynamics over time) reflects the changes in political structure. As the (geographically and politically) more decentralized structure, typical of the early 2000s, gave way to a less pluralistic one controlled by the federal executive, the real economic value of specific political ties changed accordingly. The findings we present confirm our knowledge of Russia’s political dynamics in the 2000s.

Substantively, this means that by far not all kinds of political connection work to the businesses’ benefit, and their effectiveness depends on the actual distribution of power in the system. This puts into question the existent belief in the effectiveness of political connections as such, and calls for research informed by a more nuanced understanding of the benefits that various political structures may bring to connected businesspeople.

Methodologically, our analysis validates the procedure of mapping the actual political structure by inferring it from the businesses’ worth depending on their owners’ position within this structure—a method that can be used for research into political settings where nominal structures and institutions observed cannot be taken at their face value (for instance, electoral autocracies that sometimes use quasi-democratic institutions without endowing those institutions with any real meaning).

Supplemental Material