Barnabás Gerő - Balázs Vedres: Interlocking Comrades

An analysis of the structure of interlocking directorates
in post-communist Hungary

Post-communist Eastern-Europe is characterized by a cacophony of values and voices as new groups emerge and try to institutionalize their power by a variety of narratives. This cacophony is present at the level of individual strategies as well as at the level of scholarly literature on the region. So far, there had only been a handful of attempts to cut a clear path across the heterogeneity of diverse interpretations. Our aim in this paper is, then, to assess the leading theories of East-European economic transformation by analysing a particular institution, the board of directors. The novelty of our research lies in the fact that we utilize relational data depicting the network of interlocking directorates among the largest Hungarian corporations as well as the political sphere. Furthermore, our analysis will be carried out using the same dataset in assessing various theories, enabling us to test them at the same time.

Under the previous Communist regime, political power was the monopoly of the leadership of the state-party, while central planners and increasingly independent socialist managers vied for economic power. The political and economic transition of ‘89 rendered many of the assets of these groups obsolete and introduced a cognitive vacuum as to what constitutes a widely accepted asset in the post-communist era. While the introduction of formal democracy with regular elections institutionalized political power to a large extent, the source of economic power remained much more contested.

There are three leading interpretations seeking to pin down the real beneficiaries of the post-communist economic transformation. Chronologically, these interpretations appeared in an orderly sequence representing an evolution of the theory of economic transformation, propelled partly by the dynamic of events. One interpretation argues that the former socialist managers as well as the economically educated cadres of the defunct Communist Party (the Communist nomenclature) are the real winners of the economic transformations. One version of this argument claims that these groups, taking advantage of the policy of spontaneous privatization, came to own the lucrative productive assets of the country. Another version argues that using their insider information on companies the same group stayed on as leading executives of the privatized companies even after change of ownership. The culmination of the East-European events, according to this theory, will be a form of political capitalism or “crony capitalism”. A small elite will control most of the economic assets and draw huge profits from state monopolies, state subsidies, and relies on occasional bailouts. Local FIGs or chaebols dominate economic life and maintain a cozy relationship with the political sphere. This interpretation also posits a triple role for the state. These include the lasting presence of the state as an owner, the state as the lender of last resort, as well as an overbearing regulator.

The second interpretation holds that post-communist Hungary should be aptly characterized as a form of managerialism. The argument is that the lack of real owners, which is the result of an intricate web of cross-ownership between various public and private instances, shields executives from the kind of pressure that controls managers in mature capitalist countries.

Finally, the third interpretation argues that the financial sector will come to dominate economic life as banks increasingly keep their finger on the veins of capital in time of severe capital shortage. Empirically, this trend is reflected in the overdrawn material compensation of financial sector professionals. This theory draws on the well-developed theme of financial capitalism.

Although we agree with the main points of these three interpretations, in their current form they do not amount to more than insightful sketches of broad trends. Although former Communist cadres are frequent fixtures of the powerful posts of economic life, managers command preposterous salaries, and the long arm of the state pops up in bank bailouts and consolidations, the most interesting issue concerns the manoeuvres and institutions these groups use in establishing conditions advantageous for them.

As our aim then is identify these strategies, we decided to focus on the role of one particular institution, the board of directors. Our reasons were that since these boards supervise corporate executives, this focus provides a ready access to the structure of corporate governance, that is, to economic power. Beyond being the de facto institution of control over executives, we also assume that among the volatile economic and political conditions of post-communism, these boards were an important source of income as well as status for directors and sources of information and control for firms. A focus on the board interlocks rather than ownership ties represents a double advantage. First, focusing exclusively on ownership ties would not enable us to include corporate connections to the political sphere. Second, ownership ties represent a far larger commitment than a connection through board members, thus it cannot grasp the more subtle changes reflected in the latter.

The first part of our analysis contains a descriptive investigation of the theories of political capitalism, managerialism, and financial hegemony. After that, we will carry out a nested blockmodel analysis of our relational data. In the first phase, we will generate blockmodel images of firms based on the pattern of their ties within the whole intercorporate directorate network as well as to political organizations. Subsequently, we will delete political organizations from our data set to access the pattern of intercorporate relations nested within the larger structure analyzed in the first phase. Finally, in the third phase, we will delete financial institutions from the data to analyze the bare skeleton of corporate interlocks found in the shadow of financial and political relations.

In this section, we will review three interpretations of the fundamental changes occurring in the social structure and form of domination in Hungary. These lines of research include research on the composition of the new economic, political, and cultural elite, the fate of the former communist nomenclature and the continued economic role of the state reflecting a variant of political capitalism; a theory of managerialism as the fundamental form of domination in the post-communist world; and finally, the most recent theory of East-European economic transformation, the theory of financial hegemony.

Inquiry into the formation of the new elite of post-communist countries and the trajectory of former communist cadres was carried out in a large comparative research project conducted in seven Eastern-European countries. The findings of the research with respect to Hungary showed that there was considerable reproduction as well as circulation in the elite positions. Generally, former Party apparatchiks experienced downward mobility or retired, while the managerial and technocratic strata of the communist regime was quite successful in converting political position into economic leverage. In the political and cultural elite positions, however, there was a much more pronounced presence of new elements. These new groups were made up of the clientele of the victorious opposition party, "deputy managers" of former elite positions who used the transition to oust their former bosses, and humanistic intellectuals conquering numerous political and cultural elite positions. One of the most interesting finding of the research was that the presence of private entrepreneurs from the second economy was almost negligible in the economic elite, thus contradicting Szelényi's earlier prediction that this group will be the dominant force in the post-communist economic elite.

The second issue in the first theoretical group we wish to tackle concerns the economic role of the state after 1989. In order to appreciate the importance of the issue, we should realised that the Hungarian state has always played a strong role in the economy, and this was only amplified by the communist nationalisation of the economy. After the strict Stalinist command of the economy based on direct quotas and lack of independence from central redistributors weaned, from 1968 on the autonomy of managers of state-owned companies increased as decisions regarding the level of production, choice of suppliers, and investment were captured by management. The relation between the state and the economy (between central planners and redistributors and socialist managers) was doubly parasitic: the state appropriated the surpluses of companies, while companies relied on the state for crucial inputs, as infrastructure, raw materials from abroad, and educated workforce. By 1989, managers became largely independent from the state and through the process of spontaneous privatisation sought to self-privatise their companies. This process was halted by the first democratically elected government which opted for a centrally managed privatisation based on market prices. From then on the government made various attempts to regain control of the economy (which attempts then are labelled as renationalisation of the electronic media, the printed press, the health care system, strategic industries, banks, economic development agencies, state-owned investment banks, etc.).

The second interpretation of Eastern-European transition was delivered by Szelényi, Eyal and Townsley a year later. Here the authors asserted that among the conditions of dispersed ownership and the lack of a propertied bourgeoisie, there is not effective control over management. In their view of property relations in post-communist Hungary, they rely on the findings of Stark. After describing the intricate cross-ownership between public and private institutions, Stark claims that privatisation in Hungary aimed at separating assets and liabilities. While the former was passed into private hands, the latter was dumped on the public. Furthermore, Stark sees the intricate cross-ownership structures as an ingenious device of reducing uncertainty in a volatile market environment. By jointly owning enterprises as well as relying on partial public ownership, enterprises try to reduce the inherent uncertainty they face due to lack of information on other firms, an undeveloped stock market, and unpredictable state regulations. Thus, in the last analysis, these are hedging devices.

Although Szelényi et alia relies on Stark's findings of unclear ownership relations, they provide a much more cynical interpretation of these. Their initial thesis is that it is control over property, and not ownership, which is the basis of economic power in post-communist Hungary. Diffuse property relations are the very basis responsible for the lack of real owners and the buttress of unbridled management control. Furthermore, the authors claim that the new political elite legitimates the dominance of the managers by spreading the values of expertise, impartialness, and anti-politics after the overpolitisation of everyday life under the communist regime.

Unfortunately, Szelényi and alia do not provide empirical underpinnings for their insightful theory. The only data they rely on measures the housing conditions and stock holdings of managers. Since both of them seems quite modest, the authors conclude that it is not ownership and private wealth, but cultural capital that underlies dominance in post-communist Hungary.

The third theoretical attempt interprets post-socialist economic transformations as largely determined by the financial sector that occupies central positions within the economy. Gyorgy Lengyel and Attila Bartha used a survey data of the top-management of major Hungarian banks in tracing the threads of bank power. Their conclusion was that bankers as a group cannot be interpreted as the dominant group of the Hungarian economy, although they enjoy a highly privileged position based on material compensation. Balazs Vedres used relational data on the network of interlocking directorates to determine the centrality of banks. His conclusion was that banks are central in Hungary to some extent, but they are connected to companies with a significantly poorer performance.

The data was collected during the summer of 1997 in Hungary. The relational data depicts interlocks between the largest 350 firms in the Hungarian economy, based on 1996 gross revenue. In selecting firms for inclusion in our relational data set we relied on gross revenue as measure of corporate size since presently equity, profit, and employment are an unreliable index in the Hungarian economy. We distinguished two types of interlocks. One type, management-board tie, was coded when an upper-echelon manager of a firm served on the board of another firm(s). We collapsed the two types of boards distinguished in Hungarian Corporate Law: board of directors and supervisory board. A second type, “board-board” connection, meant that an individual served on the boards of two (or more firms). The corporate data comes from a CD-ROM that contains the files of the Hungarian Court of Registry. From the corporate files at the Court the following data on corporate entities are available on CD-ROM: managers, board members, designated activities, equity. This data contains the name of the officer as well as his or her address that make identification easier. This list contains almost 8,000 names.

The data on members of parliament comes from the official website of the Hungarian Parliament, which currently has 365 members. For double-checking our findings, we also relied on a data set that depicted the economic functions of MPs. The data on high-level central bureaucrats comes from the inner governmental telephone book of the Hungarian Government, effective as of May, 1997. This latter list includes the ministers, the political, economic and administrative secretaries of state, department heads and deputy department heads of all the ministries as well as central administrative agencies. This list contains 1,500 names.

After a labor-intensive data manipulation period, we created two matrices of the size 245*245. Out of the 350 firms, 19 ministries, 7 central administrative agencies, and 6 political parties we eliminated those which have no interlocks with any of the other entities in the matrices. One of the matrices, the management-board matrix, includes directed relations from the management of a given firm to the board of the interlocked firm. In the matrix, this means that a “1” was put in the intersection of the row of the sending firm and the column of the receiving one. Government bureaucrats and MPs were serving on the boards of corporations we also included in this matrix using similar notation. The another matrix, “board-board” matrix, contained a similarly coded tie in the lower diagonal between those firms that shared an individual as their board member.

As a final step, we collapsed the separate matrices of directional management-board and non-directional board-board connections. This way, the final matrix contained all the possible connections found within the largest firms as well as central state institutions and parliamentary political parties.

In the initial analysis, we sought to test the theories of managerialism, crony capitalism, and elite reproduction propounded above. At first glance, we could see that the MPs are not particularly attractive targets for companies. Of the 365 MPs only ten served as directors on the board of companies. Of the ten MPs, five belonged to the current heir of the former Hungarian Communist Party while the other five MPs belonged to different parties. Although this bias would have supported the thesis of elite reproduction, the overall low number of MPs indicates that generally MPs are not widely present on directorate boards. We further double-checked this finding by examining the firms on which MPs served as officers. These were usually small firms, focusing on developmental efforts in a particular region or promoting a single goal (culture, education, information diffusion).

The statist version of crony capitalism is much more visibly supported by our data. Among the ministries the presence of the officials of the Finance Ministry is clearly the most visible, but there is a strong presence of administrators from the Industrial, the Transportation and Communication, and finally from the Agricultural Ministries as well. Furthermore, officials enjoy directorships on firms that sectorally clearly fall within the purview of their home ministries.

Finally, the proportion of managers on company boards is strong, but it is impossible to simply assess the managerial thesis by simply looking at the proportion of managers on corporate boards. Generally, nothing comparable to the American corporate culture of performance supervision by owners, especially as it is exercised in the ‘90s, exists in Hungary. The scope of the stock exchange is much smaller, commercial loans represent the bulk of investment capital, and no institutional owners (funds, venture capital, a developed bond market) is yet present. Furthermore, neither quarterly reports, nor a fascination with the bottom line are part of the Hungarian corporate culture. Nonetheless, the companies whose board are most characterized by the presence of managerial accomplices, are the usual suspects of the Hungarian economy. For example, leading the pack is a Hungarian, which, while the Hungarian bank sector generally is posting improving results after four lean years, almost foundered a year ago after a desperate run on its assets by its depositors. The bank was finally bailed out by a handful of companies still connected to the Hungarian state as well as the deficit-ridden but freely spending national social security fund that routinely receives huge state subsidies to prevent its collapse.

Throughout the paper our aim was to go beyond simplistic descriptive analysis and substantiate the various theories of post-communist economic transition. To this end, we searched for an analytic strategy that let us investigate these theoretical interpretations at the same time. Leaving behind the terrain of descriptive analysis, we have decided to proceed as if peeling of successive layers of an onion. The purpose of this analytic strategy was to grasp the more minute structure of typical company board compositions nested within the larger interlocks running through the financial sector and the state. This, then, means a nested positional analysis of the interlocking structure of the Hungarian economy. More specifically, after generating a positional image of the top of the Hungarian economy, based on ties between firms as well as firms and state institutions and political parties, we decided to delete the political connections. In a further round, we decided to delete the interlocks running through the financial sector. This way, we hoped to seize the skeleton of ties between the firms without those generated by financial and political connections.

In the first part of the analysis we have used our full matrix with political and financial nodes included. Our aim was to trace the diverse patterns of board compositions, thus we have run a CONCOR analysis to distinguish groups of firms with different patterns. We have partitioned our dataset to eight blocks. In the following we will describe the patterns of this first layer. (See figure 1.)

The first block represents the board composition of companies tied to the state as partly or fully state owned companies, with a representative of the State Privatisation Agency (SPA) present on their board. These companies are also tied to the Department of Industry. Members from the Department of Finance and the Department of Transport and Communication are also present occasionally. Finally, there are several companies with board members from Budapest Bank as well as FOTEX and Muszertechnika holding companies.

The second block represents a type of board composition with a representative of the MOL Rt - the monopoly Hungarian oil company - as well as the Department of Finance. The companies here have in most cases connections to the SPA companies of the first block.

The third block is overwhelmingly dominated by the only Hungarian investment bank MBFB. Once again, these companies also have connections to the SPA companies.

The fourth block consists of companies with members from the Postabank, one of the largest Hungarian bank, on their boards. These companies also have dense connections to the SPA companies.

The fifth block is tied to the Department of Agriculture and consists of mainly agricultural companies. There are no connections to the SPA companies.

The sixth block is in the gravity field of the MSZP, the majority coalition party in the parliament. These companies have considerable number of connections to the SPA companies.

The seventh block is a block of agricultural companies that are tied to the fifth agricultural block. These companies have no board members from the political or financial sphere.

The eighth block is consisting of separate diads or triads of companies isolated of the other blocks.

Figure 1: the schematic graph of the first layer of the analysis. Blocks are labelled as B1 to B8.

In the second phase of the analysis we removed the ties to political actors. On the remaining matrix we were attempting to find the typical pattern of board compositions while keeping constant the effect of the state. (See figure 2.)

The first block is organised around Kereskedelmi Bank, the second largest Hungarian bank. Some of these companies also have board members from Budapest Bank, another large bank.

The third block consists of dyads and triads of industrial (mostly metallurgical) and transport companies. A large aluminium company and a bank have considerable influence.

The forth block is dominated by MBFB, an investment bank. Most of the companies here have connections to the first block.

The fifth block has OTP Bank and Corvinbank in its center. This is also a dense, cohesive group of industrial companies.

The sixth block is an agrarian block, in the center of which there is an agraricultural company.

The seventh block is also a cohesive agrarian block with another central agricultural company. This block is closely related to the sixth.

The eighth block consists of either isolated or dyadically organised companies.

In this phase of the analysis we have used only the ties between non-financial corporations.

The first block is a block of isolated nodes. These are the companies that have ties only to the political and/or financial sphere.

The second block is a cohesive subgroup of industrial companies (mostly heavy industry, power plants) and trading companies.

The fourth block is a sizeable field of dyadically organised companies with some outgoing ties to the second block. The only point of centralisation is a large aluminium factory.

The fifth block is a dense and centralised graph of mostly chemical and other heavy industrial companies with the MOL Rt in the center.

The sixth block is a small and dense agricultural block closely coupled with the third agricultural block.

The seventh block is also a dyadically organised industrial block with very little centralisation around three holding companies.

Our aim in the preceding blockmodelling analysis was to grasp changes in the overall structure of patterns of board composition. Blockomdelling the complete set of ties results in a robust picture of distinct board patterns. Out of the eight blocks, only one was characterized by an unorganized dyadic structure with one fifths of the nodes. Running CONCOR after eliminating political interlocks, the resulting blocks were still robust and it was easy to recognize a distinct pattern based on financial ties At this point though, roughly one third of the firms shows an unorganized diadic pattern. Finally, the blocks generated after the elimination of financial ties gets even less organized, although it is still possible to recognize patterns. More than half of the firms are isolated or part of isolated dyads.

We have attempted to quantify the processes described above. We followed the changes in three related measures as we subsequently deleted political and financial ties. These measures are the overall centralization of the network, the relative number of cliques, and the proportion of ties.

	overall centralization		relative number of cliques	proportion of ties	N
	Out	In
Phase 1	16.80%	5.70%	0.42	3.41	245
Phase 2	9.05%	5.92%	0.25	2.72	227
Phase 3	9.73%	3.28%	0.11	1.89	189

All three measures capture the same trend, that is, all these proxies for structural order fall twice as much as would be proportionally expected.

The findings of our nested blockmodel analysis provide support to some extent for the thesis of political capitalism as well as that of financial hegemony. Beyond these findings, we are disinclined to discard the theory of managerialism as even after successive deletions of financial and political ties, our blocks show a certain level of orderliness.

The findings of our research prompts us to carry out further investigation of the emerging Hungarian regime of corporate governance. The extent to which this issue could be resolved using quantitative methods requires a further elaboration of our data. The most important issue here concerns the exact process of board delegation, more specifically, whether we should see boards members as actively recruited by firms or sent specifically by owners. Including ownership ties in our data increases the explanatory power of this analysis, as that way we would be able to control for the situation where owners send their managers to serve on the board of companies owned. After that point, we are compelled to continue our investigations following a more qualitative style of research. Another interesting point guiding this research focuses on the implantation of a foreign institution, the board of directors, in local context.