The Political Role of Business Magnates in Competitive Authoritarian Regimes

This contribution examines the role of business magnates (“oligarchs”) in political transitions away from competitive authoritarianism and towards either full authoritarianism or democracy. Based on 65 cases of competitive authoritarian regimes named in the academic literature, 24 historical cases with politically active business magnates are identified for further investigation. The analysis shows that in about half of those cases business magnates do not have a distinct impact on political regime change, as they are tightly integrated into the ruling elites. If they do have an impact, they hamper democratization at an early stage, making a transition to full democracy a rare exception. At the same time, a backlash led by the ruling elites against manipulation through business magnates makes a transition to full autocracy more likely than in competitive authoritarian regimes without influential business magnates. JEL-Codes: D 72, D 74, N 40, P 48


Introduction
Wealthy businesspeople who are able to influence political decisions acting on their own are in many present and historical cases treated as "grey eminences" who hold considerable political power. While it is obvious that they influence specific policies, their impact on political regime change is less clear. This question is especially relevant for so-called "soft", i.e. electoral or competitive au-thoritarian, regimes in which business elites have more autonomy than under either full authoritarian control or democratic rule of law.
This analysis thus examines the impact of wealthy businesspeople on the development of electoral or competitive authoritarian regimes. The question addressed in this contribution is whether and under which circumstances business magnates in such a regime promote or hinder a transition of the political regime away from competitive authoritarianism towards either democracy or full authoritarianism.
There exist a number of terms for wealthy businesspeople. In relation to their political role, the term "oligarch" is most common. However, in the classic definition by Aristotle, as well as in a recent prominent conceptualization by Jeffrey Winters, the term refers to wealthy people in general, while a still broader definition sees oligarchy merely as the self-serving rule of the few in any context 1 . Accordingly, concepts relating to oligarchs focus on how material wealth translates into political power. This analysis, however, seeks to examine how a specific social group, wealthy businesspeople, affects political regime change.
To clearly denote this social group, the term business magnates will be used. As especially wealthy businesspeople, business magnates form a subgroup of business or economic elites. 2 This analysis starts with a brief overview of theoretical concepts addressing the relationship between business magnates and specific types of political regimes. Based on this review, an analytical framework will be developed to analyse the impact of business magnates on change in competitive authoritarian regimes. Then, historical cases are identified and examined in order to gain a better understanding of the role of business magnates in political regime trajectories, which is summarized in the conclusion.

Business magnates and political regime contexts
The political regime of a country obviously influences how businesspeople represent their interests in political decision-making processes. On the most basic level, Winters makes a distinction based on the ability and will of the state to guarantee property rights through a properly functioning and impartial judicial system. If such a system is not provided, oligarchs, whom he defines as the extremely wealthy, engage in "wealth defence" in which the major aim is to ensure that wealth is not taken away by a predatory state or rival businesses colluding with corrupt law enforcement agencies. Promising strategies for wealth defence include hiding ownership, e.g. by transferring companies to offshore holdings, and creating political networks to ensure support against predatory attacks. If, however, property rights are secure, business magnates focus on "income defence." In such a situation, the main aim is to keep as much of the earnings generated by their businesses as possible. The most promising strategies for income defence relate to the minimization of tax payments and restrictions on competition from rival businesses. 3 Winters then constructs a typology of political regimes based on the roles oligarchs play. However, of his four types -warring, ruling, sultanistic and civil oligarchies -only the latter two refer to peaceful modern states. Sultanistic oligarchies -for which Winters uses the Philippines under Marcos and Indonesia under Suharto as prime examples -require wealth defence, while in civil oligarchies, such as the USA or Singapore, oligarchs can limit their political activities to income defence, as property rights are secure. 4 As one book review notes, Winters sees oligarchies as inherently stable, threatened only by "exogenous shocks to the otherwise machine-like oligarchic system." 5 Accordingly, Winters does not address the impact of oligarchs on political regime change, nor does he relate his four types of oligarchies to established political regime typologies. With Singapore, he has chosen a clear outlier combining authoritarianism and the stable rule of law, and, including this exceptional case as one of only two examples of civil oligarchies, blurs the link to political regime types.
Income defence as a major interest is most likely to occur in fully functioning democracies, as they guarantee secure property rights. Additionally, the rule of law, which is the basic precondition for secure property rights and is associated with a focus on monitoring and transparency, makes engaging in illegal methods of political influence riskier. Accordingly, oligarchs can influence stable democratic politics by lobbying for their business interests, but they are unlikely to promote a change of political regime. Even in not properly func- || tioning democracies, there are strong counterweights to systemic corruption and takeovers of political power by business interests. 6 In contrast, in a stable hegemonic or closed (i.e. fully) authoritarian regime, corruption may be endemic, but business representatives cannot act independently in order to influence politics in their favour. Instead, business elites become a part of the ruling elite and, thus, of a clientelist system, which offers them business promotion in return for political loyalty. Accordingly, most businesspeople engage in wealth defence by supporting the ruling political elites and hoping to win patrons for their business. Examples of such situations are socialist economies where a communist party dominates 7 or the fully authoritarian oil-and gas-producing countries of the Persian Gulf and the Caspian Sea, where domestic oil companies are economically important but have not become independent political actors. 8 However, in electoral or competitive authoritarian regimes, which combine a formally democratic constitution with strong features of authoritarianism, businesspeople have more leeway to influence political decision-making informally. In a large empirical analysis of smaller companies in Russia and Ukraine after the end of socialism, Markus describes how businesspeople face this challenge by forming alliances with various societal actors. 9 Due to their broader power, business magnates can engage in wealth defence on their own by forging direct links with relevant politicians. Thus, they often become major political actors themselves. For the post-socialist regimes in Eastern Europe, it is claimed that business magnates have been able to "capture the state" in order to ensure long-term access to large-scale rent-seeking opportunities; 10 thus, they are not just defending their wealth but creating it through corrupt connections with political elites.
The conceptualization of such regimes best suited to understanding the role of business magnates is provided by Levitsky and Way: "competitive authoritarian regimes are civilian regimes in which formal democratic institutions exist and are widely viewed as the primary means of gaining power, but in which incumbents' abuse of the state places them at a significant advantage vis-à-vis their opponents. Such regimes are competitive in that opposition parties use democratic institutions to contest seriously for power, but they are not democratic because the playing field is heavily skewed in favour of incumbents. Competition is thus real but unfair." 11 Big business can play an important role in such a system, as it can use financial power and media ownership to help the ruling political elite create an uneven playing field. 12 Business magnates can adjust to such a system easily. Gill argues that the broader group of an industrial bourgeoisie in the democratizing societies of the 19 th and 20 th centuries "sought to fit into the hierarchy of power as it found it, using existing processes and structures to press its concerns and widen its influence." 13 Similarly, an analysis of 296 Russian businesspeople for the period from 1995 to 1999 leads Braguinsky to conclude, "new oligarchs appear to have by and large accepted the rules of the game […], simply using political influence to their own advantage in the fight for control of valuable assets." 14 Looking at political developments in such a context, Hale states that research has to shift "from a logic of regime change to a logic of regime dynamics, a logic that […] can capture how the moving parts of highly patronalistic polities (such as oligarchic networks and regional political machines) arrange and rearrange themselves in regular, even predictable ways that might on the surface look like a regime "change" but that in reality reflect a stable core set of infor- mal institutions and operating principles." 15 Accordingly, competitive authoritarian regimes can remain stable for longer periods of time. 16 In a numerical analysis of competitive authoritarian regimes for the period from 1975 to 2004, Brownlee finds "that competitive authoritarian regimes are not especially prone to losing power but are significantly more likely to be followed by electoral democracy" than by full authoritarian regimes. However, he does not examine the impact of business magnates on these developments. 17

Business magnates and stability or change in competitive authoritarian regimes
The literature offers three distinct hypotheses on the role of business magnates in competitive authoritarian regimes. The classical view, especially in the neo-Marxist tradition, assumes that economic elites in general prefer fully authoritarian states in order to promote their interests over those of, e.g. workers. 18 The state capture hypothesis, developed by Hellman et al. 19 based on the post-socialist countries of the 1990s, claims that business magnates aim to stabilize competitive authoritarian regimes as it maximizes their rent-seeking opportunities. This hypothesis is in line with current concepts of authoritarian regimes, which as quoted above, highlight the stability of competitive authoritarian regimes.
In contrast, the secure property rights hypothesis argues that once business magnates have become rich, their interests will switch from what Winters calls wealth defence to income defence; i.e. they will favour the establishment of rule of law and thus a transition to democracy. Schimpfössl summarizes this shift as the "familiar story of raw robber barons beginning the journey to becoming gentlemen." 20 Fry argues that this transformation of business magnates also has the potential to increase the legitimacy of their property rights, thus giving them some agency in the establishment of secure property rights. 21 Albertus and Gay develop a formal model to demonstrate that uncertainty about the future development of a dictatorship can cause economic elites to prefer the predictability of rules-based democracies. 22 In line with this political economy perspective, North, Wallis and Weingast argue that -in 19 th and early 20 th century Western societies -elites started to favour impersonal rights, property rights prominently among them, over privileges because the resulting "open access order" was perceived to be more favourable to their personal well-being in the longer run by guaranteeing rights independently of a specific person, similar to a king. After elites had agreed on impersonal rights, the concept was extended to society at large. 23 While business magnates may have their own preferences in relation to political regime types, they may also simply behave opportunistically. Based on this assumption, Markus/Charnysh distinguish between a logic of flexibility and a logic of commitment compensation. While the first logic is based on switching to the winning side in a political power struggle, the second logic makes business magnates committed members of a specific political camp, thus linking their fortunes and requiring a much more active role in politics, often including the assumption of formal office. 24 As this study focuses on the most influential business magnates, there are also cases where they do not commit to a preexisting political camp but create their own.
The transition literature on cases of democratization has highlighted the relevance of elite pacts for political regime change. 25 This literature has largely ignored the role of business magnates, thus implicitly promoting the null hy-pothesis that business magnates do not have a discernible impact on political regime change.
Following the approaches presented above, the analytical framework employed in the following analysis focuses on the power position and the interest of business magnates as explanatory factors for their impact on political regime development in competitive authoritarian regimes. As outlined, the authoritarian preference hypothesis sees a stable coalition of dictators and business magnates based on the latter's interest in support against other societal groups, the state capture hypothesis claims that business magnates promote the stability of competitive authoritarian regimes in order to secure their opportunities for rent seeking, while the secure property rights hypothesis sees business magnates favouring a transition to democracy in order to promote a more orderly and stable position of their fortunes. Obviously, the impact of business magnates does not determine political regime change (or the lack thereof) but yields an impulse in a specific direction that can be outweighed by other factors.
To identify the respective impulses, the trajectories of all historical cases of competitive authoritarian regimes with politically influential business magnates have been examined in this analysis. Following the distinction of Markus/ Charnysh, they are grouped into three scenarios based on the position of business magnates vis-à-vis political elites.
-Scenario 1: Business magnates integrating into one existing political camp.
In this scenario, business magnates support the status quo and thus contribute to regime stability, which in a competitive authoritarian regime, would be in line with the state capture hypothesis. If, however, business magnates dissolve into the broader group of ruling elites, i.e. they do not play a distinct role, this would support the null hypothesis.
-Scenario 2: Business magnates switching political camps. In this scenario, business magnates confer additional power on the supposedly winning side, thus acting as catalysts of regime change. It is important to note that this logic of flexibility implies that business magnates do not have their own agenda but simply act opportunistically. The secure property rights hypothesis, on the other hand, proposes that business magnates are more likely to support a specific type of regime change, namely, a move to a regime with secure property rights, while the authoritarian preference hypothesis sees them promoting full autocracy.
-Scenario 3: Business magnates creating their own political camp. In this scenario, business magnates set up their own political camp in order to challenge the ruling political elites; i.e. they do not set up a political party in order to join a ruling coalition -e.g. as several business magnates did in Ukraine -but to try to take power from the ruling elites. Only cases where business magnates have a real chance to do so are included in this scenario. Accordingly, business magnates can independently try to initiate the kind of regime change they prefer. The secure property rights hypothesis sees potential for democratization, the authoritarian preference hypothesis the opposite. However, if the state capture hypothesis is correct, business magnates in a competitive authoritarian regime will simply use their political power for rent seeking, not promoting any change in a political regime from which they benefit.
These scenarios are based on the one initial condition (the power position of business magnates) that can be easily determined empirically. Accordingly, the result of interest (political regime stability or change) will differ inside each scenario. These scenarios are, therefore, only a first attempt to group a larger number of empirical cases. Based on the empirical results, the conclusion will present specific regime trajectories linking initial conditions and regime developments.

Case selection
In this analysis, only politically active business magnates are included. The selection criteria are as follows. 26 -Business as the main interest: To define business magnates as a group that can be analysed separately from political elites, this criterion draws an analytical distinction between business magnates who engage in politics to promote their business activities and fully-fledged politicians who aim for political power as an end in itself. This means that if business magnates assume formal political office, they focus on their own narrow business interests. When business magnates start to develop a broad political agenda and engage in full-time policy-making beyond their business focus, they are no longer treated as business magnates in this analysis. Similarly, politicians or civil servants who use their political influence to control economic activi-|| 26 For a more refined country-specific definition along these lines, see H. Pleines, Oligarchs and Politics in Ukraine, in: Demokratizatsiya 24/1, 2016, pp. 105-127, here pp. 114-116. ties but continue to focus on politics are not defined as business magnates, even if they become rich. 27 -High level of personal wealth based on entrepreneurial activities: To restrict the analysis to richer entrepreneurs, this criterion draws an analytical distinction between business magnates, the definition of whom implies financial weight, and the broader social group of businesspeople. The amount of wealth required to be counted as a business magnate is case specific. 28 -Political activity at the national level: The aim of this criterion is twofold.
First, it restricts the analysis to businesspeople who are politically activethe key aspect examined in this analysis. Second, it restricts the analysis to the national level, as political regime dynamics are usually determined at the national level. Regional and local politics are more diverse in most countries. Political activity at the national level can be formal or informal, but it should be clear that the entrepreneur has the intention and potential to influence political decision-making processes at the national level on a regular basis.
The relevant cases of politically active business magnates in competitive authoritarian regimes were identified in three steps. First, a list of competitive authoritarian regimes was compiled. The major source for this exercise is Brownlee's global (though incomplete) dataset. 29 Additionally, the list of competitive regimes provided by Levitsky and Way to support their case studies was included. 30 As these lists only cover the time from decolonization to the first decade of the 21 st century, several earlier and three more recent cases were added to fully cover the period from the Industrial Revolution in the 19 th century until the present time (i.e. up to 2018). 31 In sum, 65 countries were identified as || 27 This criterion creates a distinction from Winters' concept of oligarchy (introduced above), which focuses on material wealth as the source of political power, whereas this analysis focuses on the role of business magnates in politics. To avoid missing a relevant case, all cases identified by at least one source as competitive authoritarian were included at this stage. In step 3, all cases including business magnates were scrutinized to avoid false classifications.
having had a competitive authoritarian regime for at least one period lasting not less than five consecutive years.
Second, the existence of rich businesspeople in all cases of competitive authoritarianism was checked. 32 The starting point was the 35 country studies that Levitsky/Way present in their analysis of competitive authoritarian regimes. 33 For recent cases, the global list of billionaires published by Forbes since 1987 34 was used to identify the number of billionaires in competitive authoritarian regimes. 35 Business magnates, as understood here, do not necessarily have to be billionaires; thus, additional national lists were consulted to identify the richest businesspeople in the respective countries.
Third, a systematic literature review 36 was used to confirm the relevant cases and to collect information on the cases. 37 As a result, 24 cases of politically influential business magnates in competitive authoritarian regimes have been identified. A list of all cases included in the analysis is given in Table 1. Most, but not all, come from three distinct geographic regions and periods.
|| 32 There are several cases of competitive authoritarian regimes that did not provide businesspeople with the ability to rise to the status of business magnate and/or to act independently in politics, e.g. because of state or ruling party control over the economy, a low level of socioeconomic development or the very small size of the national economy. Moreover, in a few cases, foreign investors who are not included in this analysis dominated the national economy. This primarily concerns Latin America in the final third of the 20 th century, most often in Central America. There is also a diverse group of African countries with competitive authoritarian regimes, mainly in the 1990s and 2000s, which lack business magnates. 33 Levitsky/Way, Competitive Authoritarianism, parts II and III. 34 Since 1996 available at http://stats.areppim.com/stats/links_billionairexlists.htm, 02.05. 2019. Latin America for earlier years based on https://www.forbes.com/sites/doliaestevez/2017/ 04/05/brazil-tops-latin-billionaires-on-lists-3rd-decade-while-carlos-slim-remains-regions-richest/ #783cc5c53ac6, 02.05.2019. 35 The Forbes list, similar to Winters' concept of oligarchy, covers all wealthy people including, e.g. monarchs. This study, on the contrary, is restricted to businesspeople, as they are assumed to follow a specific logic of action (which clearly differs from the roles, strategies and available instruments of, e.g. monarchs). 37 Google scholar (https://scholar.google.com/) and the online network of library catalogues (www.gbv.de) were searched for references using "business" or "oligarchs" (as these terms are often used in the literature) in relation to "politics" or "political regime" for each country included in the list.
The first set of case studies concerns the influence of economic elites, landowners and merchants in late 19 th and early 20 th century Latin America, namely, Argentina, Brazil, Chile, Colombia and Uruguay. Their influence was largely based on rural clientelism and exports of natural resources. The second set of cases comes from Southeast Asia, where powerful industrial holdings emerged in the wake of rapid industrialization since the 1970s. This set includes one of Winters' exemplary cases of sultanistic oligarchies, Indonesia, as well as Malaysia, Thailand and Taiwan. South Korea and Turkey are similar cases from other regions of Asia. Finally, the third set of cases stems from the post-Soviet region, i.e. the countries of the former Soviet Union. In this region, most prominently in Russia and Ukraine but also in Armenia, Georgia and Moldova and partly in Azerbaijan and Kazakhstan, business magnates managed to assemble large business holdings when the socialist planned economy was dismantled in the 1990s.

Scenario 1: Business magnates integrated into an existing political camp
In a number of cases, politically active business magnates have been an integral part of the ruling elites as outlined in this scenario, namely, in some Latin American countries (5 countries in the 1860s-1920s and Mexico in the 1980s-1990s), Malaysia (1970s-2010s), Taiwan (1990s) and the post-Soviet region (Armenia, Azerbaijan and Kazakhstan). The case studies show two different ways in which the integration of business magnates can develop. On the one hand, integration can result from long-term socialization, as in the early Latin American cases or in Malaysia a century later. On the other hand, integration can be achieved by control and pressure, as in the cases from the post-Soviet region, where business magnates newly emerged after the end of the Soviet planned economy.

Example 1: Latin America (1860s to 1920s)
In the early 19th century, the newly independent states of Latin America adopted distinctly republican forms of government, with the exception of the constitutional monarchy of Brazil. The resulting models of government, though regularly challenged and adjusted, dominated until the 1870s, mostly resembling illiberal or chaotic democracies. As Sabato argues, "Despite the controversies often raised around Spanish American elections, for most of the period, these were the main legitimate road to government posts, and they offer a key to understand the politics of the republic." 38 Additionally, active civil societies were developing, creating the societal foundations for democracies. 39 However, political challenges and armed revolts against the constitutional order were a common feature of Latin American politics during that period. 40 Starting in the 1870s, Latin America experienced a phase of rapid economic growth based on commodities exports. 41 In this period, several Latin American countries developed political systems that have been described as oligarchic rule or oligarchic republicanism in the literature. In this context, the term oli-  81-137. garchy refers to rule by a small elite circle, which is not necessarily characterized by material wealth or a prominent role of business magnates. In these countries, elections were still the major means of assigning political power, and there were real challengers. However, political competition was heavily skewed in favour of incumbents.
The basis for this was the creation of a strong presidency with far ranging powers of interference in democratic processes and constitutional rights. The major weapon against the political opposition was manipulation of the electoral process. In the 19 th century, this regularly included restrictions on voting rights based on gender, ethnicity, property and literacy. As a result, on average, less than 10 percent of the population actually took part in elections. Moreover, vote buying and pressure on voters from state representatives, including via the police and employers, were common. While political liberties were heavily restricted, civil liberties were much better developed and protected. 42 Chile had already adopted such a regime in the 1833 constitution. In later periods of the 19 th century, Argentina, Brazil, Columbia and Uruguay also developed forms of oligarchical democracy. 43 The oligarchic republics of Latin America can be interpreted as the competitive authoritarian regimes of their time. As this assessment is based on earlier democratic standards and refers to largely rural societies, the oligarchic republics of 19 th century Latin America could also be called early modern competitive authoritarian states.
In these countries, the traditional economic elites who meet the criteria for business magnates, i.e. business interests as the core activity, vast personal wealth and political activity at the national level, were large landowners. Their political power rested on a traditional system of clientelism, also called caciquismo, which gave them control over the people living and working on their land. Clientelistic political power was thus regionally fragmented. 44  the wake of the 19 th century, merchants gradually accumulated wealth and political power -usually within the span of three to four generations. Both landowners and merchants were acting in the context of extended families and formed integral parts of the ruling elites. 45 It is important to note that business magnates were not acting individually or with a distinct identity but as part of broader elite groups in the context of family networks.
At the same time, "the image of an overall concentration of wealth and economic power in the hands of a small and closed class of powerful families may apply to some specific cases and periods, but in many others -and especially in the larger countries -it fails to account for the dynamism of these societies where the powerful of one day could be successfully challenged the next. Within the context of a nonlinear transition from a basically traditional mercantile structure of production to one increasingly marked by capitalism, the upper classes were subject to the risks and hazards posed by the new demands of the system." 46 Around the turn of the century, major challenges arose from public pressure created by a growing urban educated middle class and strong workers' movements.
In Uruguay (1903), Argentina (1912), Chile (1920) and Colombia (1936), increasing political competition challenging the hold of conservative parties on power led to a change of political regime towards unstable democracy. However, the regime trajectory did not move to consolidated full democracy because of strong resistance from the elites, including traditional economic elites and new ones based in industry. Additionally, political violence remained strong. Especially during periods of economic recession, political regimes came under pressure. Populism and violent left-wing forces were further major challenges to democracy. In response, conservative forces regularly aimed to maintain dominance by means of military coups. This was the case in Argentina (first of several in 1930), Uruguay (1933and 1973), Colombia (1953) and Chile (1973 democracy was established in Brazil in 1945, which was again terminated by a military coup in 1964. 47 As a result, the regime trajectory leading away from competitive authoritarianism did not move beyond an unstable democracy, populism and military rule for at least several decades in all five countries. It is, therefore, not possible to argue that there was a direct path from competitive authoritarianism to full democracy. Instead, the regime trajectory runs from competitive authoritarianism to unstable democracy through the rise of an urban middle class and workers' movement, where conservative political forces, which are backed by the old business elites, loose political power but remain influential enough to prevent the consolidation of full democracy.

Example 2: Post-Soviet Region (1990s to 2010s)
While in 19 th century Latin America, business magnates were socialized into the ruling elites over generations, they were newcomers in the post-Soviet region where socialist planned economies had banned private ownership of capital goods until the late 1980s. The end of socialism offered entrepreneurs the chance to make vast sums of money quickly with the help of political connections. The first future business magnates of the region started in the early 1990s with trade and financial operations. In both cases, large gains were only possible with political support. Regulatory and inspection authorities turned a blind eye to new entrepreneurs' activities. National banks provided preferential credit. State enterprises became clients. 48 The most successful new businesspeople used their profits from financial and trading activities to build industrial holdings. They took over state enterprises over the course of the privatization process, and their trading companies exploited customers' debts during bankruptcy proceedings in order to bring them under their control. Here, too, state support was indispensable. The pri-  , 1930-1970, in: Comparative Political Studies 13/2, 1980 vatization process was, in many cases, manipulated by the state officials in charge. Bankruptcy proceedings were also frequently steered in favour of the trading companies. 49 This extensive rent seeking was only possible in political regimes that gave business people enough leeway for independent action, i.e. which were not fully authoritarian, while lacking the rule of law to curtail many of the dubious deals that allowed business people to become billionaires in less than a decade. 50 The Forbes list did not include a single billionaire from the post-Soviet region up to 2000. However, over the following 18 years, the highest number of billionaires reported was 111 for Russia (2014), ten for Ukraine (2013), six for Kazakhstan (2008) and one for Georgia (since 2012). Politically influential business magnates in the competitive authoritarian regimes of the post-Soviet region exploited first their money and then their networks to support specific political actors. Many also took over mass media outlets in order to influence public opinion, on several occasions with the obvious aim of silencing critical reporting on their political allies. In some countries, business magnates have also run for or been appointed to formal political offices.
In Armenia, all business magnates were fully integrated into the ruling elites of a competitive authoritarian regime, being a part of the electoral machine not only through the provision of financial support but also by pressuring their employees into voting for the ruling party. 51 Levitsky and Way quote an Armenian legislator who claimed, "you cannot be a leader of an opposition party and have a businessman as a partner -he would be eliminated as a businessman." 52 The most prominent business magnate actively involved in Armenian politics has been Gagik Tsarukyan, leader of the Prosperous Armenia party. After coming second in the 2007 parliamentary elections, his party informally supported the ruling party. When Tsarukyan declared his intention to run in the 2013 presidential elections, state pressure soon led him to rethink the move. 53 For the cases of Azerbaijan and Kazakhstan, Radnitz has argued explicitly that the strategy of the ruling political elites to keep business under control, namely, refraining from large-scale privatization, 54 is the key factor behind the political stability in these countries as compared to the "colour revolutions" experienced by neighbouring countries (Georgia, Kyrgyzstan, Ukraine). 55 This stronger control over potentially independent business actors allowed political elites in Azerbaijan and Kazakhstan to move towards stronger authoritarian control in the 2000s. A similar argument for political regimes in sub-Saharan Africa is made by Ariola. 56

Summary
A lack of independence on the part of business elites can be due to very close relationships between political elites and business magnates based on socialization and networking -which makes defection difficult because of actual commitment to the ruling regime and a lack of credibility for defection as in the cases of Latin America (1860s-1920s) or Malaysia (1970s-2010s). 57 Business magnates' lack of independence can also be due to relatively strong authoritarian control, as in the cases of Armenia (1990s-2010s), Azerbaijan (1990s) or Kazakhstan (1990s).
In the cases of Azerbaijan and Kazakhstan, this led to the introduction of a fully authoritarian regime. However, as business magnates simply move in line with the ruling elites, this scenario also allows for democratization from above, as in the cases of Mexico (1970s to 1990s) 58 or Taiwan (1990s) where business magnates whose business holdings were deliberately restricted by the government 59 did not inhibit the transition from a full authoritarian regime to competitive authoritarianism in the 1990s or later to full democracy. 60

Scenario 2: Business magnates creating their own political camp
While the cases in scenario 1 point to the lack of a distinct impact of business magnates on political regime change, this is clearly not true for the cases in scenario 2 wherein business magnates create their own political camp. Obviously, they had to feel strong enough or consider the ruling political elites to be weak enough to make independent political action by business magnates look feasible. Accordingly, there are only four cases where business magnates managed to become major contenders for political power with the political camp they created. 61 In two cases, Thailand (2010s) and Madagascar (2000s), they failed; in another two, Georgia (2000s-2010s) and Moldova (2010s), they were successful. 61 It has to be noted, though, that according to the definition of business magnates used here, the only cases included in this scenario are those where business magnates themselves -or at least some of them -do not assume important official positions in politics but remain businesspeople who exert influence mainly in an informal capacity.

Example 1: Thailand
Thailand is an interesting case of more proactive business magnates in politics opting for what Markus/Charnysh call the logic of commitment compensation. The political regime of Thailand started to move from fully authoritarian military rule to competitive authoritarianism during the mid-1970s. Businesspeople were an important factor in the creation of an uneven playing field which marks the politics of competitive authoritarianism. As Phongpaichit and Baker summarize, "local businessmen invested heavily in getting elected, ensuring money was a key qualification. In the 1990s, approximately 70 percent of MPs were drawn from the ranks of male business owners who make up less than 3 percent of the population. In effect, parliament was a mechanism for drafting business leaders into the ruling oligarchy. The generals withdrew from the political front line but retained considerable influence, status, and independence. Top bureaucrats negotiated a new distribution of power with elected politicians for mutual benefit." 62 In this system, entrepreneurs with political connections could engage in large-scale rent seeking. Thaksin Shinawatra, for example, served different posts in the state administration and government while his companies simultaneously benefited from state procurement and highly profitable state licences in the telecommunication and TV business. By 1996, Forbes listed ten billionaires in Thailand, with Shinawatra in fourth place. According to Sen and Tyce, "these years saw the emergence of unrestrained 'money politics', with money becoming vital for bringing parties to power and keeping them there. Elections were won by those who could buy the most votes -a trend established in 1992, when the military ploughed vast sums into buying the previously untapped, but now increasingly engaged, rural vote. Elections became little more than bouts of 'horse-trading', as parties sought not only to purchase the support of voting blocs, but MPs from other parties, signalling the increasing strength of lower-level actors. Prior to the 1995 election, for example, MPs could command 'transfer fees' of 20 million baht [at that time, 0.8 million US dollars]. Once in office, the logic of competitive clientelism guided politi- cians; they had to recoup expenses and start building a coalition for the next election by diverting resources to key allies and constituencies." 63 When public discontent rose in the wake of the Southeast Asian financial crisis of 1997, Thaksin Shinawatra tried to create a rival political camp around his TRT party. The party managed to win the 2001 elections, and Shinawatra became prime minister. He was cleared of corruption charges by the constitutional court by a narrow vote and for technical reasons. As prime minister, he continued to promote his business interests, e.g. changing legislation to make the sale of his company tax free. 64 The old establishment, which opposed him, was supported by another business magnate, Sondhi Limthongkul, who had amassed huge media holdings. He used the media to attack the Thaksin Shinawatra government after the latter won a second term in office in 2005.
In the resulting power struggle, the business magnates sought not only to control media reporting but also to provide financial resources for protests actions. Both camps arranged long-term disruptive mass protests in their struggle for political power. In 2006, this led to the resignation of Thaksin Shinawatra in the wake of a short-lived military coup that left Shinawatra's proxies in power, and in 2008, the old establishment managed to gain control of government after bloody protests. This in turn provoked mass protests from the Thaksin Shinawatra camp in 2009/10. After an election victory in 2011, Yingluck Shinawatra, his sister, became prime minister. Mass protests were again used to gain or defend political power in 2013: "Many of the street protests were financed by oligarchs, although this time the money has been more difficult to trace to specific individuals.
[…] A Bangkok-based political think tank estimated that the 2013-2014 round of street protests averaged 'around five million baht [approximately 153,000 USD] a day on sustaining the protesters.'" 65 However, in the longer run, the mass protests helped neither of the two camps of politicians and business magnates to power. Instead, the military took over in a putsch in 2014. 66 Thus, Rhoden identified the following political regime trajectory: "oligarchic funding -paid-in-part protests -security concernmilitary takeover." 67 That is, the attempts of rival business magnates to gain political power provoked a return to full authoritarianism in the form of military rule.

Example 2: Georgia and Moldova
A similarly dominant position of business magnates in politics emerged in Georgia in the early 2000s when businessmen were among the first to challenge President Eduard Shevardnadze. 68 Prominently among them was Arkadi "Badri" Patarkazischwili, who had made his fortune in Russia. Another Georgian business magnate who had made his fortune in Russia, Kakha Bendukidze, became the most prominent oligarch in the Georgian governments of the 2000s, serving as a minister for five years. The most influential business magnate in Georgian politics, however, is Bidzina Ivanishvili, the only Georgian who has been included on the Forbes list of billionaires to date. In 2012, he founded the oppositional Georgian Dream party, which won the parliamentary elections that year and nominated him as prime minister. After his business came under scrutinyincluding in the wake of offshore leak scandals -he resigned as prime minister and party leader. When the Georgian Dream candidate won the 2013 presidential election, Ivanishvili was generally seen as the grey eminence in Georgian politics. After he forced the resignation of the government, he announced his return to the party leadership in 2018. 69 In Georgia, business magnates are thus a major force stabilizing the political regime between competitive authoritarianism (the classification by Levitsky and Way) and illiberal democracy (the classification by Brownlee).
In Moldova, the dominance of Soviet nomenclatural networks in the 1990s and Communist Party rule in the 2000s, in combination with the small size of the national economy, delayed the rise of business magnates. 70 Only when the Communists lost power in 2009 did business magnates assume a prominent role in politics. Anatol Stati, founder of the Ascom Group, supported an opposition party founded in 2007 and was seen as a major force behind the removal of the Communists from power. In 2009, business magnate Veaceslav Platon entered parliament; in 2010, Vlad Plahotniuc, who had acquired four of five national TV stations, followed. While Platon was neutralized by a long-term prison sentence, by 2015, Plahotniuc had emerged as the grey eminence of Moldovan politics, similar to Ivanishvili in Georgia. He managed to form a broad coalition of political parties. Although his official positions -parliamentary deputy from 2010 to 2013 and again from 2014 to 2015 -look minor, he in fact indirectly controlled the government through his position as leader of the Democratic Party until 2019. 71 Again, a coherent attempt to change the political regime is not visible. Instead, the ruling elites seem to use informal means to create what Levitsky and Way call an uneven playing field for political competition. Moldova, too, is stuck on the border between competitive authoritarianism and illiberal democracy.

Summary
The creation of a new political camp (as opposed to a new political party which can support the existing political camp) clearly places a business magnate in opposition to the ruling political elites. As the rule of law is not observed in competitive authoritarian regimes, such a strategy is a very risky endeavour, which endangers not just political perspectives but also business interests and personal freedom. The combination of the considerable resources needed to implement this strategy and the high risk attached to it explain why this scenario is exceptional.
In the two cases where magnates were successful, Georgia (2010s) and Moldova (2010s), the political strategy of business magnates can continue for a longer time. However, if the army or security forces oppose the business magnates, they can intervene, as they did in Thailand (2006 and2014) and Madagascar (2009). 72 If the army returns to the barracks (as in Thailand in 2006 and in Madagascar), the chaotic politics of competitive authoritarianism manipulated by business magnates continue. If the army stays in power, as in Thailand after 2014, the result is the establishment of full authoritarian rule.

Scenario 3: Business magnates switching political camps
In many cases, business magnates follow a logic of flexibility and align with the winning side in a struggle for political control. If they are successful, they manage to preserve their rent-seeking opportunities. Examples are Indonesia (1990s-2000s), South Korea (1980s-90s), Ukraine (2000s-10s), Kenya (since 1990s) and Malawi (since 1990s). However, flexibility clearly marks these business magnates as unreliable allies. Accordingly, political elites who want to increase their grip on power are likely to sideline politically influential business magnates, as has happened in Russia (2000s), Turkey (2010s), Venezuela (2000s) and Nicaragua (2018).

Example 1: Indonesia
Under the authoritarian rule of Suharto, which was established in 1967 and relaxed to competitive authoritarianism in the mid-1970s, Indonesia's business elites were firmly integrated into the ruling political elites. However, by the mid-1990s, some business and state administrative elites were alienated by Suharto's increasingly personalist leadership -e.g. promoting his family members to influential positions. The 1997 Asian financial crisis, which caused the GDP to decrease by over 10 percent in the following year, was a trigger of widespread public protest. While the literature on the case does not agree on the importance of elites or civil society in bringing about the end of Suharto's rule and the start of democratization in Indonesia, both clearly played a part and reinforced each || 72 Levitsky/Way, Competitive Authoritarianism, on Madagascar pp. 276-282. other. Business magnates retained their political influence after the regime change and became more independent from political and state administrative elites. Informal manipulation of politics remained an important part of their strategy, thus preventing Indonesia from moving beyond an incomplete democracy. 73

Example 2: South Korea
In South Korea, the military government had, since the 1960s, supported the development of huge industrial holdings called chaebols even more proactively than other Southeast Asian countries. By the 1990s, the turnover of the ten largest holdings equalled about half of the country's GDP, while the same indicator stood at about one-fifth of GDP for Taiwan and one-fourth for Indonesia. 74 In South Korea, the result was a very close relationship between the political elites and the families and managers of the chaebols. 75 Many high-ranking managers had worked in the state bureaucracy or with the military before joining a chaebol. 76 When the military regime moved towards competitive authoritarianism in the mid-1970s, this relationship was not challenged. Even after the end of military rule in 1987 and the establishment of democracy, changing political elites and chaebol leaders remained close. In the 1990s, the chaebols continued to "channel their resources into media, academia, and bureaucracy, shaping public opinion to their advantage. […] Politicians take the chaebol's threats of an investment strike seriously because of its consequences on economic performance." 77 The chaebol, thus, promoted the stabilization of the political regime as an incomplete democracy. However, organized civil society, trade unions and increasingly strong anti-chaebol sentiment in the population acted as a counterweight. In the 1990s, "as a result, the government's chaebol policy lacked coherence and principles. The government sometimes used its authority to retaliate against chaebol leaders who did not 'behave'. A typical example is that of Chung Ju Young, whose group, Hyundai, had to endure several years of government sanctions after he ran for the presidency in 1992." 78 Over the longer run, repeated scandals concerning chaebols, related civil society activism and court sentences handed down for prominent chaebol representatives reduced their political impact while the country seemed to move towards full democracy. 79

Example 3: Ukraine
A classic example of business magnates switching political camps regularly is Ukraine. Business magnates managed to create a majority in parliament -although a fragile one -for President Leonid Kuchma after the 1998 and 2002 elec-tions. 80 With their informal influence over a significant share of parliamentary deputies and their control of important mass media outlets, they clearly contributed to the creation of the uneven playing field in a competitive authoritarian regime. However, when the competitive authoritarian regime was threatened by public protests in 2004 and 2014, many business magnates started to hedge their bets and gradually switched sides.
A telling indicator of political flexibility is the political affiliation of the business magnates who gained seats in the national parliament. When power changed after the presidential elections of 2004 and the related "Orange Revolution", the number of business magnates in factions of the old power holders decreased from 13 to 4, while the number increased from 2 to 8 for the new ruling parties. The presidential elections of 2010 reversed the balance of power, and within two years, 11 business magnates had joined the parliamentary faction of the new ruler, President Viktor Yanukovich. After he fled the country in the wake of the EuroMaidan protests in spring 2014, that number was reduced to two in less than a year. 81 Thus, business magnates in Ukraine have never initiated a change in government or in the political regime, but they have acted as catalysts for an ongoing change by providing additional support to the winning side. However, as their manipulations of political decision-making processes continue, they impede a move to full democracy. 82

Example 4: Russia
In Russia, too, business magnates were seen as informal power brokers in the 1990s, though they were much less likely to assume political office themselves. Their financial support and control over media reporting are generally treated as decisive contributions to Boris Yeltsin's victory over the communist candidate in the 1996 presidential elections, despite approval ratings that were usually significantly below 25 percent in the two years prior to the election. In return, the business interests of the campaign sponsors received preferential treatment from the state, most prominently in the privatization auctions around the elections. 83 Towards the end of Yeltsin's second term, business magnates and members of the political elites, e.g. regional governors, began to position themselves for a potential change in power as Yeltsin's chosen successor seemed unlikely to win. However, when that chosen successor, Prime Minister Vladimir Putin, won the presidential election in 2000 in a landslide victory in the first round, he immediately started to implement what he called the "vertical of power", i.e. the subordination of all politically relevant actors to the national state executive with him as president. Business magnates with control over important mass media outlets, such as Boris Berezovsky and Vladimir Gusinsky, as well as business magnates who challenged Putin politically, most prominently Mikhail Khodorkovsky, were neutralized with the help of manipulated judicial proceedings. They lost their core business assets and were forced into exile. 84 Braguinsky finds that "more than half of the postcommunist oligarchs who rose to prominence during the Yeltsin era did not survive in the ranks of the oligarchy until 2006." 85 Other potentially independent political actors, such as regional governors, political parties or mass media, were also integrated into the vertical of power. Loyal entrepreneurs were then heavily favoured by the state. The number of Russian billionaires increased from 25 in 2004, i.e. around the time when Russia moved from competitive to fuller authoritarianism, to 87 in 2008. This number fell drastically in the wake of the international financial crisis, which hit Russia hard; however, it soon stabilized at approximately 100. 86

Example 5: Turkey
Another example of strong state actors starting to neutralize politically influential business magnates perceived as contributing to political instability is Turkey in the 2010s. In Turkey, business magnates started to emerge through the creation of large family-based holding companies in the 1960s, most prominently those of the Koç, Sabancı, later Şahenk (with Doğuş holdings) and Doğan families. Their rise was only possible with informal state support. Similar to the case of Russia in the 1990s, business magnates offered financial support for campaigns in return for preferential treatment of their business interests by politicians. As business magnates had not been part of the old elite and the Turkish political system was marked by political competition, alliances between business magnates and political actors changed regularly. As a result, business magnates became increasingly important power brokers in the 1970s. In the 1980s, they and some new holding companies, most prominently those of the Uzan family, benefited from state promotion as "champions of economic modernisation" in an attempt by the Turkish government to follow the East Asian model. 87  As a result, the business magnates of large family-based holding companies were deeply entrenched in Turkish politics, offering -in competition with each other -financing for political campaigns and support from the mass media outlets under their control to political elites, democratic politicians and military rulers alike, in return for state support on a macro-economically relevant scale. Karadag concludes that "these political competition structures seriously overburdened the financial capacities of the state, and neither politicians nor businessmen had any incentive to fundamentally alter these rules." 88 A change came with the 2002 parliamentary elections when the AKP of Recep Erdoğan was able to form a government: "These elections present a clear break in the […] politico-economic development in Turkey as, for the first time, the socio-economic counterelite [to] the influential oligarchs has been elected to power without being prevented from consolidating its position in the political process." 89 The AKP government used judicial proceedings to limit the political influence of actors without democratic legitimation, including the army and business magnates, most prominently, the Uzan family. As a result, important mass media outlets were assigned to new owners who were close to the government.

||
With the crackdown on the Gezi Park protests in 2013 and even more after the failed coup in 2016, it became increasingly clear that Erdoğan was systematically side-lining all oppositional forces, including those subject to democratic legitimation. The political regime, thus, became increasingly authoritarian. To create an ever more uneven playing field, the state put increasing pressure on business magnates, especially those with relevant media holdings, 90 while promoting its own loyal group of business magnates. 91 The resulting rise of new loyal business magnates is visible in the Forbes list, as the number of Turkish billionaire families rose from six in 2002 to 38 in 2018.

Summary
If business magnates are flexible in their political alliances, they can change sides when they expect that the ruling political elites will lose power. This does not necessarily lead to regime change, as they can misjudge the situation (losing out or returning to the old alliance). However, this is rarely the case, with the only examples being Kenya (1990s-2010s), Madagascar (1990s-2000s) and Malawi (1990s-2000s). 92 In unclear situations, business magnates likely prefer to stay with the ruling elites, as described in scenario 1.
If the regime becomes more authoritarian, business magnates who are not flexible enough in their political alignments (or fast enough, e.g. when giving up their media assets) will end up in the opposition. As the political regime becomes more authoritarian, political pressure on these business magnates will increase until they are neutralized, as seen in the cases of Russia (2000s), Turkey (2010s), Venezuela (2000s) and Nicaragua (2018). 93 In the case of attempted democratization, business magnates are likely to change sides regularly as the best way to remain influential in politics. As a result, several political factions will receive support from business magnates, which means that they partly neutralize each other and thus offer a safeguard against a permanent takeover by one political force and a return to more authoritarian politics. At the same time, their reliance on political corruption and manipulation hampers a move towards full democracy, as demonstrated by the cases of Georgia (2000s), 94 Indonesia (1990s-2000s), South Korea (1980s-1990s) and Ukraine (2000s and2010s

Conclusion
The first conclusion concerning the role of business magnates in regime change in competitive authoritarianism is that, in most cases, they do not matter. Of the 65 cases of competitive authoritarianism identified in the literature, 41 are not marked by politically active business magnates (which does not exclude the activities of other business actors); in another eleven cases, business magnates do not play a distinct role in regime change.
In these eleven cases, i.e. in all cases included in scenario 1, business magnates have been integrated so tightly into the ruling political elites that they do not have a discernible impact of their own on political regime change. Of course, they influence specific policies to their advantage, so their political influence can be traced. However, in regard to political regime change, business magnates do not have their own agenda in scenario 1; they simply go along with any regime change, including military coups -as in several Latin American cases -and full democratization -as in the cases of Mexico and Taiwan.
In the 13 cases where business magnates have a distinct effect on political regime change, the outcomes are very diverse, ranging from full authoritarianism to stable competitive authoritarianism, unstable or incomplete democracies and full democratization. Therefore, it is vital to differentiate the roles of business magnates in political regime change and in the actual regime trajectory.
The logic of action seems very similar for business magnates in all cases where they engage in political activities without full integration into the ruling elites. The easiest way to stay close to political decision-makers is to switch to the winning side in any political power struggle; thus, scenario 3 (switching sides) describes nine of 13 cases. At the same time, business magnates who have built their businesses on close and partly shady relations with the state are interested in neither the potential repression associated with full authoritarianism nor the rule of law brought about by full democratization. Accordingly, they prefer a political regime that is pluralistic enough to allow room to manoeuvre and corrupt enough to give them an advantage in the fight for state support and rent-seeking opportunities. There is no case where business magnates in a competitive authoritarian regime have proactively promoted a move to either of the two alternative models.
As competitive authoritarian regimes revolve around elections with real but unfair competition, business magnates with vast financial resources and mass media holdings command vital means for engaging in political competition on the uneven playing field of such regimes. Accordingly, those who come to power with the help of business magnates are likely to continue to count on their support. As a result, the competitive authoritarian regime remains in placepotentially with a new configuration of ruling elites.
Pressure for political regime change arises only if the winning side is a popular protest movement. If business magnates manage to switch to the winning side, they face demands from the public and foreign states to promote democracy. Moreover, with increasing political pluralism, business magnates are likely to change sides regularly to stay influential in politics. As a result, several political factions receive support from business magnates, which means that they partly neutralize each other and offer a safeguard against a permanent takeover by one political force and a return to more authoritarian politics. At the same time, the reliance of business magnates on political corruption and manipulation hampers a move towards full democracy.
The central development in such a case is that elections become free and fair, while political decision-making remains prone to established forms of undemocratic manipulation, including procedural tricks related to parliamentary decision-making processes, pressure on courts and mass media, and discriminatory action against opposition forces. In such a regime, the political strategy of business magnates can continue unchanged, as in the cases of Georgia (2000s and 2010s), Indonesia (1990s-2000s), Moldova (2010s), South Korea (1980s-1990s) and Ukraine (2000s and2010s).
Only further democratization would change the political strategy of business magnates from wealth defence and corrupt manipulation to income defence and more formalized lobbying. However, their political engagement inhibits democratic consolidation, since they represent their own individual rather than collective (entrepreneurial) interests and, even more importantly, because they use undemocratic means to promote these interests. This undermines democratic decision-making processes and delegitimizes the existing democratic constitutional order in the eyes of both the public and the political and economic elites. The regime trajectory from competitive authoritarianism to incomplete democracy is, therefore, likely to end there if business magnates remain a strong political influence.
Accordingly, a move to full democracy is only possible if the political power of business magnates is tamed. Compared with the number of competitive authoritarian regimes moving towards incomplete democracy, further development to full democracy is exceptional in cases where business magnates are independent political actors. The only case identified here is South Korea since the 2000s, which shows that such a development requires strong opposition to business magnates from the population in combination with an independent coup or by gradual repression of the most politically active business magnates, the result can be a stronger authoritarian regime where business magnates cease to be independent actors.
Although the total number of relevant cases is too small to allow for quantification, the analysis shows that the regime change towards stronger authoritarianism and towards an incomplete democracy occur equally with five cases each, while there is only a single case of a further move to full democracy. 96 Table 2 provides an overview. Compared to the overall trend for competitive authoritarianism, which favours democratization as established by Brownlee,97 the existence of independent business magnates seems to increase stability and lessen the chances of democratization in competitive authoritarian regimes.