Corporate Political Ties and Firm Value: Comparative Analysis in the Korean Market

Abstract

Although prior studies have viewed corporate political ties as unethical, this study investigates whether changes in the political environment can significantly affect the net value of firms over time in emerging and transitioning economies. Analysis of a panel of South Korean listed firms in the Gyeongsang and Jeolla provinces from 1998 to 2013 reveals that political network ties to the regime currently in power provide significantly more positive benefits to politically connected firms than to non-connected firms. However, political network ties to the rival of the regime in power have significantly more negative effects on politically connected firms than on non-connected firms. More importantly, unexpected political regime changes strongly affect the shift from positive to negative effects on firm value over time, and political network ties continue to significantly and positively affect firm value over time during regime changes after political and economic liberalization in South Korea. These results line up with various robustness tests.

1. Introduction

Prior researchers have generated a rich set of empirical results demonstrating the positive benefits of political connections for firm value in emerging and transitioning economies. Many studies have found that political connections can increase firm value through official government support such as favorable regulatory conditions, the cut-rate credit, import licenses, financial bailouts, and preferential access to debt financing. All of these factors help firms secure positional advantages over their competitors and helping to build strong market power even if overall productivity is lower [1,2,3,4]. Among these studies showing the benefits of political connections, some have focused on external factors, considering higher corruption, weaker institutional constraints, and ineffective legal protection as the sources of these benefits, whereas others have focused on the characteristics of firms such as poor governance structures, low research and development (R&D) intensities, and firm size. However, studies have also found growing evidence of a negative effect of political connections on firm value. As Shleifer and Vishny [5,6] note, it cannot be assumed that a politician is always benevolent. Politicians’ machinations to win elections or achieve other personal objectives may be harmful to firms in terms of weaker managerial practices, distortions of incentives, misallocations of investments, more rent-seeking behavior, and, finally, underperformance on an ex-ante basis. Faccio [3] explains that, although these connections provide considerable benefits, because of the overall inefficient use of resources, firms often underperform in the long term. Other studies also support these arguments [7,8,9,10]. Given the conflicting empirical results regarding the relationship between political connections and firm value, it is not surprising that many cross-disciplinary studies have been conducted in the social sciences. Theory and research have shown that the value of political connections depends on the environmental context in emerging and transitioning economies. In their seminal work, Shleifer and Vishny [5,11] suggest that the benefits of political capital enhance firm value only when their marginal benefits outweigh their marginal costs, arguing that connections are a type of game between the public, the political actors, and the economic actors, which is always controlled by incomplete contracts. In other words, although a rational economic actor can obtain benefits from political capital by appointing politically connected directors, given their knowledge and experience with government procedures [12], their tacit and timely information regarding industrial policy changes [13] or even their efforts to indirectly lead influential institutions to change economic policies to his firm’s advantage [14], he considers all the corresponding economic costs (e.g., lobbying, bribery, contribution to political campaigns, and other forms of direct and indirect payments to politicians) [3] and then decides whether he should form political ties [15]. Similarly, Yeh, Shu, and Chiu [16] find empirical evidence that Taiwanese firms with poor governance structures are more likely to engage political connections to increase firm value due to high marginal benefits and low marginal costs. However, there are several reasons why it may not be as easy to detect or explain the relationship between political connections and firm value for emerging and transitioning economies.

First, Hillman, Zardkoohi, and Bierman [17] distinguish “relational approaches” from “general transactional approaches,” arguing that many firms pursue political network ties over a long time horizon rather than issue by issue. Peng and Luo [18] also emphasize that the overall objective of political connections is to produce governmental policy outcomes that benefit the firm’s continued survival and positive performance. The primary goals of political actors are re-election [19], political survival [20], and political persistence [21] from a long-term perspective. A series of anecdotes proposed by Leuz and Oberholzer-Gee [22] describe Indonesian firms who were treated preferentially in financial allocations because of their close ties with regimes. However, when the patron of these firms unexpectedly fell from power, they faced sudden difficulties, and, thus, underperformed under the new regime. Goldman, Rocholl, and So [23,24] show empirical evidence that political connections of Standard and Poor’s (S&P) 500 firms in the United States were responsive to the 2000 presidential election and find that firms that were well connected to the Democratic Party surprisingly experienced decreases in firm value. Thus, each of these studies stresses that the mutual connections [3], gift exchanges [25], and embeddedness [26] of political and economic actors are not single episodes but, rather, have more relevance to changes in the structure and processes of the relations over time [17].

More importantly, studies point out that changes in political and economic circumstances are not always contemporaneous events [27,28,29] and that the extent of regime change and the processes of economic privatization or liberalization differ across emerging and transitioning economies [1,4,30]. Prominently, Walder [29] focuses on categorizing four types of transitional economies according to their political and economic environments. Some countries (e.g., the Czech Republic, Poland, Russia, and Ukraine) have experienced extensive regime change including the collapse of hierarchies whereas others (e.g., China, Vietnam, Kazakhstan, and Uzbekistan) have had one party remaining in power. Some countries (e.g., China, Hungary, Poland, Tajikistan, and Vietnam) have well-regulated monitoring systems to control asset appropriation by powerful elites. Others (e.g., Russia, Ukraine, and Uzbekistan) have poorly regulated privatization programs. Furthermore, Walder [29] believes that the great cross-national heterogeneity in the pace of political and economic reforms can create both opportunities and regulatory constraints, which, in turn, lead to strikingly different reactions and behaviors of political and economic actors during transitional periods. Other authors agree that researchers must examine the target countries not with “a standard cookbook approach” but rather by considering their distinctive historical paths during political-economic transitions [31,32,33,34].

Overall, many scholars have theoretically suggested that, because countries vary widely in terms of the extent, order, speed, and phase of political and economic reform over time, using “a contingency perspective” as a tool for the analysis of political capital is suitable [15,18,26,35]. Burt [15] studies 170 managers at one of the largest electronics firms in the United States and the senior investment banking officers at a large American financial organization and demonstrates that the value of socio-political capital to these groups is contingent on their network, background, and performance. Peng and Luo [18] show that, in China, the effect of a manager’s social ties, contacts, and networks varies across firms with different ownership types, business sectors, sizes, and industry growth rates. In a recent study, Zhu and Chung [13] show that the positive effects of social connections between politicians and firms on market entry are contingent on firms’ external competitive environments, the organization of the government, and the internal resource profile of each firm as measured by data from the 250 largest Taiwanese business groups from 1998 to 2004. As Peng and Luo [18] mention, understanding the notion of the contingent value of political capital provides a frame for a specific prediction to capture the interplay between political and economic actors, and, thus, to eventually evaluate the impact of political connections on firm value in emerging and transitioning economies.

To date, although much has been written, no definitive empirical study exists on the aspects of the dynamic between political connections and firm value in emerging and transitioning economies. The most closely related studies are those by Do et al. [36] and Siegel [26]. Do et al. [36] use a regression discontinuity design for 35 gubernatorial elections from 1999 to 2010 in the United States to demonstrate that social network-based political connections provide positive and significant cumulative returns to firm value. They use a novel method to identify political connections based on networks of university classmates among corporate directors and politicians and expertly find that rival business groups have opposite effects on overall firm value over time. Again, this study does not consider the dynamics of firm value and does not examine the possibility that political connections can also lead to negative consequences as a punishment following regime change. Using data on 665 South Korean firms from 1987 to 2003, Siegel [26] performs the only comparative study in which sociopolitical network ties to the regime in power are positively correlated with the rate of cross-border strategic alliances while ties to the opposing regime are negatively correlated with that rate. From a contingency perspective, his novel study examines how an unexpected event regarding the political regime can quickly change a political liability (asset) into an asset (liability) in the long term. However, he only focuses on cross-border strategic alliances as a tool for assessing political connections rather than on firm value.

In addition, corporate political ties (activities) and their strategies are more sensitive to ethical issues than their purely market-driven issues [37]. First, from the perspective of neoclassical economics, organizations in the public sector are a counterforce to firm actors and both parties are self-interested to the same extent. This counterforce is undermined if firms manage to obtain a substantial influence in the political system. Hence, the balance of the whole economic system will be in jeopardy in the long term. The studies on regulatory capture [38,39], in its critique on neoclassical views on regulators, has argued the danger of regulation coming under the influence of various corporate interests. Second, corporate political ties (activities) have ethical implications, which stem from the argument that the discourse of strategy itself is highly ideological and promoting in latent form, has ethically problematic inclinations toward managerial hegemony, aggressive competition, psychological egoism, and the viewing of economic capital as the highest form of capital [37]. All of these are presented in the guise of rationality and foresight in decision-making [40,41]. It is one thing to promote such ideology in the economic marketplace, but when done so in the public arena, it becomes paramount to consider societal implications [37].

However, the ethical aspects of corporate political ties in the Korean market have so far received little attention. In contrast with the extravagant claims about the economic benefits of political ties for firms, we have not fully understood their dark side. The question is, are corporate political ties ethically acceptable? This depends on the nature of the corporate political relationship and the context in which it occurs. While having family and helper ties are generally perceived in Korea as a good practice that is completely moral or even desirable, political ties in a corporate setting are widely regarded as a bad practice or a grey area that is inevitably associated with favoritism, nepotism, unfair competition, and fraud. In the majority of corporate scandals related to political ties, swindlers deliberately chose people whom they had close business relationships with relatives, friends, or classmates as unsuspecting easy targets. They betrayed trust for profit. Ironically, when the victims discovered the fraud, most of them tried to recruit newer victims in the same fashion to recoup their losses. Such scandals have exposed the pernicious side of corporate political ties.

There is no doubt that there are advantages of corporate political ties in the Korean market. These advantages may bring cost savings, vital resources, and special treatment to an organization that employs individuals who have political connections. However, firms often achieve these benefits at the expense of other individuals or firms since corporate political ties create significant disadvantages for parties outside the network, which stifles competition to the detriment of society. In this way, such ties sacrifice the greater social good for personal gain. A simple acid test to judge the ethics of corporate political ties is to identify whether there are any victims because of such a deal. These victims may be known, such as competitors of the firm or customers. They may even be unknown in some cases. However, the truth remains that such corporate political ties produce gains for individuals or firms but cause wider social loss. They may be considered benign from the ethics point of view only if there is no third party, either known or unknown, that is adversely affected because of this action. Unfortunately, the vast majority of corporate political ties would fail this test.

Given the ethical aspects of corporate social ties, to investigate the changes in the value of political connections over time after a regime change from a contingency perspective, this study examines the effect of political capital between businesses and the government on firm value in South Korea from 1998 to 2013. South Korea is ideal for exploring dynamic changes in firm value for several reasons. First, South Korea is representative of many emerging economies and experienced a rapid transformation of its political and economic environment toward liberalization from 1960 to 1990. This remarkable turnaround has inevitably generated a unique relationship-based society in which in-group favoritism reinforces the formation of reciprocal social networks and out-group exclusionism wields strong influence over social cleavage. This low interpersonal trust-based society provides an optimal environment for observing dramatic changes over time. Second, the study particularly analyzes panel data on firms listed on the Korea Stock Exchange that are located in Southeast (Gyeongsang province) and Southwest Korea (Jeolla province). These regions have had a longstanding rivalry since 1961, and intense regional competition (even within subregions) over political and economic resources has prevailed. This dataset is appropriate for identifying the conditions under which political capital has a positive or negative effect and the change in this effect over time within the political and economic environmental context. Third, and more importantly, the conflict between these two regions has strongly influenced the attitudes of political and economic actors in these regions. These actors have evolved from only maintaining the status quo to constructing reinforced regimes or establishing solid government–business relationships as well as from passive acceptance of government policies to actively coping with discriminatory treatment and punishment and greater participation in forming political connections. Thus, both sides aggressively struggle to maintain or change the incumbent regime. This series of behavioral changes can help investigate the impact of political connections on firm value in the context of regime changes over time.

The rest of the paper is organized as follows. Section 2 briefly discusses the related literature. Section 3 describes South Korea’s historical background with a particular focus on regionalism. Section 4 develops testable hypotheses. Section 5 describes the research design including the data, variables, and empirical methods. Section 6 presents the empirical findings and reports robustness tests and Section 7 concludes.

2. Literature Review

Previous empirical studies have found that the nexus between firms and a political connection is mostly based on corporate governance. Therefore, a relational governance perspective highlights the importance of network activities in improving overall firm performance. However, empirical results on the environmental context are somewhat mixed. Some researchers have found a positive association between political connection and firm value. Agrawal and Knoeber [12] use a sample of Forbes 800 firms between 1987 and 1990 in the United States to show that politically experienced external directors on corporate boards play a significant role. They can deter the agency problem, provide crucial information regarding the steps and processes of government procedures, provide consultation for formulating political corporate strategies, and, eventually, increase overall firm performance. Hu and Leung [42], using data on China’s listed state-owned enterprises (SOEs) between 2001 and 2005, also find that poorly performing firms have a strong tendency to replace their chairmen or CEOs with politically connected top executives from the government, who serve in disciplinary or monitoring functions to improve overall firm performance. However, other authors document that the political connections can have a negative effect. Fan, Wong and Zhang [9] report that, in China, politically connected CEOs and directors have a negative effect on the three-year post-initial public offering (IPO) stock returns of 617 A-share listed companies on the Shanghai Stock Exchange and the Shenzhen Stock Exchange between 1993 and 2000. Chang and Wong [43] also empirically find that political connectedness often has a negative effect on a firm’s decision-making in China. Thus, it is more likely to have a harmful effect on overall firm performance. Similarly, Liu, Luo, and Xu [44] confirm that, for a sample of A-share listed firms in China from 1994 to 2010, politically connected firms tend to overinvest in the establishment of political connections but then reduce their investments to the level of their peers, and that overinvestment often has a negative effect on firm performance. As the relationships between states and firms are mostly heterogeneous both across and within countries in terms of the structure and quality of corporate governance, the degree of government intervention, and the extent of economic reform [45], these combined factors can strongly increase or decrease overall firm value. For example, using 71,069 individual bank loan contracts of listed firms in Taiwan from 1997 to 2009, Shen, Wang, and Lin [46] empirically show that firms with weak corporate governance have a greater tendency to seek political connections while firms with strong corporate governance perform fewer activities related to political connections. Therefore, corporate governance and political connections may be substitutes.

Another strand of the literature focuses on corruption. Essentially, Khwaja and Mian [2] consider political rents in banking in Pakistan and empirically find that political firms receive positive benefits related to the access of government bank loans through relationships between politicians and their directors. They analyze loan-level information for more than 90,000 firms from 1996 to 2002. In addition, Wong [47] shows that government-business collusion through shareholders or directors concurrently holding seats on the Election Committee in Hong Kong yields positive benefits to firm value, as measured by the returns on equity and market-to-book ratios of firms listed on the Hong Kong Stock Exchange from 2003 to 2008. Studies even investigate the relationship between transparency and firm value. In the context of post-IPO firms in Singapore from 1998 to 2006, Ang et al. [48] find that ties between political (former or current ministers and senior civil servants) and economic actors (chairmen, CEOs, and directors) have little effect on firm value in a low-corruption environment. However, they significantly benefit firm value in specific industry areas that are strictly controlled and regulated by the government. Similarly, Niessen and Ruenzi [49] investigate the introduction of new transparency laws in Germany and find that firms with political connections to the Bundestag, which is elected by German citizens, tended to significantly outperform unconnected firms on the stock market between 2006 and 2007. Hellman, Jones, and Kaufmann [7] introduce the theoretical concept of a “captured economy,” which is related to corruption in a transition economy through which “public officials and politicians privately sell underprovided public goods and a range of rent-generating advantages to individual firms.” They also contrast state capture (i.e., “firms shaping and affecting formulation of the rules of the game through private payments to public officials and politicians”) with influence (i.e., the same activity but without recourse to payments) and administrative corruption (“petty forms of bribery in connection with the implementation of laws, rules, and regulations”), and empirically prove that political ties to public officials and politicians can improve firm performance except in the case of “administrative corruption.”

Third, several authors explicitly show that, in the case of China, an emerging and transitioning economy, the connections between or behaviors of political and economic actors can be changed to adapt to changes in the political and economic environments. Haveman et al. [34] find that, in China, links between state bureaucrats and CEOs or all top firm executives are still beneficial to overall firm performance through preferential access to bank loans by analyzing data on all listed firms from 1992 to 2007. Qin [50] also shows that firms’ direct or indirect connections to the State Council and Central Committee of the Communist Party of China had increasingly positive effects on firm performance from 1998 to 2007, using panel data for 86,827 manufacturing firms and 350 politicians. However, Nee and Opper [51] find that the effect of political connections was positively correlated with firms’ competitive advantages before 2002 but that those benefits no longer exist for economic transactions after 2002 due to competitive market conditions. This result is based on a large-scale national survey conducted by the World Bank in 2003, covering 2400 firms in 18 municipalities in China. Sheng, Zhou, and Li [52] specifically distinguish business capital from political capital and show that business network capital has more positive effects on firm performance than political network capital, which relies on institutional and market environments in China. To summarize, considering the shift from state socialism to market-based capitalism in China, political and economic actors must also evolve their thoughts, behavior, and strategies to gain more benefits as the market develops. As market competition gradually replaces the state’s role of allocating and redistributing resources, the power of political actors progressively diminishes. However, these actors can effectively develop new sources of power such as the regulation of market entry and exit, control licenses, subsidies, taxes and capitals, and the shaping of government policy [34]. Consequently, political capital remains valuable with the development of new aspects of contingencies for both political and economic actors as the market develops.

Fourth, several studies explicitly investigate crony capitalism. Johnson and Mitton [53] define crony capitalism as the “capital controls that enable politicians to support the financing of particular firms.” Faccio [3] also explains that, “dominant political leaders use their power to the advantage of their families and friends, who benefit from government–created rents.” A primary aspect of crony capitalism is a reciprocal “gift exchange” between governments and politically connected firms [25]. Other scholars depict the “interdependence” between political and economic actors [54]. On the one hand, political actors can certainly provide special support to specific firms by reducing regulatory requirements, awarding government contracts, or providing financial bailouts since it is assumed that an economic actor’s ultimate goal is to improve firm performance. On the other hand, economic actors can consistently support specific political actors via campaign contributions, donations, or other activities, as it is assumed that a political actor’s ultimate goal is to win an election. However, empirical evidence presents conflicting results across countries. In analyzing the five election cycles between 1983 and 1992 in the United States, Kroszner and Stratmann [19] confirm that the link between interest groups and Congress is a type of contract under which firms contribute to achieving the goals of legislators so that the latter can build a system of political action committees that promote repeated interactions, reputation building, and long-term relationships between legislators and firms. Sapienza [55] also finds that, in Italy, politically connected firms are more likely to receive lower interest rates from state-owned banks than from other privately owned banks. This result is significantly responsive to the electoral results of the party. Ferguson and Voth [56] specifically show that the effect of political connections between the German industry and the Nazi party in Nazi Germany had a positive association with stock returns for a sample of 789 firms listed on the Berlin Stock Exchange from April 1932 to May 1933. Based on the top 2000 companies in Taiwan, Bunkanwanicha and Wiwattanakantang [30] conclude that business tycoons often pursue public office and that, when they are elected, their firms outperform others in terms of firm value. Imai and Shelton [57] focus on the 2008 presidential election in Taiwan, analyzing Taiwanese firms listed on the Taipei Political Exchange from 18 December 2007 to 21 March 2008. They find that the stock returns of Taiwanese firms closely related to the mainland have a significantly positive association with the electoral victory for KMT in the context of the unresolved issues between Taiwan and the People’s Republic of China. In recent research, Bandeira-de-Mello et al. [58] find that political network ties between businesses and politicians in Brazil can directly benefit firm performance through mutual contributions between the two parties by considering a sample of the major elections for president, governors, senators, federal and state deputies, and multilevel listed firms from 2003 to 2006. On the contrary, some studies find different results. Bertrand et al. [8] showed that, in France, politically connected CEOs have a net negative effect on the performance of publicly traded firms based on firm-level panel data set from 1987 to 2002. Hung, Wong, and Zhang [59] also demonstrate that firms’ strong political ties have a negative and statistically significant association with firm performance relative to non-politically connected firms using a data set of Chinese SOEs listed on overseas and domestic stock exchanges in 1992. Extraordinarily, using anecdotal evidence in 2001, Jayachandran [60] demonstrates that politically connected firms experience significantly positive or negative changes in market value following an unexpected political changeover between the Republican and Democratic parties in the United States. This finding is called the “Jeffords effect.”

Lastly, some scholars examine the overall relationship between political capital and firm performance, which they predict based on worldwide research using meta-analysis and systematic review. Boubakri, Cosset, and Saffar [10] confirm that political connectedness and firm value are positively correlated for a unique data set of 234 politically connected firms from 12 developed and 11 developing countries based on an international event study over three years. Liedong [61] uses a systematic review of 56 studies to show that corporate political activities can strongly benefit firm performance, depending on the institutional context across countries. In addition, Stam, Arzlanian, and Elfring [62] provide evidence that the relationship between entrepreneurs’ personal networks and the firm performance of small firms is positive and significant based on a meta-analysis of 59 studies across countries. However, as Faccio [3] shows, for a sample of 42 countries, including 19,884 firms and 17,033 politicians, connected corporations do enjoy significant benefits, such as debt financing, lower taxation, and stronger market power, but they also face the possibility of poor performance due to rent-seeking activities by politicians.

Overall, the net value of political capital is heterogeneous depending on a combination of factors from the political and economic environments that affect both political and economic actors. The results of most studies suggest that political capital is highly contingent on the extent of corruption, the intensity of institutional constraints, legal protections, judicial independence, government intervention, structure of corporate governance, firm-specific characteristics (R&D intensity, size, and leverage), and more.

3. Historical Background

Western scholars describe East Asian capitalism as “network capitalism,” “relational capitalism,” “Confucian capitalism,” and “crony capitalism,” and the notion of “Asian values” reflects a unique set of institutions and political environment in terms of successful economic development in emerging and transitioning economies [63,64,65,66,67]. Apparently, all the above concepts highlight the importance of social network ties for corporate strategic activities, government policy decision processes, and government–business relationships in East Asian countries, where generalized interpersonal trust is relatively low [66]. Fukuyama [68] categorizes South Korea, China, Taiwan, and Hong Kong as familistic, low-trust societies in which social trust is often restricted to kinship and gradually diminishes outside of networks. Therefore, these societies tend to develop a “two-tier system of ethical values” in which in-group favoritism is more prevalent than out-group favoritism.

Kim [66] mentions that Korean social networks are especially characterized by generalized exchanges in which multiple actors take active parts in network formation and control exchanges in which network members promote exchanges among the “information, institutional resources, rights, or power they own or control.” The Korean social network yonjul, similar to guanxi, is the most significant factor determining the allocation of all kinds of resources in Korea. It is mostly comprised of kin ties, school relations, and regional ties. For example, in Korea, family members connected to conglomerates called chaebols and small and medium sized companies inherit stocks, properties, or even firms within these conglomerates. Siegel [26] empirically shows that the quality of high school and university network ties between firms and the government has a significantly positive correlation with firms’ international alliances. Regionalism also normally prevails in presidential and parliamentary elections since the conservative party has been firmly rooted in the southeast area of Korea since 1960.

Scholars demonstrate that the characteristics of social networks in South Korea generally rely on regionalism. Kim [66] defines regionalism as “a set of ideologies (or prejudices) and institutions that encourages persons to favor others within their region above those outside.” Most of the social capital in South Korea is exclusively distributed through regional networks. These networks affect personal networks, the promotion of economic and political elites, inter-firm ties, and even recruitment in the labor market.

Historically, Korea has two distinct rival regions: the Gyeongsang province (southeastern area) and the Jeolla province (southwestern area). Kim [66] suggests two theories to explain the origin of Korea’s regionalism: Theory H and Theory P. Theory H focuses on the aspects of “historically rooted prejudice” from the Koryo dynasties (A.D. 918–A.D. 1392). Regionalism in South Korea can be traced as far back as the Three Kingdoms period (18 B.C.–A.D. 668) when the Korean peninsula was split into the rival kingdoms of Goguyeo (north area), Baekje (southwestern area), and Silla (southeastern area). After a long war, Silla (i.e., the Gyeongsang province) conquered the other two rival kingdoms, and they eventually unified in A.D. 676 [69]. The Koryo dynasty began, and its founder, Wang Geon, openly discriminated against anyone from the southwestern area (i.e., the Jeolla province). As Kim [66] notes, Wang Geon’s Ten Commandments (A.D. 943) is the earliest historical document that explains Korean regionalism based on Theory H. Specifically, Wang Geon warns, “Do not promote southwestern area (Jeolla region) people to higher government positions, for their minds resemble the rugged mountains surrounding them.” The impacts of these social prejudices on regionalism continued through the Koryo and Choson dynasties (A.D. 1392–A.D. 1910) and eventually became deeply rooted in minds, thoughts, and behavior, as an implicit social value influencing job opportunities, promotions, marriage preferences, social networks, and other aspects of everyday life [70].

In contrast, Theory P focuses on the value introjection process for anti-Honam prejudice and a political objective of Park Jung Hee’s regime (1963–1979). As Kim [66] describes, when institutional predictability and accountability are uncertain, people tend to rely on specific social network ties to eliminate potential risk factors, maintain a regime, and achieve the prolonged one-man rule. Because Park Jung Hee held power though a coup d’etat, he faced a lack of popular legitimacy and justification. Therefore, he strategically allocated political and economic resources through personal network ties in his home base in the Gyeongsang province in exchange for the loyalty to pursue stabilization, handle possible counter-coups, and secure his position in the long term [66,69,70]. Politically, for example, 160 government-appointed ministers were from Gyeongsang under the military regime (1963–1987), whereas just 50 were from Jeolla [66]. Similarly, other political elites, such as generals and politicians, were also largely from the Gyeongsang province. Economically, Park Jung Hee concentrated on the Seoul (central)-Busan (Gyeongsang) corridor to allocate resources for economic development. Thus, those regions became heavily industrialized while other regions, especially Jeolla, were relatively neglected [69]. Kim [66] emphasizes that the authoritarian regime created regional economic and political inequalities and conflicts, particularly between the Gyeongsang and Jeolla provinces, as a “divide-and-conquer strategy” to maintain despotic power. Then, Park Jung Hee’s successors, Chun Doo Hwan and Roh Tae Woo, who also hailed from the Gyeongsang region, sustained these strategies or policies of providing Gyeongsang with preferential treatment while Jeolla lagged. Residents of Jeolla felt this relative deprivation, and regional conflicts have continued to date. In summary, Theory H and P both state that Korean regionalism across the Gyeongsang and Jeolla provinces played a decisive role during the rapid advance of industrialization and democratization, including in the formation of social networks, information and resource channels, shaping of government policies, and bureaucrat-business networks.

Regionalism has continuously been a dominant force behind the social cleavage in government policies toward the Gyeongsang and Jeolla provinces and through significant economic changes during the industrialization and marketization period since 1960. At the start of industrialization, Park Jung Hee aggressively focused on export-led and outward-oriented policies based on the free trade zones in Busan and Masan in the Gyeongsang region. Under central government-planned economic development policies, all resources were strategically distributed to the light, heavy, and chemical industry sectors by encouraging the development of large conglomerates or chaebols and forming close interrelations or collusions with them [70,71]. Many of these corporations received special favors from the government in exchange for donations, gifts, bribes, and other payments, and enjoyed permits for regulatory market entry, preferential bank loans, and a fixed-wage system. Through cronyism, this government-businesses cooperation led to unprecedented economic growth with GDP growing by an average of 10% annually between 1962 and 1994 [71].

However, Park [70] notes that government-guided rapid industrialization also involved a “growth first and distribution later strategy” that resulted in a bipolarized developmental environment. This environment brought about regional disparities that generated specific groups and social stratum, and particular industry sectors, corporations, and regions received benefits while others did not. Thus, these unequal distributions of resources found in central government policies and regional economic development created discrimination between Seoul, Gyeongsang province, and other areas especially Jeolla province [70,71]. In 1960, all resources were largely located in the central metropolitan areas. By the mid-1990s, roughly half the labor market resided in metropolitan regions, working in manufacturing occupations, expert skill occupations, administrative management, and office occupations, among others. In particular, Gyeongsang holds nearly 30% of administrative management and office occupations and a relatively high share of manual and skilled labor while Jeolla has only a 9% share of these occupations and is mostly engaged in agriculture and relatively small enterprises [70]. Thus, the continued economic backwardness of the Jeolla region escalated regional animosity toward government policies and the Gyeongsang province, which monopolized economic resources. In response to these discriminatory policies, Jeolla residents initiated a movement for democracy called the Gwangju Democratization Movement, which resulted in the death of nearly two thousand civilians because of the government’s military repression operations in 1980 [72]. Since every president was from the Gyeongsang region, the economic development gap between the Gyeongsang and Jeolla provinces widened remarkably from 1960 to 1997. Later, when Kim Dae Jung was elected president in 1997, Jeolla challenged Gyeongsang for economic resources [66].

Similarly, political capital also had a significant effect on firm survival during the industrialization period. Because the government built the overall business environment, allocated economic resources, and regulated market transactions, well-connected firms could improve performance with the full support of the government while poorly connected firms barely survived [73]. More recently, the most well-known examples are the bankruptcy of the Daewoo conglomerate, founded in 1967, and the survival of Hyundai, founded in 1946, during the Asian financial crisis in 1997. Hyundai supported the government’s unification policy toward North Korea by investing a substantial amount of money, establishing close connections, and providing subsidies of a few billion. However, when Daewoo’s chairman met with government bureaucrats and Kim Dae Jung, he demanded changes to government policies to benefit his firm, exacerbating the conflicts between them, and thus, the government took no action when Daewoo had trouble with financing [66].

4. Hypotheses Development

From 1961 to 1992, three generals (Park Jung Hee, Chun Doo Hwan, and Roh Tae Woo), all from the southeastern (Gyeongsang) region, held power in South Korea. In 1987, a massive pro-democracy movement broke out and Chun Doo Hwan agreed to the citizens’ demand for democratization by accepting a direct presidential election system in a breakthrough for South Korean democratization [69]. Then, from 1993 to 1997, Kim Yong Sam, from the southeastern region, was the first civilian president since the military dictatorship began in 1961. In 1997, during the Asian financial crisis, Korean voters elected former dissident Kim Dae Jung from the Jeolla region, which was the first peaceful transfer of presidential power to the opposition [26,69]. In 2002, the pro-democratic candidate Roh Moo Hyun, who was born in the Gyeongsang region, was elected president, which was another milestone in the history of democratic transitions in Korea. Because the driving forces of regional rivalry, the “Three Kims,” all retired and because Roh Moo Hyun’s home base was the Jeolla region, which favored the democratic party, regionalism faded somewhat [69]. Then, in December 2007, Lee Myong Bak from the Gyeongsang region won the president election, which resulted in another political regime change between the ruling and opposition parties.

Social conflict between the Gyeongsang and Jeolla provinces has existed for a long time in Korean history. Given how political and economic resources are allocated, regional competition has become stronger to secure better positions. As Kim [69] notes, regionalism is a major factor in the social cleavage that is “embedded in local and regional ties rooted in Korean society.” The presidential election witnessed in-group solidarity but out-group animosity across regions [66]. When a regime changes, the value of political network ties also tends to change and no firm has the assurance that a future government will have a positive or negative effect on firm performance or value.

In December 1997, pro-democratic candidate Kim Dae Jung from the opposition Millennium Democratic Party (MDP) won the presidential election. The most crucial factors for Kim Dae Jung’s victory were the formation of a regional coalition with Kim Jong Pil, whose home base was the Chungcheong region, and the internal division of the ruling Grand National Party (GNP) [69]. The prominent opposition candidate was Lee Hoi Chang, whose home base was the Gyeongsang region. He performed well in Gyeongsang, receiving nearly 55% of the votes (55.1% in Gyeongsang Namdo (North) and 61.9% in Gyeongsang Bukdo (South)) while he received roughly 3% of the votes in rival Jeolla (3.2% in Jeolla Namdo and 4.5% in Jeolla Bukdo). By contrast, Kim Dae Jung received an overwhelming 92% of the votes from Jeolla, his base (94.6% in Jeolla Namdo and 92.3% in Jeolla Bukdo), whereas he received roughly 11% of the votes from the rival Gyeongsang province (11% in Gyeongsang Namdo and 13.7% in Gyeongsang Bukdo). Because the third candidate, Rhee In Je, whose home base was Chungcheong, took a share of the votes for Gyeongsang Namdo from Lee Hoi Chang. The latter failed to win the election.

These stark differences between the Gyeongsang and Jeolla regions’ votes reflect their hostile relationships. From 1961 to 1997, Jeolla residents suffered discriminatory treatment from three former generals (Park Jung Hee, Chun Doo Hwan, and Roh Tae Woo) and one civilian leader (Kim Yong Sam), who all came from the Gyeongsang province. Gyeongsang monopolized political and economic resources for nearly 35 years, excluding the Jeolla region. Consequently, in the Jeolla region, over 90% of votes were for Kim Dae Jung while only 3% were for Lee Hoi Chang. In the same context, in Gyeongsang, only 11% of votes were for Kim Dae Jung, whereas 55% of votes were for Lee Hoi Chang. Later, Kim Dae Jung appointed individuals from Jeolla to key government posts, such as the public enterprise president and even to the Ministry of Foreign Affairs, even though he had promised to resolve the regional cleavage during his presidency [69]. The proportion of government posts appointed to individuals from Jeolla increased remarkably from 14% to 27% while the proportion of appointments from Gyeongsang decreased from 45% to 25% [69]. These regime changes lead us to propose the following hypotheses.

Hypothesis 1a (H1a).

In Gyeongsang, political connections are negatively correlated with firm value during the Kim Dae Jung administration.

Hypothesis 1b(H1b).

In Jeolla, political connections are positively correlated with firm value during the Kim Dae Jung administration.

In December 2002, South Korea chose a liberal candidate from the ruling MDP, Roh Moo Hyun, whose home base was Jeolla, as the next president. Roh Moo Hyun’s victory was attributable to two factors: his electoral alliance with Chung Moon Joon—a family member of the Hyundai conglomerate—to form a unified candidacy, and the younger generation’s preference for MDP (Kim, 2009). The opposing GNP candidate was Lee Hoi Chang. He performed well in Gyeongsang, receiving over 67.7% of the votes (67.5% in Gyeongsang Namdo and 73.4% in Gyeongsang Bukdo). However, he received only 4% of the votes in the rival Jeolla province (4.6% in Jeolla Namdo and 6.1% in Jeolla Bukdo). On the contrary, Roh Moo Hyun received an overwhelming 91% of the votes in Jeolla (93.3% in Jeolla Namdo and 91.5% in Jeolla Bukdo), and approximately 21% of the votes in the rival Gyeongsang province (27.0% in Gyeongsang Namdo and 21.6% in Gyeongsang Bukdo).

The results of the 2002 presidential election, like those of the 1997 presidential election, clearly show that regionalism or regional cleavage remained as important a determinant as ever. Even after five years, over 90% of the votes in Jeolla were for Roh Moo Hyun while only 4% were for Lee Hoi Chang. Similarly, in Gyeongsang, 21% of the votes were for Roh Moo Hyun while an overwhelming 67.7% were for Lee Hoi Chang. When elected, Roh Moo Hyun stated, “It is truly disappointing that we were unable to tear down the wall of regionalism this election” [26]. Nonetheless, although Roh Moo Hyun received fewer votes than Lee Hoi Chang in the Gyeongsang region, his 21% share of votes (27.0% in Gyeongsang Namdo) was a somewhat outstanding achievement compared to his 11% share in the 1997 presidential election. There were several reasons for these subtle changes in regional rivalry during the 2002 presidential election. On the one hand, because the major drivers of regional rivalry, the “Three Kims” (Kim Yong Sam, Kim Dae Jung, and Kim Jong Pil) all formally retired from office. Regionalism was slowly attenuated even though it maintained a major effect on voters. On the other hand, because Roh Moo Hyun’s area of support was the democratic-based Jeolla province, but his hometown was in Gyeongsang (specifically Gyeongsang Namdo), these dualities of political ideology and birthplace-based regionalism influenced the election [69]. Because government personnel leaned towards the Gyeongsang province during the administration of the three generals and one civilian leader and towards the Jeolla province during the Kim Dae Jung administration, the complexity of Roh Moo Hyun’s background provides an exploratory environment for the relationships between regionalism, social network ties, and firm value. In fact, after assuming office, Roh Moo Hyun carried out personnel changes considering the regional arrangement. Although he received over 90% of votes from the Jeolla region, 35% of his government personnel were from the Gyeongsang province and 27% were from the Jeolla region. He also pursued a government reform agenda toward more decentralization and balanced development between regions. In 2004, the Special Law on Decentralization Promotion was enacted that included transferring all central government agencies, most government ministries, and even the capital city, for the purpose of local autonomy, regional politics, and a balanced national development policy [70]. These regime changes lead us to propose the following hypotheses.

Hypothesis 2a (H2a).

In the Gyeongsang province, political connections are positively correlated with firm value during the Roh Moo Hyun administration.

Hypothesis 2b (H2b).

In the Jeolla province, political connections are positively correlated with firm value during the Roh Moo Hyun administration.

The presidential election of 2007 was marked by the victory of the opposition candidate from the GNP, Lee Myung Bak. Lee Myung Bak scored a landslide win of 48.7% of the vote against the opposing candidate from the United New Democratic Party, Chung Dong Young, who received only 26.0% of the vote [69]. Lee’s victory was the second shift in political power from the democratic towards the conservative party, after the first shift in the opposite direction in 1997. This second shift resulted from discontent with Roh Moo Hyun’s weak presidency and with the lingering economic uncertainty at home during his administration [69]. Korean voters decisively chose Lee Myung Bak hoping for economic progress, stability, and a better quality of life, based on his career as the president of Hyundai Engineering and Construction and as the former mayor of Seoul [69]. He received over 55% of the vote from his base in Gyeongsang (55.0% in Gyeongsang Namdo and 72.5% in Gyeongsang Bukdo) and only 9% of the vote from the rival Jeolla province (9.2% in Jeolla Namdo and 9.0% in Jeolla Bukdo). In contrast, Chung Dong Young received an overwhelming 78% of the vote from his base in Jeolla (78.6% in Jeolla Namdo and 81.6% in Jeolla Bukdo), and only 6% of the vote in rival Gyeongsang (12.3% in Gyeongsang Namdo and 6.7% in Gyeongsang Bukdo).

As Kim [69] discusses, although the major themes influencing constituents’ voting behavior changed over time from democratization, liberalization, equality of opportunity, policy changes, and resolving polarization to economic stabilization, individual well-being, and advancing the overall country, regionalism strongly affected the 2007 presidential election. In the Jeolla region, over 78% of votes belonged to Chung Dong Young while only 9% of votes were for Lee Myung Bak. However, in the Gyeongsang region, 55% of votes (72.5% in Gyeongsang Bukdo) were for Lee Myung Bak while only 6% of votes were for Chung Dong Young. The contrast between the two rival regions during the presidential election became sharper during the Lee Myung Bak administration. After taking office as president, Lee Myung Bak was often criticized for his personnel appointments to key government positions. These appointments were referred to as “Ko So Young,” which stands for Korea University, the Somang church, and the Gyeongsang region, where Lee received strong constituent support in the presidential election [74]. In fact, 45.2% of Lee Myung Bak’s appointments to important posts, such as senior presidential secretaries, ministers, National Police Agency Commissioner, and government-affiliated institutes. In his initial years in office, there were individuals from Gyeongsang while only 16.2% were individuals from Jeolla. The corresponding figures for Roh Moo Hyun’s administration were 35% for Gyeongsang and 27% for Jeolla [75]. The regime change from a strongly liberal Jeolla province towards a strongly conservative Gyeongsang province after a decade resulted in favoritism to Gyeongsang and discrimination against Jeolla during the Lee Myung Bak administration. These regime changes lead us to propose the following hypotheses.

Hypothesis 3a (H3a).

In the Gyeongsang province, political connections are positively correlated with firm value during the Lee Myung Bak administration.

Hypothesis 3b (H3b).

In the Jeolla province, political connections are negatively correlated with firm value during the Lee Myung Bak administration.

Political connections still positively affect firm value over time after political and economic liberalization remains open. Some scholars argue that the value of political connections gradually diminish because firms no longer require government intervention but mainly focus on their own capabilities, strategies, management, and corporate governance structures relative to their competitors. Many empirical results support this perspective in terms of market-based competition [3,8,47,49]. However, others argue that the government still plays a significant role in market economies over time as a director of markets, regulator of industry policies, and even as a patron of specific firms [12,55,57,58]. In the three decades since 1963, South Korea has experienced a drastic transformation of its political and economic environment towards liberalization. This remarkable turnaround has not only led to democratization and economic development but has also reinforced reciprocal social network ties, regional cleavage and conflict, exclusion, and rivalry competition for scarce resources during regime changes. Although collusion between politicians and businesses was once deemed a necessary evil for economic growth, it was also deemed a necessary resource for the survival of firms over time. For example, the countless scandals such as illegal political fundraising, arrests of CEOs, and investigations against former politicians covered on the front pages of newspapers after the presidential election reflect that political connections have become a common phenomenon after political and economic liberalization and that political and economic actors must fiercely battle to capture political and economic benefits [74,76,77,78,79,80,81].

Hypothesis 4 (H4).

Political connections will continue to be highly positive and significant determinants of firm value as time passes and regimes change after political and economic liberalization in South Korea.

Overall, testing the proposed hypotheses allows for an investigation into whether corporate political ties can create significantly positive or negative firm value depending on political regimes/regions, which previous literature has failed to study.

5. Data and Measurements

5.1. Sample

This study focuses on Korean firms listed on the Korea Stock Exchange in the Gyeongsang and Jeolla regions. The number of firms increases over the study period from 89 (Gyeongsang 75, Jeolla 14) in 1998 to 177 (Gyeongsang 154, Jeolla 23) in 2013. First, it excludes 25 firms in the transport equipment-renting sector in the Gyeongsang province because they are paper companies based on the Ship Investment Company Act of 2002. Doing so balances the Gyeongsang sample with the sample from Jeolla, where no such companies exist. Second, several IPOs and delisted companies with tenures of less than 10 years during the study period are also excluded. Lastly, although Korean firms are obligated to report their financial data and information to the Ministry of Trade, Industry, and Energy and the Korea Stock Exchange, some firms lack available data because of omitted or noisy information before 2000. Thus, a sample of 93 firms with 77 from Gyeongsang (83%) and 16 from Jeolla (17%), are constructed, spanning 1998 to 2013. This sample covers the majority of firms from each region. Reliable financial data and socio-political information for the 93 firms are extracted from the KIS–VALUE database, the KOCOINFO database, the Korea Investor’s Network for Disclosure System (KIND) channel, and Korea Stock Annual reports. In particular, comprehensive data, such as total market value, total assets, total liabilities, and R&D intensity as well as information about CEOs, chairpersons, and directors, are obtained from KIS–VALUE, which is operated by NICE Investor Service, a leading credit rating agency in Korea. Other sources are utilized for a robustness check between databases.

5.2. Dependent Variable

Tobin’s Q, which is the ratio of the market value of a physical asset to its replacement value, is well known and widely accepted as a proxy for firm value and is commonly estimated as follows: (market value of common stock + book value of preferred stock + book value of debt)/(book value of total assets) [49,82]. Following Black, Jang, and Kim [82], we use the book value of total assets as a proxy for the replacement value because the replacement value is not easily obtained from firms and Korean firms reflect the value of their tangible assets to the current market value. Many studies focus only on the stock returns of politicians, CEOs, board members, and more. Thus, this novel dataset sheds light on more macro-level explanations for the overall firm value of political capital [4,13,23,26,42].

Table 1. Descriptive statistics.

Regime 1
Variable Name	Obs.	Mean	S.D.	Min	Max
Region G Tobin’s Q	847	0.833	0.268	0.035	2.488
Region J Tobin’s Q	176	0.874	0.308	0	2.175
Regime 2
Variable Name	Obs.	Mean	S.D.	Min	Max
Region G Tobin’s Q	770	0.895	0.387	0.258	3.275
Region J Tobin’s Q	160	0.926	0.345	0.408	2.204
Regime 3
Variable Name	Obs.	Mean	S.D.	Min	Max
Region G Tobin’s Q	770	0.900	0.386	0.210	4.729
Region J Tobin’s Q	160	0.970	0.393	0.538	4.454

Regime 1 = 1998S1–2003S1, Regime 2 = 2003S2–2008S1, Regime 3 = 2008S2–2013S1. Region G = the Gyeongsang province, or the southeastern region, Region J = the Jeolla province, or the southwestern region.

provides the descriptive statistics of Tobin’s Q for 93 firms in the Gyeongsang and Jeolla regions. For example, the number of firms for Region G Tobin’s Q under Regime 1 is 847, which means that, during 1998 to 2003, the Tobin’s Q of 847 firms is used in Gyeongsang.

5.3. Independent Variables

Many previous studies suggest definitions of political connections. Prominently, Faccio [3] defines them as “one of the company’s large shareholders or top directors, such as CEO, president, vice–president or secretary, is a member of the parliament, a minister, or the Chief of the State.” Similarly, Boubakri, Cosset, and Saffar [10] consider that, “at least one member of its board of directors or its supervisory board is or was a politician such as a member of parliament, a minister or any other top appointed-bureaucrat.” In Germany, Niessen and Ruenzi [49] define them as “at least one member or delegate of the Bundestag, supervisory board or advisory council.” In a recent study, Haveman et al. [34] consider them as “set equal to one in years when the focal firm’s CEO, at least one other top manager, or at least one member of the board of directors had served as the chief officer or deputy chief officer at the county level (chu ji) or above.”

Overall, considering the general definitions across studies, a firm is defined as politically connected if one of the following criteria is met: (1) a CEO, chairperson, director, large shareholder, or top officer is/was a member of parliament, a minister, or top government official or senior civil servant, (2) the CEO or chairperson publicly supports a presidential candidate, or he or his employees participate in the presidential campaign, (3) the CEO or chairperson is reported by at least one of the major newspapers as supporting a certain political party, or (4) other social relations in terms of the Korean social network yonjul, kin ties, and school relations exist. This definition is very close to that of Faccio [3] but specifically adds a fourth yardstick reflecting the socio-political environment of Korea.

For example, one of the directors of the Samsung Fine Chemicals Corporation from Gyeongsang was the senior civil servant of the Fair Trade Commission in 2008, and, thus, the company is politically connected, according to criterion (1). Similarly, the chairperson of Kumho Asiana from Jeolla served as the Presidential Secretary in 2005, and another director was a judge of the Constitutional Court. Thus, again, this is defined as a connected firm. Furthermore, in 2007, the executive members of the union of nearly all of Hyundai’s affiliates, including HyunDai Mipo DockYard and HyunDai Hysco corporations from Gyeongsang, declared their support for Lee Myung Bak in the 2007 presidential election [83] and are politically connected, according to criterion (2). A major shareholder of DaYou Smart Aluminum firm from Jeolla is related by marriage to a politician, satisfying criterion (4). Other relations also include “indirect political connections,” such as lobbying, sponsorships, and well known relationships with the media. For example, the chairperson of Bohae Brew from Jeolla was arrested and investigated for suspicious lobbying for politicians in 2012, which also reflects connectedness [84].

To carefully capture the sources of political connections, a list of all CEOs, chairpersons, directors, and large shareholders between 1998 and 2013 is first extracted from the KIS-VALUE database. The data are then compared with the other sources using the list of annual and semi-annual business reports obtained from KIND and collected data on the individuals’ work histories, specific social networks, kinship, and more from major websites, newspapers, and each firms’ homepage. Lastly, the study searched for news about their political ties in the form of participation in presidential campaigns and public declarations of support for specific candidates or parties. For instance, the study looked at news in three major newspapers called Chosun, Joongang, and Donga. Correspondingly, a dummy variable, PC, was constructed, which takes a value of one if a firm is viewed as politically connected and zero otherwise. To address these potential problems of unobserved political connections between businesses and government, a new dummy variable, PC_new, was constructed, which takes an additional criterion into consideration, criterion (5). The CEO or chairperson was born in Gyeongsang or Jeolla. As a result, 18 additional firms (fourteen firms in Gyeongsang and four firms in Jeolla) are labeled as politically connected firms during the study period (

Table 2. Political connection variables.

Regime 1	Region G		Region J
Variable name	Obs. (=1)	Obs. (=0)	Obs. (=1)	Obs. (=0)
Political connection dummy	48	29	11	5
Political connection dummy_new	54	23	13	3
Political connection dummy_area	45	32	9	7
Regime 3	Region G		Region J
Variable name	Obs. (=1)	Obs. (=0)	Obs. (=1)	Obs. (=0)
Political connection dummy	41	36	8	8
Political connection dummy_new	44	33	10	6
Political connection dummy_area	36	41	5	11

Regime 1 = 1998S1–2003S1, Regime 2 = 2003S2–2008S1, Regime 3 = 2008S2–2013S1. Region G = the Gyeongsang province, or the southeastern region, Region J = the Jeolla province, or the southwestern region.

). In addition, a new dummy variable, PC_area, was constructed based on the birthplaces and supporting regions of presidents in the four sub-regions (the Gyeongsang Bukdo, Gyeongsang Namdo province, the Jeolla Bukdo, and Jeolla Namdo province) and their relations to each former president during the regime changes. Table 2 provides the numbers of politically connected firms and non-connected firms in each province depending on different measurements.

5.4. Estimated Equation

The baseline equation to be estimated is shown below.

$$\begin{array}{l}Valu{e}_{i,t}=\alpha +{\beta }_1·PC\left(\frac{O}{}4 IO\right)and ship between istic ger incentives to affectholders. ve strong incentive to inholderse should consider other f_{i,t}\\ +{\beta }_2·Firmsiz{e}_{i,t}+{\beta }_3·Leverag{e}_{i,t}+{\beta }_4·R\&D_{i,t}+Year Dumm{y}_t+{\mathsf{\epsilon }}_{i,t}\end{array}$$

where the subscripts i and t represent the firm and half-year, respectively. The dependent variable, Value, denotes Tobin’s Q. Control variables are used in the regression models as follows: Firmsize is calculated as the log of total assets, Leverage is measured as the total liabilities divided by the total assets, and R&D refers to R&D intensity, computed as the ratio of R&D expenditures to annual sales. The independent variable, PC, is a dummy variable for political connections. The year fixed effects are controlled using year dummies from 1998 to 2013. This model is based on those of Nee and Opper [52] and Ang et al. [49]. However, their models focus on the effects of market structure and corporate governance on firm value. Instead, this study focuses on the effects of political regime change and regional competition or discrimination on firm value over time.

The study includes all variables for 93 firms with dummies for each of the three political regimes and the two regions. The first regime (1998S1–2003S1, Regime 1) covers the last year of civilian leader Kim Young Sam and the five-year presidency of democratic president Kim Dae Jung. The second regime (2003S2–2008S1, Regime 2) includes the last year of Kim Dae Jung’s presidency and the presidency of his like-minded successor, Roh Mu Hyun. The third regime (2008S2–2013S1, Regime 3) includes the last year of Roh Moo Hyun’s presidency and the second regime changeover from liberal to conservative during the Lee Myung Bak administration. These half-year lagged regime dummies resolve the potential endogeneity problem.

In addition, the study constructs regional dummies for the southeastern (the Gyeongsang province, Region G) and southwestern regions (the Jeolla province, Region J). Because the regime changes are unexpectedly exogenous events for the firms in the sample in both regions, this dataset of regime and region variables can provide firm-level dynamics to cleanly identify “the changing role of these political capital variables over the passage of time” [26,53].

All the data for 93 firms in Gyeongsang and Jeolla from 1998 to 2013 are observed semi-annually. The pooled estimated generalized least squares estimation with seemingly unrelated regressions, proposed by Zellner, was used, which addresses both period heteroskedasticity and general correlations of observations within a given cross-section.

6. Results

6.1. Basic Result

The results of the baseline model estimation are presented in

Table 3. Estimation results: basic results.

	Regime 1		Regime 2		Regime 3
	Region G	Region J	Region G	Region J	Region G	Region J
	−0.029 **	0.014 *	0.275 ***	0.041 *	0.313 ***	−0.084 **
PC dummy	(0.010)	(0.008)	(0.017)	(0.024)	(0.019)	(0.022)
	−0.065 ***	−0.138 ***	0.026	−0.107 **	0.081 **	−0.163 **
Firm size	(0.015)	(0.024)	(0.023)	(0.038)	(0.025)	(0.035)
	0.781 ***	0.768 ***	0.522 ***	0.639 ***	0.345 ***	0.684 ***
Leverage	(0.024)	(0.019)	(0.052)	(0.051)	(0.062)	(0.079)
	−0.000	−0.000	−0.000	−0.012	−0.000	0.000 **
R&D intensity	(0.000)	(0.001)	(0.000)	(0.008)	(0.007)	(0.000)
Year fixed effects	YES	YES	YES	YES	YES	YES
R²	0.62	0.91	0.44	0.62	0.36	0.41
Number of Observations	847	176	770	160	770	160

Standard errors are given in parentheses. *, *, and *** denote significance at the 10%, 5%, and 1% level, respectively. Regime 1 = 1998S1–2003S1, Regime 2 = 2003S2–2008S1, Regime 3 = 2008S2–2013S1. Region G = the Gyeongsang province, or the southeastern region, Region J = the Jeolla province, or the southwestern region.

. Most results fully support the basic concept from many other studies of emerging and transitioning economies that the political connections have significantly positive or negative associations with firm value especially in relation to regime changes.

The first two columns show the results for the Gyeongsang and Jeolla provinces during the Kim Dae Jung administration (Regime 1). As predicted by H1a, political connections in the Gyeongsang province have a negative correlation with firm value, which is statistically significant at the 5% level during Regime 1. Moreover, H1b, which predicts that political connections in the Jeolla province are positively correlated with firm value during Regime 1 holds, and the result is statistically significant at the 10% level. These results are consistent with the study’s expectation that, during president Kim Dae Jung’s administration, firms in Jeolla received benefits to firm value that were supported by the government’s political and economic resources despite the limited number of samples. More importantly, this includes the prolonged discrimination and exclusion from government policies. However, those in Gyeongsang saw a decrease in firm value resulting from discriminatory treatment by the government in terms of the first regime changeover from conservative to liberal parties. Furthermore, because the empirical results show that politically connected firms have significantly positive or negative associations relative to non-connected firms in both regions, the results also partially support H4, which states that political connections will continue to be significant determinants of firm value over time throughout regime changes after political and economic liberalization in South Korea.

Columns 3 and 4 report the results for the Gyeongsang and Jeolla provinces during the Roh Mu Hyun administration (Regime 2). In accordance with H2a, political connections in Gyeongsang have statistically highly significant and positive associations with firm value during Regime 2. Similarly, in Jeolla, as predicted by H2b, politically connected firms have a positive and statistically significant association with firm value during Regime 2. These results also indicate strong statistical support for the argument that, after Roh Mu Hyun assumed office, firms in Jeolla continued to be offered the special privileges by the government that took office during Kim Dae Jung’s administration and, thus, the estimated coefficient also sharply increased because of the cumulative effect over time. Surprisingly, although politically connected firms in Gyeongsang have a negative association with firm value during Kim Dae Jung’s administration after the regime change to his successor, Roh Mu Hyun. Political connections are also positively associated with firm value in this province. This result is primarily because Roh Mu Hyun was from Gyeongsang and by his reform agenda of decentralization and balanced regional development. Once again, consistent with H4, the impact of political network ties with the government is positively and significantly correlated with firm value and becomes a significant determinant of firm value over time throughout regime changes after political and economic liberalization in South Korea.

Columns 5 and 6 present the results for the Gyeongsang and Jeolla provinces during Lee Myung Bak’s administration (Regime 3). As predicted by H3a, political connections in Gyeongsang have a positive and highly statistically significant association with firm value Regime 3. On the contrary, in accordance with H3b, political connections have a significantly negative correlation with firm value in Jeolla during Regime 3. As above, the empirical results fully support the hypotheses regarding the second regime change from liberal to conservative parties in the 2007 presidential election. Firms in Gyeongsang enjoyed more positive benefits than in the former regime, which resulted from the preferential treatment of Lee Myong Bak’s administration. Conversely, the effects of political connections on firms in Jeolla changed drastically from positive to negative, which is explained by the severe exclusion from political and economic resources during Regime 3. Moreover, as predicted by H4, the political connections in both regions are clearly important and statistically significant determinants of firm value over time throughout the regime changes after political and economic liberalization in South Korea.

Lastly, H4 received strong support from the overall results. Political connections continue to be significant determinants of firm value and to have highly positive effects in the Gyeongsang and Jeolla provinces during 1998S1-2013S1 except under special circumstances in which retribution, discrimination, or exclusion were driven by the government due to regime changes even after political and economic liberalization. In particular, although the number of politically connected firms decreased (Table 2), the absolute values of the coefficients in both regions increased over time, which reflects bitterly intense competition among firms to secure new political capital. Some firms lose the game while others actively invest in forging political connection over time. This result is also similar to that of Siegel [26].

Several control variables are also important for explaining firm value. Overall, firm size has positive or negative values and statistically significant correlations with firm value, with one exception. These results are also similar to the finding of many prior studies that firm size may affect firm value depending on the extent of corruption, the quality of corporate governance, or the institutional environment, and more [16,42,52,62]. Specifically, in the Gyeongsang province, the value of this coefficient changed from positive to negative. However, in the Jeolla region, firm size has a consistently negative association with firm value over time. Furthermore, leverage has a positive and significant correlation with firm value in all regions across the study period, which is also supported by other studies [3,42,48,49,85]. However, the R&D intensity variable is not significant in general. However, it is only marginally significant during Regime 3 in the Jeolla region.

6.2. Robust Results

To check whether the above results and inferences are robust, the study performs a further test by adding a new criterion to distinguish between PC and non-PC firms. Under the previous definition of PC, the study considered the social relations between CEOs or chairpersons and political figures, excluding their birthplace regions. As Kim [66] notes, the formation of the Korean social network yonjul is significantly affected by whether individuals have the same region of origin. Birthplace-based regionalism extensively affects the building of interpersonal social networks including the promotion of political and economic elites and, especially, presidential and parliamentary elections during political and economic liberalization in South Korea. Prominently, Siegel [26] generates rich empirical evidence that having a CEO or chairperson who was born in Jeolla, Chungcheng, or even North Korea has a significantly positive or negative correlation with forming cross-border alliances when South Korea liberalizes.

Table 4. First robustness test results: New definition of political connection.

	Regime 1		Regime 2		Regime 3
	Region G	Region J	Region G	Region J	Region G	Region J
	−0.040 ***	0.019 **	0.246 ***	0.049 **	0.278 ***	−0.051 **
PC dummy_new	(0.010)	(0.007)	(0.017)	(0.024)	(0.018)	(0.022)
	−0.064 ***	−0.148 ***	0.025	−0.108 **	0.083 **	−0.152 **
Firm size	(0.017)	(0.025)	(0.023)	(0.038)	(0.027)	(0.066)
	0.767 ***	0.771 ***	0.525 ***	0.635 ***	0.335 ***	0.716 ***
Leverage	(0.023)	(0.020)	(0.052)	(0.051)	(0.063)	(0.078)
	−0.000	−0.000	−0.000	−0.012	−0.000	0.000 **
R&D intensity	(0.000)	(0.001)	(0.000)	(0.008)	(0.007)	(0.000)
Year fixed effects	YES	YES	YES	YES	YES	YES
R²	0.62	0.91	0.44	0.62	0.36	0.41
Number of Observations	847	176	770	160	770	160

Standard errors are reported in parentheses. *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively. Regime 1 = 1998S1–2003S1, Regime 2 = 2003S2 2008S1, Regime 3 = 2008S2–2013S1. Region G = the Gyeongsang province, or southeastern region, Region J = the Jeolla province, or southwestern region.

presents the estimation results using the same empirical strategy as in the baseline equation, but the independent variable is now PC_new rather than PC.

The results of the robustness check, based on the added definition of political connection, such as the basic results in Table 3, firmly support all hypotheses. Columns 1 and 2 show that political connections in Gyeongsang have a negative effect on firm value, which is statistically significant at the 1% level during Regime 1 (compared with the 5% level in Table 3). However, those in Jeolla have a positive effect on firm value during Regime 1, which is statistically significant at the 5% level (compared with the 10% level in Table 3). In columns 3 and 4, political connections in Gyeongsang still have a positive and statistically highly significant association with firm during Regime 2 even though the value of this coefficient is slightly smaller than that in Table 3. Additionally, in Jeolla, political connections have a positive and statistically significant association with firm value at the 5% level during Regime 2 (compared with the 10% level in Table 3). In columns 5 and 6, political connections in Gyeongsang still have positive and highly statistically significant effects on firm value during Regime 3 while they have a significantly negative correlation with firm value in Jeolla. Yet, the value of the coefficient is slightly lower than the corresponding value in Table 3. Lastly, all the control variables exhibit very similar results to those in Table 3.

Furthermore, the study tests for the robustness of these estimations using a more elaborate condition. There are cross-sectional subregions even within the southeastern Gyeongsang and southwestern Jeolla provinces: Gyeongsang Bukdo (North) and Gyeongsang Namdo (South)and Jeolla Bukdo (North) and Jeolla Namdo (South). Historically, there is longstanding sub-regional competition within Gyeongsang between the elites from Gyeongsang Bukdo, which are dominated by Daegu (TK), and those from Gyeongsang Namdo, which are dominated by Busan (PK). Among Korean presidents, Park Jung Hee, Roh Tae Woo, Lee Myung Bak, and current president Park Geun Hye are all from TK, whereas Chun Doo Hwan, Kim Young Sam, and Roh Mu Hyun are from PK. As Siegel [26] describes, the competition between the TK-based networks and the PK-based networks for resources has been fierce since 1980, influencing the presidential and parliamentary elections, social relations, and business-government relationship. In fact, the dichotomous results of the presidential elections in 1997, when 61.9% of TK voted for Lee Hoi Chang and 31.3% of PK voted for Rhee In Je, and in 2003, when 73.4% of TK voted for Lee Hoi Chang and 27.0% of PK voted for Roh Moo Hyun, reflect that this sub-regional competition remains influential within Gyeongsang. In contrast, such competition has not generally been observed within Jeolla.

To consider this condition, the study utilizes the political connection dummy_area. First, during Kim Dae Jung’s administration (Regime 1), the dummy takes the value of one for politically connected firms from both the Gyeongsang Namdo and Jeolla Namdo provinces, given that Kim Dae Jung was born in Jeolla Namdo and the previous president, Kim Young Sam, was born in Gyeongsang Namdo. Otherwise, the dummy takes a value of zero. Second, during Roh Mu Hyun’s administration (Regime 2), the dummy takes the value of one for politically connected firms from both the Gyeongsang Namdo and Jeolla Namdo provinces, considering that Roh Mu Hyun was born in Gyeongsang Namdo and the previous president, Kim Dae Jung, was born in Jeolla Namdo. Otherwise, it takes a value of zero. Third, during Lee Myung Bak’s administration (Regime 3), the dummy takes the value of one for politically connected firms from both Gyeongsang Bukdo and Jeolla Namdo since Lee Myung Bak was born in Japan but his supporting subregion is Gyeongsang Bukdo and the previous president, Roh Mu Hyun, was born in Gyeongsang Namdo but his supporting subregion is Jeolla Namdo. Specifically, the top 500 firms listed on the Korea Stock Exchange in the Gyeongsang Namdo region are considered because only 18 of the 77 firms are located in Gyeongsang Bukdo. Otherwise, the dummy takes a value of zero. Thus, 20 firms (13 in Gyeongsang and 2 firms in Jeolla) are removed from the set of politically connected firms during the study period (Table 2).

Table 5. Second robustness test results: New definition of political connection.

	Regime 1		Regime 2		Regime 3
	Region G	Region J	Region G	Region J	Region G	Region J
PC dummy_area	−0.032 **	0.067 ***	0.211 ***	−0.005	0.293 ***	−0.159 ***
	(0.011)	(0.007)	(0.017)	(0.026)	(0.022)	(0.034)
Firm size	−0.066 ***	−0.112 ***	0.048 *	−0.114 **	0.048	−0.208 **
	(0.017)	(0.025)	(0.025)	(0.037)	(0.030)	(0.062)
Leverage	0.777 ***	0.785 ***	0.536 ***	0.655 ***	0.387 ***	0.770***
	(0.022)	(0.020)	(0.054)	(0.051)	(0.063)	(0.078)
R&D intensity	−0.000	0.000	−0.000	−0.013	−0.009	0.000
	(0.000)	(0.001)	(0.000)	(0.008)	(0.007)	(0.000)
Year fixed effects	YES	YES	YES	YES	YES	YES
R²	0.62	0.92	0.38	0.62	0.30	0.50
Number of Observations	847	176	770	160	770	160

Standard errors are reported in parentheses. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively. Regime 1 = 1998S1–2003S1, Regime 2 = 2003S2–2008S1, Regime 3 = 2008S2–2013S1. Region G = the Gyeongsang province, or the southeastern region, Region J = the Jeolla province, or the southwestern region.

shows the estimation results using the same empirical strategy as in the baseline equation but using PC_area rather than PC as the independent variable.

The results of the robustness check based on the sub-regional competition, like the basic results in Table 3, also strongly support all hypotheses. In columns 1 and 2, political connections in Gyeongsang still have a negative and statistically significant correlation with firm value during Regime 1 even though the value of the coefficient is slightly higher than in Table 3. In addition, those in Jeolla have a positive correlation with firm value during Regime 1. The result is statistically significant at the 1% level (compared with the 10% level in Table 3). In columns 3 and 4, political connections in Gyeongsang still have a positive and statistically highly significant association with firm value during Regime 2 even though the value of the coefficient is slightly lower than that in Table 3. Unexpectedly, in Jeolla, political connections have a slight negative association with firm value, and the statistical significance disappears, which differs from significance at the 10% level in Table 3. This result is not very surprising because of the reduced sample size (Table 2). In columns 5 and 6, political connections in Gyeongsang still have positive and highly statistically significant effects on firm value during Regime 3 even though the value of the coefficient is slightly lower than it is in Table 3. Furthermore, political connections still have a negative correlation with firm value in Jeolla during Regime 3, and this relationship is statistically significant at the 1% level (compared with the 5% level in Table 3). All the results for the control variables are similar to those in Table 3.

7. Conclusions

Prior studies simply argue that political network ties have only positive, negative, or no effects on firm value in a given political and economic environment, but no definitive empirical study has addressed the impact of changes in the environment on changes to net firm value over time in emerging and transitioning economies. Fundamentally, this study finds that political network ties to the regime currently in power have significantly positive effects on politically connected firms but not on non-connected firms and that political network ties to the political opposition to the regime in power has significantly negative effects on politically connected firms but not on non-connected firms. Moreover, from a contingency perspective, unexpected political regime changes have a sharp effect on the relationship between political ties and firm value over time. Lastly, political network ties continue to be highly positive and significant determinants of firm value over time throughout regime changes after the political and economic liberalization in South Korea.

Several prior studies only focus on relationships between political and economic actors as having positive effects, negative effects, or no correlation with firm value in emerging and transitioning economies, but they do not investigate the influence of changes to the political and economic environments on overall firm value as time passes from a contingency perspective [1,2,3,4,7,9]. Thus, the findings of this study further extend the focus of the existing literature by arguing that political capital is not immovable but rather that a firm can enjoy positive benefits from reciprocal social network ties for a certain period, even though these ties can conversely contribute to negative treatment by a rival network later on. Similarly, negative treatment from a rival network may change at another time. Clearly, these positive cascades of favorable treatment and negative cascades of out-group exclusion reinforce regionalism, social cleavage, conflict, exclusion, and competition for political and economic resources during regime changes. Since the study findings provide evidence that corporate political ties could have significantly positive or negative impact on firm value depending on political environments, corporate and political risk management should consider the characteristics and set time-varying policies.

Another significant finding of this study is that political network ties remained significant determinants of firm value even after South Korea achieved political and economic liberalization. Whether political connections gradually diminish in value [3,47,49] or still play a significant role [57,58] after political and economic liberalization is a heated debate in the field of political economy. The empirical results of this study strongly support the notion that business-government network ties between CEOs, chairpersons, or large shareholders and politicians, top government officials, or senior civil servants generate positive benefits from 1998 to 2013, which is the sample period of this study. This reflects that the government still controls resources in market economies as the director of the market, the regulator of industry policies, and a patron of specific firms after liberalization.

This study also raises further questions that cannot be readily answered with the current data. First, although this study shows that the value of political connections changes within and between rival regions as time passes, depending on regime changes, an unanswered question is how different politically connected firms are and how many of these differences change over time across firms in the two regions. As previously mentioned, this study was conducted based on semi-annual data and could not employ control variables for city-level economic development due to the relatively small sample size in Jeolla. One possible alternative is to increase the sample size in that region using listed firms on the Korean Securities Dealers Automated Quotations, which is similar to the NASDAQ, and to perform a quantitative comparison within and between those two regions.

Second, although this study uses the standard definition of political connection, it manually collected information from databases, government websites, and newspapers. The micro characteristics of political connections cannot easily be identified from such sources, which may distort the results. We need strategies to precisely detect political connectedness between firms and the government, reduce the number of sampling errors in constructing dummy variables, and effectively alleviate the effect of observed heterogeneity across firms. Recent studies employ the propensity score matching method to control the conditions on potential confounds, which eliminates spurious results [34,45]. Introducing these statistical strategies enhances the estimation results considerably.

Lastly, one of the main findings of this study is that political capital is not immovable, which reflects the positive and negative dynamics of political capital. The results reveal a series of positive cascades of favorable treatment and a series of negative cascades of out-group exclusion. However, the conditions under which these positive and negative connections are most advantageous or harmful remain unclear, including the reasons for change in the direction of their effects. Specifically, it remains to be determined if discriminatory treatment immediately increases after the opposition gains power and if the positive connection between political and economic actors is stronger. Similarly, it remains to be seen whether favorable treatment immediately increases after the opposition network gains power if the negative connection between the actors is stronger. From the firm’s perspective, an important question is whether only one-sided political ties to the government provide more benefit or whether multiple-deal connections more stably improve firm value over a long time horizon. It is not possible to answer these questions with the current data. Eminently, Zhu and Chung [13] show that a firm’s portfolio of political ties, including ruling and opposition parties, affects its performance, which provides theoretical insight into a promising task using the concept of diversification strategies.

This study has several limitations, which can be addressed in future extensional research. First, this study investigates the contingent value of political connections for firms in South Korea, which is a representative emerging economy. However, as Walder [28,29] mentions, because countries vary widely in terms of the extent, order, speed, and phase of political and economic reform as time passes in emerging and transitioning economies and even in developed economies [8,24,50], the determinants, aspects, and consequences of political capital dynamics also have vast cross-national heterogeneity depending on the institutional environment. Different types of political systems (i.e., multi-party or single-party democracy, communism, military dictatorship, and so on), different economic structures (i.e., planned, market, or mixed economies), and different strengths of regulatory constraints from government policies may have complex effects on the contingent positive and negative values over time. Thus, various types of favor exchanges and discrimination or exclusion may operate between rival political and economic actors. These contradictory interactions may influence overall firm value differently. Second, the current empirical estimation is not immune from the endogeneity problem. One may address the potential endogeneity problem between corporate political ties and firm value based on the recent remedy in Li [85]. Third, small firms are more likely to be connected to small rival parities not in power and larger firms to be connected to large parties in power, and, thus, an alternative measure for firm size could be considered [86]. Lastly, the study could have obtained better results if it had considered the channels through which political ties affect firm outcomes. One important channel could be corporate governance mechanisms such as the effect of board composition or market competition. These limitations are left for future studies to address.