Introduction

1The role of business organization in business influence has divided scholars. A dominant perspective suggests that when business elites form a united opposition front, prospects for influencing policy tend to be stronger than if each faction acts independently (Fairfield 2015; Vogel 1989). Other scholars disagree and show that business cohesion can allow the state to have some oversight over businesses and that fragmentation of business elites reduces their engagement in national politics (Cárdenas 2020; Mizruchi 2013). Whether businesses are fragmented or cohesively organized is a discussion about their internal organization and a way to interpreting the structure that enables a business to influence the state.

2To contribute to this debate, we turn to the field of corporate networks. That literature has shown that the internal connections amongst business elites can impact corporate political actions. Studies evidenced that corporations linked through interlocking directorates - corporate ties formed when directors sit in several boards - are more likely to donate to political parties ( Mizruchi 1992; Murray 2017) and that well-connected firms in the corporate network are those with a former politician in the board (Aragón-Falomir and Cárdenas 2020; Salvaj and Couyoumdjian 2016). However, this line of research has focused chiefly on the firm level and not on the network level. Therefore, it is still ignored which corporate network structure (cohesive, based on clusters, or fragmented) enables corporations to influence government policy.

3The link between the organization of business elites and business influence can contribute to better understanding the conditions of state capture. A consolidated line of research argues that institutional weakness is one of the main conditions of state capture. Where governance institutions are poorly developed, states are more likely to be captured by business private interests (Hellman, Jones, and Kaufmann 2003; Innes 2014). Whereas this hypothesis appears successful in explaining some cases in Central America, such as Costa Rica, it fails with others, such as Panama and El Salvador. In the World Bank’s worldwide governance classifications that measure the degree of institutional development between 2014-2018, Panama is above the 60^th percentile in voice and accountability, government effectiveness, political stability, and regulatory quality (where higher positions indicate better performance). However, in the control of corruption (which measures the extent to which public power is exercised for private gain as well as capture of the state by elites and private interests), Panama is below the 35^th percentile (World Bank 2021). El Salvador scores above the mean of Latin American and Caribbean countries in government effectiveness and regulatory quality but far below the mean in controlling corruption. This puzzling discrepancy between strong regulatory government institutions and high state capture indicates that other approaches are necessary to understand state capture conditions.

4This paper explores the relationship between corporate networks and business influence to figure out state capture conditions. To do so, first, we analyze corporate networks through interlocking directorates in three Central American countries - Panama, Costa Rica, and El Salvador - using network analysis; second, we pinpoint the relationship between the structure of the corporate network and two corporate political actions: contributions to presidential campaigns and revolving doors. Based on an original database on elite networks, this explorative study contributes to providing new insights into the conditions of state capture and the political role of business networking. This paper is the first comparative study on networks of interlocking directorates in Central America and combines unique data on contributions to electoral campaigns and revolving doors. Therefore, it can be a referential turning point to inspire studies of Latin American elites.

5The rest of the paper is structured as follows. In section 1, scholarship on state capture, business organization, and business influence is reviewed to specify our argument. Section 2 introduces the three cases, their commonalities, and differences. Section 3 outlines data collection and sources. Section 4 shows the results of our analysis of interlocking directorates, campaign contributions in presidential elections, and revolving doors. Section 5 discusses the findings focusing on the mechanisms between corporate networks and corporate political actions. The last section presents conclusions and avenues for future research.

1. State capture, business organization, and business influence

6To what extent business elites must be connected or united to influence policy has been a challenging question in business power studies. The elitist perspective assumes that business cohesion generates business influence (Mills 1956), whereas a more pluralist approach distinguishes between the potential for unity and the potential for influence. Based on that, a business’s ability to influence politics depends on its ability to organize and speak with one voice (Dahl 1958). Afterward, Smith (2000) criticizes the premise that unity leads to strength and argues that it depends on the issues at play. Some macro issues generate cohesion in business, such as labor policies, but they also generate the involvement of political parties, unions, and NGOs, whereby the influence of business in politics is limited, and the chances of state capture are reduced. Other more particularistic issues do not generate cohesion since they only affect a small fraction of business elites, such as the tax status of a product. These particularistic issues are more likely to generate state capture since business companies involved do not encounter any opposition.

7Stigler first coined the term capture in the 1970s to describe a process in which specific business groups use resources and public powers to improve their economic situation (Stigler 1971). Large corporations gain advantages over other actors to extract government benefits such as public contracts, tailor-made regulations, or tax benefits. Since the beginning of the 2000s, international financial institutions such as the World Bank have been interested in measuring the processes in which private companies arrange policies and legal environments for their benefit at the expense of the rest of society (Hellman, Jones, and Kaufmann 2003; World Bank 2021). The concepts of state capture and regulatory capture became popular to indicate a type of corruption in which companies exert influence and collude with public officials to obtain particular advantages.

9Along with these corporate political actions, business elites organize their interests. Here, research on corporate networks showed extensively that connections among top corporations’ directors are a means to organize business interests and generate opportunities for political action (Carroll 2004; Mizruchi 2013; Murray 2017; Useem 1984). Through corporate ties, business elites reduce the uncertainty and transaction costs of donating to political parties (if my peers trust this political candidate, I trust it too). Moreover, cohesive corporate networks can generate shared beliefs and class-wide interests, giving them a larger ability to raise funds and coordinate joint actions. Therefore, corporate networks can become a condition that enables, shapes, and constrains business influence.

10Corporate networks are uncovered by examining the presence or absence of interlocking directorates -ties created by directors of large corporations sharing board membership. Business elites may connect by other means -joint investments, family ties, membership in associations, exclusive schools, think tanks, and non-governmental organizations-, but interlocking directorates involve the large corporations and their top executives directly, ensure face-to-face meetings, two-way communication, and configure relatively stable structures. Therefore, networks of interlocking directorates are a proxy for the study of the business organization, as they show the structure of opportunities and constraints of business elites for coordination and collective action. In countries where large corporations share many directors, cohesion among business elites emerges, and where the largest corporations opt for not sharing directors or very few, fragmented networks are built.

11Interlocks are commonly used to control firms within business groups, but they become especially relevant for building consensus among businesses when corporations of different owners are connected. While the interconnection of companies within business groups was assumed, interlocks among business groups were hitherto unknown in Central American economies. Research on networks of interlocking directorates has concentrated chiefly on North American and European countries, and that carried out on Latin American countries emerged but focused mainly on the large nations of the region (Cárdenas 2016). While business associations are an alternative way of networking for building political cohesion among business elites (Luna 2020; Schneider 2004), in small economies such as those of Central America, business associations have become less relevant for political action. The high level of market concentration and transnationalization of big business groups have discouraged the direct participation of top businesspeople in traditional multi-sector business associations (Robles-Rivera 2021).

2. The Central American countries and the 2014 presidential elections

12When presidential elections threaten the privileged position of the business elites, it is a crucial time to investigate the organization of business elites and corporate political actions because control over the state is at stake. In the 2014 presidential elections, business elites faced the possibility of losing privileged access to the state in Panama, Costa Rica, and El Salvador. The former Panamanian President raised taxes on tycoons and threatened to keep powerful business groups out of government. In Costa Rica, a country traditionally governed by business-friendly parties, a leftist force led the polls threatening the country’s status quo. In El Salvador, business elites were excluded from the outgoing left government with the possibility of being ruled out for five years more in case of a new victory of the left.

13In general, during the 1990s, Panama, Costa Rica, and El Salvador experienced similar changes that might have strengthened business influence over politics. Firstly, these countries have moved from economies whose primary source of income was agriculture to service economies. The value of agriculture in Costa Rica decreased from 15.8% of the GDP in 1990 to 4.3% in 2019, in El Salvador from 16% to 5.1%, and from 8.1% to 2.1% in Panama (World Bank 2020). Business elites whose wealth is based on services and trade require more significant state intervention and specific regulations, including public procurement, flexible labor markets, promotion of tourist resorts in public lands, and low taxation of export-import goods. Secondly, the privatization of state-owned enterprises during the 1990s encouraged the emergence of larger business groups in services: banking, energy distribution, pensions, construction, and infrastructure management (Bull, Castellacci, and Kasahara 2014; Schneider 2012; Segovia 2002). Besides, it facilitated the entry of large foreign multinationals into telecommunications, banking, and industry (Robles-Rivera 2014). This privatization process meant a reduction in the size of states and, therefore, an increase in some business elites’ economic power.

15Regarding state-business relations, the three cases present some distinguishing characteristics. Costa Rica rarely faced political turmoil. Business elites have quietly managed friendly relations with political parties and governments. Businesses have supported political parties indistinctly as an insurance policy (Casas-Zamora 2005). However, presidential elections in 2014 produced a new framework for business elites that could result in losing access to influence government due to a potential electoral victory on the left. For the first time, the leftist party, the Frente Amplio (Broad Front), was ahead by 22% of total voting intention, followed by the officialism party, the National Liberation Party (PLN) at 19%, and the right-wing party, Libertarian Movement, at 14%.

16In El Salvador, although business elites were united through multi-sector business associations and a business-oriented right-wing party (Schneider 2012), three main changes suggest a drastic change in this cohesiveness during the 2000s and 2010s. First, large foreign transnational financial corporations took over major Salvadorian banks. Second, old business elites lost control over governmental institutions since the left-wing party Farabundo Marti National Liberation Front (FMLN) took power in 2009. Finally, the main right-wing governing party, ARENA, experienced a division. Dozens of leaders left ARENA and founded another party, the Great Nationalist Alliance (GANA). The 2014 presidential elections implied the possibility of the election of a former guerrilla commander, the FMNL candidate. In the perspective of Salvadoran business elites, a government led by a former guerrilla commander would involve a more radical leftist political agenda.

17In Panama in 2009, businessman Ricardo Martinelli (Democratic Change) became President in a political alliance with the pro-business right-wing party, Panameñista Party, led by Juan Carlos Varela. The coalition between Martinelli and Varela implied the incorporation of prominent figures of the Panameñista Party and business leaders to Martinelli’s cabinet. Martinelli passed a fiscal reform that directly taxes businesses and the properties of Panamanian tycoons to fulfill his campaign promises of significant investment in infrastructure and monetary transfers. Business elites felt betrayed by those policies, and the alliance between Martinelli and Varela ended in 2011. President Martinelli expelled Varela and prominent former business leaders from the government, and presidential elections in 2014 represented a dispute within powerful elites to control the state. A few supported Martinelli’s candidate, José Domingo Arias, and tycoons affected by Martinelli’s policies backed Juan Carlos Varela, who obtained the victory in the 2014 elections.

18Therefore, Costa Rica, Panama, and El Salvador’s economic and political characteristics are similar, but state-business relations are not so much, so the analysis of the organization of business elites can shed light on this issue.

3. Methods and data

19We explore the relationship between corporate networks and business influence in Panama, Costa Rica, and El Salvador in the context of the 2014 presidential elections when the business elites felt threatened. To do so, we analyze three variables - the network of interlocking directorates, contributions to presidential campaigns, and appointments of businesspeople to government positions -, and the relationships among them. This explorative study provides associational (not causal) evidence for the role of the structure of the corporate network in corporate political actions.

20Social network analysis was used to examine the networks of interlocking directorates. This methodology is well-suited to analyze large numbers of relationships and map the distribution and structuring of power (Scott 2012). Network analysis uses measures based on graph theory to identify the degree of cohesion of the whole network and internal clusters. The networks were analyzed in two phases. First, network cohesion measures were calculated to figure out whether business elites were more cohesive or fragmented at the national level. Second, components - a group of nodes connected by any path - were identified and analyzed to detect business clusters and identify which business groups are integrated in these clusters.

21Data collection and processing were carried out in three stages. First, 90 large corporations in each country were selected. Academics usually use between 50 and 250 companies in cross-national studies. A size of 90 firms was chosen following Cárdenas’s study on networks of interlocking directorates that we use as a point of comparison (Cárdenas 2016). The selection of top corporations included state-owned, national private, and foreign-owned corporations. The lack of regional or national corporate rankings led to building different lists of top corporations for each country. For Costa Rica, the ranking published by EKA Magazine for 2012 on income and number of employees was used. For El Salvador, the 2015 Industrial Ranking was taken from the Salvadoran Association of Industries. This classification considers ten sectors (textiles and clothing, food, beverages, plastic, paper and cardboard, pharmaceutical, metalworking, furniture, footwear, and banking) and measures the economic activity of corporations based on their export and import records. For Panama, since there was no single ranking, a classification was made based on the 2015 export and import rankings from Legiscomex, a business intelligence platform that analyzes the data of the National Institute of Statistics and Censuses of Panama.

22Moreover, top corporations of banking, insurance, and pension sectors were included from the ranking of the 500 largest Central American companies produced by Summa Magazine for 2015 (Revista Summa 2015). The second stage involved listing the members of the boards of directors of these large corporations. This information was taken from the Public Registry (Costa Rica), the National Registration Centre (El Salvador), and OpenCorporates (Panama) in 2016 and, if it was not available, from the companies’ web pages and their annual reports. Thirdly, those directors who belonged to several boards were identified, and the data were processed using UCINET network analysis software (Borgatti, Everett, and Freeman 2002). Networks were visualized using the Net-Draw application, where nodes represent corporations and ties denote directors who sit on several boards.

23Several sources of information were used to analyze the contributions to 2014 presidential campaigns: the Electoral Court (Tribunal Electoral), disclosure of candidates and journalist reports in Panama, the Supreme Electoral Court (Tribunal Supremo de Elecciones) in Costa Rica, and the Salvadorian Supreme Electoral Court (Tribunal Supremo Electoral de El Salvador) and Acción Ciudadana (2019). Revolving doors - appointments of businesspeople to government positions - were analyzed by investigating biographies of ministers who were members of governments from 2014. We reviewed whether the ministers were or had been directors/owners of large corporations and the economic sector to which firms belonged.

4. Results

4.1 Network analysis of interlocking directorates

Analysis of network cohesion

24Network cohesion was analyzed through the measures of density, average degree, and compactness. Density is the number of existing ties of the total possible. The average degree is the total number of ties divided by the total number of nodes. The compactness index measures how directly connected nodes in a network are. It ranges from 1 to 0. If all nodes were directly connected, compactness would be perfect (1); the closer to 0, the less compact the network. The measures of average geodesic distance and diameter, which also refer to the network’s compactness, were not representative due to the high number of nodes with no ties (isolated) and the network’s fragmentation. More than 70 percent of the corporations were isolated in the three corporate networks. To precise the specificities of the Central American corporate networks, we use external points of comparison. Building on Cárdenas’ analysis of elite corporate networks in five Latin American countries - Mexico, Chile, Peru, Brazil, and Colombia (Cárdenas 2016), we identified that density, average degree, and compactness values of the three Central American corporate networks were the lowest, and the number of isolated nodes were the highest (Table 1). Compared to other countries in Latin America, corporate networks in Panama, Costa Rica, and El Salvador are highly fragmented.

Table 1. Comparison of network cohesion measures

	Density	Average degree	Compactness	Isolated (%)	Size
Mexico	0.084	7.12	0.302	19.8	86
Chile	0.040	3.53	0.216	16.7	90
Peru	0.028	2.47	0.100	32.2	90
Brazil	0.018	1.64	0.096	32.2	90
Colombia	0.015	1.36	0.035	35.0	90
Panama	0.006	0.58	0.012	72.3	90
Costa Rica	0.004	0.33	0.004	71.1	90
El Salvador	0.005	0.42	0.006	73.3	90

Source: own research and based on Cárdenas (2016) and Cárdenas and Robles-Rivera (2020)

Analysis of components

25A component is a part of a network where all nodes are connected directly or indirectly by some path. The main component is the one with the largest number of nodes. Analysis of components allows the identification of business clusters, namely sets of connected corporations. The largest component in size was found in Panama, as 14.4% of the large corporations were in the same component, with an average degree of 1.46. In Costa Rica, components are very small. There are three components of 4, 3, and 3 corporations. The average degree of the largest main component in Costa Rica was 0.75. In El Salvador, the largest component was composed of only 7.7% of large corporations, with an average degree of 1.14. The largest business cluster in the Panamanian corporate network was strongly connected. There were almost no business clusters in Costa Rica, only three very small, poorly connected components. The main business cluster of the Salvadorian corporate network was connected, but it was small in size. Figures 1-3 display the corporate networks in each country and evidenced that the corporations that made up the main business cluster were more interlocked in Panama than in El Salvador and Costa Rica.

Source: own elaboration.

Source: own elaboration.

Source: own elaboration.

26To analyze whether these clusters of corporations were shaped by the joint action of several business groups, we identify the largest shareholder of each corporation belonging to the main component. The Panamanian corporate network's main component integrates corporations from eight national family business groups (Motta, González-Revilla, Humbert, Eleta, Quijano, De la Guardia, Galván Rodríguez, Lewis), two foreign owners (Calox International and First Quantum International), and one state-owned enterprise. The main component of the Costa Rican corporate network connects state-owned enterprises and a foreign-owned corporation. The main component in the Salvadorian corporate network integrates corporations owned by national business groups (Regalado, Romero Belismelis, Catani Papini, Calleja), and foreign owners (Grupo Empresarial Antioqueño and Imperia Continental). This evidence shows that interlocking directorates are a means to connect different business groups and were used more in Panama than in Costa Rica and El Salvador.

4.2 Contributions to presidential campaigns

27Analyzing contributions to electoral campaigns is a black box with varying degrees of transparency depending on countries. Instead of legislation in all three countries requiring keeping records of political parties’ funding, it depends on the goodwill of parties to make them public. Funds from the private sector to electoral campaigns between 2009 and 2010 were larger in Panama (56% of total spending) than in Costa Rica (42% of total spending) (Bull, Castellacci, and Kasahara 2014), which suggests that Panamanian political parties were the most dependent on the business sector and the most vulnerable to corporations’ will.

28According to data from the Electoral Court, between 2004 and 2014, the number of private donations tripled in Panama. In 2004, US$ 18.7 million was received, which increased to US$ 25.7 million in 2009 and doubled to US$ 59.2 million in the 2014 elections. While the winner of the 2014 elections, Juan Carlos Varela’s Panameñista Party, received about US$ 10 million, outgoing official party Cambio Democrático (Democratic Change) received about US$ 35 million. Journalist reports suggested that the majority of these funds (US$ 21 million) came illegally from Brazilian construction company Odebrecht (EFE 2017). The Democratic Revolutionary Party (PDR) even received donations for US$ 14 million, 4 million more than President Varela’s party.

29In Costa Rica, most private donors contributed to the National Liberation Party (PLN) in the 2014 elections, according to the records of the Supreme Electoral Court (see Table 2). In addition to receiving the most substantial amount of private money, the PLN also attracted the most significant amount of individual donations. While contributions included small and mid-sized amounts in the other parties (between US$ 10 and US $5,000), donations were concentrated above US$ 10,000 in the PLN. The other major historical party, the Christian Social Unity Party (PUSC), received the least contributions, probably due to the party’s collapse in the 2006 elections following the corruption scandals that involved several leaders. The party that won the 2014 elections, Citizen Action Party (PAC), received thirteen times less funding than the PLN.

Table 2. Contributions to political parties in Costa Rica between July 2013 and January 2014

Political party	Amount in US$
National Liberation Party (PLN)	1,227,367
Libertarian Movement	305,612
Broad Front (FA)	98,540
Citizen Action Party (PAC)	89,650
Christian Social Unity Party (PUSC)	32,469

Source: own elaboration, calculated based on the Supreme Electoral Court records

30The amounts for the July-September quarter were not available on the Supreme Electoral Court webpage.

31In El Salvador, some of the largest family business groups contributed to the right-wing ARENA party, whereas others supported the conservative GANA party, led by Antonio Saca, former leader of ARENA and President of El Salvador (2004-2009) (Labrador 2016). Companies grouped around ALBA-Petróleos, a large Venezuelan public oil company, financed the governing left-wing party FMLN, especially in the purchase of electoral advertising.

32The corporate networks of interlocking directorates can help to understand the different patterns of contributions. Tables 3.1 to 3.3 display corporations belonging to the main clusters of the corporate network in each country and their contributions through their business elites to the 2014 presidential campaigns. In Panama, 8 of the 13 corporations belonging to the main cluster contributed through their directors/owners to presidential campaigns, and all of them funded the same party, the winner Panameñista Party. In Costa Rica, the three small clusters comprised 12 corporations, but only one contributed directly to the presidential campaigns. Only 2 out of the eight companies of the cluster in El Salvador contributed to the electoral campaign. These results suggest that when business elites form a cohesive business cluster, such as in Panama, most of them contributed to electoral campaigns and the same political party.

Table 3.1 Corporations belonging to the single largest cluster in Panama and contributions made by the owners or directors of the corporations

Corporation	Family business group	Owner origin	Donor	Amount in USD	Party
MHC Holding	González-Revilla	Panama	Emanuel González Revilla	100,000	PRD
			Juan Raúl Humbert Arias	4,500	Panameñista Party
Banco General	Humbert	Panama
			Federico Humbert	15,000	Panameñista Party
ASSA	Motta	Panama	Felipe Motta Jr.	2,500	Panameñista Party
Copa Airlines	Motta	Panama
Cable Onda	Motta	Panama	Yolanda Eleta de Varela	1,600,000	Panameñista Party
Nutrición Animal	Eleta	Panama
Promarina	De la Guardia	Panama	Alfredo de la Guardia	53,500	Panameñista Party
Caja del Seguro Social	State-owned	Panama
Envases del Istmo	Galván	Panama		---
Panamascrap	Ardaiz	Panama		---
Empaques Colón	Lewis	Panama		---
Calox Paname	Fontana	Venezuela		---
Minera Panamá		USA		---

Source: own elaboration based on documents made public by Juan Carlos Varela and Juan Carlos Navarro.

Table 3.2 Corporations belonging to the single largest cluster in Costa Rica and contributions made by the owners or directors of the corporations

Corporation	Owner	Owner origin	Donor	Amount in USD	Party
Banco de Costa Rica	State-owned	Costa Rica		----
Holcim	LafargeHolcim	Switzerland		----
INS	State-owned	Costa Rica		----
Cosevi	State-owned	Costa Rica		----
Purdy Motors	Quirós family	Costa Rica	Javier Quirós Ramos	20,000	PUSC
Florida Ice & Farm	Jiménez family	Costa Rica		---
La Nación	Jiménez family	Costa Rica		---
Café Capris	Plate family	Costa Rica		---
HSBC.Aval	Cortés family	Colombia		---
BAC	Sarmiento family	Colombia		---

Source: own elaboration based on the Supreme Electoral Court records in 2014

Table 3.3 Corporations belonging to the single largest cluster in El Salvador and contributions made by the owners or directors of the corporations Aumentar Original (png, 82k) Source: own research based on Acción Ciudadana (2019) data and the Salvadorian Supreme Electoral Court.

4.3 Revolving doors

33In Panama, the Panameñista Party won the 2014 elections, and Juan Carlos Varela was elected President. With his family, Varela is the owner of the leading Panamanian alcoholic beverage company (Varela Hermanos S.A.). Varela was Minister of Foreign Affairs between 2009-2013, as part of Ricardo Martinelli’s government. With the arrival of Varela to the Panamanian presidency, numerous government positions were granted to members or business partners of three families: Motta, Humbert, and González-Revilla. Federico Humbert, who was elected Comptroller General of the Republic thanks to Varela’s endorsement, owns a stake in the Banco General and holds a position on the board of directors of Caja del Seguro Social. Emanuel González-Revilla, the owner of MHC Holding, was appointed ambassador to the United States. And several people linked to the Motta business group came into the Varela government. Table 6 displays a list of individuals related to the Motta business group who obtained a post in Varela’s government.

Table 4. Varela government’s public officials linked to the Motta business group

Individual’s name	Relationship to the Motta family	Government position
Dulcidio de la Guardia	Vice-President of the Banca de Inversión del Banco Continental	Minister of Economy and Finance (2014-2018)
Melitón Arrocha	Anna Boyd Motta’s husband	Minister for Trade (2014-2015)
Rodolfo Aguilera	Guillermo Henne Motta’s partner in several companies	Minister of Security (2014-2016)
Mirei Endara	Wife of Grupo Motta’s Treasurer (Miguel Heras Castro)	Minister of Environmental Affairs (2015-2017)
Joseph Fidanque III	Director of Copa Airlines	Manager of Tocumén S.A. (International Airport)
Francisco Sierra	Executive Financial Vice-President of Banco General	Minister-Counsellor (ad-honorem)
Rogelio Donadío	Seller of Global Brands (Motta International)	Vice-Minister of Security (2014-2016)
Carlos Duboy	General Manager of Grupo Motta International	Presidential Secretary
Jorge Motta	Stanley Motta’s cousin	National Secretary for Science and Technology
Emanuel Gonzalez-Revilla	Member of the Board of Directors of Banco General	Panamanian Ambassador to the US

Source: own elaboration based on Torrijos Legazpi (2015)

34Out of the 14 President Varela’s ministers, six came from large private companies. These former business directors and managers held positions in strategic ministries: Economy, Agriculture, Commerce, Transport, Housing, and the Panama Canal. These businesspeople were appointed to ministries associated with their business interests. For example, Jorge Arango, the owner of Compañía Agrícola Industrial, the largest pork producer in Panama, was elected as Minister of Agricultural Development. Mario Etchelecu, a tycoon in the real estate sector, was appointed Housing Minister. The Minister of Public Works, Ramón Arosemena, was a construction sector businessman. Dulcidio de la Guardia, the Minister of Economy and Finance, was Vice-President of Investment Banking at Banco Continental, a bank acquired by Banco General, which the Motta and Humbert business groups own. Table 7 shows the businesspeople recruited to ministries in Panama, Costa Rica and El Salvador between 2014 - 2018.

Table 5. Ministers in government of Panama, Costa Rica, and El Salvador between 2014 and 2018 and their affiliations to business sectors

Name	Ministry	Private Company/Sector
Panama
Roberto Roy	Canal Minister	R-M Engineering/ Construction
Dulcidio de Guardia	Minister of Economy and Finance	Banco Continental/ Finance
Jorge Arango	Minister of Agricultural Development	Compañía Agrícola Industrial, S.A. (Caisa)/ Agriculture
Ramón Arosemena	Minister of Public Works	American Engineering Group/ Construction
Augusto Arosemena	Minister of Trade and Industry	Arias Fabrega & Fabrega/ Legal Services
Mario Etchelecu	Housing Minister	Bienes Raíces COT/ Real Estate
Costa Rica
Manuel González	Ministry of Foreign Affairs and Religion	Facio & Cañas/ Law Firm
Alexander Mora	Ministry of Foreign Trade	TecApro/ Information technology
Wilhelm Von Breymann	Ministry of Tourism	Costa Rican Trails/ Tourism
El Salvador
Tharsis Salomón López	Minister of Economy	Grupo Rayo/ Car industry

Source: own elaboration

35In Costa Rica, historically, traditional parties have allied themselves with different sectors of business elites. The Christian Social Unity Party (PUSC) was closer to the traditionally oligarchic agricultural economic elite, while the National Liberation Party (PLN) was closer to the industrial elites that emerged in the 1950s. However, after more than 30 years of two-party governments (PUSC and PLN) and after two consecutive PLN periods (2006-2014), the Citizen Action Party (PAC) won the elections in the run-off. With the coming to power of the PAC, relations with the business elites changed. Costa Rica’s President and leader of PAC, Luis Guillermo Solís, appointed active party members, career officials, or academics as ministers. Out of the 23 Costa Rican government ministers appointed in 2014, only three came directly from the private sector and were from mid-sized companies: Alexander Mora, from an IT company, Manuel Gonzales, from a law firm, and Wilhelm Von Breyman, from a tourism company.

36In El Salvador, the first FMLN government presided by Mauricio Funes (2009-2014) recruited academics, professionals, members of the party, and businesspeople to executive posts. After the 2014 presidential elections, out of the total number of ministers appointed by President Salvador Sánchez Cerén, only the Minister of Economy came from the business sector, Tharsis Salomón López, owner of the Rayo group, a mid-sized company engaged in vehicle repair. In summary, the Panamanian administration recruited far more from business elites and top corporations when appointing ministers than the Costa Rican and Salvadorian governments.

5. Discussion: the mechanisms between corporate, campaign contributions, and revolving doors

37Corporate networks in Panama, Costa Rica, and El Salvador are not cohesive at the national level. However, a more accurate analysis of the internal components revealed different network structures. In Panama, a set of business groups formed a cohesive cluster, whereas, in Costa Rica and El Salvador, only a few corporations were connected without forming any cohesive cluster. This variety of business organizations might correspond to different types of business influence. In Panama, the internal cohesion within the business cluster could facilitate the coordination of campaign contributions. The Panamanian businesspeople connected within the business cluster funded the same presidential candidate. The chances of reaching agreements on whom to donate could have been easier because the negotiation involved a set of business groups connected amongst them. By winning the candidate they financed, they not only had more government access, but they were also able to demand government positions. Several business elite members linked to the cohesive business cluster were appointed to government posts after the 2014 elections in Panama. Thus, all business groups of the business cluster gained access to the state due to the internal connections among them. In Costa Rica and El Salvador, the absence of a cohesive cluster impeded the business groups to coordinate their electoral contributions. Given that the party that won elections got fewer contributions from the business elites, they lost their ability to influence policymaking to some extent. The number of businesspeople appointed to government posts was almost nil in Costa Rica and El Salvador.

38Therefore, corporate networks can shape business influence. A single cohesive cluster facilitates negotiations and coordination to finance political parties. This gives some business elites more power in case the party wins since it relied on this business cluster to reach the presidency. On the other side, the absence of connections between business elites makes it difficult to achieve political agreements. Although business elites carry out actions to influence policymaking, they do so in an uncoordinated way. Figure 4 summarizes the mechanisms between the structure of the corporate network and the corporate political actions.

Figure 4. Mechanisms between corporate networks and corporate political actions

39In Panama, despite President Varela’s increased transparency and prosecuted corrupt officials, the influence of business on politics persisted and upgraded. Public procurements, fiscal reforms, and judicial appointments served to a great extent particular business interest in Panama. For example, in 2010, Martinelli approved a tax law to update the property appraisals in five exclusive areas of Panama City, where several tycoons lived, but in 2015 President Varela derogated it, and the ones who paid the property tax could ask to have their money back (Bellini 2016). According to the World Bank Governance Indicator “control of corruption,” which measures the extent to which public power is exercised for private gain (used as a proxy for state capture), during the presidency of Juan Carlos between 2014 and 2018, Panama fell from the 46th to the 36th percentile rank, where lower values indicate higher levels of state capture (World Bank 2021).

40In Costa Rica, after 2014, President Solís broke up with the tradition of recruiting big businessmen and technocrats favorable to free trade policies to lead the economic and commerce administration. Solís’ government approved a new labor law despite the opposition of the former government and top business elites. The law reduced some privileges of business elites regarding collective bargaining between unions and employers, facilitated the reincorporation of dismissed employees, the payment of wages to striking workers, and grants incentives to facilitate the negotiation between workers and employers (Castro Méndez 2017).

41In El Salvador, studies on taxation evidenced how a small group of business elites historically promoted a regressive fiscal legal framework with very reduced levels of universality (Schneider 2012). However, with the election of the FMLN in 2009 and 2014, the government passed three important fiscal reforms, which taxed business elites. For example, it was implemented an increase of income tax, new taxes on sumptuary real estate, a tax on financial operations greater than US$ 750, and the retention of the tax of merchants from the informal sector (0.25% above the US$ 5,000). Another policy that substantially challenged the status quo of business elites was the new policy of minimum wages. Salvadoran business elites opposed wage increases. Nevertheless, after 2014, the government passed a new law that increased the minimum wage of agricultural workers from US$ 118 to 224 and US$ 210 to 300 for maquila workers.

Conclusions

42Earlier studies have debated which type of corporate network-cohesive or fragmented-favors business influence in politics (Mizruchi 2013; Fairfield 2015). This research revealed that it is not about either cohesion or fragmentation but rather the number and level of cohesion of internal clusters in the corporate network. A single cohesive cluster of business elites plays a crucial role in fostering coordination among different business groups and planning collective actions and thereby enhancing the capacity of some (not all) business elites to influence state regulations. Networks allow business elites to collectively recognize, coordinate and promote their interests as if they were public interests (Bernhagen, 2007). Therefore, regulatory proposals arising from cohesive business clusters can have more legitimacy and greater feasibility of implementation, even though this entails a high cost for the rest of the population.

43We found that a cluster of business groups gained larger influence over the government in Panama. They financed coordinately a businessman running for President and used revolving doors to avoid political intermediaries between business interests and public policy formulation. This business cluster could exclude intermediary politicians thanks to their undisputable control over the winning political party. State capture can provide business elites with a comparative advantage in Panama, where national business groups face high trade openness and increased competition with larger transnational firms.

44State capture is a context-dependent and path-dependent phenomenon determined by the complex institutional system in a specific country. Previous studies showed that under the conditions of high income inequality, weak law enforcement, an oligopolistic market structure, state capture becomes more likely to emerge (Omelyanchuk 2001; Durand 2019; Hellman, Jones, and Kaufmann 2003). Our results suggest that networking among business elites can be a source that precedes state capture, as it organizes access to the state through collective mobilization of resources and coordinated action planning. In addition to the conditions of institutional weakness, income inequality, and sectoral concentration for understanding state capture, the network analysis of interlocking directorates helps map the opportunities and constraints of business influence.

45Corporate networks in Panama integrate some large business groups, but without building a national integrated business elite. A concern arises as to whether the union of the entire business elite would favor state capture more or whether, on the contrary, it would allow the state to control the business elites. The formation of a national cohesive big business community would force business elites to articulate very diversified and cross-sectorial interests. In doing so, national cohesive corporate networks would generate more inclusive, broader, and legitimate demands. Political elites would be more open to incorporate those class-wide projects in the political agenda, and, in consequence, state capture might not be as necessary for business elites.

46In conclusion, the study outlined here contributes to the literature in the following ways. At the theoretical level, to incorporate corporate networks into the discussion of state capture conditions. At the methodological level, it uses network analysis to study the political strategies of business groups. At the empirical level, to map the network-organization of business elites in Panama, Costa Rica, and El Salvador for the first time and generate data on revolving doors and private contributions to political parties. Consequently, it invites further avenues of research on elite networks and their political consequences. Future studies should address whether cohesive business clusters, given their capacity for political influence, can help understand the emergence or blockage of institutions for major social problems, such as the policy responses to the pandemic and economic crisis.