A comparative analysis of the internationalization of sub-national and central state-owned enterprises: shreds of evidence from Latin America

Abstract

Purpose

In this exploratory multiple case study, we aim to compare the internationalization of two state-owned enterprises (SOEs) owned by subnational governments with three owned by central governments in Latin America. This study provides a contextualized answer to the question: What are the differences in the internationalization of subnationally owned SOEs compared to central SOEs? This study finds that the speed and diversification of these two types of SOEs’ internationalization differ because they have a different expansion logic. Subnationally owned SOEs have a gradual and diversified expansion following market rules. Central government’s SOEs are specialized and take more drastic steps in their internationalization, which relates to non-market factors.

Design/methodology/approach

This study builds an exploratory qualitative comparative case analysis that uses multiple sources of data and information to develop a comprehensive understanding of SOEs through process tracing.

Findings

The study posits some assumptions that are confirmed in the case analysis. This study finds relevant differences between sub-national (SSOEs) and central authority (CSOEs’) strategies. SSOEs’ fewer resources and needs to increase income push them to follow a gradual market-driven internationalization and to diversify abroad. CSOEs non-gradual growth is justified by non-market factors (i.e. national politics). CSOEs do not diversify abroad due to the broader set of constituencies they have to face.

Research limitations/implications

Given the exploratory comparative case study of this research, the findings are bounded by the particularities of the cases and their region (Latin America). This paper and its findings can be useful for theory building but it does not claim any generalization capacity.

Originality/value

This study adds complexity into the SOEs phenomenon by distinguishing between different types of SOEs. This paper contributes to the study of subnational phenomena and its effect in SOEs’ internationalization process, which is an understudied topic. To the authors’ best knowledge, this is among the first studies that explore subnational SOEs in Latin America.

Citation

Finchelstein, D., Gonzalez-Perez, M.A. and Salvaj, E.H. (2022), "A comparative analysis of the internationalization of sub-national and central state-owned enterprises: shreds of evidence from Latin America", Multinational Business Review, Vol. 30 No. 2, pp. 259-288. https://doi.org/10.1108/MBR-05-2020-0117

Introduction

While there has been a great interest in private firms' internationalization, the international expansion of state-owned enterprises (SOEs) has recently called global attention. For instance, almost a quarter of the firms in the 2019 Fortune Global 500 Ranking firms had complete or partial state ownership (Casanova et al., 2021). This group has several firms from China and others from the Middle East, Europe and Latin America. From oil giants such as Saudi Aramco or the Chinese Sinopec to electricity companies such as the French EDF to financial institutions such as Banco do Brazil, SOEs are among the global leaders in their industries. Thus, the relation between the international strategy of emerging markets firms and the state has recently reached more attention (Cuervo-Cazurra, 2018a; Grosman et al., 2016; Finchelstein, 2017; Lazzarini et al., 2020; Musacchio et al., 2015; Nuruzzaman et al., 2020). Within this area of study, there is a growing literature on multinationals owned by the state (Benito et al., 2016; Cuervo-Cazurra et al., 2014; Cuervo-Cazurra 2018b; Estrin et al., 2016; Kalasin et al., 2020; Li et al., 2017; Heugens et al., 2019; Musacchio and Lazzarini, 2018; Tihanyi et al., 2019; Velez-Ocampo et al., 2017).

The universe of SOEs is vast, and there are critical dimensions to understand the internationalization outcome. Most of the previous research studies enlightening distinctions between SOEs' internationalization focus on the degree of state ownership, where different behavior and outcomes depend on whether the state has partial ownership or fully controls them (Kalasin et al., 2020; Liang et al., 2015; Mariotti and Marzano, 2020; Musacchio and Lazzarini, 2014). Other scholars analyze if SOEs are publicly traded or not or how state ownership influences financial performance (Bruton et al., 2015; Grosman et al., 2016; Lazzarini and Musacchio, 2018; Tihanyi et al., 2019). The size and resources from each state and the company's autonomy concerning the state have also been analyzed (Inoue et al., 2013; McDermott, 2002; Musacchio and Lazzarini, 2014). This article focuses on the type of state ownership, which has been relatively understudied (Li et al., 2014). There are different layers and jurisdictions within the states. Thus, SOEs should not be treated as a homogenous subject as they are affected by the type of state that owns them. For instance, variation in SOEs’ ownership defines access to different kinds of resources. Moreover, each state’s influence in the decision-making process differs which, as explained below, shapes the strategy of these firms.

We provide an in-depth case analysis of how SOEs differ in their strategy to expand beyond their national borders when SOEs are owned by the central authority (CSOEs from now on) or sub-national levels of government (SSOEs). For instance, the Brazilian electricity company CEMIG is owned by the Minas Gerais's State, while the Brazilian federal government owns Eletrobras. The reasons behind their decision to operate internationally and their overall international experience differ. Eletrobras’ international endeavors are part of bilateral agreements with neighboring countries to build large hydroelectric plants or transmission interconnections. CEMIG’s limited experience abroad was done to obtain new markets and sources of revenues. We expect to contribute to SOEs literature by shedding light on the differences in the internationalization of SSOEs and CSOEs. We examine two subnational and three nationally owned state companies through exploratory comparative case analysis. Our research question is: What are the differences in the process of internationalization of SSOEs compared to CSOEs? Given that there are multiple dimensions within any internationalization process, we will concentrate on two key ones in this exploratory research. First, the internationalization speed, which is crucial to assess the capabilities' creation of any MNE (Riviere and Bass, 2019). The other relevant dimension that we consider is the sectoral focus within the internationalization process. We assess if these two types of SOEs’ internationalization are focused on a single activity or diversified into more than one sector when they expand abroad.

Only a few authors have already examined how institutions at the subnational level influence firms’ international strategy (Li et al., 2018; Ma et al., 2016). Our research contributes to this topic by studying how the international strategy of SSOEs shows unique behavior and performance in Latin America, which differs from CSOEs. Hence, we tackle an important – and often unattended- puzzle. There is very little knowledge of the reasons behind SSOEs internationalization. While each country and region differ in the distribution and availability of resources to subnational governments, in Latin America, local governments tend to have fewer resources than central ones. Thus, in this region, SSOEs budgets are not as soft (Cuervo-Cazurra et al., 2014; Lazzarini et al., 2020) as the ones from CSOEs. Moreover, subnational states do not have the same international networks and political agenda as the central ones, and thus their firms are less encouraged by geopolitical reasons to go abroad (Duanmu, 2014; Li et al., 2016). CSOEs usually follow a non-gradual growth, characterized by drastic entry, or exit decisions. The constant incentives to increase income pushes SSOEs to seek alternative activities abroad and diversify while the CSOEs examined here do not diversify abroad.

As a second contribution, our research is an addition to the study of subnational phenomena, which have been deeply examined in other disciplines – especially in political science and to a lesser degree in sociology (Falleti, 2005; Gibson, 2005; Giraudy, 2013; Gervasoni, 2010; Stark, 1996; Walder, 1995). An exciting set of studies within the management field highlight how different institutional settings shape a firm’s strategy (Cuervo-Cazurra et al., 2019a; McDermott et al., 2009; Finchelstein, 2017; Wan and Hoskisson, 2003). However, the study of SOEs' subnational ownership and its effect on their internationalization process has not attracted similar attention. Rather than focusing on how subnational institutions shape private firms, we focus on firms owned by different layers of state institutions, including sub-national ones.

Third, in terms of geographic contributions, Latin America has gained more research relevance (Aguinis et al., 2020) but the topic examined in this article has been comparatively understudied. The few articles on subnational phenomena within the international business field focus on Asia (Bruton et al., 2015; Li et al., 2018). There is a genuine reason behind this, as most internationalized SOEs come from Asia, especially China (Li et al., 2018). We show that this is also relevant in Latin America.

This research is a qualitative comparative case analysis. We do not aim to make findings that can be automatically generalized. Qualitative case studies have been largely used in international business (Bucheli and Salvaj, 2018; Cuervo-Cazurra et al., 2019b, Finchelstein, 2017; Rui et al., 2016) and they can contribute to further theory building (Eisenhardt, 1989; Runfola et al., 2017; Tsang, 2013). After a comparative case study of two SSOEs from Argentina and Colombia [Investigaciones aplicadas (INVAP) and Empresas Públicas de Medellín (EPM)] and three CSOEs from Argentina, Chile and Colombia [Empresa Colombiana de Petróleos (Ecopetrol), Empresa Nacional del Petróleo (ENAP) and Yacimientos Petrolíferos Fiscales (YPF), we offer relevant findings based on a contextualized explanation (Tsang, 2013; Welch et al., 2011). We use multiple sources of data and information to develop a comprehensive view of SOEs. A thorough process tracing (George and Bennett, 2005) of the cases, combined with data triangulation (Pauwels and Matthyssens, 2004; Ventresca and Mohr, 2002), has been applied.

The article has five sections. After this introduction, we present the literature review followed by the theoretical assumptions, which include the suggested mechanisms by which ownership influences the internationalization process. Then, we display the methodology and case selection that shapes this research. We continue with an analysis of the five case studies, followed by comparative case analysis. We finish the article with conclusions and offering some managerial implications.

Literature review

In recent years, several SOEs have gone beyond their borders becoming relevant global players, producing a wave of research on internationalized SOEs (Clegg et al., 2018; Cuervo-Cazurra et al., 2014; Estrin et al., 2016; Kalasin et al., 2020; Li et al., 2014; Mariotti and Marzano, 2019, 2020; Musacchio and Lazzarini, 2014, 2018; Musacchio et al., 2015; Tihanyi et al., 2019). Many studies signaled SOEs’ disadvantages to expand abroad compared to privately owned firms (Dharwadkar et al., 2000; Kornai, 1990; Megginson and Netter, 2001). However, more recent research has highlighted SOEs' potential advantages in the internationalization process based on their unique capabilities and resources (Benito et al., 2016). For instance, the Chinese State Grid has taken advantage of the resources and technological cooperation with other state agencies to make a significant international expansion (Casanova et al., 2021). Petrobras has also taken advantage of its large size and more significant resources to expand into other Latin American countries.

Regarding why SOEs expand abroad, many studies examine how politics, such as bilateral relations between states, play a vital role in the internationalization strategy of SOEs (Chen et al., 2018; Knill et al., 2012; Li et al., 2017; Tihanyi et al., 2019). Conversely, Choudhury and Khanna (2014), argue that SOEs expand abroad to obtain resources that could make them independent from other state actors. Other studies concentrate on how SOEs internationalize. There are significant contributions to how the institutional context defines the modes and entry strategies (Cui and Jiang, 2012; Nuruzzaman et al., 2020; Yang, 2018). Another set of research examines how the different ownership levels of state intervention affect the internationalization outcome (Kalasin et al., 2020; Mariotti and Marzano, 2020). Musacchio et al. (2015) make a pivotal contribution to this area by illustrating how there is a variety of state intervention that depends on whether the state appears as the sole owner, the main one, a minority shareholder or indirectly intervenes through the support of private firms. These authors show how firms with different kinds of state ownership differ in various dimensions. Musacchio and Lazzarini (2014) assess several types of state ownership in Brazil, which encourage distinct strategies within the group of Brazilian SOEs. We argue that there is a need to focus not just on the ownership level but also on the type of state ownership. There are different kinds of state ownership, such as local or county ownership, which have different goals than those owned by the central state (Li et al., 2014). Adding this subnational dimension to the analysis enlightens important distinctions within the universe of internationalized SOEs. As we explain in the next section, the resources, political goals and constraints of subnational governments are different from central ones, ultimately affecting these diverse types of SOEs.

To understand the role of ownership by states, subnational phenomena become crucial. Within political science, subnational phenomena have been largely studied in the Latin American political context. Gibson (2005) assess how subnational coalitions constrain the central government public policymaking process. Similarly, Gibson and Calvo (2000) highlight subnational coalitions' key role to implement economic reforms during the 1990s. Gervasoni (2010) examines how the different fiscal resources obtained at the subnational level define these governments’ institutional quality. Giraudy (2013) analyzes different types of governance existing at sub-national administrations in Latin America. The empirical work done in political science confirms that ignoring subnational factors may give a distorted diagnostic of the country (Herrigel, 1996; McDermott, 2002; Snyder, 2001) and thus it misunderstands actors' behavior (Gervasoni, 2010; Gibson, 1997; Samuels and Snyder, 2001) – including business organizations-. Like Falleti (2005), we link the process of decentralization and the evolution of the subnational administrations with our object of study. Thus, we follow this theoretical approach highlighting subnational factors' critical role in other actors and institutions. While most political science studies are interested in the effects of subnational factors in political actors and larger institutions, we concentrate on the effects it has on SOEs. Another crucial difference from political science research is that most empirical studies on political science exhibit subnational administrations' adverse effects in their respective dependent variable (Gervasoni, 2010; Gibson, 2005; Giraudy, 2013). Conversely, we observe how subnational factors can have a particular – not necessarily negative – performance on the firms. We align with few studies in international business that emphasize the role of subnational units of analysis (Cantwell, 2016; Li et al., 2018; Yang, 2018). For instance, Ma et al. (2016) contend that subnational institutions positively affect Chinese entrepreneurial firms’ internationalization. Nevertheless, most of these studies focus on subnational institutions' effect on private firms but do not analyze firms owned by a subnational government. We address this last point by comparing firms owned by subnational governments with some owned by the central state.

Theoretical assumptions

The speed and logic of the internationalization in sub-national state-owned enterprises and central authority state-owned enterprises

Similar to Li et al. (2014) seminal work, we argue that heterogeneity exist between the group of internationalized SOEs. We focus on how diverse institutions and governance structures create different incentives for SSOEs. SSOEs have other political premises and constituencies than CSOEs. The central states have broad international agendas based on international political diplomacy or the need for critical natural resources and CSOEs can be instrumentally used for these purposes (Li et al., 2017; Musacchio and Lazzarini, 2014). For instance, the expansion of Chinese SOEs in Africa and Latin America has gone in simultaneous with the Chinese State's attempts to increase its political, economic and diplomatic relations in these regions' countries (Child and Rodrigues, 2005). In Africa, Chinese investments went from representing 2% of the ones done by the USA to 55% in 2014 (Financial Times, 2019).

Similarly, Chinese investments have significant increased since 2010 in Latin America, almost doubling the FDI flow from the previous two decades (Perez Ludueña, 2017). This data is consistent with Knill et al. (2012), who affirm that international political relations are crucial to determining where states invest. In addition, Finchelstein (2013) shows how several Brazilian SOEs expanded during the 2000s following the government's goal to convert its national champions into international ones.

Suppose the central state uses CSOEs for diplomatic or geopolitical reasons. In that case, the expansion of these firms follows a particular pattern that responds to these goals displacing the relevance of organic and/or gradual growth. International political agendas become the crucial factor to consider. Thus, states encourage CSOEs to make investments aligned with these goals in a relatively short period irrespective of how organically this move is. Furthermore, states can provide CSOEs with additional funding to make acquisitions abroad. Hence, international expansion occurs through large but sporadic operations (leapfrogging), which, transform the scales and structure of CSOEs. As a result, CSOEs exhibit sudden growth explained by non-market factors such as geopolitical, or even ideological reasons.

Subnational governments usually have fewer political or international policy reasons to internationalize. In most cases, they are not interested in investments to influence a country, increase diplomacy or obtain an essential resource for the nation. Subnational governments are more focused on local goals that have different effects on SSOEs. SSOEs are used to offer better products and services or increase revenues for the local budget. Several scholars explored the SSOEs that survived market liberalization have pursued strategies to increase their competitiveness and productivity following the same premises that private corporations (Li et al., 2014). Thus, the expansion abroad follows a path that resembles private corporations pursuing market-seeking investments. The international expansion of SSOEs is more gradual and centered on market reasons.

The availability of resources is also a determining factor in expanding abroad. For instance, Wright et al. (2020) assess that the Chinese government has assisted the Chinese national champions in their internationalization process. Thus, it is essential to explore the differences between CSOES and SSOES in this dimension. The first thing to acknowledge is that the distribution of resources within each state shows a great variety from country to country. Numerous factors are defining how subnational governments manage their resources. The institutional design is probably a key variable to consider. For instance, Weingast (1995) shows how some western countries, such as the USA and UK, with a fair and efficient distribution of resources following a proper federalist system, have developed. Qian and Weingast (1997) assess how China's federalist system, which promotes provincial competition, has also encouraged a more efficient use of resources. Gervasoni (2010) analyzes subnational governments in Argentina, linking these governments' quality with their resource availability. From a more practical standpoint, the constraints that each country outlines regarding subnational autonomy to issue bonds or not, the central states’ discretionary capacity to determine resources’ distribution are crucial elements defining the availability of resources that subnational governments have. In particular, Latin America's presidential system (Linz and Stepan, 1996; O'Donnell, 1978; Zakaria, 1997) has provided the central authority with more discretionary powers regarding resource collecting and distribution, which are frequently used to obtain political agreements and support. Thus, central states’ are less constrained than subnational ones, which ultimately rely on the central state's distribution or formal legislation to issue bonds. Suppose subnational governments have more limited access to resources, such as in Latin America. In that case, their capacity to offer special funds to SSOEs to make a large investment abroad is limited.

Leapfrogging in the internationalization process is less likely if a firm lacks the necessary outward FDI funds. Kalasin et al. (2020) argue that SOEs have a helping hand, explained by soft budget constraints (Chen et al., 2018; Lazzarini et al., 2020), allowing them to dispose of more considerable sums of capital with fewer restrictions. While this can apply to CSOEs, in Latin America, SSOEs do not benefit from this helping hand as their budget is limited. The frugality of resources reinforces gradualism in the international expansion process of SSOEs while simultaneously creating a particular expansion logic. SSOEs should not produce additional expenses that would be difficult to explain beyond business-related reasons.

To sum up, we posit that fewer resources push SSOEs to pursue more market-driven strategies to increase income. Their international expansion follows this logic (Li et al., 2014) becoming more gradual. Conversely, CSOEs usually follow a non-gradual growth, characterized by drastic decisions (i.e. a large acquisition or a sudden sale that do not respond to the previous strategic moves). This non-gradual growth is justified by non-market factors such as searching for crucial resources for the country and/or political reasons (Cui and Jiang, 2012).

The focus in the internationalization process of sub-national state-owned enterprises and central authority state-owned enterprises

The market-driven orientation of SSOEs generates other interesting effects, which shape their internationalization strategy. As explained, subnational governments are more interested in obtaining resources from their firms instead of achieving an international political goal. Thus, SSOEs' management is mainly focused on increasing profits, which pushes these organizations to look for additional income sources. As a result, SSOEs become attentive to new market opportunities in which they might be competitive, and diversification appears as a viable option for growth (Hautz et al., 2014). Especially if SSOEs have some knowledge or resource created in the original sector that could be used as a potential competitive advantage in another activity. The crucial role of managerial and operational relatedness of the original activity to the new one in which companies diversify is vital for their success (De Andres et al., 2017; Markides and Williamson, 1994). Li et al. (2014, p. 991) offered examples of SSOEs diversification in their internationalization strategies such as the Vietnamese Saigon tourist or the Chinese Liugong. These firms have invested abroad in new activities in both cases but take advantage of their strengths, such as their customer needs and flexibility. In short, SSOEs' implicit logic of increasing income encourages them to be proactive in search of other sectors as long as they consider that it might be profitable. Thus, diversification is likely to occur in SSOEs' international endeavors.

Tihanyi et al. (2019) claim that SOEs are more conservative in their international strategy due to the broad set of stakeholders they have to face. These authors do not distinguish between types of state ownership as we do, and their claim refers indistinctly to SOEs compared to private companies. Still, central states are larger and integrate a broader set of stakeholders than subnational governments. Thus, following this logic and applying it to distinguish between CSOEs and SSOEs, we argue that CSOEs have strong incentives to be more conservative regarding diversification. The central authorities would probably disapprove of national funds in activities beyond these purposes to avoid opposition from a more extensive constituency set. Li et al. (2014, p. 991) describe how the Chinese government even settled a regulation in 2012 forbidding CSOEs to invest in non-core activities unless they had special authorization from the government. Hence, expanding into different unstipulated sectors is more challenging to justify and achieve.

The market-driven autonomy of SSOEs creates a flexible position in terms of sectors. However, it is more restrictive (or gradual) concerning the speed of the internationalization process (Hult et al., 2020). SSOEs are quite selective on where and how to invest given their more limited resources. However, they are more open to exploring new areas of production that might create additional income sources. Conversely, CSOEs are less conservative in expenditures to enter different countries, especially if the central state is pushing for a specific international agenda. Kornai et al. (2003) argue that SOEs have a softer output monitoring than private firms as there is less managerial control. Following this logic, the output monitoring of CSOEs is soft as long as they fulfil the planned geopolitical purpose. SSOEs' more limited resources push them to be stricter in the output monitoring and thus more conservative in their expansion. They have to be coherent with their financial resources and constraints. CSOEs only need to be more conservative regarding the industries in which the money is spent beyond the original target because it increases constituencies' opposition.

In short, the constant incentives to increase income pushes SSOEs to seek alternative activities abroad. A diversified internationalization is found in those cases. On the other hand, CSOEs do not diversify abroad as their strategy is more connected to the specific initial goal that the state decided for internationalizing. Thus, there are fewer incentives to diversify, and their international operations are mainly concentrated in the initial expansion sector. Assumptions contains a synthesis of the ideas and assumptions discussed, which will be empirically examined in the comparative case study analysis (Table 1).

Methods and sources

Methodological approach

We follow a qualitative case approach (Doz, 2011; Eisenhardt, 1989) to study how state ownership affects the internationalization of CSOEs and SSOEs. Given that our unit of analysis is SOEs, decisions might be extraordinarily influenced by non-business factors, and the political and institutional environment also needs to be introduced and contextualized (Welch et al., 2011). Furthermore, our study involves multiple dimensions and specific contexts in three countries. Thus, a qualitative comparative analysis, which relies on process tracing (George and Bennett, 2005), is a proper methodological tool to approach this subject.

Qualitative case studies have been largely used in international business (Bucheli and Salvaj, 2014, 2018; Cuervo-Cazurra et al., 2019b; Finchelstein, 2017; Rui et al., 2016; Gonzalez-Perez and Velez-Ocampo, 2014; Velez-Ocampo et al., 2021) and they can contribute to further theory building (Eisenhardt, 1989; Tsang, 2013; Runfola et al., 2017). We use multiple sources of data and information to develop a comprehensive understanding of SOEs. The complex features and governance of SOEs and the change in their internationalization demands a reading that is attentive to the business, social and political context to understand their evolution and the corporate strategic decisions. We combine process tracing (George and Bennett, 2005) with data triangulation (Pauwels and Matthyssens, 2004; Ventresca and Mohr, 2002) to have a complex holistic approach.

We follow established research using case studies (Eisenhardt and Bourgeois, 1988) and use a comparative case studies approach. Case selection is critical when developing and testing theory with case studies (Eisenhardt, 1989; Araujo and Dubois, 2004; Runfola et al., 2017). Although most international business studies still treat context as a homogeneous construct based on a single dimension (Poulis et al., 2013), we recognize its role and implication for research (idem). Qualitative research is considered context-sensitive (Welch et al., 2011). Our five selected firms owned by governments shared a context: emerging markets economies in Latin America. Therefore, we had a context-focused selection of case studies. Before conducting face-to-face (or technology-mediated) interviews with elite informants (Aguinis and Solarino, 2019), including several top executives during the internationalization process of these firms, we used direct observation, secondary data analysis and purposeful sampling to address the contextualization challenges presented prior and through case fieldwork. Because some of the research questions required an in-depth understanding of the decision-making process and organizational narratives (Solarino and Aguinis, 2020), interviewing elite informants was necessary for the research design as it enriches the study with multiple views that would not be available otherwise (Ostrander, 1993).

Case selection

We purposely selected five internationalized SOEs in three Latin American countries, three controlled by the central government (ENAP, Ecopetrol and YPF) and two by regional/city governments (EPM and INVAP). While the distinction between SSOEs and CSOEs is relatively new, the few existing studies almost exclusively focus on Asia. Moreover, to our knowledge, there is no relevant literature on the topic using Latin American cases, which increases the relevance of our regional selection.

The institutional instability existing in Latin America is also a distinctive characteristic compared to China or western developed countries. Latin America has particularity been praised for the opportunity it presents to perform theory testing and to be used as a “natural laboratory” (Aguinis et al., 2020). Thus, we argue that a comparative case analysis that aims to contribute to theory building, such as this one, is quite compatible and can take advantage of this unique environment.

As mentioned, the study of emerging markets SOEs has been geographically focused on Asia, especially China. Still, we believe that the Latin American cases can provide novel considerations and information about this phenomenon. From a political perspective, Asian countries have different representative systems than the ones in Latin America. Most Latin American countries are young democracies with recurrent rotation in the governing party, political instability, populism and societal political demonstrations (Aguinis et al., 2020). Instead, China does not have these democratic characteristics, and it is controlled by a single political party with strict control over the social, economic and political processes. This is not a minor point for the study of SOEs as in China, all these companies have a representative from the Communist Party at their boards while this rule is not present in Latin America. Thus, the influence that the government has on Chinese SOEs is much stronger. Size is another factor that distinguishes China and its SOEs from the Latin American countries and SOEs.

Additionally, the state’s broader and more robust presence in China can be observed in a large number of SOEs in this country. As Wright et al. (2020) describe the largest companies in China are still state-controlled. In Latin America, the number of SOEs that each country has differs. For example, Brazil has more SOEs than Chile, but none of the Latin American countries have as many SOEs as China and, in none of these countries, they play such an all-encompassing role. Size has an effect in the access to resources that the firms have and their revenues. For instance, in the oil sector, which is one the largest Latin American companies, the two most important firms had revenues of 72.8 billion dollars (Pemex from Mexico) and 76.6 billion dollars (Petrobras from Brazil) in 2020. In the same period, the two largest Chinese oil companies (Sinopec and China National Petroleum) had revenues of 407 and 379.3 billion dollars, respectively.

As stated by several scholars, the region’s relative homogeneity of institutional and economic conditions across countries reduces confounding effects and increases comparability (Aguinis et al., 2020, p. 618; Cuervo Cazurra, 2016). Still, we acknowledge the differences in the particular institutional context of the three countries where the studied SOEs are. Thus, we included one CSOE and SSOEs for Argentina and Colombia to check if this institutional variation was a key driver overshadowing the real results [1].

These cases also provide some exciting variation in history and evolution and sector and size, which enriches the sample variation. Nevertheless, the five selected cases are very relevant in political and economic terms for their own countries and are all well covered by the media, facilitating access to secondary sources. We follow Eisenhardt’s work, confirming that at least four cases are appropriate to develop a comparative case study analysis (Eisenhardt, 1989). The selection of our five cases allows us to control variation in our construct: type of ownership.

Measurement, data gathering and analysis

State ownership is our key construct. To measure it, we focus on the shareholders’ compositions of SOEs, classifying different state ownership types. Namely, we distinguish central state ownership versus subnational one. We collected data using different methods to enhance construct validity, allowing us to perform data source triangulation (Pauwels and Matthyssens, 2004). Our study combines written sources (financial statements, annual corporate reports, archival data and media coverage) with personal interviews to business elite informants of the selected companies (i.e. a former CEO of ENAP, the current Deputy Manager of INVAP, former Chief Economist of YPF, the Vice-President of New Market Development of EPM, former CEO of Ecopetrol or current Strategy Manager of Ecopetrol and central government executives with direct responsibilities regarding these SOEs). The researchers selected the elite informants based on their involvement and decision-making role in the firm’s internationalization. Researchers contacted the key informants both by email and telephone, referred by professional connections and personal networks. Ten face-to-face – or technology-mediated – interviews occurred between 2018 and 2021. Questions were designed uniquely for each interviewee to profit from their specific knowledge.

All interviews were audio-recorded, transcribed and immediately analyzed. The data analysis involved a deductive approach. First, we read the script thoroughly to identify common themes and to point in on essential insights. In this process, we labeled key phrases with codes. Then, as we aimed to highlight the relevant aspects of the interviews that will resonate with the study, we used thematic analysis.

We used several secondary sources to study the historical evolution of our construct and case studies. Regarding the archival sources, we surveyed annual reports of these companies on public agencies or ministries (i.e. CORFO or the Mining Ministry for the case of ENAP) and we triangulated them with third-party reports on these companies. We also studied corporate presentations, annual corporate and sustainability reports, press releases and articles in magazines and newspapers specialized in finance, politics and economy. This triangulation of multiple data sources and methods substantiates our constructs (Ventresca and Mohr, 2002). Like Rui et al. (2016), we develop a sequential analysis in which the interviews with elite informants (Aguinis and Solarino, 2019) and secondary sources are contrasted and complemented to build critical constructs of the type of state ownership. This process had different phases. First, collected data was coded and categorized to relate it to our research question (Ghauri, 2004). We followed with a within-case analysis and a cross-case pattern search. Second, we used pattern matching to compare predicted theoretical and empirically-based patterns (Sinkovics, 2018). Third, to sharp measurability and validity, we apply proposition shaping. Fourth, to build internal validity, we conducted a comparison with relevant literature. Finally, we identified theoretical saturation.

Cases and comparative analysis

Sub-national state-owned enterprises 1 – INVAP

INVAP was established in 1976, and it is among the most successful SOEs in Argentina, which is confirmed by its sustainable international exposure and advanced technology (Thomas et al., 2008). Owned by the Province of Rio Negro, the company is divided into four business units: nuclear, aerospace, medical equipment and industrial and alternative energy. INVAP has placed its products in five continents. While most of its international activities are export-based, INVAP has opened six branches to offer training and follow-up services of its products (Cuervo-Cazurra et al., 2019a).

INVAP was created to maximize knowledge and convert it into business resources and opportunities. The decision to create INVAP in Rio Negro is related to the fact that Argentina's primary nuclear research organism (Instituto Balseiro) is in that province. INVAP has built a deliberate plan to nurture its working plants with its expertise in nuclear research. This has created a virtuous cycle between INVAP and the nuclear agencies, in which they transmit knowledge between each other. Although INVAP's foundation supported Argentina's energy commission and the central government, the Province of Rio Negro has had full control of its ownership. This province decides the final destination of INVAP's profits and controls the board of directors.

The company's growth and international expansion have been driven by market demands and the company's needs within this context. INVAP has not had access to capital injections beyond its markets’ capabilities. It has never received subsidies or extraordinary resources like other SOEs. INVAP has always sought a tight economic and financial behavior to avoid becoming a weight in the provincial budget. Julian Gadano, former sub-secretary of atomic energy, confirmed: INVAP has never asked for money to the State (interviewed on 11th September 2018). The company's international expansion follows a consistent market-driven path sustained by INVAP's resources rather than from Rio Negro's financial assistance.

This fact has pushed INVAP to follow a gradual market-driven expansion path that resembles private companies (Thomas et al., 2008). First, given the nature of the nuclear reactor business, market saturation within Argentina came fast, as nuclear reactors' needs are limited. As the deputy manager of INVAP, Marcelo Basigalup, affirms: internationalization comes as a natural process to continue selling […] the salaries of our workers are paid with our income, not with money coming from a State ministry (interviewed on 13th September 2018).

Second, INVAP's capabilities position this SOE in a competitive advantage that could be leveraged internationally. When INVAP started its operations, there was only one US firm able to offer nuclear reactors for non-military purposes (Basigalup, interviewed on 13th September 2018). Hence, INVAP had vital knowledge to use it competitively in international ventures. Its first international operation abroad in 1978, selling critical components of a nuclear reactor to Peru, was done because they had the technological capabilities to offer these services. The same applies to operations in countries such as Australia and The Netherlands. The international expansion of INVAP has been defined by the search for new income sources, given that they had the capabilities to bid for these projects competitively.

While the owner of INVAP (Province of Rio Negro) has the right to use INVAP's profits, it has – until now – never taken out these profits. Instead, all this flow of capital has been reinvested in the firm. This freedom is remarkable given all the economic crises that Argentina has suffered since INVAP's creation, affecting this province's budgetary needs on several occasions. However, the Rio Negro province does not have the resources to fund a sizeable international expansion of INVAP. The freedom to use the yearly obtained funds and the lack of extraordinary resources is congruent with gradual but steady growth. The need to have stable growth and the urge to avoid losses has pushed INVAP to develop an active search for new business opportunities we need to sell our products to pay our costs (Basigalup, interviewed on 13th September 2018). While the primary international operations of INVAP have been in the manufacturing and consequent maintenance services of nuclear reactors for non-military purposes, the company has shown a proactive behavior to obtain other businesses.

Given that knowledge-based activities have been its distinctive feature and competitive advantage, INVAP has sought to apply this skill to activities beyond nuclear reactors. Medical applications linked to nuclear techniques were developed and sold by INVAP. In 1998 the first radiotherapy equipment was sold to Bolivia, followed by other contracts obtained in Venezuela during the 2000s (2004 and 2007). Taking advantage of their expertise, INVAP has also sold a lyophilization factory to treat food products in Mexico (2004). Since the middle 2000s, it has also expanded the aerospace unit abroad. Knowledge transfer has been the distinctive feature of this expansion. For instance, in 2008, the company signed an agreement to develop satellites for Brazil's research purposes. INVAP provided crucial bionic knowledge, which implied a knowledge transfer to this country.

In short, while INVAP’s international operations are mainly focused in the nuclear reactors business, the company has used its expertise and flexible labor force (Thomas et al., 2008) to diversify into new businesses in areas where the knowledge of the firm becomes a competitive advantage. INVAP has had international experience in its four units of operation. Basigalup illustrates: we sell complex technological developments […] our distinctive competitive advantage is the transfer of technology through our sales of different products (interviewed on 13th September 2018).

Sub-national state-owned enterprises 2 – EPM

EPM (Empresas Públicas de Medellín) is a Colombian SOE founded in 1955 and wholly-owned by Medellin's municipality. EPM has its most important assets in the electricity sector, but it is also involved in gas distribution, water treatment and distribution, waste management, renewable energy and telecommunications. As a result, this municipally-owned corporation has become one of the most relevant companies in Colombia. EPM is among the 15 largest SOEs in the region and the fourth largest Colombian multinational (América Economía, 2018).

EPM is a crucial part of Medellin's economic structure, and it is an essential source of income for this subnational government. EPM makes an essential contribution to Medellin's budget. According to Alejandro José Jaramillo, Vice-President of New Market Development of EPM: the company is a cash flow generator for the city (interviewed on 18th December 2018). In 2017 EPM's profit transfers to the city accounted for approximately 18% of its total expenses. In the last few years, EPM has transferred 55% of its profits to Medellin. Bad economic results of EPM have a substantial impact on the city, and thus, EPM is strongly encouraged to seek new market opportunities to increase revenues. The subnational ownership of EPM has shaped its international trajectory. Any investment needs to be largely considered so that it does not affect the city’s budget. Moderation rather than radical moves have defined EPM’s investments abroad. EPM has followed a gradual internationalization path in search of new business opportunities.

The company had its first international venture in Panama in 2003 with a US$51m greenfield investment in a hydropower project (Arias Jímenez, 2015). In 2008, EPM decided to expand internationally to continue growing and overcome the Colombian market saturation. Our core business is electricity, and we were reaching our limit to grow domestically due to regulation (Jaramillo, interviewed on 18th December 2018). The Colombian market regulations limited any company’s participation in generating electricity to 25% of the total market share. In 2010, EPM crystalized its international expansion with a partial acquisition of a power generation plant in Guatemala and then acquired of power distribution in Panama and El Salvador in 2011. In 2013, EPM acquired a water treatment company in Mexico and a greenfield investment in Chile's wind power. In 2015, EPM acquired a water company in Chile.

The diversification in the international path is also related to the type of ownership. EPM is the main economic actor in the provision of public services in the Medellin region. Water, energy and waste management are the main areas in which it intervenes, but the company has always been attentive to other activities. EPM has expertise in the management of clients. They have used this resource to expand into other complementary sectors with such skill: we expect to develop new activities that complement the ones we have (Jaramillo, interviewed on 18th December 2018). The subnational authority’s decision to expand EPM into other public services defines its corporate structure and diversification abroad. For instance, EPM created a separate unit in 2006 (TigoUNE) due to the growth of telecommunications in Colombia. TigoUNE merged with a branch of Millicom International Cellular (Tigo) a few years later. EPM owns almost half of the shares in TigoUNE, which has a small operation in the USA.

EPM also has other diversified investments abroad. EPM has waste management services in Guatemala and Panama, electricity production and distribution in Panama and El Salvador, water utilities in Mexico and Chile and Chile’s wind energy. Except for gas distribution, all of EPM’s business units have had operations abroad. Two factors explain this diversification. First, EPM already had a domestic diversification, which created the technical capabilities and experience to expand in these sectors beyond its borders. Second, the city's strong incentive to increase income encouraged EPM to seek lucrative investments abroad. The most emblematic example of the latter is the investments in Chile. The two leading businesses developments in Chile were characterized by the potential they represented within EPM. Both investments together represented the highest foreign investment of EPM and increased the diversification of its international activities and its economic resources. In 2017, Chilean activities accounted for 29% of EPM's international operations’ EBITDA (Grupo EPM, 2018). Nevertheless, EPM has recently decided to sell the Chilean assets due to local financial constraints.

Central authority state-owned enterprises 1 – ENAP

ENAP is the leading oil company in Chile in terms of revenues and market share. Founded in 1950, it is still fully controlled by the central state. ENAP plays a crucial role in a country that needs to import most of its oil. The firm has a presence both in the oil upstream and downstream businesses. ENAP exploits the only viable gas and oil well in Chile and controls all the Chilean refinery capacity.

Internationally, ENAP concentrates on the upstream business. Still, it also had a brief experience downstream in Peru and Ecuador, which was sold in 2013. ENAP's internationalization started in 1986 with the purchase of an oil well in Ecuador. Its expansion continued until 2005 when the company reached its peak international exposure with investments in eight countries from three continents. In 2006, the international presence started retracting, and by 2021 ENAP only had a presence in three countries (Argentina, Ecuador and Egypt).

The internationalization process in ENAP is significantly different from the ones observed in the two analyzed SSOEs. The first distinction arises from the internationalization pattern as it does not follow a gradual market-driven path. The company expanded into upstream, which was not a significant business in Chile. The first investments in Argentina, Colombia and Ecuador were in this area. Most of these investments came with very little knowledge transfer. ENAP did not have the expert technical knowledge to transfer. As the former VP of CORFO, the national agency controlling ENAP, affirms: we do not get any strategic business advantage from being in faraway countries like Egypt (interviewed on 29th October 2018). In several cases, the number of Chilean skilled workers in their international operations are reduced to one or two. Ricardo Cruzat (former CEO of ENAP) illustrated that: we were operating in Egypt, and there was only one Chilean there (…) the same in Argentina and Ecuador. Clearly, there is no observable market-driven competitive advantage related to knowledge or technical skills for these cases.

From a macroeconomic and political perspective, international investments are better understood. The large decisions in ENAP have always been decided by the national administration of the time […] the key decisions on what to invest are still defined by the State as a central owner (Hernan Cheyre, interviewed on 29th October 2018). Especially given the need to own critical resources for the Chilean economy. One of their most strategic investments abroad in the south of Argentina occurs during the oil sector’s deregulation in this country. The decision to expand into Argentina follows a general pattern that is not explained solely by market needs. Instead, it is the opportunity to acquire oil wells given the scarcity of them in Chile. As it happens in several CSOEs, non-market strategic reasons justify these decisions (Li et al., 2014).

Thus, the internationalization process was characterized by a high level of sectoral concentration in upstream, which can be connected to the government's non-market strategic decisions. As Hernan Cheyre confirms, the central government decided that ENAP should limit their industrial diversification: When ENAP considered expanding into the electricity sector, a new government came to power. Moreover, it said that they did not want ENAP to be in that sector as they wanted the company to concentrate in oil and gas and so the company had to cancel that project (interviewed on 29th October 2018). The only exception was the acquisition of Shell's gas stations in Peru (2004) and Ecuador (2005) through Primax, a joint venture with the Peruvian Grupo Romero. Nevertheless, even in this case – which is not a strictly diversified investment – the reasons do not follow a market-based explanation as it happens in the SSOEs cases. One could have assumed that the Chilean refineries could supply these gas stations, but this was economically unsustainable. ENAP was not self-sufficient in oil production and had to import to their refineries. The other alternative is that the acquisition would allow ENAP to transfer their expertise in the downstream and marketing areas into their new acquisition. However, ENAP never had the operational control of Primax. Grupo Romero had 51% of the joint venture and uninterruptedly controlled it. In 2013, ENAP sold its shares of Primax to Grupo Romero, leaving its only foreign investment in the downstream area.

The initial international expansion of ENAP was done with financial support from the government. ENAP's round of acquisitions exceeded the limited capacity and size that the company had in its domestic market. Additionally, the international expansion of ENAP (1989‐2005) exhibits some interesting differences concerning the reversal period (2006‐2019). During the international expansion period, the company was able to have more years of free availability of its profits, and it reinvested them to continue growing. The international reversal period coincides with an increase of the years in which ENAP transfers its savings to the central state [2]. Large capital expenditures were questioned later by the government and anticipated the higher level of intervention in the company during the reversal stage. These excessive expenditures also created a misalignment between the government and the company that eventually ended in more state intervention and the reversal. As recognized by Cruzat: ENAP's internationalization was never part of a master strategic plan. The Chilean State seemed the international expansion of ENAP as logic until the company started losing money (Interviewed on 3rd December 2018). The sale of the shares in Primax – following two years of losses and an increase in the debt rations at ENAP headquarters – concurs with this pattern as its CEO publicly states that they are selling a non-strategic asset to reduce debt (Expansión, 2013). This ratifies that the international strategy of ENAP was susceptible to the fluctuations of the central state changing perspective on the company.

Central authority state-owned enterprises 2 – YPF

Founded in 1922, YPF is the first SOE oil company in the history of America. It is still the leading Argentine producer of oil, both in upstream and downstream. The evolution of YPF has been very tightly connected to fluctuations in Argentina's political economy and has gone through several significant transformations. From consolidating President Peron's nationalized oil assets in the 1950s to a public company with the central state in the 1990s. Then, from being owned by a foreign MNE in the 2000s, to nationalize and become an SOE again in the 2010s.

These drastic changes have also influenced the internationalization of YPF. YPF started its primary internationalization process in the 1990s, after transforming the firm into a publicly-traded company under state control. In 1994, YPF signed an agreement with the Bolivian State to explore wells in this country. Most importantly, in 1995, YPF acquired the US company Maxus, which had operations in many Latin American countries. Federico Sturzenegger, former chief economist of YPF, explains that the main idea behind the international expansion of YPF was that given that the privatization reduced the domestic size of the company, if it didn’t expand abroad and increase its size it could have been acquired by a larger company (interviewed on 27th January 2021). Thus, YPF suddenly obtained extensive international exposure. By 1999, YPF had a presence in 11 countries. However, that year the Argentine State sold its shares in the company to the Spanish Repsol. The internationalization of YPF as a SOE ended during this period as the company became a foreign MNE branch with no control in the overall international strategy. In 2012, Argentina unilaterally took control of Repsol's shares, but its international operations have been neglectable since that moment. Its US subsidiary filed for bankruptcy, and it only has a few exploratory wells in South America, representing only 1.6% of the total acreage, but without generating income.

Given Argentina's long history of inflation, the state has used the public utilities that it owned to alleviate inflation by freezing prices. Potential revenues of these companies are sacrificed to aid the government. The use of YPF for political purposes has been recurrent and has prevented the firm from raising revenues that could potentially be used for expanding abroad.

Due to the vital role of energy in the Argentine political economy, the central state has paid much attention to YPF (Gadano, 2006). Argentina's historic capital restrictions explain YPF's concentration in the domestic market. An outward FDI would implicate taking the much-needed capital abroad. In particular, YPF had had significant investment troubles due to the challenges the state faced funding the large investments that exploration required. Under these conditions, there was never a plan to expand abroad until the 1990s. In this period, economic reforms created a new scenario. The burden to explore the Argentinean soil did not weigh solely in YPF, given that the oil sector deregulation attracted large sums of FDI.

Under a deregulated market, a state with no immediate budget needs and a fresh flow of capital brought by its publicly traded shares, expanding abroad became feasible. Rather than gradually, YPF internationalization happened in a brief period with large acquisitions. In four years, YPF invested in explorations in Malaysia, the Gulf of Mexico, Colombia, Venezuela and purchased shares in a Russian oil-related company (Sabbatella, 2013). The most emblematic international venture was the purchase of Maxus for US$740m, a large sum compared to other investments previously done abroad. This rapid international growth was almost exclusively concentrated in upstream. Its internationalization responds to a plan developed by a CEO that was favored by a strong presidential support and had an individual autonomy to take decisions (Gadano, 2012; Serrani, 2013) and an economic context that favored the attempt to expand abroad given the large sum of capital inflows that entered Argentina for a small part of the 1990s. This plan could not be maintained once the state had urgent economic needs – capital shortage and deficit in its budget (Gadano, 2012). Therefore, in the 1990s, the state decided to sell the company. YPF went from being one of the largest Latin American SOEs to becoming a European MNE branch. About the sale of YPF to Repsol, Sturzenegger contends that the sale was mainly done to solve the State fiscal problems […] the government needed money and Repsol was willing to pay a large sum for YPF (interviewed on 27th January 2021). The period in which Repsol owned YPF (1999‐2012) is not analyzed as the company stops being an SOE.

Since the 2012 nationalization YPF has had other priorities. Although it has some small exploratory wells – with no income so far – in Bolivia and Chile; the discovery of one of the world largest shale gas reserves in Argentina, has discouraged any attempt to expand abroad. Rather than investing abroad, YPF needs capital injections to cover the large expenditures to make the new reserve operational. Additionally, in 2016 YPF put US$200m and filed for bankruptcy of Maxus, its US affiliate, to prevent potential liabilities of a contamination lawsuit. If declared guilty, YPF will have to face a commitment of approximately US$14bn (El Cronista, 2018).

Central authority state-owned enterprises 3 – Ecopetrol

Founded in 1955, Ecopetrol is the largest Colombian company and its largest MNE. Similar to the Argentinean YPF, Ecopetrol started as a fully owned and controlled SOE. Then it was converted into a mixed capital company in which the state is the main majority shareholder. Still, in the case of Ecopetrol, the Colombian government never sold the majority shares. As Pablo Barcena, strategy manager of Ecopetrol, describes: The State is the main shareholder with the majority of the voting rights (Barcena, interviewed on 6th July 2021). The state still owns more than 88% of Ecopetrol.

Historically, the company has mainly concentrated in the oil sector and its internationalization has exclusively occurred in this industry. While it has attempted to expand into Africa with no success, its productive assets are in America. Ecopetrol’s internationalization is quite recent. Its first international experience occurred in Brazil in 2006 in a joint operation with the Brazilian SOE Petrobras and then expanded quite rapidly to Peru and the USA. In 2007, it also entered into Peru. The next year also developed a joint operation in the USA, which exposed the company for the first time into a country beyond Latin America. It is also important to note that the internationalization started before the distribution of shares that implied corporate governance changes that provided the company with more managerial autonomy and accountability. The initial decision to go abroad was made when the Colombian government had fewer checks and balances in the control of the firm. Yet, the investments abroad continued in the next few years, but mainly in these three original countries.

In the past few years, Ecopetrol has suffered some interesting challenges that have transformed its internationalization process. First, the COVID-19 pandemic has forced the company to modify its plans. While this global phenomenon affected most companies around the globe, and particularly Latin American ones (Gonzalez-Perez et al., 2021), the internationalization process of Ecopetrol has been specially affected due to the pandemic’s impact in the state’s budget. The need of the government to increase funds for the additional pandemic expenses related to health and social program shapes the investment plans and the distribution of profits of Ecopetrol.

It is important to highlight that the dividends of Ecopetrol represented about 1% of Colombia’s GDP in 2020. As stated by the specialized economist journalist Jesus A. Vargas Marin in regards Ecopetrol’s distribution of that year’s dividends: the distribution of dividends was irresponsible given the small profits of the company and the relatively low cash flow of the company at that time (Vargas Marin, interviewed on 18th July 2021). That same year, the company reduced their planned investments abroad by $900m (Portafolio, 2020). Another salient milestone occurred only a few months later when at the beginning of 2021, Ecopetrol announced the sale of its shares in OIG. Ecopetrol investments in Peru were done through OIG. Thus, after this operation, Ecopetrol exited Peru and was left with operating wells only in Brazil and the USA.

The sale of Ecopetrol’s Peruvian assets goes in simultaneous with another significant event. Namely, the announcement that the Colombian state was selling ISA. ISA is the largest transmission electricity company in Colombia, and one of the largest in Latin America, and same as Ecopetrol is mainly owned by the Colombian State. In a noticeable move by the Colombian State and its SOEs, the government signed an official pre-agreement with Ecopetrol so that the latter would buy ISA. The decision to sell ISA has also been influenced by the state needs to obtain liquid funds to sustain its budget. As Javier Gutierrez Pemberthy, former President of Ecopetrol, explains: Clearly the government needs the resources from the sale of ISA but also the government takes advantage of selling ISA without selling it […] the government took the decision to negotiate with Ecopetrol the sale of ISA (interviewed 27th July 2021). If Ecoeptrol completes this acquisition, then it would transfer approximately US$3.7bn to the state coffers. To obtain these funds, the company announced that it would discontinue investments and the emission of more shares (Ecopetrol, 2021). Ecopetrol has already started with this process regarding their international operations. As described, Ecopetrol has cut their investments in the USA and sold all of its assets in Peru. In this case, the changes in its international strategy are more connected to local needs than to a sizeable geopolitical agenda. Still, it is shaped by the urging political and economic needs of the state, which go beyond the particular demands of Ecopetrol’s business.

Regarding the activities in which Ecopetrol operates abroad, the company has three units of production. However, all of its foreign investments have been focused on the upstream oil sector. This characteristic confirms the decision to concentrate its international activities in only one sector. Furthermore, Ecopetrol conceives that the diversification should be in the places they sell rather than in the goods they produce abroad. As Barcena notes: the diversification strategy of Ecopetrol consists of selling our oil to different regions […] we have created more value thanks to the diversification of the places we sell (Barcena, interviewed on 6th July, 2021).

In short, like the cases of ENAP and YPF, the process of internationalization of Ecopetrol has also been affected by state urges and decisions that go beyond the company. While the internationalization process started right before the implementation of stricter corporate governance rules and the main countries of investments were defined in a very fast way, it is the current retraction in the internationalization process where the non-market decisions can be seen with more clarity. In the past two years the company has limited its investments in USA and sold all of its assets in Peru, which coincides with a period of increasing funding needs for the state and its decision to use Ecopetrol to make a large acquisition of another SOE. This new acquisition implies an indirect transfer of funds from the company to the state. As stated by Prof Camilo Díaz regarding this acquisition: what is being done is to create an alternative source of funding for the State […] this a well-thought movement from the Minister of Economics (Periódico Digital de la Universidad Nacional de Colombia, 2021).

Comparative case analysis

The five cases show different internationalization paths, which are influenced by the characteristics of ownership and the level of financial freedom. In Tables 2 and 3 we summarized some key indicators and the variables analyzed in this research.

CSOEs show sporadic growth and drastic changes in their internationalization process, which is mainly explained by non-market reasons. At ENAP, the need to obtain critical resources that were not available in its own country is crucial for its expansion abroad. As assessed by Hernan Cheyre: 'the internationalization of ENAP follows an inorganic growth path' (interviewed on 29th October 2018). ENAP's expansion occurred in the upstream sector, which was not ENAP's primary domestic income source. At YPF, the internationalization occurred in a period with fewer financial restrictions. As Sturznegger mentions: the IPO and the access to the capital markets allowed YPF to have resources that did not have before (interviewed on 27th January 2021). The expansion was done fast with large acquisitions, resembling leapfrog investments. When financial constraints affected Argentina again, the expansion stopped. At Ecopetrol, the internationalization process has been more moderated than in the two other cases, but the crucial needs of the country have also played a key role. As Gutierrez Pemberthy explains: the main goal of the internationalization was to increase the national oil reserves as Ecopetrol is the national oil company (interviewed on 27th July 2021). Moreover, the internationalization process of Ecopetrol has been recently altered by the financial needs of the central government. The current limits in their international expansion are a byproduct of these needs.

Conversely, market-oriented reasons were the key drivers of the internationalization of both SSOEs. At INVAP, market saturation, given the limited domestic demand for nuclear reactors, played a key role: INVAP internationalizes to sustain their revenues once the local demand diminished (Gadano, interviewed on 11th September 2018). Similarly, regulatory limits prevented EPM from expanding in the Colombian electric sector and pushed it to expand beyond the country. Due to regulatory constraints, we were reaching our limits of expansion in the local markets. However, we had financial resources, and thus we started searching for international opportunities (Jaramillo, interviewed on 18th December 2018).

The diversification found in the international investments of the two SSOEs is also consistent with our theoretical framework. INVAP has expanded in all of its business units, exploiting the competitive advantages created by its technology and the decision to complement the offered goods with critical services, including knowledge transfer. As Basigalup highlights, “for all of our activities abroad when we sell a product we do more than deliver it, we keep in contact, and the buyers end with key knowledge that we transfer […] that is crucial for expanding abroad in all of our sectors” (interviewed on 13th September 2018). In a similar fashion, EPM also has all its business units but one (gas distribution) with international operations. The conception of EPM, as a public utilities' provider, determined by its subnational owner, has pushed it to expand beyond electricity. Their expertise on clients is crucial for their diversification, as Jaramillo illustrated: “we see our strategy based on managing clients because this opens the door to expand into other sectors” (interviewed on 18th December 2018). There was an urge to increase income sources in both cases, encouraging both SSOEs to diversify abroad. None of these features was present in the CSOEs, and the internationalization was mainly concentrated in upstream oil production.

It is important to note that there are country-level differences affecting the strategy of the 5 analyzed firms. For instance, the financial constraints that Argentina has faced are larger than those in Colombia and Chile. This might affect the availability of resources of the firms. By the same token, the political orientation of Chilean and Colombian central governments has been more open to market rules than the Argentine one, especially during the 2000s and the first half of the 2010s and this could also affect the international expansion logic and its respective speed and focus. Nevertheless, despite these different institutional characteristics, the internationalization process of the five analyzed cases confirms our initial distinction between SSOEs and CSOEs. In the case of CSOEs, which are more directly involved with the central governments and some of these differences, the three companies share the same pattern. In all of these cases we observe a concentration in one sector of the international activities and the international expansion and/or retraction are driven by non-market factors linked to the particular needs of each respective central state. Still, the greater institutional instability observed in Argentina has promoted more radical changes in it CSOE than in the cases of Chile and Colombia. YPF has expanded more rapidly and its privatization and later nationalization abruptly finished its international endeavor. In the cases of ENAP and Ecopetrol the movements were defined by the financial needs of the states – specially the international contraction – but have not been so drastic and these firms still hold relevant assets abroad.

Table 4 summarizes the main findings of this study. It connects the type of state ownership with the different logics of expansion and variations in speed and focus.

Based on both the theoretical assumptions and the data of our exploratory research findings, we provide Table 5 that connects these assumptions with a few critical additional quotes and highlights obtained from the case studies.

We have illustrated how SSOEs and CSOEs diverse international strategies produce variation in the internationalization process. These distinctions are important as using the same treatment irrespective of the type of state ownership may lead to a misunderstanding of several SOEs' related topics. In a significant and influential study, Musacchio et al. (2015) argue that by ignoring the different ownership levels within SOEs universe, there are some fundamental traits of SOEs’ behaviors that are missing. Although we explore the type rather than the degree of state ownership, we argue that additional fundamental traits can only be observed when considering the difference between SSOEs and CSOEs. In particular, there might be misconceptions of what drives some SOEs to expand in more than one sector while others concentrate in a single core business. A similar issue could be found to understand the difference of speed in the internationalization process of SOEs. All of these factors are intertwined with the distinct access to financial resources and the different importance that SSOEs and CSOEs have for their direct owner. Most importantly, these factors affect the drivers and logic of SOEs internationalization process. To sum up, the international strategy of an SOEs could be misunderstood if the type of ownership is ignored. As we have just demonstrated, in these five cases, the type of ownership has played a crucial role to understand differences in the internationalization process.

Policymakers can also benefit from our findings as it enlightens the diverse challenges and capabilities that SSOEs and CSOEs face. For instance, realizing that SSOEs main challenges and incentives are related to the need to increase revenues and that this need has a strong effect in the internationalization process, might encourage policymakers to be careful about the dividends distribution demands. As the INVAP and EPM cases show, a greater financial autonomy to dispose of the profits helps to consolidate the internationalization process. Similarly, policymakers need also to be aware of the consequences behind pushing CSOEs to alter their international investment plans due to non-market reasons. These non-market pressures might have a stronger negative consequence after the internationalization started as it can create an internationalization reversal after some of the largest costs of the expansion, which usually occurred in the early stages of this process, have been done. The YPF case can be a crude example of this problem that could also be observed in a smaller degree in the more recent international reversal of ENAP and Ecopetrol.

Conclusion

This research addresses two interrelated puzzles. First, it attempts to contribute to the specific characteristics and reasons behind SOEs international strategy. Much has been written about this phenomenon concerning the ownership level in the internationalization of SOEs (Mariotti and Marzano, 2020; Musacchio and Lazzarini, 2014; Liang et al., 2015; Kalasin et al., 2020) but limited attention has been paid to the type of state ownership (Li et al., 2014). We posit that there is a need to analyze different kinds of state ownership and consider the distinction between the strategies developed by SSOEs and CSOEs. This is particularly important because recent studies of SOEs contain that these companies could have a “strategist perspective” (Wright et al., 2020), creating positive effects through patient capital, investing in industries with sunk costs, among others (Cuervo-Cazurra et al., 2014; Mariotti and Marzano, 2019). This perspective seems compatible with the characterization that we offer about CSOEs but these features do not share much resemblance with SSOEs’ resources nor international strategy. As we have shown, SSOEs can also have goods results in their internationalization process through another strategy and with a different set of resources.

The second puzzle is related to the particular dynamics of subnational phenomena. Both in management (Ma et al., 2016; Yang, 2018) and other disciplines (Gibson, 2005; Giraudy, 2013; Snyder, 2001), scholars have evaluated how different subnational factors influence development, democracy and FDI. Nevertheless, less has been written about the behavior of entities owned by subnational governments, such as SSOEs. How and why an SSOE behaves differently than other types of SOEs deserves more analyses, and we also contribute to this area. Other scholars have also affirmed that SOEs engage in higher and riskier endeavors abroad (Ramamurti and Hillemann, 2018) but we show that the type of ownership may produce different kinds of risks. In our case analysis, we find that SSOEs take greater risks in terms of diversification but have a more gradual growth while CSOEs may be more conservatives in the activities in which they expand abroad but they are willing to take more drastic decisions in their internationalization expansion and retraction. We expect that our research can help to enlighten these differences and behavior that recent studies have not explored enough.

While it has not been the core of this research, throughout our case tracing and analysis we have implicitly examined some key features of SOEs that can be used for future research projects and can also be contemplated concerning their managerial implications. Namely, SOEs are a potential prey for political and economic intervention since they can become sources of short-term funding for the governments in crisis times. We have mentioned how the Argentine government used YPF access to credit to sustain unrelated monetary policies (Serrani, 2013) and how the recent limits to Ecopetrol’s international experience seem to be related to the financial needs of the Colombian government. These findings can help ensure that managers and policymakers understand the long-term implications of their immediate decisions. Future research could explore what aspects would help increase SOEs’ autonomy. For example, the institutionalization of the rules that define how and why SOEs invest and the use and distribution of their profits or country-specific financial institutions' impact on SOEs' autonomy.

Further work could also examine the effect of corporate governance (criteria for board independence, board composition, board duties) in SSOEs and CSOEs that are partially owned by private investors (domestic and foreign). Another possible area of future research would be to study sectors’ specific features regarding regulatory frameworks and location-specific resources. Besides, our study will also serve as a base for future studies on deglobalization, risk aversion and corporate diplomacy in both SOEs and mixed-ownership enterprises.

Although comparative case studies are a valid methodology (Eisenhardt, 1989; Doz, 2011; Runfola et al., 2017; Tsang, 2013), we must acknowledge the limits of our research. As a comparative case study, our findings are constrained by each case unique evolution, and they are not widely generalizable. Our analysis of five SOEs from three countries might represent the internationalized SOEs dynamics in Latin America. However, the research question might require further exploration with more cases in this and other regions. Given that from an empirical perspective most of the studies on sub-nationalism and internationalization have focused on Asia, our research of an understudied region complements and also challenges the findings of existing literature. For instance, Jiang et al. (2010) assess that Chinese SSOEs are the ones with fewer respect for accountability and minority shareholder rights while we do not find that outcome in our analysis. Actually, an SSOE (INVAP) is the firm with fewer state intervention in the decision-making process. Additionally, some studies in China contend that efficiency in the operating control of SOEs operated by the federal government is greater than the ones managed by local authorities (Chen et al., 2009; Cull et al., 2017). Our findings in the Latin American perspective differ as SSOEs are the ones with a more careful pro-market behavior and the international retraction is more clearly noted in the cases of the three CSOEs (YPF and ENAP) and the one with a more sustained international presence is an SSOE (INVAP). While it goes beyond the goals of this article, we believe that rather than type of ownership, the previously mentioned regarding state behavior and needs might have an explanatory role in this process. Irrespective of this last matter, our research may be useful for inter-regional comparative analyses. We believe that our results are also valuable for further theory building in subnational phenomena and, specifically, different effects created by the type of state ownership.

Table 1.

Assumptions to be examined in the cases

Assumptions
CSO's have a non-market driven and sudden internationalization explained by political and ideological needs produced outside the firm
Due to the lack of international goals and subnational governments' limited budget, the international expansion of SSOEs follows a market-driven and gradual internationalization
The need for SSOEs to increase revenues encourages them to diversify into new sectors abroad
The central states are conservative regarding strategies disrupting their initial geopolitical goals, and thus they do not usually expand into new sectors abroad

Source: Authors’ own construction

Table 2.

A Summary of key indicators of the studied cases

SOE	Type of state ownership	Logic of internationalization	Characteristics of internationalization
INVAP	Subnational	Market-driven	Gradual and diversified
EPM	Subnational	Market-driven	Gradual and diversified
YPF	Central State	Non-market reasons in key moments	Drastic leapfrogging and concentrated
ENAP	Central State	Non-market reasons in key moments	Moderated leapfrogging and concentrated
Ecopetrol	Central State	Non-market reasons in key moments	Moderated leapfrogging and concentrated

Source: Authors’ own construction

Table 3.

Succinct profile of the studied companies

SOE	Revenues (in USD millions, 2019)	Employees (2019)	Business units operating internationally1	Exports/revenues abroad (2019)
INVAP	181	1,301	4 out of 4	35%
EPM	961	14,046	5 out of 7	16%
ENAP	7,628	3,354	1 out of 2	2%
YPF	11,613	19,072	N/A	N/A
Ecopetrol	19,102	10,087	1 out of 3	51%

Notes:

¹This estimation is based on the current activities abroad in relation to the current business units that each company has. In the case of EPM the company had wind energy abroad but it no longer has. Thus, we estimate five business units abroad instead of the six that used to have. For a brief period ENAP had two activities abroad but it currently has one. YPF currently does not have assets abroad so it makes the value in this cell unavailable. As a reference, YPF – through affiliates in which is a large shareholder and its own businesses – has participation in seven different activities. When it was internationalized, it only did so in oil the oil sector

Source: Own elaboration based on data from companies’ financial statements and other secondary sources

Table 4.

Main findings (construct and outcomes)

Construct	Value	Key intermediating factor	Effects on speed and focus of internationalization process
Type of State ownership	National	Existence of non-market factors driving internationalization (i.e. international policy agenda, resources, etc.)	Leapfrogging (i.e. large sporadic acquisitions abroad)
			Concentration in the original internationalized sector
	Subnational	Market orientation as main driver of internationalization	Gradual sequence/market-driven internationalization
			Search for additional sources of competitiveness (i.e. diversification)

Source: Authors’ own construction

Table 5.

Key assumptions, relevant highlights and quotes

Assumptions related to speed and focus of internationalization process	Relevant highlights obtained from cases	Significant quotes from key SOEs’ executives
Speed CSOEs: Non-market driven and sudden internationalization explained by political and ideological needs produced outside the firm	The lack of oil in Chile has been a key driver for the international expansion of ENAP. The central State also defined the reversal in the internationalization process The Argentine State has widely and repeatedly used YPF for political purposes. The international retraction was a consequence of the sale and then nationalization of the company Ecopetrol’s recent international retraction is connected to the State urge to get funds	“ENAP has been able to expand and get a debt significantly greater than its EBITDA because it has the support of the Chilean State ([…]) the indirect subsidy to the company comes through its CAPEX that is never paid” (Cruzat, interviewed on 3rd December 2018) “During the time that Ecopetrol internationalized there was a reduction in the national reserves of oil and it would become difficult in the middle-term to supply the local needs … it was this fact what encouraged the internationalization of the company” (Gutierrez Pemberthy, interviewed on 27th July, 2021)
Speed SSOEs: The international expansion follows a market-driven and gradual internationalization	INVAP’s first international experience was in the 1970s and it has continued in a gradual organic mood. It has done so without State financial support EPM developed a master plan for their internationalization. First experiences have been used to take additional knowledge for the next acquisitions	INVAP works and has a culture similar to a private firm ([…]) one of the most critical part of INVAP is their sales department, which operates like a private firm ([…]) none of the other SOEs under my jurisdiction are like that (Gadano, interviewed on 11th September 2018) EPM developed a master plan for their organic expansion at the beginning of the international expansion that it has been followed through (Jaramillo, interviewed on 18th December, 2018)
Focus SSOEs: Need to increase revenues encourages diversification abroad	INVAP has three different activities abroad (nuclear reactors, medical devices, satellites) EPM has had four different activities abroad (water treatment, electricity, alternative energy and telecommunications)	We expect EPM to develop new activities that complement the ones we have (Jaramillo, interviewed on 18th December, 2018)
Focus CSOEs: The Central States are conservative regarding strategies disrupting their initial geopolitical goals, and thus they do not usually expand into new sectors abroad	Both ENAP, Ecopetrol and YPF have had all their international investments in the oil sector. Almost exclusive focus in upstream oil production	‘During the 1990s all of YPF’s focus was in consolidating the acquired assets abroad’ (Sturzenegger, interview on 27th January, 2021) “The diversification strategy of Ecopetrol consists on selling our oil to different regions … we have created more value thanks to the diversification of the places we sell” (Barcena, interview on 6th July, 2021) “The priority in the internationalization has been to increase the oil reserves … for Ecopetrol it does not make much sense to expand on other sectors such as midstream or downstream” (Gutierrez Pemberthy, interviewed on 27th July, 2021)

Source: Authors’ own construction

Notes

1. Unfortunately, to the extent of our knowledge there is no SSOE in Chile and this is why we have not included a sixth case from this country.

2. While the difference is not that large, during the expansion period the number of years in which profits are reinvested is 35% greater than in the reversal one.