Institutional Logics and the Internationalization of a State-Owned Enterprise_ Evaluation of International Venture Opportunities by Telecom Finland 1987–1998

We contribute to the research on internationalization of state-owned enterprises (SOEs) by studying the coevolution of state governance of SOEs and SOEs’ evaluation of international venture opportunities during a shift in dominant institutional logic from state to market logic. Using a novel digital historical method to study Telecom Finland, we argue that as state governance mechanisms change due to a logic shift, rationales underlying SOEs’ internationalization can significantly change and impact SOEs’ geographical and partner preferences. However, a logic shift also affords SOEs significant influence over the formation of new state governance policies under the new dominant logic.

In this paper, we adopt the institutional logics perspective and focus on a fully-owned SOE during the transformation of Western European state capitalism during the 1980s to 1990s from state logic, i.e., SOEs as strictly-controlled extensions of the public bureaucracy operating mostly in markets considered natural monopolies, to market logic, i.e., SOEs as for-profit corporations in competitive markets (Megginson & Netter, 2001;Musacchio et al., 2015). Under this framing, we investigate how this societal-level logic shift influenced the rationales underlying SOE internationalization. Due to the interconnected relationship between the state and its SOEs (e.g., Bass & Chakrabarty, 2014;Choudhury & Khanna, 2014;Rodrigues & Dieleman, 2018), we must pay attention to the coevolution of the state governance of SOEs, which we define as the extent of state control and how this control is exercised, and how SOEs internationalize (Cantwell, Dunning, & Lundan, 2010;Cuervo-Cazurra, 2015). Coevolution refers to the multilevel, multidirectional causality, nonlinearity, positive feedback, and path and history dependence properties between the state governance of SOEs and SOE internationalization (Lewin & Volberda, 1999). Thus, we ask the following question: How does a shift in dominant institutional logic from state to market logic affect the coevolution of the state governance of SOEs and SOEs' evaluation of international venture opportunities?
We define institutional logics as supraorganizational historically embedded belief systems that provide rationales for strategic actions, such as internationalization decisions (Friedland & Alford, 1991;Greve & Zhang, 2017;Thornton & Ocasio, 1999. Extant research posits that state and private control infuse state and market logics into SOEs and that the extent of state control based on factors, such as share the balance between the different logics (Bruton, Peng, Ahlstrom, Stan, & Xu, 2015;Cuervo-Cazurra et al., 2014;Mariotti & Marzano, 2019). However, the decisions of a fully-owned SOE that operates in a competitive market are necessarily based on a constellation of logics as the SOE interacts with other market participants, such as suppliers, buyers and competitors, through market mechanisms (Durand & Thornton, 2018;Goodrick & Reay, 2011;Greenwood, Raynard, Kodeih, Micelotta, & Lounsbury, 2011). Thus, to truly understand the balance and dynamics between different institutional logics, we must investigate decision-making within an organization and the rationales used to evaluate participation or nonparticipation in international ventures (Greve & Zhang, 2017).
The home context of an SOE is an important determinant of the balance between state and market logics in SOE internationalization due to the historically contingent differences among institutions across countries that result in different policies and abilities to govern SOEs (Mariotti & Marzano, 2019). However, our understanding of the home context and its influence on the rationales underlying SOE internationalization remains static and superficial (Estrin et al., 2016;Grøgaard et al., 2019;Liang, Ren, & Sun, 2015). Additionally, the extant research has neglected the role of SOEs in shaping the institutions that govern them (Cuervo-Cazurra, 2015;Cuervo-Cazurra et al., 2019). In coordinated market economies 1 , such as Finland, SOEs and their international strategies tend to be jointly coordinated among business, labor, and government interests (Mariotti & Marzano, 2019). Thus, SOEs from such a context need to balance state and market logics to fulfill the requirements of various interests, but the context also affords them a significant role in shaping the institutions that govern them. This context is, therefore, suitable for observing the coevolution of the state governance of SOEs and SOEs' evaluation of international venture opportunities.
Hence, we chose to conduct an in-depth historical case study of Telecom Finland (TF), a state-owned telecommunications provider in Finland. Our study is based on more than 54,000 pages of primary sources from 1987 to 1998 and traces the evolution of TF from a government department to a public corporation in 1990 and a fully stateowned limited liability company in 1994. We concluded our study when TF was partially privatized and publicly listed in November 1998; i.e., our case organization was fully state-owned throughout the period of interest. During this period, TF's institutional environment significantly changed with a shift in dominant logic from state to market logic, which affected the institutional pressures, abilities, and opportunities for internationalization. As institutional logics are historically embedded (Friedland & Alford, 1991;Thornton & Ocasio, 1999 and the relative strength between state and market logic in SOEs has shifted back and forth , a historical approach allows us to analyze the gradual changes in the state governance policies over TF and TF's evaluation of international venture opportunities in a wider historical context (Jones & Khanna, 2006). Additionally, a single case study allows us to explore in depth the rationales used in decision-making regarding international venture opportunities (Siggelkow, 2007). Finally, considering a case of a fullyowned SOE, we can attend to changes in the balance of institutional logics beyond an overly simplified notion of state control based on ownership (Bruton et al., 2015;Grosman, Okhmatovskiy, & Wright, 2016;Liang et al., 2015).
We make three contributions. First, we contribute to the literature concerning SOEs in the international context through the institutional logics lens by showing how a shift from state to market logic influences the relationship between the state and SOE internationalization (Thornton, 2002;Thornton & Ocasio, 1999). Furthermore, we argue that 1) the dilution of state control is neither required nor sufficient for market logic to overcome state logic; rather, there must be changes in the rationales guiding strategic decision-making (Greve & Zhang, 2017) and 2) a shift in dominant logic affords SOEs influence over how they are governed by the state (Cuervo-Cazurra, 2015;Cuervo-Cazurra et al., 2019). Second, we contribute to the SOMNC literature by showing that although SOMNCs can internationalize to the same extent or even more than their privately-owned counterparts (Estrin et al., 2016), different rationales underlie their internationalization, and these rationales might significantly change during long internationalization processes , influencing geographical and partner preferences. Finally, we answer multiple calls in the international business (IB) literature for more historical research that can uncover contextually embedded relationships in internationalization (Buckley, 2016;Burgelman, 2011;da Silva Lopes, Casson, & Jones, 2019;Jones & Khanna, 2006;Verbeke & Kano, 2015), which can help shed light on the contemporary phenomenon of de-globalization when political objectives might become dominant again Witt, 2019).

Role of SOEs in the Transition from Old to New State Capitalism
The market-led multinationalization of SOEs has been considered to contradict the old state capitalism logic (e.g., Anastassopoulos, Blanc, & Dussauge, 1987;Vernon, 1979). In the old state capitalism logic, SOEs were created to address social and political goals or market failures in the home market, such as in the case of public goods, positive or negative externalities, and natural monopolies (e.g., Cuervo-Cazurra et al., 2014;Megginson & Netter, 2001;Millward, 2011;Musacchio et al., 2015;Toninelli, 2000). As such, SOEs' performance cannot be evaluated solely based on financial metrics and compared to their mainly profit-seeking privately-owned counterparts (Aharoni, 2000;Millward, 2011). The ever-changing goals of SOEs are diverse and intangible, including sustaining employment, supporting remote regions, and keeping prices low, often decreasing economic performance (Vernon, 1979). SOEs also face major principal-agent challenges due to the delegation of the monitoring of SOEs by society to politicians, who can use SOEs for political gain and extract financial rents for personal benefit (Inoue, Lazzarini, & Musacchio, 2013).
In many OECD countries, especially in Western Europe, prior to the 1980s, this agency problem was addressed with the following two basic doctrines: separation of the public and private sectors and procedural rules limiting the discretion of public officials and SOE managers, especially over staff, contracts, and money (Dunleavy & Hood, 1994;Hood, 1995). Since the late 1970s, the international diffusion of neoliberal policies advocating for less government intervention and the introduction of market mechanisms to the public sector resulted in broad reform of public sector bodies commonly termed new public management (Lee & Strang, 2006). This process reversed the former doctrines with the removal of the separation between the public and private sectors and shift from procedural rules to management-by-results (Hood, 1995), which influenced the laws, institutions, practices, and regulations governing SOEs (Megginson & Netter, 2001) and shifted public service providers toward private governance modes (Grosman et al., 2016). This process resulted in new varieties of state capitalism in which SOEs of various ownership levels began to operate in international and competitive markets (Cuervo-Cazurra et al., 2014;Musacchio et al., 2015).

Institutional Logics in the SOE Context
We frame the transformation from old to new state capitalism and the introduction of market mechanisms to the governance of SOEs as a gradual shift in dominant logic from state to market logic. As such, this 1 See literature concerning varieties of capitalism for a list of countries categorized as coordinated market economies and a broader typology e.g., (Fainshmidt et al., 2018;Hall and Soskice, 2001). study continues a stream of historically embedded institutional logics studies that investigate the increasing prevalence of market logic in different contexts, such as the higher education publishing industry (Thornton & Ocasio, 1999), finance (Lounsbury, 2002), M&As of Chinese firms (Greve & Zhang, 2017), the internationalization of Chinese SOEs (Tang, 2019), and public administration (Meyer & Hammerschmid, 2006;Townley, 2002).
Under state logic, the legitimacy of SOEs stems from fulfilling social and political objectives set by the government (Rodrigues & Dieleman, 2018;Thomann, Lieberherr, & Ingold, 2016), and resources are channeled through the state accordingly (Greve & Zhang, 2017). Under market logic, the legitimacy of SOEs stems from their financial performance and relative market position, which guide managerial attention to growth, shareholder value maximization, and the acquisition of resources and capabilities (Greve & Zhang, 2017;Newenham-Kahindi & Stevens, 2018;Rodrigues & Dieleman, 2018;Thomann et al., 2016;Thornton, 2002;Thornton & Ocasio, 1999). In the international context, management practices tend to be aligned with global standards, and the time horizon of investments tends to be shorter (Newenham-Kahindi & Stevens, 2018).
The institutional logics perspective evolved from the neoinstitutional perspective's strong emphasis on isomorphism and conformity to institutional expectations, leaving limited room for theorizing heterogeneity across institutional contexts (Aguilera & Grøgaard, 2019). In contrast, the logics perspective views society as an inter-institutional system of societal sectors, such as the market, state, democracy, family, and religion, and each sector has its own logic providing rationales for strategic action (Friedland & Alford, 1991;Greve & Zhang, 2017;Thornton & Ocasio, 2008).
The logics perspective takes a multilevel view of how societal logics influence the behavior of organizations and individuals (Friedland & Alford, 1991;Thornton & Ocasio, 2008), such as how changes in the balance of societal-level logics impact intra-organizational decisionmaking through changes at the international and national levels (in our case the European environment, the socio-political environment of Finland, and the Finnish telecommunications market). Thus, institutional logics are locally instantiated and enacted within the historical context, allowing heterogeneity in behavior across institutional contexts (Besharov & Smith, 2014;Thornton & Ocasio, 2008), which is central to IB research employing institutional perspectives (Aguilera & Grøgaard, 2019;Hotho & Pedersen, 2012). For example, Greve and Zhang (2017) and Tang (2019) focus on the logic shift from state socialism to market capitalism in the Chinese context since the 1980s, which is similar to the shift in the balance of societal sectors in our case, but due to the different contexts, the emerging logics and their enactment by actors differ. In comparison, variation in institutional arrangements consistent with different national contexts has been the core of IB research employing institutional economics (e.g., Cantwell et al., 2010) and comparative capitalism perspectives (e.g., Jackson & Deeg, 2008;Witt & Jackson, 2016).

Internationalization of SOEs
State ownership has been found to be both a constraining and an enabling factor of firm internationalization (Estrin et al., 2016). On the one hand, SOMNCs face stronger institutional pressure in host countries, e.g., by raising national security concerns, and adapt by taking lower equity stakes to enhance their external legitimacy (Meyer, Ding, Li, & Zhang, 2014). SOMNCs also face stronger institutional pressure from the home country due to dependence on the government for resources (Choudhury & Khanna, 2014;Cui & Jiang, 2012). Internationalization can be a way for SOEs to gain resource independence from the state, which, in turn, could trigger the state to tighten its control over SOEs (Rodrigues & Dieleman, 2018). On the other hand, a strong diplomatic relationship between the home and host governments can mitigate political risks in the host country (Duanmu, 2012(Duanmu, , 2014García-Canal & Guillén, 2008), enhance external legitimacy and alleviate the liability of foreignness (Li, Meyer, Zhang, & Ding, 2018). States can also drive SOEs to internationalize (Liang et al., 2015) and use SOEs to extract resources for exploration (Bass & Chakrabarty, 2014), especially autocratic states (Clegg et al., 2018).
Overall, SOMNCs might be able to benefit from internationalization more than privately-owned firms due to the specific assets available to SOEs, especially in the case of former state monopolies, such as our case (Benito, Rygh, & Lunnan, 2016). SOMNCs can pursue internationalization to the same extent as their privately-owned counterparts (Estrin et al., 2016). However, recent studies concerning the relationship between state ownership and internationalization provide mixed results. For example, one study proposes that SOEs fully or majority owned by the state are more inclined to internationalize (Cuervo-Cazurra et al., 2014), whereas another recent study found that firms with a medium level of state ownership internationalize the most (Kalasin et al., 2019). In contrast, Tihanyi et al. (2019) found that the extent of state ownership decreases while SOEs' political connections increase SOE internationalization. Thus, the mechanisms underlying the internationalization of SOEs remain unclear (Liang et al., 2015). Variations in the type and level of state control create different types of institutional pressures for internationalization (Wang, Hong, Kafouros, & Wright, 2012), and governance reforms can impact why and how SOEs internationalize (Liang et al., 2015). As states decrease their influence over SOEs, social and political objectives for internationalization decrease, and the focus shifts to financial performance (Cuervo-Cazurra et al., 2014). While most studies associate state ownership stake and executive political connections with state control over SOEs, states and SOEs can also adopt corporate governance structures to limit state involvement (Grosman et al., 2016), such as in our case.

METHODOLOGY
As our aim is to elaborate upon existing theory concerning how a shift in dominant institutional logic impacts SOE internationalization through changes in decision-making rationales, we conducted an abductive historical case study (Ketokivi & Choi, 2014;Mantere & Ketokivi, 2013;Rowlinson, Hassard, & Decker, 2014). First, our abductive approach is particularly suitable for confronting extant theory as we pay attention to contextual particularities and empirical tendencies while theoretically interpreting and refining our theoretical understanding (Ketokivi & Choi, 2014;Mantere & Ketokivi, 2013;Welch, Piekkari, Plakoyiannaki, & Paavilainen-Mäntymäki, 2011). Second, single case studies enable in-depth inquiry into a phenomenon, which can be used to provide a "conceptual contribution" by elaborating upon constructs and their underlying relational mechanism over time (Siggelkow, 2007). Furthermore, our case is unusually revelatory, and we have unique research access to archival sources, enabling an indepth understanding of decision-making rationales (Eisenhardt & Graebner, 2007). Finally, we use a historical approach to reveal temporal variance, path dependency, and the effect of context and contingencies over time (Buckley, 2016;Jones & Khanna, 2006;Vaara & Lamberg, 2016) and, thus, answer Burgelman's (2011) call for research utilizing historical longitudinal qualitative studies that establish a link between narrative and the reductionist approach in qualitative IB research.

Research Setting
Some of the largest MNCs worldwide emerged from previously state-owned telecommunications monopolies (Clifton, Díaz-Fuentes, & Revuelta, 2010). Since the 1980s, the telecommunications market has undergone major regulatory, organizational and technological transformation globally, including deregulation, the privatization of former national monopolies, and the emergence of new mobile and digital technology (Graack, 1996). As a result, competition combined with new types of consumer services led to rapid market growth in the 1990s (ITU, 2002, p. 19), and the telecommunications market structure changed from detached national markets controlled by a single stateowned monopoly to a global market of MNCs.
In Europe, the changes in the telecommunications market structure were closely linked to the economic integration process, and a main element was the agreement to open the market to competition in the European Union by 1998 (e.g., Clifton et al., 2010;Thatcher, 2001). In Finland, the liberalization process started in 1987, and in the early 1990s, the Finnish telecommunications market was found to be among the most competitive markets in Europe (e.g., Graack & Elixman, 1999). Competition for mobile subscriptions began as early as 1991, and all major product groups were liberalized by 1994; subsequently, the government promoted competition within the existing networks. For example, after August 1996, network operators were required to open their connections to "virtual operators" who did not build their own infrastructure but leased it from network operators to sell consumer services.
The unique market structure in Finland, consisting of state-owned TF and a group of private telephone companies operating local telephone monopolies in the most lucrative urban regions, contributed to the quick emergence of a competitive market (Graack & Elixman, 1999). In Finland, private telephone companies formed a loose interfirm structure that challenged TF in the domestic market. However, these companies were unable to concentrate their resources on internationalization until the second half of the 1990s (Nevalainen, 2018). In comparison, in Sweden, challenger Comviq (later part of TELE2) competed intensively with Televerket (the Swedish telecom monopoly) to gain access to the telecommunications market and gradually gained market share but at a slower pace than the challengers in Finland Geissinger, Laurell, Sandström, Eriksson, & Nykvist, 2019). In Europe, including Finland and Sweden, internationalization in the telecommunications sector was led by stateowned incumbents, while new entrants focused on their respective domestic markets until the mid-1990s (Alonso, Clifton, Díaz-Fuentes, Fernández-Gutiérrez, & Revuelta, 2013).
Early exposure to liberalization has often been thought to have benefited incumbent operators by giving them an advantage over others in the internationalization process (Sarkar, Cavusgil, & Aulakh, 1999). However, incumbent telecommunications monopolies from countries that were slower to liberalize could leverage their monopoly rents and political connections in their internationalization efforts (Clifton, Comín, & Díaz-Fuentes, 2011). In some cases, liberalization did not result in real competition as former monopolies could block smaller entrants (Waverman & Sirel, 1997). In our case, the gradual liberalization process that started early combined with established potential local competitors encouraged TF to proactively develop its organization (Nevalainen, 2017) and seek new markets to replace the lost monopoly profits. Coincidentally, neighboring markets opened up opportunities.
The collapse of the Soviet Union in the early 1990s created a vacuum in Eastern Europe that offered business opportunities for internationalizing telecoms (Martin, 2002). For TF, the area was especially important in the beginning of its internationalization process as the Soviet Union was Finland's largest trading partner with up to 20 percent share of exports in the 1980s (Gorodnichenko, Mendoza, & Tesar, 2012), and expectations of economic growth in the region were high. TF established operations in Estonia and Northwest Russia even before the Soviet Union formally ceased to exist as the Soviet government encouraged the establishment of joint ventures with Western companies, especially in the telecommunications sector (Borg & Emmert, 1989). Subsequently, TF expanded beyond Eastern Europe to several markets, such as Turkey, Hong Kong, and Germany. The change in dominant logic influenced the choice of geographical markets and partners, but the entry mode remained largely unchanged.

Data Collection
This study is based on extensive archival data. We obtained full access to TF archival material from 1981-1998 in the National Archives of Finland and Telia Finland (previously TeliaSonera Finland, Sonera, TF, and PTL Tele), a Finnish subsidiary of Telia Company. This archival material mainly consisted of board minutes, management minutes, strategy documents, correspondence, and regulatory documents; this material along with annual reports and internal magazines from several libraries formed the basis of our data (Table 1).
All archival material was originally in paper format, and we digitized such material over a period of two years. We employed an iterative approach through which sources led to new sources, and we collected an extensive archival set consisting of over 54,000 pages from 95 archive folders. With the help of four research assistants, all original sources were first photographed and then digitized into machine readable documents using optical character recognition software. Then, with the help of research assistants, we added the documents to our relational database (cf. Murmann, 2010) and coded the documents with relational data regarding individuals, organizations, times, and international ventures. Through this digital history method, we were able to examine TF's evaluation of international venture opportunities from its first international venture in 1987 to when TF was publicly listed and partially privatized in 1998. This approach allowed us to gain a new understanding of the contextual embeddedness and historically contingent nature of the internationalization process (Vaara & Lamberg, 2016).

Analytical Process
We conducted our analysis in three distinct phases by combining process study methods (Langley, 1999) and a digital history approach exposing primary archival sources to constant source criticism and triangulation (Buckley, 2016;Kipping, Wadhwani, & Bucheli, 2014). During each phase of our analysis, we ensured the credibility of our findings with the following steps. First, we documented our entire analysis process within our relational database. With this audit trail, we are able to trace our findings back to the original archival sources. Second, we searched all documents in our database with several keywords for each venture to ensure that we identified all types of relevant sources. We linked all relevant archival sources to each venture and subsequently coded all identified decision-making rationales for each source. This approach enabled us to triangulate our codes across different source types created for different audiences and purposes . Third, to ensure intercoder reliability in this interpretive process, all authors cross-checked the venture decisionmaking rationales for each venture.
In our theorizing, we gradually progressed to a higher level of abstraction while iterating our emerging findings along with archival sources and contrasted this abstraction to the extant general theoretical understanding. Our theorizing process closely followed what is commonly known as abductive reasoning (Dubois & Gadde, 2002;Mantere & Ketokivi, 2013). Abductive case studies are typically used to investigate, elaborate and refine existing theoretical frameworks based on novel empirical insights and particularities (Ketokivi & Choi, 2014;Mantere & Ketokivi, 2013). During this process, we allowed simultaneous iteration between extant theory and our empirical context (Dubois & Gadde, 2002) by seeking multiple competing theoretical interpretations for the empirically-identified tendencies (Mantere & Ketokivi, 2013). This approach enabled us to gradually build upon extant theory in a way that reconciles our empirical insights with the contextual particularities stemming from our in-depth inquiry of decision-making rationales (Ketokivi & Choi, 2014). Next, we describe the three phases of our analysis leading to our theoretical model in greater detail.
During the first phase, we focused on 764 top management team meeting minutes. We separated each management team meeting minute issue between 1981 and 1998 using the relational database and inductively coded each meeting issue related to internationalization. During this process, of 3,097 issues in the database, 357 meeting minute issues broadly related to the international activities of the case organization were identified. Then, we interpreted each identified management team meeting issue and its relevant attachments and coded the issues with empirical codes, such as international cooperation, joint venture, venture capital, foreign currency loan, and international traffic.
At this point, the prominent role of international ventures emerged from the data. We chose to focus on the 199 minute issues related to international ventures with the first venture appearing in 1987. This approach resulted in 69 identified ventures, including ventures that went into operation, those in which TF refused to participate and those for which TF lost the license bid and were consequently shut down. We elaborated upon these initial empirical findings based on the extant understanding of how SOEs internationalize. We noticed that TF's internationalization substantially changed during our period of interest, even though the state ownership stake remained the same. This initial insight led us to focus more closely on the rationales used in decisionmaking with regard to international venture opportunities.
During the second phase, we coded each venture in the database separately and searched all documents in the database using particular venture keywords. All identified archival sources related to a particular venture were coded in relation to the venture. This approach enabled us to triangulate between various source types created for different purposes and audiences , such as management team meeting minutes and their attachments, annual reports, draft decisions, venture business plans, venture contracts, strategy documents, board minutes, correspondence, and internal magazines. On average, there were 10 sources per venture, with a median of 7.5 sources, a maximum of 39 sources, and a minimum of 1 source.
Then, we interpreted all archival sources related to each venture and gradually formed an understanding of the decision-making rationales underlying the establishment of each venture and the development of the ventures. Furthermore, the rationales for participating in a particular venture were linked to each source and venture. Through this analysis, we identified the first-order rationales used to evaluate the ventures and grouped the rationales in second-order themes ( Fig. 1). Gradually, a pattern emerged from the data showing that there was a shift in the rationales used to evaluate international venture opportunities. Then, we aggregated the changes in rationales into the following two dimensions: national and multinational rationales. In contrast to the extant understanding of SOE internationalization, which emphasized the state ownership stake and executive political connections, we noticed the importance of shifting state governance mechanisms influencing the decision-making rationales with regard to venture opportunities and how TF's decision-making rationales gradually coevolved with the changing state governance mechanisms. This discovery led us to investigate the coevolving relationship between these two aspects in-depth and focus on contextual elements that could explain this phenomenon.
Thus, during the final phase of the analysis, we conducted temporal and cross-case comparisons of all ventures. We separated each venture by the business type, entry mode, geographical location, entry and exit rationales, and partners and drew timelines to temporally bracket the ventures and decisions in relation to the contextual development. This approach enabled us to form a list of the rationales and partner and market preferences in each venture across the period of interest. Then, we engaged in case narrative writing to produce the contextually Z. Cheung, et al. Journal of World Business 55 (2020) 101140 embedded internationalization process of TF, including TF's international venture opportunities in relation to the development of the governance of TF, the Finnish telecommunications market, the Finnish socio-political environment, and the European environment. Finally, we progressed to our final theoretical model in which the key concepts were rooted in the data but constructed theoretically (Alvesson & Kärreman, 2007;Mantere & Ketokivi, 2013) by positioning our conceptual findings and results based on the extant theoretical understanding of institutional logics in the SOE internationalization context. We compared our conceptual findings via a cross-case comparison and contextually embedded case narrative. This approach enabled us to identify the initial coevolution mechanism between the state governance of SOEs and TF's rationales used to evaluate international venture opportunities. We noticed that the change in the evaluation of international venture opportunities coevolved with the partner and geographical market preferences. Furthermore, the logic guiding TF decision-making coevolved with the logic guiding state governance of SOEs. After several rounds of iterations between our findings and data, and positioning our concepts and results against multiple theoretical interpretations, we noticed that the evolving relationship between the state governance of SOEs and TF's decision-making rationales were related to the gradual shift in dominant institutional logic from state to market logic.

FINDINGS
First, we present our findings regarding how the shift in the dominant institutional logic from state to market logic in the telecommunications market changed the relationship between the state and TF. Second, we show the change in rationales used to evaluate international venture opportunities in detail. Third, we show how the changing rationales impacted existing markets and the choice of new geographical markets. Fourth, we show how changing rationales impacted existing partner relationships and the choice of new partners.

Shift in Institutional Logics in the Telecommunications Market
The shift in the dominant institutional logic from state to market logic influenced the telecommunications market at an international market, state, and organizational levels ( Table 2). Fig. 2 summarizes the key processes that created international opportunities and that pressured and enabled TF to internationalize. The market structure changed both internationally (Clifton et al., 2011;Sarkar et al., 1999;Thatcher, 2004) and in Finland (Nevalainen, 2017;Turpeinen, 1996) by moving from local monopolies to competition between public and private actors that are regulated similarly (e.g., through equal telecommunications licensing conditions). Especially in Finland, where the market was divided between TF and local telecommunications providers, the threat of both local and international competition was imminent ( Fig. 2: 1). Private telecommunications providers established strong positions in urban areas and were equipped with the resources to penetrate TF's profitable monopoly market areas, such as their mobile and long-distance telecommunications businesses.
Internationally, due to the relatively small size of the dispersed Finnish market, the top management of TF considered the Finnish market easy to penetrate for large foreign telecoms. Specifically, in the late 1980s, the Finnish government was already committed to the European Community's plan to create a common telecommunications market ( Fig. 2: 2a, 2b, 3a, and 3b). Furthermore, given the severe recession in Finland in the early 1990s due to the banking crisis reinforced by the collapse of the Soviet Union, TF had to search abroad for opportunities ( Fig. 2: 4a and 4b). During the same period, the Finnish government encouraged TF to help modernize the underdeveloped telecommunications infrastructure in Russia and the Baltics as it was politically important to support Finnish trade to the East and form ties between the East and the West after the collapse of socialism in 1989 ( Fig. 2: 5 and 6). Subsequently, liberalization at the European Union level opened up opportunities in the Western European market (Fig. 2: 7) and exposed the Finnish market to international competition ( Fig. 2: 8a and 8b). In turn, internationalization drove the top management of TF to become proactive in developing its organization, which required governance reforms that would give TF more independence from the state.
"American, Japanese and European telecoms are investing billions in Eastern European networks, the Baltic states, Russia and the former Soviet Union. The danger is that when the domestic telecom companies focus on competing with each other, one or more strong foreign telecom -Memo on TF's business in the 1990s, 1992 However, before 1990, TF was a government agency under bureaucratic resource control by ministries and the Finnish Parliament, which hampered TF's responses to changes in the environment. For example, the budget of TF heavily depended on government financing and, thus, experienced constant resource scarcity that rendered longterm investments challenging, and each international investment required government approval regardless of its size. The state also imposed public service obligations on TF, such as providing telecommunications infrastructure to rural areas and employment quotas, which required cross subsidization between profitable business areas and public service obligations to meet the state mandated budget. Human resource management was under especially strict control as civil service employment limited hiring and government steering was at the level of individual employees. As a result, there was a resource mismatch with excess resources in rural areas, staff with lower educational levels and the lack of resources needed for international operations.
"Under the current organizational model of Post and Telecom, operations and finances are organized like those of a state agency. This does not give enough room for the quick and effective decision-making needed for business." -Post and Telecom Act Working Group Report, 1987 The bureaucratic resource control faced increasing pressure from two types of actors as market liberalism became more prevalent in the political environment and was influenced by the examples of the US, UK, and Sweden, which were important points of comparison for Finnish policy makers and industries. First, policy makers, especially government officials and the newly elected pro-liberalization government in 1987, actively promoted the reorganization of public sector bodies (Fig. 2: 2b and 3b). Second, the politically appointed incumbent management, which was increasingly influenced by the market logic promoted by external advisors, such as management consultants, investment bankers, and business schools, engaged in significant political activity to gain decision-making independence from the state (Fig. 2: 9). This process resulted in a gradual shift from strict resource control by the state to management-by-results.
The process was punctuated by the following two milestones: "public corporation reform" in 1990, which separated TF's budget from the government budget, and incorporation in 1994, which rendered TF a fully state-owned limited liability company. The public corporation reform, which transformed TF to a business and government agency hybrid organization, provided the top management of TF significant flexibility to compete in the domestic market and to some extent in the international market ( Fig. 2: 10). However, debt financing and subsidiary management, which were critical for international operations, still required time-consuming political approvals. Thus, by 1991, the top management of TF started to prepare and lobby for incorporation with the hope of obtaining an equal status to its privately-owned competitors ( Fig. 2: 11).
The shift to management-by-results signified that instead of strict resource control, the state set targets for TF, such as profit and repatriation (i.e., the amount needed to be paid to the state) targets. This shift in how the state enforced control over TF impacted all aspects of the organization with three major implications that enabled TF to shift towards market logic. First, the top management of TF gained the right to nominate managers, and consequently, professional managers were favored over managers with political backgrounds. Second, TF gained discretion over internationalization decisions, and in 1991, the management of international ventures was centralized to a unit in Belgium operated by Finnish and foreign professional managers. Third, direct reporting to policy makers was changed to direct reporting to a professional board with directors mostly from the private sector who emphasized shareholder value over public service goals (Fig. 2: 12).
In summary, the international diffusion of neoliberal policies reflecting market logic created international opportunities and drove the Finnish socio-political environment and government to adopt promarket reforms. This shift resulted in an increasingly competitive telecommunications market in Finland, which pressured TF to seek international opportunities, and changed how the state governed its SOEs, which enabled TF to internationalize. The state and TF coevolved and mutually reinforced the dominance of market logic over state logic. The government pressured TF to internationalize for political reasons, and TF pressured the state to implement reforms that could enable TF to

From National to Multinational Rationales
We found that as the dominant institutional logic shifted from state logic to market logic, the strategic rationales used by TF in evaluating international venture opportunities shifted from national to multinational rationales. National rationales refer to evaluation criteria rooted in the home country ranging from wider social and political objectives to considering how an international venture could benefit the home market in Finland. Therefore, national rationales are closely related to the state logic, which implies public services and resources channeled through state control with an emphasis on political and social objectives (Greve & Zhang, 2017;Thomann et al., 2016) in the international context. We identified three types of national rationales: defensive, home industry, and national champion rationales (Table 3).
Multinational rationales refer to evaluation criteria emphasizing individual ventures' ability to create financial value and develop capabilities for the organization overall instead of benefitting the home market. These rationales are closely associated with the market logic, which implies managerial attention to growth, financial performance, and maximizing shareholder value in the international context ( Thornton & Ocasio, 1999). We identified two types of multinational rationales: growth and portfolio strategy (Table 3). Additionally, baseline financial rationales, such as profitability, were present to varying degrees throughout the focal period, i.e., ventures have to be profitable even if they have strong defensive rationales supporting state logic or growth rationales supporting market logic.

"The activities by the Swedes [Televerket] and the Finns [TF] in Estonia
Telephone are not development aid. […] It is not the American way of expecting payback in six months either but maybe in ten years." -Interview with Estonia Telephone's CEO in TF's internal magazine 'Tietolinja' 15/1993 The shift in rationales was gradual and punctuated by the two previously discussed changes in formal governance and organizational form, i.e., the public corporation reform in 1990 and incorporation in 1994. The increasing influence of market logic on TF's top management was gradual as they adopted business-like management practices, such as various strategic and scenario planning systems, and implemented organizational changes, such as the matrix organization and profit centers, within the limits of the existing mode of state governance of SOEs and organizational form. Additionally, TF's top management sought governance and organizational form reforms that would grant them more independence from the state to implement strategies that were increasingly aligned with the market logic. The public corporation reform enabled TF to pursue internationalization primarily based on national rationales as the state still had significant control over resource allocation to international ventures. However, experience from early international ventures showed that for TF to attract international venture opportunities, it had to be a credible partner to market-oriented international players, which revealed to the top management that public corporation reform was not sufficient in the changing market environment. The top management moved towards multinational rationales in internationalization and actively pressured for incorporation. Finally, incorporation allowed TF to pursue internationalization based on multinational rationales.
Before 1994, TF clearly emphasized national rationales when evaluating international venture opportunities. First, the strategic rationales underlying internationalization were defensive in nature, i.e., internationalization was considered a way to prevent foreign companies from penetrating the Finnish market by catering to all telecommunications service needs of Finnish companies. TF's first and short-lived international venture, i.e., Scantel, was initiated in 1987 by TF's Swedish counterpart Televerket. Participation in Scantel was justified by the competitive threat posed by market liberalization and technological development in the telecommunications industry in Europe.

"Changes in the international environment-increasing competition, changing regulations, and technological development-have led to a situation, where traditional telecoms are in danger of ending up in an unfavorable position if we don't start actively competing for new markets"
-Memo on joining a common Nordic telecommunications venture Scantel, 1987 Other examples of the use of defensive rationales include several mobile telephone ventures initiated between 1990 and 1992: EMT in Estonia, LMT in Latvia, BMTS in Poland, and North-West GSM in the northwest region of Russia. TF's consortium lost the BMTS bid, but EMT and LMT are still a part of the same organization after multiple successive mobile telephone generations and mergers. North-West GSM has been known as MegaFon since 2002 and was the second largest mobile operator in Russia when TF sold its stake in 2017.
The second type of national rationales we identified is national champion rationales, which were especially important for justifying participation in early ventures in Russia and the Baltics. National champion rationales refer to TF's ambition to assume the spearhead role as a Finnish technology exporter and how international ventures were Table 3 National and multinational rationale types and example quotes. National rationales Defensive rationales "Maintaining Finland's and especially TF's national and international post and telecom service level and selection at the level of other leading industrialized countries can be considered a general strategic goal of TF." -Memo on Post and Telecom Finland's strategies for international activities, 1989 "TF's goal is to get telecommunications licenses in other regions outside Finland. This is due to the collapse of monopolies and the resulting increased competition both in Finland and abroad." -Annex to management team meeting, Estonian mobile telephone venture, 1991 National champion rationales "As there are increasing amounts of information and goods moving between the East and West, Finland has to make sure that it positions itself as a broker. Our activities and range of services have to be adapted to this development to maintain the Finnish industry's export ability." -Memo on Post and Telecom Finland's strategies for international activities, 1989 "The project in Latvia open ups new export opportunities to the Finnish Industry." -Internal magazine 'Tietolinja' 2/1994 Home market rationales "The mobile network project in Poland also promotes TF's efforts to make the neighboring areas' telecommunications network as similar as possible as in Finland, thereby indirectly supporting TF's business in the home market." -Memo on the mobile network offer in Poland, 1991 "a) Overseas activities must be limited to businesses that TF already provides successfully. b) Overseas activities must support TF's businesses in Finland, such as by increasing the sales volume, margin, or quality." -The organization of TF's overseas business, 1991 Multinational rationales Growth rationales "Rationales: We require a 17 percent return on investment from the project.
Additionally, the project offers an opportunity to enter the fast-growing mobile telephone market in the Czech Republic." -CEO's draft decision for the Board: LevTel consortium bid for a Czech GSM license, 1996 "The market is growing at a very fast pace, and experience from NWGSM in St.
Petersburg shows that difficulties stemming from an imperfect legal environment can be overcome." -Memo on Russia GSM-1800 Moscow, St. Petersburg, 1998 Portfolio strategy rationales "The company will be a representative of TF and search for new business opportunities in Southeast Asia, especially China." -Decision draft for the Board: Establishing a subsidiary in Hong Kong, 1995 "This case fits Tele's strategy and makes it possible to strengthen our position in Sweden and enhance cooperation with Tele Danmark and Telenor." -Memo on InformationsMäklarna i Sverige AB, 1997 Z. Cheung, et al. Journal of World Business 55 (2020) 101140 required to benefit the broader Finnish society, especially by supporting the Finnish export industry.

"The importance of telecommunications in the next century will be similar to electricity and road networks in this century. Post and Telecom Finland has only two options: either be a part of the development or give up on being a part of the development of new technology, which would be a loss for the entire Finnish economy."
-Memo on Post and Telecom Finland's strategies for international activities, 1989 Such rationales were especially important for the international ventures that started in the late 1980s and early 1990s and continued to play a role until 1994. For example, participation in projects implemented to modernize telecommunications infrastructure projects in Estonia and Latvia shortly after the fall of the Soviet Union was partially justified by the increasing telecommunications needs of the Finnish export firms already active in these countries and the need to make entry to these markets easier for Finnish companies.
Parallel to the defensive and national champion rationales, which focused more on broader societal issues, there was a third type of strategic rationales that had a narrower focus on customers and the service level in the home market. These home market rationales reflect the requirement for international ventures to support business in the home market and focus on the business areas that were already successful in Finland. For example, in the 1991 memo on the organization of TF's businesses abroad, the criteria for international ventures were clearly stated as follows: international ventures must 1) be in a business that has already been proven successful in Finland, 2) support TF's Finnish operations, such as by increasing volume, profit margins, or quality, 3) be profitable without accounting for the benefits to TF's Finnish business, and 4) have a reasonable required level of investment compared to investments in the home market. Criteria 1), 2), and 4) have a clear home market focus, while criterion 3) serves as the baseline requirement for ventures to be profitable. The mobile network ventures in the Baltic countries serve as good examples of the practice of these criteria. The government's decision to incorporate TF in 1994 finalized the change in how the state governed TF from resource control to management-by-results. At this time, the remaining regulatory restrictions imposed on managerial discretion over internationalization decisions were removed, making it possible to implement internationalization aligned with the market logic. The focus of TF increasingly shifts towards financial performance over social or political objectives.
First, although the aim to grow was present in earlier ventures, we found that in the latter half of the 1990s, growth became a taken-forgranted rationale for internationalization and a legitimate measure of success (Table 3). For example, TF participated in a mobile network venture in Hong Kong in 1994 as a first step to participate in the growing Southeast Asian market and the emerging market in China. Instead of leveraging capabilities proven in Finland, the Hong Kong project was justified as an opportunity to develop capabilities in new technology, Asian markets, and extremely competitive markets.
"The objective of the project is to gain access to the growing Southeast Asian market and through that make contacts in the emerging Chinese market. It is evident that all existing operators will be applying for a license. Additionally, those who unsuccessfully applied for a GSM license two years ago are probably joining the race. According to some sources, there might be up to 20 consortia participating in the bid" -Memo on the PCN project in Hong Kong, 1994 The second type of multinational rationales is portfolio strategy rationales. We define these rationales as the evaluation of the fit of individual ventures to the overall venture portfolio. The fit of multiple projects was evaluated considering the overall venture portfolio. Specifically, the limited resources available for international ventures forced the top management to define a strategic focus. For example, the SmartRing venture in 1997 in Germany offered a banking services solution licensing business between the bank and telecommunications service layers, and the fit of this service to the overall service portfolio was questioned by a top management team member. Furthermore, traditional talk services in the Central European market were evaluated as not fitting TF's strategy and were centralized to a venture called Axxess with the aim to exit the business entirely.

"[A top management team member] noted that the planned talk business of Axxess is outside TF's strategic focus. [CEO of TF] noted that TF cannot afford to participate in every single international venture. We need a clearer focus."
-Top management team meeting minutes 39/1997 Additionally, TF employed a more strategic path to enter geographical markets considered central to the overall strategy of the organization. First, TF established subsidiaries to obtain access to target markets, such as Hong Kong, Sweden, and Germany. Second, building credibility abroad was considered critical for attracting international venture opportunities. TF lobbied the state for reforms that granted them more independence and flexibility in decision-making, which helped build credibility. TF even participated in ventures that were not related to TF's core business as such ventures could potentially provide further access to the market.

Choice of Geographical Markets
We found that as different rationales were used to evaluate international venture opportunities, the geographical markets TF considered, pursued, and entered changed (See Table 4). This finding was due to the different set of rationales that ventures must fulfill to be legitimate from the SOMNC parent's perspective. Ventures based on national rationales must offer services that are proven in the Finnish market and serve the national interest, such as infrastructure projects, and these ventures were only available in less developed countries or countries geographically proximate to where Finnish firms had a significant presence. Ventures based on multinational rationales must offer significant growth potential and fit TF's portfolio of ventures based on the international strategy of the SOMNC parent, such as entering new market segments to acquire foreign customers or capabilities. Thus, the set of criteria for foreign market evaluation and subsequent rationales for market entry influenced TF's geographical market choice. Moreover, the change in the rationales affected how TF's existing ventures were reevaluated.
We found that based on national rationales, international ventures were established in the countries closest to Finland and countries with which Finnish firms had significant trade relations. These countries included Russia and the Baltics (former Soviet Union), Eastern Europe and, to a much lesser extent, Western Europe. The national rationales directed decision-making toward specific geographic areas. First, TF's rationale for entering nearby countries reflected defensive rationales: to prevent foreign telecommunications competitors from entering Finland and enable TF to survive in the increasingly international competitive landscape. As the countries near TF's home market had an underdeveloped telecommunications infrastructure, building and exporting the standards used in Finland, such as the Nordic Mobile Telephony (NMT) network developed along with other Nordic state telecoms, represented a strategic tool that hindered the entrance of other players to the market and simultaneously enhanced TF's competitive position. For instance, the joint ventures launched in the former Soviet Union highlight the rationale that to survive in emerging international competition, TF must proactively internationalize and seize market opportunities to compete with other major players projected to enter their home and nearby markets.
"The opening up of Eastern Europe gives us a unique opportunity to expand TF's operations to nearby regions, such as Murmansk, Karelia, Leningrad, and the small Baltic states. […] Due to the region's underdeveloped telecommunications infrastructure, TF's strategy is to take advantage of the existing NMT-450 network because this way, we can get extensive coverage quickly and with reasonable costs. At the same time, we gain a foothold in neighboring areas and prevent other international operators from penetrating these markets." -Memo on the mobile network project in Estonia, 1991 Second, the national champion rationales highlight the broader societal goals of TF as a leading Finnish technology exporter and modern communications provider. The underlying goal was to advance the trade and exports of Finnish firms, which influenced the geographical market entry choice towards markets that were important for the Finnish export industry. For instance, in an international strategy memo from 1989, TF stated that they "strive to launch more joint ventures in the Soviet Union in order to advance Finnish trade to the East" 2 . This situation is evident in the case of Tele Nord, which was a joint venture undertaking to develop communications networks between the Eastern part of Finland and the Northwest regions of Russia.
Third, to support business in the home market and focus on successful business in Finland, the home market rationales directed TF to nearby countries where it could leverage those core businesses already established at home and better serve its existing customer base. The opening of Eastern Europe offered the possibility to modernize the telecommunications infrastructure in Russia and the Baltics. TF aimed to leverage its already proven competencies in the home market, increase its volume and profits in the home market, and better serve Finnish corporate customers by offering expanded geographical telecommunications links or widening the use of the technological standards used in Finland (e.g., NMT mobile standard). For example, when constructing and operating NMT networks in Estonia, TF utilized their existing NMT base station in Finland to operate the network.
"The geographically closest regions, such as Leningrad, are the most interesting. We want to offer our Finnish customers expanding access [to services]. The aim of the different foreign projects is to offer our customers better services" -Interview with TF's head of mobile telecommunications in TF's Internal Magazine 'Tietolinja ', 9/1991 We found that based on multinational rationales, the criteria for market entry and the evaluation of markets shifted to stress the growth potential of ventures and the focal market and the fit of a venture to TF's portfolio of businesses as taken-for-granted rationales. Based on our analysis of the rationales and geographic locations of the ventures, we are able to detect a notable change in 1994 in terms of which markets TF entered and the criteria for market entry. TF established ventures or participated in license bids in Western European countries, such as Germany, Belgium, and the Netherlands, and in more distant markets, such as Hong Kong, Turkey, Lebanon, Israel and the USA. Furthermore, after 1994, the internationalization of TF was increasingly guided by strategic planning and professional business management practices stressing growth as a rationale along with the fit of a Table 4 Geographical market entry choice and change in rationales used to evaluate international venture opportunities 1987-1998. National rationales 1987-1994Multinational rationales 1994-1998 Strategic rationales Z. Cheung, et al. Journal of World Business 55 (2020) 101140 particular venture to the parent company's corporate international strategy rather than the state department bureaucratic planning process and national rationales found prior to 1994.

"The strategy of Telecom Finland is to expand to international markets and make a significant part of its revenue and shareholder value from businesses abroad. […] The most attractive markets for Telecom Finland are advanced countries, like Western European countries."
-The establishment of a joint venture with Hansenet Telekommunication GmbH in Hamburg, Germany, 1997 First, our analysis shows that the ventures launched by TF after 1994 reflected growth rationales, such as market potential, customer demand and market size, in the criteria to establish a venture and the evaluation of geographic locations. Furthermore, we noted a shift in the weight and scale of the international activities of TF compared to the home market business. The strategic goal of the parent firm was to gain a significant amount of their total revenue and new growth from business abroad rather than focus on their existing customers as was evident before 1994. Growth as a taken-for-granted assumption directed TF to niche markets in Western Europe, where telecommunications service markets have been deregulated since 1998, and distant areas with promising market potential, especially in mobile communications.
Second, after 1994, TF increasingly adopted a portfolio perspective when evaluating new ventures. Thus, the rationales to launch or participate in ventures were evaluated against the corporate international strategy; these rationales included reasons, such as to acquire capabilities, customers and technology, enter new business segments, gain market access and gain synergies with the partners in the venture. The evaluation of ventures was increasingly guided by contemporary management practices. These practices directed the market entry choice to developed economies with niche markets and a fitting role for TF as a complementary partner. For example, TF launched multiple ventures in Germany to acquire new customers and enter the deregulated telecommunication service business segments with value-added service platforms. The aim of utilizing TF's capabilities in service platforms and entering niche markets was outlined in TF's international strategy.
Finally, our analysis shows that TF began to re-evaluate its existing ventures based on multinational rather than national rationales. In some markets, this evaluation led ventures to lose their strategic fit in the parent company's perspective. However, this evaluation also led to complementary investments in these markets based on a new set of criteria that supported the re-legitimation of existing ventures. This process was evident in Russia and the Baltics, where new GSM network ventures and license bids in Russia and mobile phone retail stores and marketing offices in the Baltics supported the existing telecommunications operator business.

"Sales of terminal, switches, telephones, and other related devices and services […] through brick and mortar stores support TF's profile in the region and, especially, TF's venture Lattetelekom's [TF's operator venture in Latvia] position as a full-service supplier."
-CEO's decision to the board: Founding a sales company in Latvia, 3/1998

Choice of Partners in International Ventures
We found that when defensive, national champion, and home market rationales were prevalent, similar partners were preferred. For example, all three types of rationales steered TF to partner with other Nordic state telecommunications agencies, especially their Swedish counterpart Televerket (Table 5). First, there was a shared defensive regional interest. In a memo regarding TF's first international venture with all Nordic state telecommunications agencies, TF states that due to their relatively small size, they can be more competitive together as partners. Second, reflecting national champion rationales, the need to render the Nordic NMT standard the dominant mobile technology in the Baltic rim region was shared, especially with Televerket. All NMT projects (BMTS, EMT, and LMT) were partnerships with Televerket. Among other considerations, this collaboration allowed the partners to share the financial risk of investing in countries that had only recently become market economies. Third, the home market rationale of leveraging services proven in the home market did not require complementary capabilities from the partners, and the Finnish and Swedish export industries have similar telecommunications needs and business interests in the same regions.
As the growth and strategic portfolio rationales gradually became more dominant, the partners in different projects became more diverse, an emphasis was placed on building alliances, especially for ventures related to value-added services, and complementary capabilities became more important. For example, the P Plus GSM project in Hong Kong was planned as the first step to enter the opening Chinese and growing Southeast Asian market. In the project, TF partnered with Chinese and Taiwanese firms to gain regional knowledge that could help further penetrate the market. The German virtual network operator venture with Talkline and Hansenet represents complementary partnerships closely linked to TF's strategy to build a service portfolio in addition to know-how in telecommunications infrastructure.

"TALKLINE GMBH [subsidiary of the German electricity utilities company RWE] possesses marketing, tariffing, sales billing and customer care experience with GSM mobile telephone card in Germany and other countries in Europe"
-Letter of intent between Talkline GmbH and Telecom Finland LTD, 1995 However, the transition was gradual, and the coexistence of rationales indicated the need for ventures to satisfy both national and multinational rationales, causing conflicts in the venture partner evaluation process in some cases. The Nordic alliance with Danish and Norwegian telecoms in 1997 was criticized as a partnership of "homogeneous and equally strong partners that does not generate any possibility to develop capabilities" 3 . The decision to not participate in a consortium to operate a trunking network in Belgium in 1995 is an example of national rationales winning over multinational rationales. The main argument against the venture was TF's lack of successful experience in such a business in Finland, which is clearly a home market type of rationale, whereas the excellent track record and experience of their partners and potential to further penetrate the strategically important market, which are multinational rationales, combined with the baseline requirement of profitability were used to support the venture. The decision triggered strong criticism from TF's international business unit against the decision-making process regarding new ventures.

"Looking at the decision-making process, I have to wonder how TF's internationalization strategy is being followed [as TF decided to not partner with] the most successful trunking network operator in Europe"
-Fax to the top management team from the head of the international business unit, 1995 The change from national to multinational rationales in evaluating international venture opportunities also has historically contingent implications for partner relationships. Ventures created in different periods have different partners due to different selection criteria. However, many ventures from the earlier period continued to exist in the latter period, and therefore, conflicts with important partners who were simultaneously major competitors emerged. TF's relationship with Swedish Televerket provides an example of such conflict.
Televerket was the partner of choice of TF in all mobile network ventures until 1993; subsequently, there were no new ventures between the two firms (Table 5). Furthermore, the firms were often in competing consortiums, and each firm sought to penetrate the other's home market. TF had multiple alliance considerations with other Nordic telecommunications agencies to enter the Swedish market, rendering Televerket a "common enemy" 4 . In 1995, the CEO of Televerket expressed his disappointment in losing the modernization of Latvia's telecommunications infrastructure deal to TF and threatened that unless there was closer collaboration between the two to limit competition, Televerket would consider dissolving the joint ventures in Russia and the Baltic states. However, the collaboration and competition status remained until the merger of Televerket (then called Telia) and TF (then called Sonera) in 2002.

DISCUSSION AND CONCLUSION
Based on our findings, we developed a conceptual model of how the increasing influence of market logic impacts the coevolution of state governance of SOEs and SOE internationalization and mutually reinforces the shift in dominant logic, which is summarized in Fig. 3. On the one hand, domestic market liberalization pushed TF to internationalize to replace the market share at risk in the home market and develop competitive capabilities. On the other hand, attracting international venture opportunities requires flexibility and speed in decision-making that gives TF credibility as a partner in international ventures, which procedural resource control by the state did not support. Thus, the ongoing liberalization of the telecommunications market drove TF to search for international venture opportunities, whereas internationalization led TF to become more active in pushing for reforms to obtain independence in decision-making and further internationalization. By actively pushing for reforms, TF's management played a significant role in defining the policies used to govern all SOEs in Finland.
Within SOEs, the logic shift is reflected in the rationales used for the evaluation of international venture opportunities and venture partner and geographical market preferences. Coevolution is also evident within SOEs as certain international market opportunities drive decision-making towards multinational rationales. In turn, the prevalence of multinational rationales, such as growth, creates pressure to internationalize, and the organizations adopt practices aligned with the market logic. For example, the ventures in Turkey and Hong Kong, both initiated at the time of incorporation in 1994, cannot be justified by purely national rationales. Thus, we propose that the enactment of the new dominant logic by the SOE coevolved with internationalization decision-making rationales and availability of international venture opportunities, where the shift in one factor drove the other factors in the same direction (Fig. 3).
However, the shift in dominant institutional logic within the state or SOE is also determined by aspects other than internationalization, such as the domestic competitive environment and market size, which can further accelerate or decelerate the logic shift. In our case, the relatively small market divided between multiple telecommunications operators accelerated the shift towards market logic. Furthermore, as the state tends to regulate SOEs operating in multiple markets similarly, the readiness for market liberalization in other SOE markets, e.g., postal service and railways, could impact the speed of the shift towards market logic. In the case of TF, the organizational link with postal services significantly decelerated the development of the telecommunications business.
Our study makes three key contributions to existing research. First, by showing how a shift in dominant logic impacts decision-making in an SOE and the interrelationship between the state and SOE internationalization, we contribute to the institutional logics perspective of SOEs in the international context (Bruton et al., 2015;Greve & Zhang, 2017;Thornton & Ocasio, 1999. Second, by showing that a) SOEs can internationalize for different reasons, b) the reasons can change over time with intertemporal implications and c) the internationalization process is deeply embedded in the institutional environment, we contribute to research concerning the internationalization of SOEs (Bass & Chakrabarty, 2014;Cuervo-Cazurra et al., 2014;Estrin et al., 2016;Grøgaard et al., 2019;Mariotti & Marzano, 2020;Rodrigues & Dieleman, 2018). Finally, our rigorous historical approach using a relational database enabled us to systematically analyze a large amount of archival data and, thus, answer multiple calls in international business research to apply historical methods that can strengthen the longitudinal dimension of IB research (Buckley, 2016;Burgelman, 2011;Jones & Khanna, 2006).
We show that a shift from state to market logic can alter how state control of SOEs is exercised without changes in the ownership stake. We found that in addition to the extent of state control, it is crucial to analyze how state control is exercised to understand the influence of state and market logics on SOE decision-making. Thus, state versus private ownership is not a sufficient indicator of state control (Mariotti & Marzano, 2019;Wood & Wright, 2015), and we need to investigate the rationales underlying decision-making within firms to understand changes in the balance among multiple logics (Greve & Zhang, 2017). Extant research posits that state and private ownership infuse state and market logics into SOEs, respectively, and that the extent of state control indicates the strength of state logic (Bruton et al., 2015;Cuervo-Cazurra et al., 2014). For example, state control can be exercised through resource control, politically appointed management and board of directors, and regulation. However, the state can also restrain its control, such as through indirect ownership, e.g., state ownership through a holding company or sovereign wealth fund (Musacchio et al., 2015). In contrast, state control could be exercised without ownership, such as by leveraging political connections or controlling financial resources through state-owned banks (Liang et al., 2015;Musacchio et al., 2015). For example, in China, the state can leverage control over privately-owned companies, and these companies might face significant national security scrutiny when they expand to developed markets. Furthermore, our study contributes to our understanding that SOEs can influence their home country institutions (Cuervo-Cazurra, 2015;Cuervo-Cazurra et al., 2019) by showing that the dynamic nature of institutional logics enables agency in the institutionalization process (Friedland & Alford, 1991;Thornton & Ocasio, 1999. During a logic shift, the management of an SOE can significantly influence the enactment of the new dominant market logic in the domain of state governance of SOEs by pushing for autonomy in decision-making. This, in turn, can have persistent influence on how the state governs all of its SOEs. A logic shift could even create a governance void, as old control mechanisms are dissolved and ambiguity regarding new control mechanisms emerges, leading to increased managerial autonomy (Lazzarini, Mesquita, Monteiro, & Musacchio, 2020), excess risk-taking and empire building by management (Zhu, Tse, & Li, 2019). Thus far, the focus in IB research employing the institutional logics perspective has mainly been the institutional complexity arising from contradictory logics in the home and host contexts in which logics are treated as rather static (Newenham-Kahindi & Stevens, 2018;Saka-Helmhout, Deeg, & Greenwood, 2016;Xing, Liu, & Lattemann, 2018). Such perspectives tend to emphasize the heroic ability of agents to combine and navigate multiple logics, neglecting the relationship between changes in the historical context and agency.
The SOE context also provides an opportunity to contribute to the broader institutional logics literature. We found that the shift in dominant institutional logic affected both the state and the SOE in a mutually reinforcing way. Thus, we propose that due to the interconnected relationship between how the state governs SOEs and SOEs' internationalization (Choudhury & Khanna, 2014;Liang et al., 2015;Rodrigues & Dieleman, 2018), it is necessary to analyze the shift in dominant institutional logic in both the state and the SOE, specifically, how the state and the SOE influence each other's enactment of the new dominant logic. More generally, we propose that future research using the institutional logics perspective should investigate the simultaneous enactment of a shift in the balance of institutional logics in multiple interrelated actors operating at multiple levels, such as regulators and firms in highly-regulated industries (Thornton & Ocasio, 1999).
Our study also extends our understanding of how changes in the state governance of SOEs influence the underlying rationales of SOE internationalization. In general, as the state decreases its influence over SOEs, the social and political objectives for internationalization decrease, and focus shifts to financial performance (Cuervo-Cazurra et al., 2014). In addition, we found that the changing rationales had implications for geographical market and partner choices. Although, similar to Duanmu (2012), (2014) and García-Canal and Guillén (2008), we found that when state logic dominated, TF took more political risks in its geographical preferences, and these risks were mitigated by partnering with similar partners to share the financial burden, but that when market logic dominated, TF took significantly more financial risks and partnered with smaller complementary players. Thus, there is a broad shift in the risk preferences of an SOE, rather than only less tolerance for political risks, as the state reduces its influence over the SOE.
Furthermore, we argue that state influence can have far-reaching implications for strategic decision-making even after privatization. Our findings show that the old rationales continued to impact new decisions as old rationales remain imprinted on existing ventures and have historically contingent implications (Buckley, 2016;Jones & Khanna, 2006), which can create conflicts, such as old partners in one market becoming competitors in another market. For example, Televerket was initially the preferred partner and subsequently became the greatest competitor. Ventures established based on national rationales might also lose their strategic fit when subsequently evaluated using multinational rationales. SOMNCs can also make additional investments to realign old ventures with the new strategy, such as the complementary investments in the Baltics that help existing ventures grow. Thus, we propose that there is a need to further investigate how stateness remains imprinted in the behavior of former SOEs (Marquis & Qiao, 2018;Marquis & Tilcsik, 2013) and old dominant logic continues to create conflicts even though its significance in decision-making has diminished (Reay & Hinings, 2009).
For example, state logic can remain dormant when market logic becomes dominant and re-activated by, e.g., the recent trend of deglobalization and rising nationalism worldwide Witt, 2019). Our findings suggest that SOMNCs will not simply reverse their internationalization efforts by retreating and focus more on domestic markets. Instead, we expect SOMNCs to repurpose their existing international ventures to further their home countries' political interests, which can range from extending home country sustainability requirements to host country business partners to corporate espionage in host countries. Therefore, venture partners, states, and politicians need to be cautious about existing international ventures of parent companies with state ownership or other forms of home country political dependencies. Even if the parent company has a long history of internationalization based on multinational rationales, a change in state control policies can quickly change the behavior of the company and its international ventures. This risk has mainly been ignored in extant research on SOE internationalization.
We also aimed to extend beyond the dichotomy of the restricting and enabling nature of institutions in IB research to explore how institutions actively orient decision-making in the internationalization process to one certain set of options over other sets (Cardinale, 2018). Thus, the balance between state and market logics steer SOEs to a certain option set that is further evaluated by some baseline criteria, Z. Cheung, et al. Journal of World Business 55 (2020) 101140 such as the profitability requirement. Additionally, the orienting effect is based on the actor's current and historical positions (Cardinale, 2018). For example, based on multinational rationales, TF's geographical market preferences gravitated towards growing markets, but as an embodiment of its historical position, the existing ventures steered the focus to existing geographical markets. Finally, our study utilized a historical approach (Burgelman, 2011;Jones & Khanna, 2006) and relational database method (Murmann, 2010) to study the rationales underlying internationalization and thereby uncovers previously unattended relationships in the internationalization process (Buckley, 2016;Burgelman, 2011;Jones & Khanna, 2006). Specifically, our historical method is able to provide a thick description of the changes in the institutional environment surrounding TF across an extended period, which enables the analysis of how the institutional change process affected TF's internationalization in addition to the impact of the initial and outcome institutional environment (Aguilera & Grøgaard, 2019). In the case of the deregulation and liberalization of the European telecommunications market, the outcomes of the process were similar across the region, including largescale privatization of SOEs, international competition, and similar regulatory frameworks (Graack & Elixman, 1999;Sarkar et al., 1999). However, the different starting points, paces, and paths during the process resulted in different internationalization patterns (e.g., Clifton et al., 2010;Kern & Gospel, 2020). This historical contingency is critical for the internationalization of firms and central to historical research (Buckley, 2016). To understand a particular firm's strategy, it is necessary to uncover its previous choices and the historical conditions leading to the current outcomes (Jones & Khanna, 2006;Kipping & Üsdiken, 2014).
We also advance rigor in conducting longitudinal qualitative studies (Birkinshaw, Brannen, & Tung, 2011) with historical archival data in the IB field (Perchard, MacKenzie, Decker, & Favero, 2017). Our approach grounds the analysis in a sequenced set of decisions (Buckley, 2016) coded as verbatim documented descriptions (textual) linked to archival sources. We code our data in a relational database to establish a link between evidence and interpretation that is consistently exposed to comprehensive source criticism  and the historical comprehension of events (Rowlinson et al., 2014). Our approach reduces source interpretation biases and increases transparency in the qualitative study process as each source is interpreted and then imputed as data, which can be provided as evidence to other researchers at any time. This type of theory development resonates with the calls for historical analyses in the IB field (Buckley, 2016;Burgelman, 2011;da Silva Lopes et al., 2019;Jones & Khanna, 2006;Verbeke & Kano, 2015) by enabling researchers to connect rudimentary concepts rooted in data to theoretical conceptualization.
As logics are locally enacted and our research context is an SOE from a coordinated market economy, there are limitations in the generalizability of our findings, especially to emerging market contexts in which SOEs enact state and market logics differently (Greve & Zhang, 2017;Thornton & Ocasio, 2008). For example, the credibility of the state's commitment to not impose social and political objectives on its SOEs varies between democratic and autocratic countries (Clegg et al., 2018). Thus, to reach the same level of credibility in emerging market contexts or in contexts characterized by inefficient governments, the state might have to formally relinquish control, such as by becoming a minority shareholder or improving the transparency of corporate governance by listing (Mariotti & Marzano, 2020;Musacchio et al., 2015). However, an emerging market SOE might adopt transparent corporate governance policies ceremonially, and the state could still impose social and political objectives on SOE decision-making (Meyer & Rowan, 1977;Weber, Davis, & Lounsbury, 2009). Hence, we propose that to truly uncover the balance between state and market logics in emerging market SOEs, there is a need to analyze the rationales used for decisionmaking.
As the phenomenon under study is complex, we acknowledge that there are significant historical factors that impacted TF's internationalization process in addition to those presented in the findings. First, during our period of interest, the direction of the dominant institutional logic clearly shifted from state to market logic. However, there is also a need to investigate shifts in the opposite direction as promarket reversals are increasingly prevalent worldwide Witt, 2019). Second, network infrastructure projects have very restricted market entry windows in the form of license bids (Sarkar et al., 1999). If a bid is not successful, the opportunity to enter that market becomes very limited, introducing a somewhat arbitrary geographical presence. Therefore, it was essential to also analyze ventures that were unsuccessful. Third, in the telecommunications industry, which is often considered critical for national security, it is common to impose strict entry mode and partner requirements, e.g., local actors have to own a majority stake in the joint venture. Thus, the discretion to select partners and entry mode was limited in many cases. There was no variance in the entry mode in our case, and stand-alone market entry is rare in the telecommunications industry in general during the period of interest (Sarkar et al., 1999).

Declaration of Competing Interest
None