Governmental goals and the international strategies of state-owned multinational enterprises: a conceptual discussion

State owned multinational enterprises (SOMNEs) have received extensive attention in recent research in international business and corporate governance, which demonstrates effects of state ownership on a range of international strategic decisions such as the degree of internationalization, foreign entry modes, and host country location choices. Such effects are explained by factors such as SOMNEs’ non-financial goals, corporate governance, and institutional pressures. However, results are mixed and context-dependent, and overall we still have an incomplete understanding of what governments aim to achieve through SOMNEs, and how these goals in turn lead to different international strategies. This conceptual article aims to explore how specific government goals may affect international strategies. We provide a more fine-grained view on SOMNE financial and non-financial goals and link them to key international strategic decisions such as the degree of internationalization, entry and establishment modes, and host country location choice. We review and extend previous literature and identify novel theoretical arguments, leading to an extensive set of propositions. We also sketch ideas for empirical studies of SOMNE objectives.


Introduction
The internationalization of state-owned enterprises (SOEs) has emerged as a striking feature of international business (IB) over the last decade (Cuervo-Cazurra et al., 2014;Wright et al., 2021). State-owned multinational enterprises (SOMNEs) originate from a wide variety of advanced and emerging economies, and examples can found in most industries and sectors (Kowalski, 2020;UNCTAD, 2017). The 2020 Fortune Global 500 list included five SOEs among the top 20, including three Chinese SOEs among the top five. 1 SOMNEs' growing role in the World economy has sometimes been controversial, due to supposed threats to the national interests of host countries, or distortions of international competition due to "unfair" advantages provided by the home state (Cuervo-Cazurra, 2018; Kowalski et al., 2013). It is hence not surprising that SOMNEs are currently a current hot topic in IB research (Benito et al., 2016;Kalasin et al., 2020).
Following recent literature (Cuervo-Cazzura & Li, 2021;, we define SOMNEs as enterprises, in which the state in their home country has a direct ownership, and that conduct and own business activities outside their home country. The literature has demonstrated that state ownership can influence a range of internationalization strategies such as the degree of internationalization, entry modes and host country location choices, but effects are mixed and seem to depend on contextual factors (Rygh, 2019). Overall, we still have an incomplete understanding of the effects of state ownership on internationalization strategies. In this article, we argue that a more detailed study of the different goals of SOMNEs is necessary to better understand the diverse effects that state ownership may have.
Explanations for differences in the behavior of SOEs, as compared to privately owned enterprises (POEs), often focus on the non-financial goals of SOEs that may complement or replace the purely financial goals typically ascribed to POEs (Cuervo-Cazurra, 2018;Cuervo-Cazurra et al., 2014;Negandhi et al., 1986). Economic theory suggests that non-financial goals of SOEs may include addressing market failures such as natural monopoly, externalities and public goods, or engaging in industrial policy more broadly (Putniņš, 2015). Other often-cited goals for SOEs include redistribution, providing employment, helping economically depressed regions, and even ideology or economic nationalism (Grout & Stevens, 2003).
However, despite important theoretical advances and empirical evidence in the previous literature on SOMNEs, we still know relatively little about how the specific types of goals governments have for their SOEs affect the international activities of such companies. Various types of non-financial goals of SOEs tend to be assumed to work in the same direction for international strategies, even though industrial policy, political goals and socio-economic goals could well have opposite effects. As one example, a governmental goal for SOEs such as preserving domestic employment is likely to have very different effects on internationalization than a governmental goal of securing access to natural or technological resources. Studies have therefore 1 3 Governmental goals and the international strategies of… found, for instance, both positive, negative and more complex effects of state ownership on the degree of internationalization (Hobdari et al., 2011;Hong et al., 2014;Huang et al., 2017;Kalasin et al., 2020;Majocchi & Strange, 2012;Wang et al., 2012). Most studies have associated such differences to governance characteristics, and to the benefits and costs of links to the home state, but differences in the goals actually pursued by SOMNEs could potentially also help account for the mixed results found in the literature so far. The diversity amongst SOMNEs is illustrated by data about the 20 largest ones (measured in various ways, including their multinationality; please see the "Appendix" table for details), demonstrating that such companies are found in several countries and across a range of industries. The list also reveals considerable variation in state ownership compositions and different main reasons-or goals-for state involvement in the companies.
One important reason for our current relatively limited understanding of the implications of different SOMNE goals is the difficulty of measuring such objectives in practice. Indeed, the official goals of specific SOEs often remain unclear and unspecified (Cannizzaro & Weiner, 2018), which has led to calls for greater transparency about SOE non-financial objectives in order to reduce potential distortions in international trade (see e.g., Kowalski et al., 2013). Moreover, a given SOMNE may pursue different (and possibly conflicting) objectives simultaneously, and objectives may even change over time for the same company. As pointed out by Mazzolini (1979b), SOEs have often functioned as "all-purpose" organizations utilized by politicians to address specific issues, as they arose in circumstances such as in economic downturns, and Mazzolini (1979b) argued the same logic often applied to SOEs' international operations.
While some of the earliest SOMNE studies based on qualitative research uncovered many insights on SOMNE objectives and their evolution (Anastassopoulos et al., 1987;Kumar, 1981;Mazzolini, 1979b;Vernon, 1979), the recent quantitatively oriented literature has largely relied on proxies for SOMNE goals such as state ownership shares, or characteristics of home country institutions (Benito et al., 2016;Clegg et al., 2018;Estrin et al., 2016;Grøgaard et al., 2019;Mariotti & Marzano, 2019). Such studies have often argued that partial private ownership and certain forms of market-oriented institutions tend to make the goals of SOEs more similar to those of POEs. Similarly, Li et al. (2014) provide a detailed discussion of emerging market institutional changes and argue that centrally, SOEs have predominantly political goals; while locally, SOEs have predominantly financial goals.
Although bounded rationality (Simon, 1961) suggests that even governments would find it difficult to plan for all possible contingencies and specify complete objectives for their SOEs, some countries have made attempts at clarifying the goals of SOEs and their market operations and making them more transparent. For instance, Norway, a country with a significant level of state ownership (Goldeng et al., 2008;Lie, 2016), has for more than a decade published official reports specifying the goals of its most important SOEs. 2

3
Notwithstanding empirical challenges, there is also a need to further develop theory. Therefore, the present conceptual article is an attempt to provide a more finegrained view on the particular goals of SOMNEs, as well on the implications of the various goals for international strategies. Our analysis proceeds in three steps. First, in the next section of this article, we build on previous literature, both theoretical and empirical, to build a typology of SOE and SOMNE goals. Although a wide range of different objectives have been cited, we organize the objectives into four broad categories: Financial; Industrial Policy; Socio-economic; and Political. Second, in a subsequent section we review and extend theoretical arguments for the implications of each of these categories of goals on key international strategic decisions, specifically the scale and scope of internationalization; entry modes; and host country location choice. Third, we consider further aspects such as corporate governance, private co-ownership and institutional context, linking them up to the various goals and illustrating how the particular mix of goals may vary across countries and over time. We end the article with some suggestions for more detailed empirical studies of SOMNE objectives and some concluding remarks.

A typology of SOMNE goals
State ownership has been studied within a variety of disciplines, notably in economics and business, but also for instance political science. Key questions considered in the literature, often comparing SOEs against the benchmark of POEs, have included the specific goals of SOEs, their particular governance and organizational characteristics, and theoretical justifications for state ownership. Research has also employed a variety of different theories, depending on the specific question at hand (Peng et al., 2016). Welfare economics has been the main basis for studying rationales for state ownership, exploring how state ownership can provide a second-best solution under market failures. Organizational economics (e.g. agency theory and property rights theory) has been used to qualify such rationales by also considering the specific governance characteristics of SOEs, while public choice theories focused on government failure and self-serving behavior by Government. From the management field perspectives such as institutional theory and the resource-based view have been used to explore particular institutional pressures facing SOEs, and how state ownership affects the development of the resource supporting firms' competitive advantages. Sorrentino (2020) highlights the fundamental role of the concept of organizational goals in the study of organizations, and discusses the relationship of this concept to other concepts such as values, vision and mission. Sorrentino (2020) also cites the "classic question of whether organizations, like individuals, effectively have identifiable goals and, if so, of what type" (p. 77). A related question concerns the implications for goal consistency of the fact that typically many different public sector organizations (e.g. different ministries and agencies) are involved in the governance of SOEs (Mariotti & Marzano, 2021). In this article, we acknowledge these debates but will assume that, at least as a first step, one can meaningfully talk both about a Government's goals for its SOEs, and an SOE's goals. In order to provide focus to the discussion, we will also for the most of the article assume that the Government and the SOEs goals are identical, but will relax this assumption when revisiting the particular governance features of SOEs.
Objectives and motivations of firms and their owners cannot be observed directly (Godfrey & Hill, 1995), and in this respect SOEs differ little from POEs. Despite this, it is usually assumed that SOEs pursue non-financial goals besides or instead of the financial goals that are purportedly pursued by POEs in general (Negandhi et al, 1986;Rudy et al, 2016). Such non-financial goals can be based both on theory (e.g. economic theory of state ownership being used to address market failures) or just emerge from empirical observation. In this section, building on the literature on SOEs in general and SOMNEs specifically, we present a typology of the objectives of SOMNEs. 3 Within a framework based on economics and business, but also incorporating arguments from a wider set of literature, we propose that the key goals can be categorized into four groups: Financial, Industrial Policy, Socio-Economic and Political. Importantly, we do not claim that all these types of goals will be present to the same extent for all SOEs. The specific mixture and weight of goals will depend on cross-sectional characteristics such as institutional and economic conditions in the SOEs' home country, but may even vary longitudinally in tandem with factors such as the country's economic development.
In the following, we discuss each of these four stylized goals in turn, while the next section will link such goals up to key international strategies studied in IB research. A later section will extend the analysis by considering aspects such as goal conflicts between citizens, politicians and SOE managers; the implications of private co-ownership; and the role of the institutional context, taking a first step towards the challenging question of where SOE goals actually come from.

Financial goals
Although the focus in the SOE literature is on non-financial objectives, the state enterprise as an organizational form is chosen over alternative public administration forms (Cuervo-Cazurra et al., 2014;Putniņš, 2015) for specific reasons such as increasing managerial autonomy and not being part of the State budget. Therefore, SOEs typically also have at least some form of financial goals. Indeed, SOEs can potentially represent an important source of income for a government, and could be more politically palatable than other forms of raising income such as trade taxes (Kostecki, 1981). Thus, studies on SOMNEs have also mentioned their role in gaining foreign currency for the home country (Kumar, 1981;Vernon, 1979). Recent analyses of SOEs' mergers and acquisitions (M&A) considering purpose statements and comparative study of investment patterns have also confirmed the relevance of traditional financial motives in M&A settings, such as efficiency gains, market power and diversification (Clò et al., 2015;Clò et al., 2017;Del Bo et al., 2017;Florio et al., 2018).
An important question here, which has received some attention in the state ownership literature, is to what extent the Government's financial goals for an SOE translates into their actual pursuit by the SOE; or put differently, whether a true focus on financial goals would essentially make SOEs similar to POEs. Some studies have found that SOEs with a financial focus do not have significantly different performance from POEs (Bozec et al., 2002;Kole & Mulherin, 1997). From a theoretical perspective, this is not a clear-cut question, since state ownership may also have particular implications for corporate governance, and in turn for financial outcomes. We revisit this question later in the fourth section of this article, where we also consider the recent insights on hybrid SOEs with partial private ownership (Bruton et al, 2015;Chen et al., 2019;Musacchio et al., 2015;Zhou, 2018). For now, however, we will assume that the Government can in fact specify purely financial goals for SOEs in such a way that SOE behavior approximates "profit maximization".

Industrial policy goals
Among non-financial goals, the most frequently mentioned from the perspective of economic theory include addressing market failures and pursuing industrial policies. Such goals qualify as non-financial (from the firm's point of view) since they could often require the SOE not to profit-maximize, or even to run with losses (Putniņš, 2015). Put differently, such goals refer to economic efficiency from the perspective of an economy or industry as a whole, rather than firm-level financial results. Moreover, economic efficiency in this sense is distinct from income-maximization by the Government, and therefore does not represent purely financial goals. Such goals can all be said to be related to industrial policy, albeit in a broad sense. Industrial policy itself is a contested term that has been defined in a variety of ways. We here adopt the broad definition by Warwick (2013, p. 16) according to which: "Industrial Policy is any type of intervention or government policy that attempts to improve the business environment or to alter the structure of economic activity toward sectors, technologies or tasks that are expected to offer better prospects for economic growth or societal welfare than would occur in the absence of such intervention." This broad notion of industrial policy goals captures both industry-oriented policies, where a particular industry may be supported based on an assumption that the industry has certain positive effects for an economy, and firm-oriented policies, whereby the aim is to improve the internal functioning of an industry.
Thus, we also include traditional market failures in this broad concept of industrial policy, given that industrial policy itself is often motivated by some form of market failure. The "classic" market failures studied in economics, and typically discussed with reference to a benchmark idealized market with "perfect competition", include natural monopoly, externalities, and public goods (Cuervo-Cazurra et al., 2014;Putniņš, 2015), and are each discussed in turn in the following.
Natural monopoly refers to a situation where there are very high fixed costs in production compared to the size of the market, implying that having a single firm producing for the market would be most efficient. Examples include traditional public utilities such as energy, railways and fixed-line telephony (Putniņš, 2015). In these cases, efficient pricing would require the firms to run with a loss, and hence state ownership could be an option. Note that efficient pricing here relates to economy-wide or industry-wide efficiency; income maximization by a government would instead imply restricting production in order to achieve the monopoly price, illustrating the difference between industrial policy goals and financial goals from a government's perspective. Indeed, in certain contexts where state ownership is less prevalent, such as North America, private natural monopolies are run in a way that is closer to profit maximization. 4 Externalities refer to a situation where the social cost of production is not equal to the firm's private cost (either because there are positive side-effects such as innovation, or negative side-effects such as pollution), meaning that production may either be too low or too high from a societal perspective. In theory, state ownership can be one way of ensuring the efficient level of production, balancing the overall social benefits and costs of production. Finally, public goods refer to goods for which the cost of providing them to an additional person (e.g. a TV broadcast) is zero (nonrivalry), while excluding a person from consuming them is either not possible or is undesirable (non-excludability). Again, state ownership could theoretically be one option for ensuring the efficient production and provision of such goods.
Other relevant market failures include those related to asymmetric information, or coordination problems (Putniņš, 2015). These are also often relevant for industrial policy, where the state may for instance play an important role in making complementary investments or in reducing the risk for private actors in making investments (Murphy et al, 1989;Musacchio et al., 2015). SOMNE resource-seeking or technology-seeking investments can also be seen from this perspective, as they may seek to ensure supply of natural resources to fuel the domestic economy, or acquire advanced technologies for upgrading the domestic economy (Rudy et al., 2016). 5 Based on M&A purpose statements, Florio et al. (2018) identify an "innovation" rationale for some SOEs (for instance in climate finance, renewable and environmentally friendly energy, or the development of physical and technological infrastructure). These can also be considered industrial policy goals, as the aim is to increase 4 We are grateful to an anonymous reviewer for prompting us to clarify this point and acknowledge the relevance of the institutional context. It is also interesting how some traditional natural monopolies have lobbied in international trade negotiations (Woll, 2007). 5 While market failures may establish an a priori rationale for state ownership, the economics literature has emphasized that one also needs to establish a rationale for why state production, rather than contracting out to private firms, is needed. Arguments here have focused on contractual issues relating to private firm risk aversion or financial constraints, and the ability to write complete contracts (Hart et al., 1997;Martimort, 2006;Sappington & Stiglitz, 1987). Rygh (2018) argues that similar arguments may apply, with some modification, to SOE international operations. the economic efficiency of the economy. They are also distinct from revenue maximization by the home government, as the aim will be to diffuse the resources in the knowledge in the home economy, potentially at the expense of financial results for individual SOEs. In some cases, industrial policy goals may involve supporting the development of industries that are seen to be of particular importance. One notable contemporary case is China, which explicitly promotes emerging strategic industries. According to Davies (2013;p. 36) the industries were "next generation IT, energy conservation, environmental protection, new energy, biotechnology, highend equipment manufacturing, new materials and new-energy vehicles".

Socio-economic goals
A third and frequently cited category of potential non-financial objectives of SOEs are socio-economic goals such as redistribution (which can also include helping economically depressed regions). Such goals can also include the provision of so-called "merit goods" (Sandmo, 1983), which are seen as particularly important for citizens' functioning in society (including goods such as education and healthcare). SOEs have often also been used to provide employment (Duanmu & Pittman, 2019). A relevant example following the recent financial crisis was bailouts of firms that were seen to be "too big to fail", and whose collapse could have serious consequences for social stability. Florio et al. (2018) find evidence of such a "financial distress" motivation for M&As by SOEs.
Such goals could have a direct impact on financial goals, as they may require firms to keep employment higher than at the profit-maximizing level, save lossmaking firms, or choose less profitable projects in the government's targeted regions (Mazzolini, 1979a, b). 6 In an international context, this type of goals could imply using SOMNEs for foreign aid and development purposes, of which one example is the Norwegian financial SOE Norfund (Rygh, 2018).

Political goals
Political goals is clearly a very broad term, which we will here use to cover several different types of goals. One type is diplomatic and foreign policy goals for SOMNE operations abroad (Rudy et al., 2016). For instance, Chinese SOMNEs are widely assumed to operate with diplomatic and foreign policy goals (Wang, 2002). However, there are also several examples of European SOMNEs being assigned such goals, such as Renault in Canada (Anastassopoulos et al., 1987). It is often assumed that such investments are made to establish or maintain relations between the home and host countries. However, a more negative view of political goals would suggest that SOMNEs may be used as part of ideological and nationalist strategies, and that the investments are not necessarily beneficial for the host country (Cuervo-Cazurra, 2018;Cuervo-Cazurra et al., 2014). 7 Of the categories of SOMNE goals, political goals is probably the one that is most difficult to pin down, and perhaps the most dependent on the particular context. IB and governance scholars may here be able to find relevant insights in the political science and international relations literature (Keohane & Nye, 1989;Lake, 1996;Rygh, 2018).

Degree of internationalization
Broadly, the literature has considered two dimensions of the degree of internationalization (Sullivan, 1993). The scale of internationalization refers to the extent of the firm's activities abroad (e.g. captured by aspects such as the ratio of foreign assets to total assets, or foreign sales to total sales). This concept relates among other things to the importance of internationalization for a firm's strategy. A related but distinct concept is a firm's scope of internationalization, which refers to the breadth or diversity of the foreign operations, e.g. measured by the number of host countries for foreign direct investment (FDI), or the number of foreign subsidiaries (Hitt et al., 2006). State ownership could influence both dimensions, but as explained below not necessarily in the same way. Factors assumed to affect internationalization include executive and top management team characteristics, firm resources, and international experience (Békés et al., 2021;Hitt et al., 2006;Kirca et al., 2012). However, corporate governance aspects including ownership have also received some attention (Aguilera et al., 2019;Békés et al., 2021;Bhaumik et al., 2009;George et al., 2005;Papenfuß, 2020;Tihanyi et al., 2009).
All else equal, SOEs for which the home government has financial goals should internationalize based on similar strategic reasons as POEs (assuming that POEs generally pursue financial goals), and hence have a similar scale and scope of internationalization. 8 Studies considering SOEs that were assumed to be in this category have found some supporting evidence for this. Following deregulation and liberalization leading to increased international opportunities and competition, telecom SOEs internationalized to the same extent as their private counterparts (Alonso et al., 2013). Similarly, Collins (1986) noted that European SOEs in the automotive industry, where explicit government social goals were generally less relevant, internationalized in more or less the same way and for the same reasons as European POEs in the same sector.
However, several other studies have argued that social and political goals can reduce the propensity of SOEs to internationalize. Based on case studies of a large number of European SOEs, Mazzolini (1979b) noted that since SOEs ostensibly pursue social goals, and since such goals will tend to be linked to domestic outcomes, international activities are less relevant for SOEs unless they are used to support domestic social goals. This is in contrast to POEs, where international activities will often have strategic importance for their financial goals. For this reason, SOE managers may pay less attention to international opportunities and be less able to exploit them. Besides this, politicians concerned with re-election may tend to emphasize projects that directly and visibly benefit domestic voters (Boyd, 1986), which would also usually work against international expansion.
As for political goals, the effect is less clear-cut. Although the above section cited examples of how SOMNEs have been used to establish or strengthen cooperation between countries, it is likely that this goal by itself generally has a relatively modest effect on the overall scale of internationalization. Possible exceptions may include extensive initiatives such as the Chinese Belt and Road Initiative.
The domestic bias generally found in Mazzolini's (1979b) case studies of European SOEs is largely corroborated by recent statistical studies using European samples. Majocchi and Strange (2012) find that state ownership is negatively related to a measure of international diversification for Italian firms, while Benito et al. (2011) report that state ownership has a negative effect on the propensity of Norwegian firms to locate divisional headquarters abroad. Relatedly, Hobdari et al. (2011) find that state ownership reduces exporting by Slovenian and Estonian firms.
Nevertheless, SOEs' relative domestic bias is sometimes found to be reversed, and this seems often to be linked to industrial policy goals that require international operations. As a key example, SOEs from a wide range of home countries have frequently been involved in resource seeking in the petroleum sector (Deng, 2007;Franko, 1975;Khandwalla, 1986;Mazzolini, 1979b;Noreng, 1981). More recently, it has been argued that SOMNEs from emerging markets such as China engage in strategic asset-seeking FDI (especially M&A in developed economies) in order to access advanced technologies that can be diffused in the home economy and upgrade the capabilities of local firms (Kowalski et al., 2013;Liang et al., 2021;Rudy et al., 2016). China has also encouraged internationalization in certain "strategic" sectors of importance for the home economy. This is consistent with the broader notion of industrial policy discussed above, as the goal seems to be economy-wide or industry-wide development rather than individual firm financial performance. Providing some supporting evidence, Wang et al. (2012) find that state ownership promotes the volume of FDI by Chinese firms. Hong et al. (2014) also find that state ownership promotes Chinese FDI, although this effect is moderated by various firm level and industry factors. On the other hand, Hu and Cui (2014) report no significant effect of state ownership on internationalization of Chinese firms. It may be that the role of SOMNEs in China's overseas industrial policy is becoming less important, as the private sector gradually develops, suggesting that it is also relevant to look at the stage of development of the SOMNE's home country (Panibratov & Klishevich, 2021).
Overall, assuming all else equal, we can state our first set of propositions 9 : A related, but distinct issue is the scope of internationalization. Theoretically, being present in many different foreign markets can provide benefits in such areas as innovation, as MNEs can combine a diverse set of ideas and influences from these different markets (Oh & Contractor, 2012). On the other hand, being present in a diverse set of markets could amplify coordination costs (Fisch, 2012;Richter, 2014). Privately owned MNEs would be expected to make this trade-off based on expected financial results. However, SOMNEs may base their decisions about the scope of internationalization on other considerations. First, industrial policy objectives could theoretically lead to a greater scope than if based on financial reasons alone. For instance, technology-seeking SOMNEs may seek out a broad range of foreign markets in order to potentially access a wider range of technologies and knowledge, which can generate positive externalities and be widely diffused in the home economy. As such, the strategic assets would have an even higher value for SOMNEs than for a private MNE (Rudy et al., 2016). Such SOMNEs would be less concerned about financial performance, and might hence also be less concerned about any costs associated with having a very broad scope of internationalization.
SOMNEs with important political goals are likely to have a lesser scope of internationalization than privately owned MNEs. Since such SOMNEs are aiming to establish or maintain political relations between the home and host country, they will likely want to focus their resources on particular host countries, rather than spreading over many different countries. Similarly, SOMNEs with socio-economic goals are likely to have a narrower scope of internationalization. A broad scope of internationalization typically entails that activities in many different foreign countries replace domestic activities, including employment. Similarly, there is no clear reason for why a broad scope of internationalization would support (domestic) goals of redistribution. Hence, again assuming all else equal, we provide the following set of propositions.

a. SOEs with financial goals have a similar scope of internationalization to that of POEs. b. SOEs with industrial policy goals have a broader scope of internationalization than POEs. c. SOEs with socio-economic goals have a narrower scope of internationalization than POEs. d. SOEs with political goals have a narrower scope of internationalization than
POEs.

Entry and establishment modes
Another major international strategic decision relates to the modes used to enter a foreign market (Brouthers & Hennart, 2007;Slangen & Hennart, 2007), which includes decisions such as joint or full ownership of the foreign operation, and whether to establish the operation as a greenfield project (i.e. from scratch) versus acquiring an already existing operation. Some previous studies have considered how ownership and corporate governance may affect foreign market entry choices (e.g., Musteen et al., 2009), but not systematically linked them to the goals of SOEs. Again, abstracting for now from other corporate governance features of SOMNEs, it would be expected that SOMNEs with financial goals make entry mode decisions based on similar reasons as private MNEs. In contrast, different nonfinancial goals of SOMNEs could have varied implications. First, industrial policy goals of SOMNEs could include such goals as ensuring supply of energy resources in the longer term (Bass & Chakrabarty, 2014), or acquiring technologies that have positive externalities for the home economy and could be diffused widely (Kowalski et al., 2013;Rudy et al., 2016;Rygh, 2018). This might lead SOMNEs to prefer high-control (full ownership) modes. However, as suggested by Grøgaard et al. (2019), SOMNEs motivated by such goals might also be less concerned with value capture, and hence high-control modes may be less essential.
In terms of establishment mode, technology-seeking SOMNEs are likely to seek acquisitions of firms that possess advanced technologies. Based on M&A purpose statements, Florio et al. (2018) identified an "innovation" rationale for some M&As by SOEs. For natural resource-seeking SOMNEs, the picture is less clear since such SOMNEs may above all be seeking to ensure long-term resource access, which could involve the development of resource assets through greenfields (Bass & Chakrabarty, 2014;Grøgaard et al., 2019).
Second, political goals such as establishing or maintaining relations with a host country are likely to lead to a greater attention to collaborative modes, especially with host country enterprises, due to political considerations. Thus, in order to improve legitimacy (Meyer et al., 2014), SOMNEs with political and diplomatic goals are likely to be more prone to choose joint as well as greenfield ventures.

Governmental goals and the international strategies of…
Finally, socio-economic goals, to the extent that these also extend to the host country (Rygh, 2018), could also imply a preference for collaborative modes such as joint ventures, as well as greenfield projects that provide new economic activity, employment and economic development in the host country (Mazzolini, 1979b;UNCTAD, 2013).
The above discussion can be summarized in the following two sets of propositions, again assuming that other factors such as corporate governance are held constant.

Host country location
Given that SOEs take the step into locating business activities abroad, another question is whether their location choices differ from those of privately owned MNEs. Non-financial goals of SOEs that have been argued to potentially affect location choices include diplomatic and foreign policy goals, resource-seeking, and technology-seeking. China's outward FDI is often assumed to be driven both by resourceseeking motivations and by political and economic cooperation (e.g., Kaplinsky & Morris, 2009;Sanfilippo, 2010). Earlier European SOEs' internationalization was also sometimes found to be related both to resource-seeking and foreign policy motives (e.g., Anastassopoulos et al., 1987;Mazzolini, 1979b). A useful starting point for this theorization is Dunning's FDI motives (Benito, 2015;Dunning & Lundan, 2008), with recent extensions including geopolitical influence-seeking motives (Moghaddam et al., 2014). In general, financial objectives would be linked to market-seeking motives, and conversely firms with non-financial objectives are likely to have a lesser emphasis on market-seeking.

a. SOEs with financial goals make location choices on similar considerations as
POEs. b. SOEs with industrial policy goals make location choices that are less motivated by market-seeking and efficiency-seeking, and more motivated by resource-seeking and strategic asset-seeking. c. SOEs with socio-economic goals make location choices that are less motivated by market-seeking, efficiency-seeking, resource-seeking, and strategic asset-seeking, and more motivated by non-business related motivations, such as development concerns. d. SOEs with political goals make location choices that are less motivated by market-seeking, efficiency-seeking, resource-seeking, and strategic asset-seeking, and more motivated by geopolitical influence-seeking.
These propositions can be operationalized and tested in terms of variables commonly used to proxy market-seeking motives (e.g. market size or purchasing power), efficiency-seeking motivations (e.g. factor costs), resource-seeking motives (e.g. natural resource abundance), strategic asset-seeking motives (e.g. education levels or innovation levels), and geopolitical influence-seeking (e.g. relations between home and host countries). Thus, for instance Ramasamy et al. (2012) find that Chinese POEs are mainly market seekers, while SOEs' FDI is associated with natural resource availability and politically risky environments. Similarly, Amighini et al. (2013) find that Chinese private MNEs are attracted by large markets and hostcountry strategic assets, while being averse to economic and political risks. In contrast, Amighini et al. (2013) report that SOEs invested more in natural resource sectors and were less concerned about political and economic conditions in the host countries.

Discussion and extensions
State owned MNEs play an increasingly important role in the global economy, and one of the most salient aspects of such enterprises is their potentially greater weight on non-financial objectives than the private MNEs that have hitherto been the focus of governance-related IB research. The main aim of this conceptual article has been Governmental goals and the international strategies of… to provide a more fine-grained analysis of the non-financial objectives of SOEs, and their potentially diverging implications for international strategies. We have reviewed and extended previous theoretical arguments and developed a number of propositions, many of which we believe to be original to the IB literature. Table 1 below summarizes these propositions.
However, in order to focus on the variety of SOMNE goals the analysis has been based on certain important simplifications. In particular, we have assumed away goal conflicts between the Government owner and SOE managers, and the role of corporate governance and institutions more generally. Moreover, we have so far not engaged with the difficult question of how to identify and measure such goals in empirical analyses. In the remainder of this article, we revisit some of the simplifications made, and offer some suggestions for empirical research.

SOMNE corporate governance
So far, we have abstracted from the fact that SOMNEs may not in practice act in a manner that is completely consistent with the objectives of the state owner. Obviously, such goal congruence may be too strong an assumption in many cases, and the peculiarities of SOE corporate governance have indeed been a key issue in the state ownership literature (Papenfuß, 2020). Studies based on agency theory highlight the more complex agency chain from voters as principals via several layers of agents such as politicians, bureaucrats and finally SOE managers (Benito et al., 2016;Martimort, 2006;, each of whom may pursue their personal interests. Moreover, SOE corporate governance may also involve different ministries (OECD, 2005) in a multi-principal structure, with the potential presence of private co-owners making governance even more complex (Chen et al., 2019;Chen & Young, 2010;Musacchio et al., 2015). Moreover, studies have pointed out that certain corporate governance mechanisms that are in operation for POEs (at least in the liberal market economies that have implicitly been the focus of the agency-based studies of SOEs) such as sale of shares or takeovers are partly or fully deactivated in SOEs. Finally, SOEs tend to make less use of high-powered incentives typical of private firms (Bruton et al., 2015;Peng et al., 2016;Tirole, 1994).
It is notable that many agency-based studies explicitly or implicitly assume that the goal of SOE corporate governance is to maximize financial performance, although SOE corporate governance should be seen in relation to the particular goals that the SOE is expected to pursue, whether financial or non-financial (Le & Buck, 2011). The agency-based literature provides important insights into the question of whether purely financial objectives would in fact lead SOMNEs to behave similarly to POEs. However, to better understand the relationships between various goals of SOMNEs that have both financial and non-financial goals, a multitask agency perspective is likely to be needed (Bai et al., 2006). Theorization within such a perspective may benefit from a fine-grained categorization of non-financial goals pursued by SOMNEs.

SOMNEs as hybrids and private co-ownership
As highlighted by many recent studies, an important difference between today's SOEs and those of the past is the prevalence of shared ownership with private investors as either minority or majority owners (Benito et al., 2016;Bruton et al., 2015;Chen et al., 2019;Musacchio et al., 2015;Zhou, 2018). Such private co-ownership is argued to address many of the corporate governance issues traditionally associated with wholly SOEs (Gupta, 2005), although it also opens up for new principal-principal conflicts between the state and private owners (Chen et al., 2019;. Although many studies have discussed such goal conflicts in general terms, few studies have considered the implications of specific SOE non-financial goals. One hint at such an argument is provided in Lazzarini and Musacchio (2018), who argue that industrial policy-related objectives could also be a source of rents (and as such, potentially less a source of conflict between state and private owners). In contrast, Lazzarini and  argue that social objectives are likely to be a source of principal-agent and principal-principal conflicts in partially state-owned enterprises. A more detailed study of various SOE objectives, is hence also likely to provide a better understanding of the extent and nature of goal conflicts between state and private co-owners. Here, the SOMNE literature can also build on insights from literature on political capital and its potential economic value for firms (Sun et al., 2016).

SOMNE goals, firm-specific advantages and risk preferences
So far, the analysis has not considered indirect effects of different SOMNE objectives on international strategies via factors such as SOMNEs' resources (or firm-specific advantages (FSAs) as they are often termed in the international business literature). However, one might expect that SOMNEs' goals also affect the development of their FSAs. Financially oriented SOMNEs might develop quite similar FSAs as private MNEs. An interesting question is how political goals affect FSAs. It is likely that SOMNEs with such goals get quite extensive political support from the home state, which could help with handling political risk (Buckley et al., 2007;Duanmu, 2014;Knutsen et al., 2011). This could be conceptualized as political FSAs, which might in turn affect international strategies such as location choices or entry modes. However, governments may also offer political support for SOMNEs pursuing other important objectives, such as industrial policy. Related to FSAs, different forms of objectives for SOMNEs could also imply specific forms of moderation effects of state ownership on much studied relationships in IB literature. For instance, various types of non-financial goals could have divergent implications for the propensity of SOMNEs to internalize FSAs and for the preferred entry modes (Pan et al., 2014).
Another unresolved question relates to how different government goals affect SOEs' risk preferences. In theory, SOEs could be more risk tolerant than POEs, given that the state owner is highly diversified (Arrow & Lind, 1970;Charreaux, 1997;Vernon, 1979). Indeed, in theory this is one possible argument for state ownership (Sappington & Stiglitz, 1987), and may for instance support a role for SOEs in industrial policy, that can involve large-scale and risky projects (Florio et al., 2018). In contrast, other studies have argued that SOEs' social goals are likely to make them more risk-averse (Boubakri et al., 2013;. Again, these theoretical and empirical differences illustrate that a more fine-grained view of SOEs' objectives could help us better understand the effects of state ownership on variables that are of importance for internationalization.

Institutional contexts
SOE corporate governance is closely related to the institutional context, and this link constitutes one of the key areas where the literature is currently being advanced (Benito et al., 2016;Estrin et al., 2016;Mariotti & Marzano, 2019. On the one hand, a core aspect of a strong institutional environment such as a well-functioning government bureaucracy will make it more likely that SOEs pursue the objectives they have been tasked with, among others by insulating bureaucrats and SOE managers from (undue) political interference (Evans & Rauch, 1999;Grøgaard et al., 2019;Musacchio et al., 2015). Whether SOEs are publicly listed may also matter, with listed SOEs argued to be more likely to have a financial focus due to, among other things, pressures from private co-shareholders (Benito et al., 2016;Estrin et al., 2016;Kalasin et al., 2020).
On the other hand, the intent of a government to use SOEs for socio-economic and political purposes may itself be related to aspects of the home country context, such as its (liberal or coordinated) market orientation (Grøgaard et al., 2019;Hall & Gingerich, 2009;Mariotti & Marzano, 2019) or even cultural characteristics (Boubakri et al., 2016). More liberal market economies may be more focused on "market preservation" and have a lesser willingness to assign a broad set of social and political goals to SOEs, than will coordinated market economies or state-influenced market economies (Mariotti & Marzano, 2021, p. 116). In terms of industrial policies, these differences between the types of market economies might be less clear, to the extent that they are aiming at improving the functioning of markets and address market failures. There may also be a difference between democratic governments and more autocratic governments, where the latter may be better able to override other social actors to promote governmental goals (Clegg et al., 2018). Geopolitical goals may be consistent with an aim to use SOMNEs to upgrade the economy and increase international economic power.
Institutional context may also matter when reconciling governmental goals for SOEs with the goals of other actors, notably private co-investors. For instance, Marzano (2020, 2021) find that in more coordinated market economies, there are stronger mechanisms for building coalitions between dominant state owners, minority private owners and other stakeholders taking a longer-term view on goals and strategies. However, when ownership is shared with strong co-owners such as MNEs, there is also an increased potential for principal-principal conflicts. In contrast, in liberal market economies the combination of resources from state and private owners may be easier, as the state is relinquishing an active role.

Governmental goals and the international strategies of…
Overall, diversity of SOE goals, governance and resources are likely to reflect the diversity of their home countries and governments (Rygh, 2019). Along with a more fine-grained view of SOMNE objectives, it is necessary to study how institutional contexts support or hinder the pursuit of these various objectives, whether financial or non-financial. However, more attention to institutional context will also help answer the broader question of why SOMNEs have different goals, and where these goals originate from. This perspective also suggests that integration of economic and institutional theories can provide additional insights. Welfare economics implies the assumption of a benevolent "social planner", while institutional perspectives may inform studies of how the goals of politicians for SOEs actually arise and are negotiated within a particular institutional context.

Empirical study of SOMNE objectives
A key challenge in studying non-financial goals is that an exact specification of such goals is typically not available to researchers, and indeed the goals of a particular SOMNE may be mixed and change over time. In many cases, IB researchers will have to continue to rely on proxies for SOMNE goals such as home country institutional variables, the extent of private ownership, and even industry factors. However, IB and governance researchers should also look out for additional sources of relevant information for particular SOMNEs. One example of such information is the Norwegian Government's official reports on state ownership (NOU 2004:7), which since 2005 have classified SOEs into four categories: (1) commercial objectives; (2) commercial objectives and ensuring head office functions in Norway; (3) commercial objectives and other specific, defined objectives; and (4) sector-specific objectives. 10 This classification of particular SOEs is periodically updated by the Ministry of Trade, Industry and Fisheries, and indeed SOEs have occasionally moved between categories. Other data sources that could be harnessed for largescale analyses include the Orbis cross-border investment database 11 that also provides "deal comments and rationales" that might be amenable to coding and content analysis, for instance in order to compare the rationales of SOMNEs and POMNEs. Studies such as Florio et al. (2018) demonstrate the value of considering such data, allowing to identify both financial and non-financial motives of SOEs.
While exploiting such large-scale data will be useful, it is also likely that IB and governance researchers will to a greater extent need to turn to the type of comprehensive primary qualitative research that characterized the early literature on SOMNEs, as exemplified by the work of Mazzolini (1979b), Anastassopoulos et al. (1987), as well as to carry out more longitudinal case studies and business history work (Amdam, 2009;Buckley, 2009;Jones & Khanna, 2006).
A more recent example of such research is Rodrigues and Dieleman (2018) longitudinal study of Brazil's Vale and its internationalization. Another interesting case is Norway's Statoil (now Equinor), that was established to play a key industrial policy role in the development of Norway's oil resources, but has since been assigned predominantly financial goals, been publicly listed and embarked on extensive internationalization (Gordon & Stenvoll, 2007;Ryggvik, 2015). Such work can provide crucial insights into how the goals of SOEs evolve over time, and the implications of this for internationalization strategies. Again, such longitudinal qualitative research may also include study of the political context where the actual goals for SOEs arise.

Conclusion
State-owned MNEs (SOMNEs) are an important current phenomenon in the global economy, and represent a research area intersecting international business, state ownership and corporate governance. This article has explained how a more detailed study of SOMNE objectives and their implications is likely to represent an important next step for the evolving literature on SOMNEs, by helping to resolve mixed and contradictory findings in the literature. The present article has taken some first steps towards promoting such a research agenda, by providing theoretical arguments, developing a comprehensive set of testable propositions, and discussing some ideas for empirical research.
Although further theoretical and empirical research is needed, our analysis already indicates some research, managerial and policy implications. Firstly, we have demonstrated how different, but plausible assumptions regarding SOMNE's goals can imply different international strategies. In contrast, previous literature has often cited different goals, but subsumed all of them into a single predicted effect of state ownership. From a research perspective, our analysis points to the need for further theoretical development, possibly further breaking down the broader categories of goals here into more specific objectives (e.g. different types of industrial policy or political goals), as well as exploring the implications of combinations of different goals. For the viewpoint of managers, a more fine-grained understanding of goals beyond the broad notion that SOMNEs have social and political goals will be relevant for the management of and interaction with such enterprises. Finally, from a policy perspective knowledge on how specific goals actually affect internationalization outcomes will be very valuable both to understand the costs of pursuing such goals, as well as the likely attainment of such goals in practice. A better understanding of the goals of SOMNEs also has potential implications for host-country and international policy responses to SOMNEs' international activities. Industrial *NR = not ranked; **In addition, the government holds so-called 'Golden Shares'; ***TNI (transnational index) is calculated as the unweighted average of the following three ratios: (i) foreign assets to total assets, (ii) foreign sales to total sales, and (iii) foreign employment to total employment.