It is widely acknowledged that multinational enterprise (MNE) decision-making is sensitive to the institutions and policies of the local context. Surprisingly, the level of engagement of the IB literature with the contemporary concerns of international political economy (IPE)Footnote 1 literature has been modest. Subsequently, the analysis of the role of the state and its relationship with the MNE remains underdeveloped, and the heterogeneity across political systems and this might influence the stance of political actors towards MNEs has not been adequately considered.

In exploring the MNE–state nexus, IB scholars have focused on developing a nuanced understanding of the different underlying motives for MNE investment (Cuervo-Cazurra & Narula, 2015). That these motives significantly influence the MNE’s interaction with the state (and other stakeholders) in the host country is well developed; for example, in studies about the impact of the quality of host-country institutions on MNE choices (Driffield, Mickiewicz, & Temouri, 2016; Jackson & Deeg, 2008). The literature also recognises that MNEs lobby host-country governments for specific changes to regulations and policies (Desbordes & Vauday, 2007; Hillman & Ursprung, 1993). Thus, it has become conventional wisdom that MNEs seek particular location advantages that match their objectives, and optimally complement their ownership advantages. Simultaneously, it is also understood that the nature of the location advantages, and the MNE’s access to these resources, is shaped by the specific policy stance of each nation state.

Our aim is to extend IB analysis to allow for a more balanced account of the two sides in the MNE–state nexus. On this basis, we proceed to examine the impact of heterogeneity in the political arrangements of the different states where MNEs locate. We seek to acknowledge that governments and their policy stances – like MNEs and their motivations – are heterogeneous and diverse. The policies of each nation state reflect different evolutionary processes, resource endowments and path dependence (Estrin & Grigoric, 2022; North, Wallis, & Weingast, 2009). This is important because the traditional IB approach to understanding how MNEs bargain with states has tended to work with a simplified abstraction of the state.

To illustrate these issues, consider the global auto manufacturer, Daimler, and its investments directed towards two of its largest markets: the US and China. The size of these two markets means that there are few comparable “outside options” for Daimler. Its entry strategy and organizational decisions in either market was fundamentally different, because of the differences in the objectives of the two governments. During the 1980s, the Chinese state required commitment by foreign investors to advance their national objectives of fostering innovation, technology transfer, nurture a domestic Chinese supply chain, and to actively collaborate with domestic partners. In other words, Daimler’s China operations were shaped not just by its commercial considerations, but also by the state’s priorities. Daimler has been expected to further Chinese industrial policy, and relies disproportionately on domestic suppliers and local expertise (Côté, et al., 2020). In the US, by contrast, they established a wholly owned subsidiary, entering the US market after a fairly standard negotiation over employment targets in exchange for relevant incentives and subsidies. Since entry, their operations have evolved organically, developing their supply chains based on economic factors and past ties. Their ongoing engagement with the state has been primarily to do with lobbying for location advantages (or mitigating potential disadvantages), to influence regulatory constraints, and to improve market access.

We make two different but related contributions to the literature on the interaction between host states and MNEs. First, we explicitly introduce the heterogeneity of states into the discussion, in particular, differences in the objectives of the various types of states. Drawing on Grossman and Helpman (1994) and the related political economy literature, we consider two key characteristics that shape the state’s objectives: the priority the state places on improving the welfare of its citizens, and/or the private benefits of the political elite (or other interest groups). We consider the likelihood that some states – depending on their political systems – will prioritise benefits to the political elites (and/or other interest groups), over the welfare of its general population. This enables us to add micro-foundations to the characterisation of the state. In turn, this takes us beyond the economic considerations of the host-country government and allows us to focus on the state’s purely political objectives; in contexts characterised by economic nationalism and populism, for example, these have special salience (Devinney & Hartwell, 2020; Lubinski & Wadhwani, 2020).

Our second contribution is to enrich the stream of IB research on bargaining between MNEs and the host-country governments. We extend the current approach by capturing the process of bargaining between the MNE and host country using the Nash bargaining framework (Nash, 1950). This rich framework allows us to emphasise the existence and nature of “outside options” that are available to both the MNE and the state. These are shown to significantly influence the outcome of the bargaining process,Footnote 2 enabling us to account for the possibility that domestic firms may offer benefits to host countries that MNEs are unable to match. For instance, domestic ownership offers political benefits to the host countries (Chari & Gupta, 2008),Footnote 3 and the state may stipulate the enhancement of the domestic sector as a priority. By extension, it enables us to consider factors such as path-dependent relationships between the state and domestic economic agents (Commander & Estrin, 2022) that can set the context for bargaining between MNEs and host-country governments.

In the next section, we review the existing literature on the interactions between MNEs and the state, and the interface between the IB and IPE literatures. We go on to present a novel framing of the MNE–state bargaining framework which explicitly considers heterogeneity of the state’s objectives in different political systems. We then formally analyse MNE–state bargaining using the Nash model and emphasise the impact of the outside options of each party. We conclude by discussing the broader ramifications of our analysis.

MNEs and political economy: A brief review of the literature

The political economy literature has a long tradition of considering scenarios where economic agents with competing interests lobby to persuade a government to design policies and institutions that serve their own interests. Indeed, lobbying by interest groups has been used to explain a variety of outcomes such as trade policy (Grossman & Helpman, 1994), pollution taxes (Fredriksson, 1997), barriers to new entry in industrial sectors in general (Perotti & Volpin, 2004), and inward investment in particular (Chari & Gupta, 2008), as well as government support for ailing sectors within an economy (Baldwin & Robert-Nicoud, 2007).

Within this line of inquiry, the literature on MNEs has discussed the use of MNE–state (micro-level) bargaining to alter state policies to which they are especially sensitive (Desbordes & Vauday, 2007; Hillman & Ursprung, 1993). For example, they may attempt to influence key input costs, from labour standards and minimum wages to incentives and subsidies, or strengthen local intellectual property rights to protect the technological sources of their competitive advantages. In addition, the widespread institutional voids that characterise many developing economies may widen the scope for MNE influence on policy, sometimes leading to regulatory capture, which significantly alters the potential impact of MNE investments.

The IB literature specifically discusses three elements of this negotiation process in detail. First, the literature on MNE–state negotiations (see Boddewyn, 2016; Eden, et al., 2004) focuses heavily on the relative bargaining power of the actors; the lower (higher) the bargaining power of the MNE (host-country government), the less (more) it is enabled to achieve its objectives. Power imbalance between an MNE and host-country government arises from a number of sources such as monopoly control over resources and technology (Kobrin, 1987), resource complementarity between host-country resources and ownership advantages of the MNE (Luo, 2001), timing of the MNE’s entry into the host country (Eden & Molot, 2002), whether or not the MNE belongs to strategic network partners (Nebus & Ruffin, 2010), and the relationship between the MNEs and their home-country governments (Li, Newenham-Kahindi, Shapiro, & Chen, 2013). Irrespective of the source of power (or prominence), the relative power of the players (or actors) may change over time (Boddewyn & Doh, 2011; Kindleberger, 1969; Mullner & Puck, 2018).

Second, in contexts where these negotiations are undertaken ex ante, i.e., lobbying before the MNE makes an investment, there is potential for a hold-up problem. Prior to investing, MNEs may require assurances from the government about access to markets and resources, and for the host country, about technology transfer from MNEs (Eden et al., 2004). The MNE often has the upper hand at this point in the negotiation, because of their ownership advantages, such as proprietary technologies, or privileged market access. This may lead the host-country government to offer major concessions in order to secure the investment. However, once the MNE has established operations in the host economy, the relative bargaining position of the two parties changes in favour of the state because the MNE has now sunk the assets, rendering withdrawal complex and expensive. This may lead the host-country government to renegotiate the concessions originally offered, even if acting in bad faith carries reputational repercussions for governments. Indeed, even governments that otherwise actively seek FDI may be tempted to renegotiate the terms of investment in such situations. The attempted renegotiation of the contract between Enron and the local stakeholders of its Dabhol Power Station in India is such an example. This hold-up problem is termed “obsolescing bargaining” in the IB literature.Footnote 4

The obsolescing bargaining framework can be extended to treat the relationship between individual MNEs and host countries as a multi-period and dynamic bargaining process (Ramamurti, 2001).Footnote 5 In this two-tier bargaining framework, tier-1 (macro-level) negotiations reduce the scope for host countries’ opportunistic behaviour, that underlies the obsolescing bargaining problem. They also establish boundaries that constrain the tier-2 (micro-level) negotiations between the MNE and host country in their pre-investment agreements. International investment treaties can be gauged through this lens. These treaties are negotiated to ensure that the property rights of inward investors are peremptorily protected, prior to project-specific negotiations between MNEs and host-country governments (Egger & Pfaffermayr, 2004).Footnote 6 However, such treaties are themselves subject to renegotiation, making them part of a dynamic political economy process.

Third, the literature that examines the interaction between MNEs and host-country governments involves discussions of non-market strategies employed by MNEs to establish and manage their relationships with the respective host-country governments (Mahini & Wells, 1986). Specifically, these strategies are used to mitigate the challenges associated with the liability of foreignness and outsidership; these may be higher for certain types of MNEs such as state-owned enterprises from emerging market economies (Cuervo-Cazurra, 2018; Meyer, Ding, Jing, & Zhang, 2014). To a significant extent, the literature focuses on the use of corporate political activities and corporate social responsibility activities (Mbalyohere, et al., 2017). However, there may also be other forms of engagement with host-country governments such as soft money contributions to political parties, and leveraging the soft power of home-country governments (Rajwani & Liedong, 2015).

As suggested by our discussion, the otherwise rich literature on MNE–host-country government relationships explores this relationship largely from the strategic perspective of MNEs. The objectives of the host-country government are certainly not completely ignored, but largely viewed through the economic perspective that emphasises potential benefits like technology transfers that accrue from MNEs. However, as stressed in the literature on political economy, the political objectives of governments are a significant factor in shaping how rents and surpluses are shared by the state, depending upon its economic and political priorities. In addition, political stances can result in the adoption of ideologically motivated policies despite their potentially adverse economic impact; examples include the US–China trade war and Brexit. Indeed, the importance of heterogeneity in “government objectives” has been acknowledged in the IB literature (Cuervo-Cazurra & Li, 2021; van der Straaten, Narula, & Giuliani, 2023) and the wider management literature (Jacobson, Lenway, & Ring, 1993), but is yet to meaningfully feature in the discussion about MNE–host-country government relationships. We develop this perspective below.

A novel MNE–state bargaining framework

International political economy and the objectives of the state

States (or governments) can have a variety of objectives which may be conditional on the political system and constitutional arrangements. This issue has already been considered in the IB literature with efforts to classify systems based on institutional arrangements (e.g., Fainshmidt, Judge, Aguilera, & Smith, 2018). Instead, our analysis draws on political theory studies which conduct the analysis and categorisation of political systems (e.g., Acemoglu & Robinson, 2023).

In this study, state objectives refer to the extent to which the state prioritises the welfare of its population (as opposed to the welfare of its ruling elite, or other interest groups within the economy). Though imperfectly, this aligns with the traditional distinction between democracy and autocracy (Acemoglu, 2008; Przeworski, 2004). We follow Przeworski, et al., (2000) in focusing on competition for power, or contestation as the defining characteristic of democratic political systems. To quote Przeworski et al., (2000: 15) “democracy…is a regime in which those who govern are selected through contested elections”. In contrast, autocracies are characterised by a single political party, ruling out the possibility of government alternation from elections. This implies that democracies are more sensitive to the welfare of the population, at least to the extent that voting is associated with individual welfare. Of course, this is an oversimplification: autocratic states have been known to prioritise general welfare improvement, while democracies often also emphasise the welfare of political or economic elites over that of the general population. We will revisit this conjecture later, both here and in the discussion section.

Indeed, democracy versus autocracy is rarely viewed as a binary classification of political systems, but as a continuum of possibilities, which include hybrid democratic–authoritarian regimes (Knutsen & Nygard, 2015). As we have noted, the IB literature considers institutions a crucial element of the contextual factors influencing MNE choices. However, political systems as used here, are distinct from these institutional variables. This can be seen in Fig. 1, which plots the correlation between autocratic and democratic political systems in a continuum, and standard institutional indicators like the rule of law and corruption, for 163 countries in 2016. While there is a positive correlation between “good” institutions and democracy, there is also considerable heterogeneity.

Fig. 1
figure 1

Rule of law versus democracy

Identifying the objectives of the political elite requires an understanding of the distinction between democratic and autocratic political systems in a political economy framework. By establishing this as a continuum, the development of a simple heuristic framework is enabled, wherein the objective of the state can be viewed in terms of its utility (U), which comprises two elements. The first is the welfare of each person (voter) in the society, which we think about in a utilitarian way by simply adding up the utility (u) of each individual i in the society (ui). Hence, when a state is “perfectly” democratic, we assume that politicians maximise the welfare of the society by maximising the sum of individual welfares. That is, when the market for political power is fully contestable, political elites (who form governments) cannot earn rent, and the entirety of the surplus goes to improve the welfare of society.

The second component of U are “private benefits” (PB). These accrue to the political elites (and other interest groups, on whom the government depends for political power). In a purely authoritarian state, these private benefits may accrue to a single person, though he or she may then distribute them more widely to their political supporters and networks (Przeworski et al., 2000). In such an authoritarian regime, the private benefits of the political elites are maximised, and the welfare of society is zero.

The objective function of the state can then be characterised as:Footnote 7

$$U \, = \, a\left[ {\sum_{i} \left( {u_{i} } \right)} \right] + \left( {1 - a} \right)PB$$

hence for a perfectly democratic regime, a = 1 and since private benefits do not enter the state’s objectives U = ∑i (ui). In a “perfect autocracy” where power is not contestable and citizens do not have meaningful voting rights, surpluses accrue entirely as private benefits. Hence, in a “perfect” autocracy, a = 0 and U = PB.

However, such extreme circumstances are rare. Most actual countries have a values somewhere between zero and one. In practice, autocracies usually pay a certain amount of attention to the aspirations of their populations, while political elites and other interest groups in most democratic states get some private benefits, legally or otherwise.Footnote 8 In general, democracies will have a noticeably higher a value than autocracies. This conjecture finds support in numerous estimations of the parameters of the Grossman–Helpman model; the relative weight assigned by the government to citizen welfare is higher for countries like the US, than for countries such as China and Turkey (Goldberg & Maggi, 1999; Mitra, Thomakos, & Ulubasoglu, 2002). Further, in countries like Turkey that have experienced years of autocracy followed by years of democracy, “the weight on welfare relative to [political] contributions [that increase the likelihood of the incumbent government holding on to power] was higher in the democratic regime than for dictatorship” (Mitra et al., 2002: 507).

It follows that when a lies in the (0, 1) continuum, the government will attempt to generate some rents by way of control over a range of factors that affect an MNE, such as resources, regulations, and policies. This rent can then be redistributed according to its objectives. For example, the host-country government may press the MNE for the transfer of technology to domestic firms that may otherwise not be made available, or to employ more workers. In the first case, such technology transfer may benefit specific domestic industrialists and in the second case, greater MNE employment helps improve general welfare. The MNE, in turn, will lobby the host-country government to help create locational advantages (such as infrastructure investment) that also result in rent: a return that is greater than the MNE might expect in the absence of those locational advantages. However, the rent would be created only if both parties make the necessary investments; we elaborate on this in the next section. In order to capture the structure of this interaction between the MNE and the host-country government, we propose the use of a variation of the Nash bargaining framework (Nash, 1950).

The Nash bargaining model

We build on the non-cooperative bargaining model developed by, for example, Sutton (1986) and Binmore, Shaked and Sutton (1989). We consider two actors,Footnote 9 the MNE and the host-country government, and make the meaningful assumption that left to its own devices, each party wishes to make the largest possible gain (defined in terms of their own objectives) from the MNE’s investment. Figure 2 offers a simplified visual of the processes underlying our model. The two actors each undertake a series of observable actions that are underpinned by unobserved strategies that affect both their own payoffs and those of the other party. For example, the MNE might choose to upscale its business activities from exporting to FDI, or it can decide to exit. Which actions it chooses depends upon a variety of factors, including changes in its portfolio of global assets and operations, and/or changes to formal institutions, regulations and policies in the host-country context. The host-country government may, similarly, choose to implement regulations and promulgate policies in line with its objectives. The key issue is that the payoffs to each of these two parties will be affected by the actions of both the parties and, by extension, the objectives of the other.

Fig. 2
figure 2

A schematic representation of factors shaping MNE–state bargaining in a Nash framework

Consider, for example, the strategy of an MNE undertaking a particular acquisition in the host economy. There are a range of possible outcomes depending on the actions of the MNE and host-country government. At one extreme, the MNE may make the decision not to invest, for example because the host country has announced a strategy to highly tax any firm from a particular home country. In this case, the payoff to each side is (0, 0). At the other extreme, the host-country government may provide initial subsidies (S) but then tax the MNE (at a rate t), and on that basis the MNE may decide to make an initial investment, and earn profits (P) after initial setup costs. If the objectives of each party are entirely financial, then the payoffs to the MNE and the host-country government in this case are (S + (1-t)P: tP -S). It is easy to see that the MNE’s payoff, whether 0 when it decides to stay out of the host country, or {S + (1-t)P} when it enters, depends on the government’s actions around tax and subsidies which, in turn, depends on the government’s objectives. As outlined in the previous section, he MNE’s actions and subsequent payoff will be influenced by the relative weights the host-country government attaches to two factors: its private benefits and the prosperity of its citizens.

In order to formalise the Nash bargaining framework, let us presume that the payoffs of the MNE are denoted by a set of utility pairs of all possible deals (upon which they are able to agree) between each party: X1 for the MNE and X2 for the host-country government, as well as by a “disagreement pair,” D1 and D2, respectively. The disagreement pair represents the payoffs that the government and MNE will each receive if one or the other simply walks away from the deal and there is no agreement. Thus, for each party, the disagreement payoff represents their respective outside option.

If the MNE decides not to invest, its outside option is not necessarily zero, but rather the return it will earn by investing into the next-best available location. This will clearly vary according to how unique the locational advantages of the host economy are (Fig. 2). For the host-country government, the outside option may be the return it would obtain by diverting the resources originally set aside for the MNE, to the next-best investor (foreign or domestic). From a purely economic perspective, this is contingent on how many other MNEs or domestic actors with comparable competences and resources are interested in investing in the country. However, as we discuss later, for a host-country government that attaches significant weight to its private benefits, or believes that the interests of the citizens are best served by pursuing an inward-looking industrial policy, the outside option may come from domestic industrialists (Fig. 2). The key insight is that the gain from any particular bargain is contingent on the range of alternatives available to each party.

To take the analysis further, we need to consider how the payoffs are actually allocated between the two parties. In the original Nash (1950) model, the solution to the bargaining problem is very simple; the two parties split the returns equally. Thus, the bargaining outcome is the pair of payoffs to each party, (X1*:X2*), which maximises the product (Z) of the rents of each party, that is to say their net gains (gains net of the disagreement payoff) are:

$$Z = \left( {X_{1} {-}D_{1} } \right)\;\left( {X_{2} {-}D_{2} } \right)$$

The Nash bargaining solution therefore assigns to each party their disagreement payoff plus one-half of what remains of the total amount generated by the bargain after the disagreement payoffs have been made, though their own net gains may be higher or lower, depending on the relative return on their best strategy as against the outside option. The framework is essentially one of rent-sharing, when rent is the excess payoff each of the two parties receive over and above the payoffs associated with their respective outside options. For instance, in a resource-rich host country, if an agreement between an MNE and the host-country government results in exclusive access to a resource by the MNE, this will create such a “rent” for the MNE. The rent for the host-country government can be similarly explained.

Note that the rent will be generated only if both parties make the necessary investments in context-specific assets that are necessary to generate rent. For example, a host country may have to invest political capital and resources for the creation of formal institutions, subsidies and incentives, and physical infrastructure needed by the MNE. Similarly, the MNE may have to ensure that its technology can be adapted to host-country conditions, and it invests in building relationships with local supply chains to facilitate technology transfer and improve efficiency. The bargaining process can, therefore, be viewed as a two-part process (Fig. 2). In the first, the MNE and the government make context-specific investments in their respective capabilities. In the second, they decide whether to cooperate, conditional on their respective outside options. If they decide to cooperate, the combination of institutions, resources and technology, among other things, creates rents for both the MNE and host-country government. The rewards to each depend on what strategies are played, i.e., the instruments or policies chosen, as well as the alternative possibilities available to each side. To reiterate, the objectives of the government will be reflected in their strategies and associated actions or policy/legislative choices.

A comparison of existing frameworks with the proposed one is provided in Table 1. To recapitulate, in extant IB literature, the MNE–host-country relationship and the associated negotiating process were viewed through the lens of obsolescence bargaining. More recent literature discussed the role of pre-entry negotiations between the governments of the home and host countries of the MNE, and those between host-country governments and international organisations to ensure that formal institutions were in place to pre-empt post-entry renegotiations. Upon entry, the MNE itself might have used non-market strategies to reduce the likelihood of the contract renegotiation, especially where it concerned the weakening of property rights and/or other forms of expropriation of the MNE’s assets and associated profits. The analysis takes into consideration the (exogenously determined) differential bargaining power of the MNE and the host-country government, and focuses on MNE strategies. The state, by contrast, was assumed to focus on economic gains alone. Further, while the existing literature discussed the role of outside options, the discussion was not linked to net gains from the bargain. In the traditional framing, the disagreement pairs (D1, D2) are normalized to zero and the bargaining power, which we denote as β, is identified through a single parameter. Hence, the two parties maximised a simple payoff function, Z = (X1)β (X2) (Nebus & Rufin, 2010). On this basis, IB scholars have considered numerous factors determining the return to the MNE (X1). However, since the pioneering work of Stopford and Strange (1992), there has been little explicit analysis of the objectives of the state in the MNE–state bargaining framework.

Table 1 Comparing the existing framework(s) with the proposed framework

In our proposed model, bargaining power is endogenous, depending on the outside options for each party. Outside options are central to our proposed framework. In addition to permitting explicit analysis of the state’s objectives and strategies, our bargaining framework encompasses many of the extant ideas in the IB literature on bargaining. In particular, denoting player 1 as the MNE and player 2 as the host-country government, the costs that could be imposed by host governments on MNEs (Vernon, 1971) can be captured through X1 – the payoffs to the MNE. As discussed earlier, the choice of this action will depend on the objectives of the host-country government. The literature on non-market strategies (e.g., Doh, McGuire, & Ozaki, 2015) can also be interpreted in terms of MNEs seeking to improve both their set of potential strategies (X1) and increase the cost of the outside option (D2) for governments. This analysis represents an organising framework into which we can reframe the existing rich literature about MNE–government relations, and augment this analysis with the explicit treatment of the objectives and outside options of the government.

Expanding on the heterogeneity of state objectives, strategies and outside options

Figure 2 succinctly summarises the key aspects of our revised model. The key contribution to the literature on MNEs and development is that to understand how MNEs might affect the welfare of the host location, scholars and policy makers need to be clear about whose welfare is prioritised: whether that of the larger population, or the private benefits of specific subgroups. In a democratic state, with governments typically attaching a higher weight to the welfare of the country’s citizens, investment policies tend to focus on economic growth and development. In such an environment, the government looks to MNEs to help achieve key economic objectives, such as employment and income growth, reducing inequalities, building up the domestic firm sector, and enhancing national competitiveness.

What strategies might such states bring to bear to increase returns to the MNE, thereby increasing payoffs from FDI? First, the host economy needs to differentiate itself from other countries as a location for FDI. The greater the attractiveness of the location to MNEs, the fewer will be the outside options for the MNE. The state may seek to differentiate itself by offering more specialised location advantages, in terms of the availability and quality of science and technology infrastructure, and the provision of specialised human capital. Depending upon the state and its resource constraints, it may engage in providing location advantages such as favourable incentives and subsidies to improve its bargaining position, thereby reducing the outside options available to the MNE. The IB literature abounds with examples of tax policies, special economic zones, and discretionary favourable treatments designed to increase the returns to FDI (Côté et al., 2020a, b; Narula & Zhan, 2019).

Weak location advantages act to deter MNEs, and these are often associated with weak institutions in developing countries. States that emphasise the welfare of their citizens will be motivated to improve the quality of their institutions, both to increase payoffs to MNE strategies in their location, and differentiate themselves from competitor countries offering alternative host locations. Thus, the state can also choose the quality of the institutional arrangements facing the MNE in the host economy, being aware that stronger and more effective institutions are associated with lower transactions costs that are deemed desirable by MNEs. This leads us to expect that states which prioritise their citizens’ welfare will improve institutional quality as a policy response to the expectations of MNEs, and that the pressures from competing states will lead democratic countries to attempt to enhance institutional quality more than their neighbours: a “race to the top.” MNEs can, of course, facilitate the process by explicitly bargaining with the host-country government (Papageorgiadis, McDonald, Wang, & Konara, 2020).

However, states that emphasise the private benefits of the ruling elites over the welfare of its population (which are often, but not always, autocracies) behave differently (Shleifer & Vishny, 1998). The political elites are largely interested in the private gains brought to themselves and the people within their networks, sometimes in the form of personal income and/or non-pecuniary privileges that are embedded with political power. These states, therefore, have limited incentive to improve location advantages, especially where improved institutions cause greater contestability of economic resources (and the political space), and reduce the rent for the political elite and their networks. This will make such states reluctant to credibly enter the (Nash) bargaining process (Ramamurti, 2001).

All types of states along the spectrum from democracy to autocracy have an alternative to MNE investment in the form of investment by domestic firms and investors, which may be a component of an active industrial policy that fosters the creation of national champions, reflecting a political economy that emphasises building up domestic industrial competitiveness (Fig. 2). Even where the MNE has superior ownership advantages to these domestic actors, and the domestic actors are clearly unlikely (at least in terms of linkages or spillovers) to have as large an effect on development, there may be non-economic and strategic reasons that generate a preference for domestic investors over MNEs. There is also a case to be made that domestic actors can make more credible commitments to augment the private benefits of the political elites (Perotti & Volpin, 2004). Indeed, there is evidence to suggest that in many contexts, domestic firms are willing and able to provide the autocrat with the relevant (private) political benefits, perhaps in return for domestic protection, state grants and subsidies and soft loans from state banks (Commander & Estrin, 2022). The outside option for governments that place a higher weight on private benefits can be characterised as entry barriers for MNEs (Acemoglu, 2008). It is, therefore, unsurprising that relatively more autocratic systems such as in Russia, tend to operate with large domestic monopolies and oligarchic business groups, operating for the most part behind protective trade walls and with limited FDI.

Discussion and conclusions

Broadly defined, the field of IB has arguably come into its own over the last 50 years. Two of the hallmarks of this period have been the globalisation of the world’s economy (taken to mean the growing cross-border interdependence of markets for goods, services, and capital), and, more recently, the broader acceptance by policy makers and academics of MNE-assisted development as a legitimate means to promote economic growth and reduce inequality (Narula & Pineli, 2019). Indeed, as argued by Boddewyn (2016: 13), the emergence of the “Washington Consensus” was characterised by states perceiving MNEs as a means to generate and sustain national competitive advantages.

It is not always clear, however, whether all states consider economic development a primary goal, or raising the income levels of the general populace to be their desired outcome (van der Straaten et al., 2023). While policy choices may be influenced through lobbying by organised interest groups such as domestic industries and MNEs, in democratic contexts they may also be influenced by electoral considerations and voter preferences (Glazer, Gradstein, & Konrad, 1998; Quinn & Toyoda, 2007). We have explicitly introduced political choices that shape the economic priorities and associated policy frameworks that states decide to implement. Further, even in contexts where economic development and raising the income levels of the general populace remains important, there is increasing emphasis on wider objectives such as supply-chain resilience, geopolitical considerations and (reshoring of) good jobs for the local population (Rodrik, et al., 2023), and reducing inequality at multiple levels (van der Straaten et al., 2023). This introduces significant heterogeneity in the objectives of the state, and alters the basis of MNE–state negotiations, and by extension, the possible outcomes. A more comprehensive discussion about the bargaining between MNEs and host-country states requires considering both cross-sectional heterogeneity in state objectives and changes in state objectives over time.

Policy stances are not unmoving or carved in stone. Plebiscites, elections, domestic upheavals, or revolutions can result in radical shifts from one policy stance to another as seen in countries as diverse as the US, India, UK, and China in the last decade. The economic priorities of states can also be part of a deliberate recalibration of policy. States can prioritise a domestic focus that emphasises job creation and the building up of national champions in one period, switching to a focus on MNE investment and international competitiveness in a subsequent one. States can also shift either gradually or suddenly from democracy and MNE-friendly policies towards populism, techno-nationalism and autocracy, and from within those two stances, from a policy emphasis favouring broad economic prosperity, to (or away from) policies that focus on improving the welfare of a narrow political and economic elite (van der Straaten et al., 2023). In several European countries there has been a substantial increase in the vote share of populist parties over the past 20 years.Footnote 10 Populism has also emerged in other parts of the world,Footnote 11 in addition to a rise in economic nationalism (de Bolle & Zettelmeyer, 2019). This has implications for MNE strategies (Lubinski & Wadhwani, 2020).

In comparison to states, MNEs are less given to major swings in strategy and objectives. MNEs rarely exist (nowadays) as a loose federation of affiliates, and work most often as a network of affiliates with common corporate goals and objectives, and a unified set of norms that oblige them to be predictable, aware that any large deviation would adversely affect their reputation as reliable partners. Indeed, while earlier IB literature modelled MNE investments during the import-substitution era as “miniature replicas” of the home-country operations (Pearce, 1999), with liberalisation, MNEs have increasingly resorted to complex network structures with a fragmentation of the value chain that optimises scale, scope and intensity of activity to avail of the most appropriate global location advantages. This has been further exacerbated by the growing use of global value chains (Strange & Humphrey, 2019). In the context of bargaining between MNEs and host-country states, therefore, the heterogeneity of states, as opposed to heterogeneity of MNE objectives, may now have greater salience.

The more recent development-oriented literature has taken a more nuanced view, acknowledging that the state and its actions can be driven by domestic priorities related with increasing welfare and reducing inequalities, just as much as building and sustaining the competitiveness of economic actors (Cárdenas, et al., 2016; van der Straaten et al., 2023). However, this literature does not fully account for decisions and positions that host-country states might take purely on grounds of political imperatives that have implications for the private benefits of the political elite.

Our analysis and the model lend some structure to understanding this heterogeneity and how the choice of policy might be influenced by the relative weight the host-country government attaches to the welfare of its citizens. It also tells us how the outcome of the bargaining between an MNE and the state is affected by the outside options available to the state, including options offered by domestic interest groups (Fig. 2). The literature on political economy tells us how and why states are likely to change their minds about MNE investments; this necessarily changes the attraction and location advantages of a particular country, a critical issue for IB. The economics literature has long noted that it is the stability of policy that matters, almost as much as the policies themselves. The uncertainty created by policy instability can have a significant adverse impact on investment and, by extension, economic growth (Rodrik, 1991). MNE investments are contingent on the states’ provision of location advantages, and their availability for what is a long enough period for the MNE to achieve its objectives. There are implications for the state’s development agenda as a result of policy instability: an MNE’s increased embeddedness in the local economy is made on the assumption that the policy stance of the host government stays stable and predictable. Our framework provides a micro-foundation that accounts for policy (in)stability in host-country states.

The framework can accommodate specific types of states that have been discussed in the literature. For example, the “developmental state” (Slater, 2010), a term often associated with East Asian economies that are, or have been, fairly autocratic from a political perspective but still place considerable emphasis on rapid economic development and, by extension, on the welfare of their citizens. Examples include China, where the state’s legitimacy and political power stems from its ability to industrialise and create jobs (Commander & Estrin, 2022), or South Korea since the 1970s, with a (then) autocratic regime staking its future on export-led growth (Amsden, 1989). While these governments are generally keen to engage with MNEs, they are also keen to manage the process by liberalising policy regimes for the “right kind” of investment, conditional on the transfer of specific types of technology, imposing restrictions on MNEs to maximise long-term spillover and linkage effects. In a populist state, by contrast, governments generally pursue policies that are popular for an electoral majority in the short run, even if these may be unsustainable in the long run (Devinney & Hartwell, 2020). The emphasis on local political and ideological objectives may lead them to seek disengagement from global supply chains and emphasise domestic employment targets over international competitiveness and trade integration. The trade war with China begun by the Trump administration is an example of a set of policies that fall into the populist category. Arguably, an MNE's liability of outsidership is much higher in a populist state: such governments are suspicious of inward FDI, especially when it is associated with market dominance by foreign firms. Importantly, in both these types of states, government policy stems from factors that affect the political survival of these governments and this, in turn, has implications for what an MNE can achieve through bargaining with the host-country government.

We have highlighted the role of outside options. The principle of outside options from an MNE’s perspective is hard-baked into the concept of location advantages, which are always relative to the location-specific assets of other spatial configurations. Different regimes (or the change of regimes) further highlight the MNE’s outside options, particularly for factor inputs such as labour. Labour costs rise only by small percentages over time, and their implications for MNE productivity remain tiny. Conversely, a simple policy reversal (say, restricting FDI in a strategic sector) due to regime change can wipe out the MNE’s gains for several years into the future.

Certainly, the outside options of both parties vary over time, and bargaining stances will naturally also change. For instance, a greenfield industrial park in a remote province has fewer outside options in its early years, compared with one that is established and is already home to several turnkey investors. However, states are also resource constrained, and in this age of footloose firms, the MNE typically has many more outside options in terms of locations, especially when it comes to “generic” inputs and location advantages. The rarer and more specialised the kinds of location-specific assets, the more limited the outside options for the MNE (Narula & Zhan, 2019).

MNEs often actively seek to shape policy. Therefore, their non-market strategies do matter. However, the literature arguably assigns too much importance to lobbying. Indeed, it is reasonable to suggest that lobbying matters only on the margins. The “big” issues, the immoveable planks of policy orientation are largely shaped by fiat, and lobbying simply helps alter the orientation, the details of the broad planks, but not the planks themselves. To permit or exclude foreign capital in a particular sector, for example, is ordinarily in the executive purview, and it is in the implementation of such policies that exceptions may be made through lobbying. Broad policy decisions are often exogenous to the MNE because they are shaped by political factors that affect the likelihood of a government’s survival (e.g., Fatas & Mihov, 2003).

Directions for future research

It is immediately obvious that our framework can help enrich the literature on how MNEs can use non-market strategies and corporate diplomacy (Li, Meyer, Zhang, & Ding, 2018; Mellahi, et al., 2016) to shape host-country institutions. Such analyses would have to account for the boundary conditions of non-market strategies that are dependent on the objectives of the host-country state. Given the heterogeneity of state objectives and changes to these objectives over time, our framework paves the way for rigorous novel empirical analysis in this domain. The scope of future analyses can be further widened and enriched when we take into consideration the increasing heterogeneity in the nature of MNEs themselves, with family-owned (and often business group-affiliated) firms and state-owned firms from the emerging market economies and commodity-rich countries joining the ranks of the privately owned and, generally speaking, widely held MNEs of the industrialized world.

We have not sufficiently engaged with the issue of the state’s resource constraints, which in our framework influence its potential strategies and payoffs. Resource constraints matter as much as political constraints when it comes to bargaining with MNEs, for the simple reason that there are cognitive limits to the kinds of location advantages the state can offer. This can be in terms of incentives and subsidies, where developing countries are rarely able to compete for critical anchor investments. Location advantages can be “exogenous” since they necessarily build upon the extant economic structure of the country, its natural resources, and comparative advantages. Meaningfully and realistically assessing these resource constraints is essential for governments to tailor the nature of their de facto location advantages to the kinds of investments they seek to attract. By extension, recognising the resource constraints limits the outside options available to the state. Understanding the associated trade-offs has considerable implications for policy.

The literature on MNE-led development has largely assumed that states are developmental in nature, or, at least, consider that the economic prosperity of society at large matters to the state in MNE–state negotiations. The evidence indicates, instead, that despite widespread economic liberalisation and rising engagement with MNEs, there has been plenty of export growth, but little evidence of overall economic development in the form of reduced within-country inequality. This continues to perturb scholars and academics (Narula & van der Straaten, 2020). The heterogeneity of state policies, as well as the dynamics of these policies, driven by evolving domestic priorities and politics will help address this puzzle.

Our acknowledgement of the heterogeneity of the state enables a richer appreciation of the policy choices of governments. Although our model has taken a stylised view of how states prioritise private benefits versus general welfare, in reality. states can be seen to share surpluses beyond just the binary choice of political elites and the wider population. This may reflect industrial policy; for instance, the creation (or strengthening) of domestic industrialists as a pathway to building international competitiveness, or it may reflect regulatory or political capture by traditional landowners, as we highlight in Fig. 2. This is a fruitful area for future research, to reflect on the conscious and unconscious calculus of rent-sharing within different economies, and at different points in their histories (Fishwick, 2019; Jenkins, 1991).

How states assign weights allows researchers in this domain to focus on well-established measures of social welfare and subjective well-being (Fleurbaey, 2009), specific elements of government policy that are reflected in, for example, the budgetary commitments of governments to health, education, social insurance and regional development, and its commitments to addressing various forms of inequality. The move to a continuous measure of the state’s objectives will enable researchers to explore a wider range of questions such as potential threshold effects, whereby some bargaining outcomes are only possible if the government has a certain minimum level of commitment for citizens’ welfare.

We have not discussed inter-regional differences within a state. From an empirical perspective, heterogeneity in the objectives of regional governments as well as their location advantages can potentially be measured, subject to availability of budgetary and/or other data. Research on inequalities has highlighted that regional differences in location advantages can constrain development, and where states intentionally emphasise one region over another in terms of resource allocation, this can lead to long-term consequences in terms of one region’s ability to leverage MNE investment relative to another. This forms a valuable area for further research.

Our characterisation of the state’s objectives also has implications for the way in which the state’s outside options can be modelled within an empirical framework. While recent literature has acknowledged that the state’s objective may have a considerable political element, and how outside options play a role in the context of bargaining (Li, Assche Van, Li, & Qian, 2022), our analysis brings to the fore the possibility that the “outside option” for a host-country government may come from within, especially if the political elites attach a higher weight to their own private benefits than to the welfare of the citizens. This, in turn, suggests that there may be considerable within-country variations in the bargaining outcome. Of course, this depends on sectoral differences in market structure – competitive, oligopolistic, (quasi) monopolistic – and the presence of large domestic private and public sector players who earn rents that can be shared with the government.

Finally, an important dimension which we have not addressed here are the obligations of states to comply with legally binding supranational, multilateral and bilateral treaties, and the extent to which these treaties shape the degrees of freedom that states have in their negotiations with inward investors. For instance, some states find ways to selectively comply with multilateral treaty obligations (such as within the WTO) to protect intellectual property rights, maintain labour standards and reduce emissions. MNEs seeking to lower costs look for such locations that are willing to only lightly enforce environmental, social and governance (ESG) regulation. However, the growing engagement of civil society and other stakeholders means that states have less leeway to offer lower ESG standards as a location advantage. These external obligations also act to constrain the bargaining position of states, by altering their outside options.