Introduction

The award-winning paper by Goerzen et al. (2013) was a pioneer in taking geographic scale down to the city level at a time when the majority of location decision research remained stubbornly at the country level (Iammarino & McCann, 2013; Meyer et al., 2011). It was against this backdrop of aggregate-fixated studies on location choices (Goerzen & Beamish, 2003) that researchers’ poor grasp of the determinants of location decisions at lower scale, especially cities, was becoming increasingly problematic. In the world of practice and policy, it had never been doubted that investment policy and promotion could be far more granular (Lewis & Whyte, 2022; Sanchiz & Omic, 2020) and that firms conceive their location strategies in terms of cities, whether they be global cities or not, as identified in the original Goerzen et al. (2013) study.Footnote 1 Their paper on the determinants of subsidiary locations inside or outside global cities arrived just at the right time, bearing a novel question that had been rarely addressed in the IB literature as of 2013.

Drawing on economic geography and international business strategy, our Decade Winner was able to take the first steps to combining higher-order locational qualities with the strategic logic behind firms’ location decisions. By shedding light on “exotic” features of global cities—global connectedness, cosmopolitanism—along with the abundance of advanced producer services and the international business staple “liability of foreignness” (Zaheer, 1995), the authors served us an augmented model and a deeper explanation of observation and understanding of received theory. We learned that firms localize their subsidiaries in global cities to help overcome the liability of foreignness associated with the host country location. The reader immediately appreciated that it was necessary to break with the past and think about market entry and production decisions on more than a single level.

A decade on, technological, geopolitical, natural, and other disruptions justify reassessment of the 2013 findings. Salient disruptions include the UK’s fraught withdrawal from the European Union, “Brexit”, techno-nationalism and the rivalry between the US and China, global warming and the Sustainable Development Goals’ (SDG), COVID-19, and war in Europe. At the time of publication, no challenge was so acute, but now these demand an update and redefinition of the relevant characteristics of global cities (Glaeser, 2022). These changed contexts have created barriers and frictions across national borders, economic blocs, between and even within global cities, lowering the degree of interconnectedness. Global cities have become more closed during the COVID period (Ai et al., 2020), sacrificing their cosmopolitanism.

In our view, when re-reading this paper today, more important than world events are factors largely latent 10 years ago that have now blossomed. Transitions within cities and their regions have become more visible, as global cities have shifted to be characteristically multifaceted while firms have become more ecosystem-like (Iansiti & Levien, 2004). These profound changes demand recognition and problematization. Without in any way pretending we know the answers, in this commentary we try to set the ball rolling by asking some relevant questions and sketching out the avenues we believe future researchers will find fruitful. We suggest the starting point for inquiry is that changes at the city and firm levels render ambiguous our current criteria with which to explain firms’ location decisions in global cities. We suggest that the origin for this is that firms are increasingly adopting an approach of localizing within overseas environments through networks of formal and informal partnerships. These strategies supplement the conventional foreign direct investment (FDI) on which Goerzen et al. (2013) was founded. The growth of these new patterns of localization, even if they do not overturn or materially change the findings of the original 2013 paper, have important implications for research. They demand data on the full range of ownership and collaborative value-adding strategies, including those that do not require equity ownership.

Our intention is to build on Goerzen et al. (2013) by putting forward the argument that, on top of the changing character of global cities, we must understand this shift in firms’ strategies toward internationalizing using what we term an “ecosystem approach.” This we define as firms’ entering a local environment through partnership and informal collaborations, without necessarily establishing any subsidiaries (Rong et al., 2015). We posit that it may be the growth of this approach that is at work in assisting firms to alleviate and overcome the liability of foreignness they are likely to encounter in foreign locations. We also present the topic of sustainability as a tangible symbolic move toward ecosystem thinking and discuss how sustainability might affect firms’ location decisions regarding their subsidiaries or, more generally, their operations inside or outside global cities. After offering some light and shade on Goerzen et al. (2013) contribution to the field, we propose some future research agenda items of promise for the next decade. We wish to leave the reader with the abiding conclusion that the work of Goerzen et al. (2013) has acted as a catalyst and a springboard to motivate new questions. These are not only for investment location but also the quality of investment and collaborative value adding operations, and for the impacts of these on a wider canvas than was envisaged in 2013.

Global cities and the location decision

Building on and extending the contribution of Goerzen et al. (2013)

The paper was innovative in analyzing the determinants of a MNE’s foreign subsidiary location decisions—conceiving of them as a choice between inside or outside a global city. For years, international business researchers had been tied to treating countries as internally homogeneous, with FDI driven by country-level determinants. This was largely because only aggregate universe data on FDI stocks and flows (usually official estimates) was readily available, particularly if the research question required an investigation across countries (Dunning & Lundan, 2008).Footnote 2 To move further down the geographic scale was a “holy grail” for researchers when a majority of research on the location decision was rooted at the country level (Goerzen & Beamish, 2003). While more niche data were available to unlock the sub-national level, this was invariably at the cost of loss of generality. This makes Goerzen et al. (2013) all the more remarkable, possibly unique, in securing the optimal trade-off between data limitation—choosing to be limited to a single home country and to conventional FDI—to gain the prize of theoretical insight and implications. As a rule, sadly, a general poverty of available data means empirical research persistently lags developments in theoretical knowledge, itself hampered by the inertia inherent in conventional IB theories. Yet the winner managed to buck this dismal trend by being in step with emerging thinking on geographic scale as a focused agenda item within the IB literature (Mudambi et al., 2018).

In the second decade of the 21st century, this paper was one of a select number—at the time dominated by conceptual or very focused empirical contributions—which viewed geographic scale at the subnational level (Chan et al., 2010; Ma et al., 2013; Nachum & Wymbs, 2005). The 2013 Special Issue of JIBS containing our award winner focused on the role of variations in the subnational geographic context in IB research (Beugelsdijk & Mudambi, 2013). It ushered in a new emphasis on scale, led by Goerzen et al.’s (2013) explicit focus on global cities (Sassen, 2012), which was a new high in terms of granularity.

Their study drew upon Kaigai Shinshutsu Kigyo Soran (2001 version) published annually by Toyokeizai, which includes comprehensive data on Japanese MNEs and their overseas subsidiaries globally, to support their reasoning that ‘global connectedness, cosmopolitanism, and abundance of advanced producer services’ assist MNEs in overcoming the liability of foreignness. With the passage of 10 years, the concepts of connectedness, cosmopolitanism, and advanced producer services appear as one-dimensional surface regularities atop a pile of hidden relationships. This is clearly an area on which to build. Yet, even in this seeming limitation, the 2013 winner is instructive, as it employed multi-level modeling to capture the nested nature of subsidiaries within MNEs. This methodological contribution bringing empirical inquiry into line with theoretical advances in modeling on different levels remains important today.

Innovation in measuring and examining the concept of inter-city international connectivity had to wait for separate projects by the co-authors a few years later to reach maturity (Asmussen et al., 2019; Belderbos et al., 2017). Given that the subsidiary’s liability of foreignness (LOF) increases with remoteness, a control variable for home–host distance again later introduced by one of the co-authors (Belderbos et al., 2017), would have been beneficial. As the authors themselves noted, variation between global cities also merits further attention, as not only the commonalities should be expected to influence firm behavior (Sassen, 2012). Despite the “wish list” above, the paper has clearly stood the test of time theoretically, methodologically, and empirically.

Re-reading the Decade Winner today, factors that were embryonic 10 years ago have acquired great salience. We should ask “would we get the same results today on the same—but updated—data, employing superior techniques, and what would those techniques be?” Yet more important for our scientific research: what more sophisticated data might be employed? Recontextualizing the findings and insights of the award-winning paper 10 years later demands we explore novel, relevant questions for future research.

The changing faces of global cities and firms

The award-winning paper gave us an excellent scientific account, but at a single point in time. The then purpose was to highlight location determinants of firms’ foreign subsidiaries in global cities as a snapshot. The evolving nature of global cities was beyond the authors’ gaze, though briefly invoked in their discussion. The common presumption that global cities will “always be there” remains unchanged and intact, and is belied by the reality that they are dynamic and multifaceted (de Visser et al., 2021). We cannot focus on the spatial dimension at the cost of ignoring the temporal dimension. At the same time, firms’ patterns of location choices also exhibit some significant changes. In the following section, we summarize some of the transitions that appear to be taking place both at the global city level and at the firm level.

Global cities in transition: Getting multifaceted

Cities, including global cities, are evolving to be multifaceted in nature (de Visser et al., 2021). The global city is absorbing various elements of differing concepts of modern cities, mirroring changes in technological, social, and natural environments. For example, the worldwide trend toward digitalization affects all of our lifestyles, and global cities are naturally among the most affected given their roles as financial, industrial, and cultural centers; digitalization makes global cities increasingly wired and interconnected internally and externally (De Dutta & Prasad, 2020). Such a rapid trajectory means global cities absorbed the characteristics of the “smart city”—the forerunner of digitalization and information and communication technologies (ICT) (Albino et al., 2015). We can infer that the global city then encompasses some characteristics of the smart city (Tura & Ojanen, 2022), as well as other related concepts such as the digital city (Anthopoulos & Tsoukalas, 2005), intelligent city (Komninos, 2007), wired city (De Dutta & Prasad, 2020), creative city (Scott, 2006), among other categories (Albino et al., 2015; Nam & Pardo, 2011). While the global city and its adjacent concepts, such as the smart city, have been taxonomically distinct and were not usually discussed together (Albino et al., 2015; Nam & Pardo, 2011), the acceleration of digitalization and the advent of the knowledge society contributes more digital, high-tech, learning, and creative elements to major global cities (De Dutta & Prasad, 2020). The re-bundling of these various formerly distinct civic characteristics under a single umbrella category of global city is partially underway. We can surmise that there are many facets of a global city that may sometimes be fragmented and decomposed into much smaller segments. Inevitably, this blurs the boundary of the global city and obscures its differences with the relevant adjoining concepts of cities (Albino et al., 2015; Rossi, 2017). The consequence is that the “inside-outside” dichotomy with regard to the boundary of global cities has less relevance today. Such a blurring trend of global city boundary, in turn, provides multinationals with wider options to locate their activities, beyond the traditional boundary of global cities.

We can translate from these shifts to portraying global cities in their variety of differing natures as in Fig. 1. As we noted above, in the wake of rapid digitalization and the development of information and communication technologies (ICT) infrastructure, global cities have come to manifest similar elements to the smart city (Dutta & Prasad, 2020). Global cities now embody some of the “technological” characteristics that are no longer the monopoly of the smart city and the digital city. Global cities also exhibit the characteristics of “community,” in which the civic and institutional and policy domains work in partnership to facilitate the diffusion of digital knowledge and learning practices (Berardi, 2013). Also, when a global city needs to be more digital and knowledge-intensive, it requires policy-level support (Albino et al., 2015). With reference to global concerns, global cities embody “environmental” characteristics that foster sustainable environment. No longer detached from the sustainability agenda, global cities clearly incorporate environmental characteristics, so offering scope to prioritize the SDGs.

Fig. 1
figure 1

Multifaceted nature of global cities

Full size image

Our discussion suggests that we cannot regard global cities in any monolithic sense. Global cities are the aggregation of different elements, as illustrated in Fig. 1, and it is these elements, among others, which confer the multifaceted character of global cities. What we can suppose is that, in the wake of digitalization, knowledge economy, and the dimension of global sustainability, global cities come to encompass features of other city type. In other words, characteristics similar to those prevailing inside global cities are now found outside these same cities and vice versa. While we here reference the smart and digital dimensions of global cities, the multifaceted nature of global cities is not confined to these dimensions.

Firms in transition: Metamorphosis into an “ecosystem”

Firms often enter a local environment through partnership and informal collaborations, without necessarily establishing any subsidiaries (Rong et al., 2015). This is characteristic of an “ecosystem.” Firms deploy various modes of entry—foreign direct investment (FDI), equity/non-equity joint ventures (JV), alliances, and informal collaborations—to localize their operations (Rong et al., 2015). In this way, many firms are proactively defining and delineating their own “ecosystems” (Iansiti & Levien, 2004; Moore, 1993) that straddle the edges of global cities. In today’s open-network context in which firms engage in extensive collaborations around the world, ecosystem thinking has come to prevail. Thus, multiple objectives may simultaneously coexist when making location decisions. Goerzen et al. (2013) hinted at this trend based on their empirical investigation of subsidiaries’ own joint ventures. Despite facing limited data, the authors possessed the foresight to signal the need for future research to account for subsidiaries’ relations with buyers and suppliers.

The idea of an ecosystem is borrowed from ecology, in which it is a biological community of interacting organisms within their physical environment (Concise Oxford English Dictionary, 2006). Now adopted into the business context, an ecosystem is a large number of loosely interconnected participants who depend on each other for their mutual effectiveness and survival (Iansiti & Levien, 2004). A business ecosystem is “an independent economic community with different stakeholders, including direct industrial players, government agencies, industry associations, competitors, and customers, who mutually benefit each other and face similar outcomes” (Rong et al., 2015: 294). Adner (2017) defined an ecosystem in proactive terms as “the alignment structure of the multilateral set of partners that need to interact in order for a focal value proposition to materialize” (Adner, 2017). And ecosystem strategy is “the way in which a focal firm approaches the alignment of partners and secures its role in a competitive ecosystem” (Adner, 2017). Regardless of the definition (Adner, 2017; Iansiti & Levien, 2004; Moore, 1993), in today’s open economy the ecosystem approach has now become prevalent. It is a satisfactory way to explain how firms strategize to gain benefits (Williamson & De Meyer, 2012).

When firms enter a foreign market, they have to contend with, and then negotiate, complex challenges by following the business ecosystem approach (Ronget al., 2015). Thus MNEs can overcome those disadvantages encountered within a foreign environment through actively nurturing the business ecosystem (Rong et al., 2015). Firms may limit and manage uncertainty through flexible coordination with other actors (Williamson & De Meyer, 2012) when, in choosing to enter foreign markets, they actively assimilate into the ecosystem through extensive collaboration with partners (Hult et al., 2020). Invoking the three types of LOF laid out by Goerzen et al. (2013), namely uncertainty, discrimination, and complexity, we can surmise that firms may overcome the liability they generate, at least in part, through adopting an ecosystem approach rather than by relying entirely on the properties of global cities per se. Our discussion therefore suggests that a MNE may practically overcome LOF through an ecosystem-oriented strategy rather than through a strategy premised on entering global cities in their own right, à la Goerzen et al. (2013). In proposing this, we recognize the tradeoff between rising coordination costs associated with managing such fluid, informal, and formal inter-firm relationships as well as the risk of excessive dependence on complex inter-firm systems (Nambisan et al., 2019) and any reduction in LOF. Such a possibility was indeed flagged by our Decade Winners when finding that a local JV partnership strategy (Hennart & Zeng, 2002) combined with a global integration approach (Goerzen et al., 2013) would reduce LOF harm, e.g., from uncertainty. The authors surely sparked further discussion and empirical investigation to follow.

Ecosystems blur the criteria for location decisions

At the firm level, the determinants of foreign subsidiary location appear to be getting more complex and multidimensional. Goerzen et al. (2013) associated MNEs’ decisions to locate their subsidiaries inside or outside global cities with demand-driven (competence-exploiting) or supply-driven (competence-creating) investment motives (Cantwell & Mudambi, 2005). However, just as the global city itself is getting more multifaceted in nature, the MNE subsidiary location decision is surely complicated by ecosystem logic, supplanting a discrete choice between inside or outside the global city.

It is quite possible that firms with demand-driven motives may be attracted to locations outside global cities. If so, in the absence of any of the three types of LOF, this would leave little benefit to locating within a global city. And MNEs able to take advantage of digitalization may secure the required connectivity outside rather than inside the global city, thereby avoiding higher costs and unwanted proximity to competitors (Anthopoulos & Tsoukalas, 2005).

Be the motive supply side or demand side, when entering a foreign location, MNEs are free to locate subsidiaries inside or outside cities as they wish, utilizing an extensive network of open collaboration with local partners (Lavie & Miller, 2008).Footnote 3 Entirely removing the need to establish foreign subsidiaries, inter-firm collaboration is making internationalization accessible to growing numbers of firms (Freeman et al., 2006). An example of the decline in pure supply-driven and demand-driven motives for location behavior is the shift away from upfront R&D investment. Across many sectors, no longer is this a prerequisite for conducting global innovation (Doz & Wilson, 2012). Even firms with supply-driven R&D motives may be more light-footed and choose a location inside the global city; as long as they are part of the ecosystem they do not need to own every technology. This contrasts with the pre-ecosystem, in-house R&D internationalization approach, in which firms created and owned all their proprietary technologies, typically in overseas competence-creating R&D subsidiaries (Cantwell & Mudambi, 2005; Song et al., 2011).

Today, many firms within an ecosystem at various levels—local, regional, or global—are part of multiple global value chains. They may operate on the supply side or demand side, depending on the roles they play (Morris et al., 2022; Van Assche, 2020). The fine-grained variation in firms’ entry motives arises from their roots within multiple global value chains, no longer clear-cut and far less easy to classify.

The innovation ecosystem intertwines the supply side/demand side motives of location decisions. Inside the ecosystem, it is even harder for firms to identify whether their motives are supply or demand-side. For example, lead users on the demand side are likely populated in large global cities, yet their input is essential for product development and design on the supply side (von Hippel, 2006).

We present Fig. 2 to illustrate how criteria for the location of MNEs’ subsidiaries inside or outside global cities have progressively become less clear-cut. The changing nature of the internationalizing firm, its flourishing external partnerships within the ecosystem coupled with the increasingly multifaceted global city with blurred boundaries, widens the ripples. In the transition from its origin, the first ripple reflects the subsidiary location decision based squarely on the firm’s motive, as in Goerzen et al. (2013). Subsequent ripples capture the behaviors of the whole cast our focal firm’s collaborators within the ecosystem—whether equity-based partnerships or not. The effect is to significantly liberate the firm’s location choice. As the MNE subsidiary location decision is less cut and dried, researchers are now on a quest for more comprehensive information on relationships and improved operational data. They hope to meet the challenge of explaining where a firm will locate its value adding activity, ideally to transcend the coarse “supply-side or demand-side” distinction.

Fig. 2
figure 2

Firms’ ecosystem-based location choices

Full size image

Sustainability within the ecosystem

Global sustainability is a challenge for the new reality of business ecosystems. Extrapolating from our above reasoning, sustainability appears unattainable without the formation of partnerships within a context of collaboration. Generalizing the principles of internationalization and location decisions, we discard dominant reactive logic to embrace the idea that demand- and supply-side reasons connected with sustainability will spur MNEs to proactively become part of an ecosystem, either within or beyond a global city. The qualitative difference between the conventional internationalization account and ecosystem thinking is that the firm, the city, and its hinterland become active agents, interacting and shaping each other, as outlined in the previous section. Our research question then must be “What difference does adopting an ecosystem perspective make?” to sustainability when we consider the drivers of internalization in the context of global cities and MNE location strategy. Our objective is ultimately to consider the potential directions for future inquiries, given the thin research base addressing this question. While scant research necessarily impoverishes our ability to cite studies that directly support our reasoning, we can invoke the logic set out in this commentary together with the latest thinking and knowledge relevant to international business and sustainability. Our main purpose is to consider how a pro-sustainability agenda could be woven into a framework in the line of descent from Goerzen et al. (2013).

An augmented ecosystem comprising sustainability should cover the three received categories of: (1) economic sustainability; (2) environmental sustainability; and (3) social sustainability (van Tulder & van Mil, 2023). While the 2013 Decade Award winners did not consider sustainability, neither the paper nor the authors were antagonistic to it.Footnote 4 Indeed, there is much in Goerzen et al. (2013) that is receptive to the sustainability view of international business. Sustainability in the context of internationalization demands that we focus on sustainable development, as operations abroad necessarily imply the likelihood of economic growth with economic, environmental, and social impactsFootnote 5. The Brundtland Commission (UN, 1987) defined sustainable development as: “Development that meets the needs of the present without compromising the ability of future generations to meet their own needs” (van Tulder & van Mil, 2023, p. 38). This “Brundtland definition” heralded the fusion of the three dimensions of economic, social, and environmentally sustainable development within the agendas of national and international organizations, corporations, states—and cities. This assimilation has been progressive and experienced a step change with the publication of the Sustainable Development Goals (SDGs) (UN, 2015).

Today, economic, social, and environmental sustainability might qualify as novel drivers of internationalization, with the possibility of beneficial sustainability “feedback” effects. When we adopt ecosystem thinking, as argued in the section on Sustainability of this commentary, taking Goerzen et al. (2013) as our reference point, we can conjecture that sustainability goes beyond a straight choice between demand- or supply-driven motives over the location of international operations. Limited to the economic sphere, the 2013 paper found that “competence-exploiting” (i.e., demand-driven) subsidiaries with a focus on market servicing locate preferentially within global cities. However, sustainable economic development cannot be divorced from environmental and social sustainability (van Tulder & van Mil, 2023; 477), which may impact the location decision.

Extending from Goerzen et al. (2013) let us consider the creation and exploitation of sustainability-related technologies. The locational determination of supply-driven competence-creating subsidiaries with “a focus on enhancing production and research and development competencies” (ibid., p. 433) were found less likely to be attracted to global cities “whether they locate in foreign markets in order to establish production or seek inspiration for new product development.” (ibid., p. 434). With the benefit of 10 years’ hindsight, and taking an inclusive view of the equity and non-equity operations of the MNE, we can infer that routine production may well take place outside cities, possibly under novel partnership models promoting sustainability. The “inspiration function” for new product development may more likely locate inside cities, given the importance of ‘open source’, ‘open innovation’, and the relevance of distributed innovation processes across organizational boundaries, again benefiting sustainability through using society’s resources more effectively. The boundary between demand- and supply-driven motives truly becomes blurred not only locally but internationally when we incorporate sustainability.

There is a long-held view that FDI is attracted to environmentally sustainable locations (Globerman & Shapiro, 2002) but only incomplete evidence. There is some recent support for the proposition that MNEs locate in cities on the basis of civic environmental sustainability credentials. Here, Pisani et al. (2019) is novel in theorizing and finding evidence that host environmental sustainability may be a locational attraction at the city level. Thus, there is now some tentative support for the hypothesis that MNEs may seek locations on the basis of environmental sustainability. Additionally, these same MNEs might in turn exert a positive effect on sustainability in those places where they locate (De Marchi, Cainelli, & Grandinetti, 2022). This aligns with the notion of an ecosystem of multinational enterprises and global cities yielding environmental sustainability with feedback effects. We can conjecture that if it does indeed “pay” for cities to be “green” (Pisani et al., 2019), then greenness becomes tractable as a locational factor and national green policy, and green cities, may be effective in attracting both the equity and non-equity operations of MNEs.

Taking as broad a view of social sustainability as we have of other sustainabilities, we might infer that social sustainability also attracts MNE operations; but how would this fit within the original 2013 study? Here, the IB literature has for many years associated states’ good governance with locational attractiveness to MNEs (Dunning & Lundan, 2008) and a key element of social development is the role of institutions. i.e., property rights, the rule of law, and social infrastructure. Since Rodrik et al. (2002), these have been recognized to exert a dominant empirical role. Locational attraction meshes with MNEs’ desire to incorporate all forms of risk, including tightening regulatory risk (re-sustainability), into corporate social performance (CSP) beyond financial return and shareholder wealth maximization. This is mirrored in the literature on CSP highlighting the agency of MNEs choosing to acquire environmental, social, and corporate governance (ESG) credentials (Napier et al., 2023).

Yet, there is a darker possibility, of eventual bifurcation in the world economy—between locations that pursue sustainability and those that reject it, with a race to the top at the top, and to the bottom lower down. This would create a very divided world in terms of sustainability. One area for investigation must be any parallel relationship with ongoing geopolitically generated bifurcation in IB. De-risking, reshoring, friend-shoring—are all signs of bifurcation that must bear some relation with and consequences for sustainability—its distribution and all-important global level. International trade and investment agreements may signal greater prospects for sustainable development (Vertinsky et al., 2023), but against a backdrop of rising tension. The relation between, for example, the FDI-pollution haven hypothesis (Singhania & Saini, 2021) and geopolitics may indicate scope for added leverage to positively impact sustainability. In the absence of urgently needed policy levers, merely scientifically explaining the variation between outcomes is, sadly, of very limited value for the sustainability agenda, particularly in the timescale of the climate change crisis in which global levels are paramount. Here, our framework of coevolution through city-level and firm-level transitions may be instructive in thinking about how to develop the impetus for accelerating the attainment of the SDGs perhaps using city-level policy initiatives, harnessing cities and their hinterlands together with MNEs as proactive agents via FDI and in partnerships.

Some consequences of an ecosystem approach

The impact of disruptions

Today’s context of natural and geopolitical disruptions challenges researchers on how best to handle MNEs’ locational stance in general, with further attention required specifically on how disruptions impact firms’ location decisions (Romanello & Veglio, 2022). What kind of disruptions would facilitate or hinder firms’ decision to enter global cities? Geopolitical and other types of disruption—war, pandemics, earthquakes, and climate change, among others—appear to have slowed the pace of globalization (Linsi, 2021) arguably compromising connectedness. Thus, the precise degree and nature of various kinds of disruptions likely impact the global city decision (Chen et al., 2022; Doh & Benischke, 2022). A much more fine-grained understanding of disruptions and the location decision is required—one capable of analyzing how business ecosystems accommodate and change when exposed to short- and long-term shocks. While there is evidence that firms readjust their global value chains as a consequence of natural and manmade disasters (Gereffi, 2020; Oh & Oetzel, 2022) research has not yet “joined the dots” to take an ecosystem view of the mechanisms involved.

A pandemic profoundly affects the popular meaning of cities. The exodus from big cities following COVID-19 has been much debated (Nathan & Overman, 2020; Smith, 2020) as has the impact of global warming on international business and on firms’ location decisions vis-à-vis cities. We know only of the possibility of shifts in populations in complex manners (Kolk, 2016). One conclusion we can propose is that geopolitical disruption does influence firms’ location decisions to choose or to avoid certain global cities. For example, Brexit obliged some firms originating outside the European Union to reconsider the location of their regional headquarters (RHQs), e.g., to decide to relocate from London to other European cities (Glückler & Wójcik, 2023).

The fundamental properties of global cities—international connectedness, cosmopolitan environment, and advanced producer services—appear susceptible to today’s profound changes in the geopolitical domain, the legacy of the pandemic, and other crises. The modern context of disruption demands resilience as an essential quality needed by global cities and firms alike (Sim et al., 2003). An ecosystem view offers hope that resilience will be all the greater, drawing on a network of long-term relations built at the firm level that in turn fortify the city and its environment. Conversely, fragile relations undermine resilience, causing the system to collapse like a house of cards when a shock occurs. Which of these eventualities prevail requires research on the quality of relations.

Coevolutionary approach

While transitions at the firm and the city take place on separate levels, we see these changes as somewhat related. Goerzen et al. (2013) implied the possible coevolution of firms’ location strategies and the emergence of certain locales as centers of specific economic activities. Building on the notion of a coevolutionary approach between the firm and its environment (Cano-Kollmann et al., 2016), we see an opportunity for researchers to explore how firms and global cities co-evolve within an open ecosystem.

Taking forward what Goerzen et al. (2013) started a decade ago, we anticipate theoretical development and fresh insights into location decisions in the context of firms’ ecosystems. These comprise networks of horizontal inter-firm partnerships and vertical GVCs, inside or across the boundary of global cities. Inside firms’ ecosystems, increasingly complex and nuanced motives interact with the dynamic, multifaceted nature of global cities. We can be sure that the actual range of scenarios for coevolution of the firm and the city will exceed whatever we have been able to set out here. Therefore, we hope to encourage researchers to identify and pick up on any other dimensions of the city and the firm that will unquestionably become salient.

Performance implications of differing location patterns

In focusing on the choice of location, Goerzen et al. (2013) gazed ahead in suggesting that later studies seek to connect location choice with corporate and subsidiary (economic) performance. Today, researchers may readily draw on corporate, subsidiary, and ESG performance data to test the performance outcomes of firms’ location decisions inside or outside the global cities.

However, the standard performance implications of firms’ decisions to go inside or outside global cities is less of a burning research question than it once was. It falls short of delivering on the need to take an ecosystem approach to global city location. To illustrate this, our Fig. 3 explores relevant location patterns. The vertical dimension distinguishes subsidiary-based versus ecosystem-based corporate localization approaches. The horizontal dimension distinguishes connectivity within a global city from that across multiple global city locations. Our reasoning maps to differing performance implications across these distinct approaches.

Fig. 3
figure 3

Performance implications by location patterns

Full size image

The lower-left quadrant labeled as the Local-Internal type represents the firm entering a global city location by establishing subsidiaries. Goerzen et al. (2013) falls into this category. This is the traditional entry pattern in which a firm leverages its firm-specific advantages (FSAs) and its home-country advantages (HCAs) (Rugman & Verbeke, 1998) to capitalize on its “intra-firm” connections within a multinational corporation (Kogut & Zander, 1993). The lower-right quadrant, labeled as the Global-Internal represents a sequence in which a firm enters multiple global city locations by establishing subsidiaries. This pattern extends the previous pattern, leveraging firm-specific advantages (FSAs) across locations; and facilitating intra-firm knowledge transfer between distant subsidiaries (Asakawa et al., 2018; Kafouros et al., 2012).

In contrast, an ecosystem approach will have a different impact on performance and firm effectiveness. Although our knowledge to date is scant, we reason that the upper-left quadrant, labeled as the Local-External, may stand for a firm entering a global city ecosystem, i.e., through partnering with organizations within the local environment, obviating the need to establish subsidiaries. The upper-right quadrant, labeled as Global-External, represents a sequence in which the firm enters multiple global city locations through partnerships with organizations inside the local ecosystems, again without establishing subsidiaries. Without doubt, the consequences of the ecosystem approach to entering single or multiple global city locations need scientific investigation. We might envisage that the ecosystem approach is likely to promote sourcing of local or global knowledge from a single or multiple global city location(s) drawing on rich external networks (Song et al., 2011) while local or global partnerships may also facilitate inbound/outbound innovation (Casiman & Valentini, 2016). The Global-External category represents the most complex pattern, carrying high levels of potential both in terms of cost and benefit: It incurs the highest resource cost owing to the complexity of managing informal networks with external parties in multiple global city locations. This category also promises the greatest potential to source, leverage, and commercialize globally dispersed knowledge and innovation via extensive globally reaching external networks. The firm is both the agent and the beneficiary of interconnectedness between global cities (Goerzen et al., 2013) just as innovation clusters are globally linked (Turkina & Van Assche, 2018). Thus, firms employing partnerships enhance their ability to identify opportunities for innovation (Doz & Wilson, 2012) and gain scope to leverage their innovations for commercialization (Casiman & Valentini, 2016)—all propositions that merit further investigation of this most complex of entry types.

Exploring additional data and approaches to capture the ecosystem

Toyokeizai’s "Overseas Japanese Companies Data"—the “Kaigai Shinshutsu Kigyo Soran” database—literally translated as "Overseas Activities of Japanese Companies”, was a good fit with the empirical objectives of Goerzen et al. (2013). It is one of the most widely used datasets on overseas Japanese companies, offering consistent panel data for parent companies and overseas affiliate companies, though lacking data on subsidiaries’ external networks other than subsidiaries’ joint venture partners. The database captures only Japanese firms’ localization through affiliate companies, i.e., subsidiaries, and is blind to localization via non-equity relations. In view of our discussion on the theoretical importance and empirical growth of partnerships, many not requiring equity, this blind spot must be considered a severe limitation for future research on firms operating within ecosystems and their relationship with global cities.

Japan has served as an excellent test bed for research to date (Goerzen et al., 2013). To extend this inquiry, the Recof M&A database, compiled by the Recof Data Corporation in Japan, offers the most comprehensive data on M&As and collaborations of Japanese firms by industry, domestically, and internationally since 1996. While even Recof falls short of capturing firms’ other-than-formal collaborations prevalent in the ecosystem, we hope that future studies may collect or employ alternative and more complete localization data, at last to do justice to the fluid and informal nature of the overseas entry decision.

Researchers must grapple with how to go beyond the simple anatomy of an ecosystem to capture how it operates. There are no “off the peg” data, not for Japanese or any other firms. More specifically, we recognize the following opportunities to which future researchers can contribute. We would suggest innovation, utilizing data on internationalization as a proxy for the anatomy of ecosystems, combining it with geographical data on the multifaceted nature of global cities. Novel approaches to capturing the characteristics of global value chains (McWilliam et al., 2020) can be harnessed along with trade in value-added data (TiVA).Footnote 6 In principle, social network analysis (Kurt & Kurt, 2020) can capture formal and informal networks of overseas relations, while other novel methods, such as machine learning and AI tools, deserve further attention (Delios et al., 2023). Merely imaging complex ecosystems is insufficient, we need research methods that can handle a wide range of data (Delios et al., 2023), including “big data” and Geographical Information Systems (GIS) data.Footnote 7 Relatedly, sustainability is an archetypal example of an ecosystem view. Thus, van Tulder and van Mil (2023) write “New (digital) technologies for capturing novel sources of (big) data are emerging that may fill in important blanks in the near future" regarding the measurement of progress toward sustainable development (ibid., p 82). This may alleviate the vacuum in data on a global basis that are needed for scientific evaluation of sustainability in relation to IB and MNE location choice in particular. If combined with corporate ESG rating data, this will open a new avenue for research. Finally, the multifaceted nature of global cities, their blurred boundaries, as well as their interconnectivity merit further empirical investigation, and some promising first steps are underway (Acuto & Leffel, 2021; Leffel et al., 2023).

Conclusion

In revisiting Goerzen et al.’s (2013) contribution to the field, our commentary has sought to sketch out the way ahead, from the “big picture” challenge of re-casting the 2013 study’s concept of a world of direct effects knowable with some degree of certainty,Footnote 8 toward one in which a web of relationships and grand challenges seem to confuse us all. Set against this lofty aspiration, we also identify ways in which a number of more pragmatic aspects require updating if we are to inch our way towards meeting these grand challenges. This is because the world has changed in the intervening years since 2013, and re-contextualization within today’s circumstances, at a practical level is due. City and firm transitions are underway, though there are no clear-cut changes which can yet be conclusively demonstrated or evaluated empirically. We currently stand on the threshold, and are only glimpsing the new directions of firms’ location decisions and the nature of global cities. Without claiming we know the answers, we have advocated an ecosystem perspective on cities and firms, and suggested some future research directions which we regard as promising for the next decade. At the same time, we recognize the escalating demands these new directions place on empirical researchers to source data.

We conclude that the authors’ identification of connectedness, cosmopolitanism, and advanced producer services presaged the growing importance of digital infrastructure, human factors and how cities can—and should—be considered as aggregations of smaller economic units that are the building blocks of international and national location decisions. In this sense, the paper was indeed ahead of its time with regard to the transitions underway for global cities and for firms. Recognition that global cities are increasingly multifaceted in nature, and that firms are accelerating towards the norm of adopting an ecosystem approach, spawns new research questions. This new approach is characterized by extensive collaboration and partnerships with various organizations external to the firm that are in reality a form of internationalization. We have argued that such changes at the city and firm levels would make firms’ extant criteria for locating in global cities ambiguous. The advent of an ecosystem approach is led by firms themselves, involving partnerships and collaboration with external organizations such as customers, suppliers, universities, and competitors, amongst others. This may signify that multinationals can overcome the disadvantages they face within a new international environment through actively nurturing the business ecosystem, while acknowledging the additional cost associated with such a fluid, informal form of relationship outside the firms, as well as the risk of excessive dependence on complex inter-firm systems (Nambisan et al., 2019). If our logic is well founded, such a strategy will help these firms overcome the uncertainty they face through adopting an ecosystem approach and not necessarily by relying entirely on the properties of global cities per se.

Our reasoning put forward in the section on sustainability, above, shapes the focus of our conclusions. We picked sustainability as a symbolic—if still somewhat esoteric—example of an ecosystem view, and discussed how sustainability might affect firms’ location decisions inside or outside global cities. Given sustainability’s salience this is unlikely to be a poor choice for future research to push forward the efforts initiated by Goerzen et al. (2013) a decade earlier. Staying with sustainability as an illustration of how ecosystem thinking may intervene in our research agenda, we suggest that diversifying into the environmental and social domains adds new layers of blending between supply- and demand-driven motives. The ecosystem perspective encompasses the city, its hinterland, regional and global supply chains and, indeed, the globe. Conceptually as well as computationally, this greatly increases the “complexity in the action”.Footnote 9 Operations in one place are correlated with operations in other places, either within the MNE or, as we have suggested, through the tissue of operations that any MNE is likely to have that are not conventional investments, as well as via the supreme level of the ecosystem of the natural world itself (e.g., regarding emissions, stakeholder demands, and the SDGs themselves). At first blush, this might seem to place anything practical for our research agenda beyond reach but, we believe, it does not. For societal impact, new research incorporating sustainability must strive for simplicity in the principles that underlie the complexity of interactions, to make the Sustainability Agenda tractable. A practical way is by complementing research in the line of descent from Goerzen et al. (2013) with, for example, management research on “positive IB practices” (Tung, 2023) on executives’ decision-making processes with regard to location, collaboration, policy responsiveness, and feedback into policy—all with an eye on sustainability and impact issues. Research on this, much of it likely to be qualitative, promises to cut through the modeling complexity, building new theoretical insights for the direction of causality. At the grander scale, co-determination of location by regional within-country and international variation (as in the 2013 paper) still offers potential for harnessing sustainability policy, to test propositions yielded by new theory. Looking into the future, we can glimpse that research is needed on whether MNEs may become the proactive agents of mainstreaming greenness as a desirable locational attribute. Sustainability is just an example. If we take any of the SDGs’ impacts within a single location, they are likely to have causes and further effects that span the globe. Having said all that, however, even though we inevitably move away from a narrow focus on global cities alone, they will continue to anchor research.