Business groups are widely recognized as a prominent feature of East Asian economies. However, how to define the boundary of a business group remains a question. Deviating from previous studies that defined business groups by ultimate control, this study approaches the issue from an angle of community network structures. Using director interlocking networks among Taiwanese publicly listed companies from 2010 to 2020, we apply the method of community detection to delineate corporate communities and compare the results with those by traditional definition. Our finding suggests that while there are substantial overlaps between two methods, community structures have better power of prediction for outcomes. Firms associated with corporate communities exhibit greater performance advantages than to the business group defined by ultimate control. Given the rapid transformation of East Asian capitalism, we believe community structure is a better method to analyze business groups and more studies need to be done.

Recent commentaries (notably popularized by Robert Reich) have referred to the political economy in the United States today as the “Second Gilded Age.” While America’s first Gilded Age was characterized by a tremendous concentration of wealth and power, it also represented an era during which exclusive private clubs provided meeting spaces and served as indictors of their members’ prestige. This paper examines the role such prestigious social clubs played in integrating the corporate elite at the end of the first Gilded Age in the United States. Utilizing board interlock data from 1904 and the social club memberships that were listed in Rossiter’s Club Men of New York (1901-2), the initial analyses demonstrate that membership in the 61 social clubs listed in the Social Register (1896) greatly increased director centrality and reduced social distance. The second portion of the paper interrogates further a potential division of labor among social clubs as a mechanism of cohesion among the directors of the nation’s largest corporations. By focusing the effects of membership in clubs emphasizing wealth, ancestry, culture or politics, the results show that these clubs make unique contributions to the integration of the social network of interlocking directors – that is, the four different types of clubs form non-redundant ties among these corporate elites. However, just because an individual lists a membership in these exclusive clubs does not necessarily mean substantive interactions occur between fellow club members or on the premises of these clubs. To assess the class formation effects of these social clubs, the final set of results moves beyond the structural conditions for interaction, to analysis of the how one prominent member used his memberships. Through an analysis of J. Pierpont Morgan’s Day Books, these data support the assumption that elite social clubs mattered. These results bolster the argument that ties formed through mutual affiliations in these bastions of exclusive power and prestige did in fact contribute mightily to the cohesion of the corporate directors at the end of the United State’ first Gilded Age.

Research has examined the relationship between formal bureaucratic structures and informal network structures within organizations. However, bureaucracy is not the only set of rules that governs organizations. Kinship has an important influence on how many small firms are organized and from a global perspective it also has an influence in large firms. In this paper we examine how informal organizational structures—specifically advice seeking networks—combine with kinship and bureaucracy in organizations.

Informal advice relations across units in an organization have been shown to incorporate knowledge that leads to greater productivity, innovation, and competitive advantage. However, boundary spanning ties require time and energy to create and maintain. Research has focused on understanding the mechanisms that underlie cross-unit ties in the context of bureaucratic formal rules and structures. In this paper we expand this focus to examine what occurs when there are two potentially competing sets of rules within an organization.

Our baseline hypothesis is that advice seeking is more likely to be seen within organizational boundaries and less likely to cross organizational boundaries. However, reciprocity within the informal advice network creates a level of trust that facilitates advice seeking across organizational boundaries. We then theorise that shared kinship membership facilitates advice seeking across organizational boundaries.

We gathered data on 604 employees in a financial institution located in West Asia. We use an exponential random graph model (ERGM) to analyse our data. Our initial findings suggest support for our hypotheses.

Collaborative R&D projects tackle major societal challenges, such as sustainability and the circular economy (CE). However, as projects progress, day-to-day operations can overshadow strategic goals. While leaders are expected to focus on long-term objectives and employees on execution, it remains unclear how hierarchy and network structures influence engagement in big-picture discussions. This study examines whether strategic goals remain central in project interactions or become confined to specific parts of the network—and how hierarchical position and network structures shape these discussions.

Using longitudinal network data from two interdisciplinary R&D projects in CE (N₁ = 79, 29.1% leaders; N₂ = 54, 40.7% leaders), we analyze conversational patterns during project meetings. Participants reported whom they spoke with and whether their discussions focused on operations, future projects, or big-picture topics (e.g., sustainability, CE).

Preliminary results suggest that big-picture discussions were limited to small, isolated cliques rather than widely shared across the network. Surprisingly, hierarchical position and network centrality had little impact, indicating that formal leadership alone does not ensure a sustained strategic focus. Instead, strategic discussions often remained disconnected from broader collaboration.

These findings underscore the need for leaders to monitor when strategic goals fade and actively reinforce the project’s broader transformation efforts. To ensure meaningful contributions to circular supply chains, leaders must embed large-scale objectives in daily collaboration. At the same time, strategic discussions should extend beyond central figures, empowering peripheral members to engage and ensuring that big-picture goals remain an integral part of project efforts.

With the upsurge in scholarly and policy interest in elites and inequality since the 2008 financial crisis, particular attention has been paid to finance and financiers as key structures and agents driving economic inequality. But while there have been many critical studies of financialisation, and several influential ethnographies of the financial sector, there have been few systematic empirical studies examining financial elites. This paper examines the financial elite in the UK - which has historically had a highly influential financial sector and remains an important hub for global finance - along with a comparator population of non-financial UK business elites. It details the demographic character of both these groups; the degree of cohesion of each via board interlocks in business and the third sector; the level of integration between them; as well as the nature and extent of their political engagement via an analysis of their political donations.

The paper combines the methodology of Multiple Correspondence Analysis (MCA) and Social Network Analysis (SNA) to investigate the Brazilian business elite, focusing on the interlocking directorates and boards formed by 72 publicly traded companies whose shares were traded on B3 (Brazil's main stock exchange) in 2020. To this end, information about the executive boards and boards of directors of these companies was analyzed using R Studio software, with the reference forms annually submitted by the companies to the Securities and Exchange Commission (CVM) as the primary sources. Guided by a notion of elite based on Wright Mills' positional method and Pierre Bourdieu's construct of the field of power, the study aims to demonstrate how individuals' participation in organizations, their shared educational background, and their professional trajectories form a coordinated elite, which constitutes a necessary foundation for the exercise of power. We also mapped the relationship of these agents with national political and cultural spheres, investigating their participation as members of the government, cultural institutions (such as family and private foundations and think tanks), and social clubs. We used SNA to identify groupings or communities (community detection) through clustering algorithms (stochastic block modeling), enabling the creation of an active categorical variable to be included in the MCA. This methodological combination contributes to the sociology of elites by making it possible to objectify social capital and the multi-positionality of agents and to assess the influence of these elements on the structure of social space.

Firms exchange various valuable resources to accomplish their daily tasks and achieve their goals. However, the process of tie formation in networks of business transactions between firms is still unclear. We follow and combine previous knowledge from the sociology of organizations, network science, and economic geography to inspire our research hypotheses on how we can potentially describe and infer the substructures that may better represent their network formation processes. We define our networks with firms as nodes and economic transactions between them as edges, which we observe from the 2017 Hungarian VAT declaration records. We construct our networks from transactions between firms of the specific industry-region, their economic transactions with external partners from different industries and regions, and the transactions between these partners. We focus on firms operating in the metal production sector (N=10 networks). Using Bayesian ERGMs, our preliminary results show that structural and actor-relation effects are relevant and significant in identifying the process of tie formation in these networks. Firms in this sector in 2017 relied on existing connections for their exchanges, and typically did not engage in circular trade patterns. Moreover, there was limited preferential attachment in these networks, suggesting that firms imposed capacity and risk constraints. Our contribution extends to the possibility of applying these models to other industries, as well as identifying trends that may be useful for further studies that need to set prior distributions when using Bayesian estimation as a solution for sparse network inference.

Corporate interlocks—the sharing of directors across multiple company boards—create influential networks through which cybersecurity knowledge and practices flow between organisations. These connections are remarkably common: in 2018, 78% of the top 50 S&P500 firms shared at least one director. Despite recent trends limiting multiple directorships—59% of S&P500 firms cap directors at three additional positions—these interconnections remain powerful in corporate governance.

These directorate networks serve as channels through which cybersecurity attitudes propagate across organisational boundaries. Beyond individual application of experience, these networks facilitate collective learning as board members exchange insights. Interlocks function as effective external knowledge sources, providing access to insider strategies that would otherwise remain confined.

Our study employs a Linear Network Autocorrelation Model (LNAM) to examine how interlocks influence cybersecurity governance across S&P500 companies. We analyse key variables including Risk and Audit committee structures, presence of CROs and CISOs, board size, women directors, governance quality scores, and ransomware experiences, while controlling for industry effects. This approach reveals how cybersecurity practices diffuse through director networks and identifies which governance characteristics most effectively enhance cyber resilience.

This study examines the impact of ethnic social capital on wages in Singapore, focusing on the contrasting effects of Chinese and Malay social ties. Using data from Wave 2 of a 2023 study of approximately 900 young adults (aged 21-39) from lower to middle-income households, the analysis reveals that in a labour market where education, occupation, and experience are the primary determinants of wages, social capital plays a comparatively minor role. However, a notable asymmetry emerges: while Chinese social capital has minimal or no effect on wages, Malay social capital is linked to a significant wage penalty. Among degree holders and PMETs, having more Malay contacts correlates with lower wages, even when these contacts are high-status. To address the possibility of reverse causality—where the negative link between Malay social capital and wages could stem from more Malays occupying lower-wage jobs—the study controls for workplace ethnic composition. Additionally, by controlling for the respondent’s ethnicity, the analysis confirms that the wage penalty is driven by the devaluation of Malay social networks, not the respondent’s ethnicity. These findings highlight the devaluation of Malay social capital and point to underlying racial biases and structural inequalities in Singapore’s multi-ethnic labour market, even among highly qualified individuals.

In the current context of globalization, this research explores the networks of multinational firms in African cities. Historically, the colonial era structured African economic relations, remaining after the countries’ independences. For 10 years, African cities have diversified their globalization, supported by the emergence of new partnerships (Martin, 2023; Chaumont, 2024).

We undertook an empirical study of the ownership networks of multinational companies based on the ORBIS-BvD 2010–2022 dataset, a sample of the top 3,000 global groups’ networks of firms (all their direct and indirect subsidiaries with ownership above 10%) in 2010, 2013, 2016, 2019, and 2022). We located 98% of the African companies in homogenized Large Urban regions (LURs) (Rogromel & Rozenblat, 2024) building weighted and oriented inter and intra-city networks.

Between 2010 and 2022, the number of firms of the sample in African cities significantly increased from 2,400 to 28,000, with links rising from 3,700 to 28,000 (Orbis-BvD-Citadyne Unil, 2010–2022). This evolution reflects not only a growing integration in globalization but overall, the growth of intra-urban, intranational, and intracontinental urban networks, while the intercontinental networks lost their domination. The African Continental Free Trade Area (AfCFTA) agreements, by simplifying regulations and opening African markets, have significantly accelerated this intra-continental integration (Obida, 2019). Cities such as Port-Louis, Cairo, Lagos, and Johannesburg became regional hubs, concentrating these intra-African interurban networks. While a vertical structure dominates with political or economic capitals, a horizontal structure emerges where specialized cities interact in a complementary fashion with the major hubs.

Drawing on insights from the political economy literature, I argue that the decline of the inner circles cannot be understood without taking economic power of the companies in relation to the economic power of states into consideration. Research that has sought to explain the decline of the corporate inner circles across nations have explained this decline by investigating structural network characteristics, such as the decline of big linkers, and to some extent changes in the political economy by first showing how the financialization led banks to withdraw from the inner circles and secondly how the globalization of the economy led big international companies to withdraw from. Yet other potential drivers such as changes in the national business structure and historical changes in for example firm size, mergers and acquisitions, or ownership structures remain understudied. By empirically investigating the development of the corporate interlock network among the hundred largest Danish companies from 1973 to 2022, I confirm previous findings, by showing a decline of the inner circle from the 1990s onwards that can be explained by a decline in big linkers in the network. The significant decline in the size of the inner circle is however not reflected in the evolution of the accumulated firm size of the inner circle. I therefore argue that the economic power of the inner circle does not fracture but remain stable over time.

The field of competitive dynamics has greatly improved existing understanding of how firms engage in competitive behavior with one another to improve their competitive position. Additionally, competitive dynamics researchers have long studied how organizations engage in dyadic inter-firm rivalry with one another through targeted competitive actions and responses. However, the outcomes of interfirm rivalry affects not only the firms' direct competitors but also evoke reactions from managers of firms beyond the dyad, leading to evolution in the competitive network. Research has long discussed the need to address how the structure of firms’ interorganizational competitive network beyond dyads affects their competitive behavior, along with how firms’ competitive networks lead them to form, maintain, or dissolve competitive ties. Adopting the manager-oriented perspective of competitor identification, this study proposes that numerous factors may impact firms’ perceptions of direct, indirect, and potential competitors, such as the structure of the interorganizational network, the competitive behavior of the firm, and other firm-level attributes. Additionally, this study proposes the usage of statistical models capable of examining change in firms’ competitive network structure, along with methods for examining potential co-evolution mechanisms between the interorganizational competitive network and firms’ competitive behavior. This study contributes to existing research in competitive dynamics by exploring structural, behavioral, and attribute antecedents of firms’ competitive tie formation, leading to an enhanced understanding of how firms identify competitors in their industry.

This study examines how being part of a tightly interconnected cluster of firms – a network community – that has a permeable boundary with bridging ties to firms outside the community influences individual firm innovation performance. Network communities characterized by high permeability—where members maintain bridging ties to external firms—enable the selective importation of diverse knowledge, thereby fostering valuable recombination and innovation. Furthermore, we theorize that the benefits of community permeability are contingent upon the extent to which a firm maintains within-community cohesion and firm-level knowledge heterogeneity. We tested our hypotheses using longitudinal data on alliances and patents from 2002 to 2023. Our findings reveal that firms embedded in network communities with high permeability exhibit enhanced innovation performance. In addition, the benefits of community permeability are amplified when firms are deeply embedded within cohesive community structures and possess highly heterogeneous internal knowledge portfolios. These results highlight that it is not just interfirm network ties that matter for innovation, but the community of network ties a firm is embedded in and the extent to which the community has bridging ties to firms outside the community relationship. This underscores the dual importance of maintaining robust internal community ties for effective knowledge integration while engaging in external linkages that provide access to diverse insights. This research advances the theoretical understanding of the benefits of open and closed networks, reconciling the debate by considering structures between and within network communities.

In a multinational enterprise (MNE), inventor collaborations are essential for the multilateral transfer of technological knowledge, leveraging dispersed expertise, fostering innovation, and maintaining competitive advantage. Previous studies on MNE intra-firm knowledge transfer and innovation applied regression-based statistical analyses to network data, which nevertheless suffers from the violation of independence assumption because of the interdependent nature of organizational members and their interactions. In this study, we adopt the Markov Exponential Random Graph Model (ERGM) approach to examine the co-inventor network within a multinational enterprise (MNE), using data from ARM plc, a semiconductor design MNE, during the 2010–2012 period. The ERGM approach explicitly addresses the interdependencies inherent in network data and accommodates the testing of node- and dyad-level variables. We test the impacts of technological similarity and geographic proximity in the tie formation process while controlling for network structural statistics. Model estimation is performed using Markov Chain Monte Carlo, and simulation results validate the model fit. The findings indicate that co-inventor collaborations are significantly affected by both technological similarity and geographic closeness. This work provides valuable insights into knowledge transfer processes and collaborative innovation, offering a robust and practical framework for analyzing complex network dynamics in MNE.

In this paper we perform a longitudinal ego network analysis of organisational networks containing multiple types of functional actors operating through various jurisdictions. Our data – various corporate debt products issued in Hong Kong – features a range of functionally-specific actors that combine into product ego-networks with rich attribute data. We are interested in how these networks' structures may differ in terms of selection of partnering actors, across product types and jurisdictions, over time. We present a protocol on how to interrogate such data using both existing and novel approaches for analysing ego-networks.

Our study examines the impact of COVID-19 on the Japanese economy through a multi-faceted corporate network analysis, including director interlock networks, director dispatch networks, and shareholding networks. Our previous research demonstrated that corporate profitability exhibited both positive and negative correlations with centralities, and that the relationship varied depending on different centralities. Motivated by these findings, we investigate how the shocks caused by COVID-19 vary depending on a firm's network position, as measured by various centralities. To conduct this analysis, we utilize data from listed companies in Japan, covering approximately 3,800 firms per year for six years (2014, 2016, 2018, 2020, 2022, and 2024). For empirical analysis, we identify structural changes using dummy variables. Firms are divided into two groups based on centrality—the top 25% and the rest—to examine differences in the impact of the economic shock. We employ an approach similar to the Difference-in-Differences (DiD) method: Before the pandemic, firms with high and low centrality had a profitability gap; after the pandemic, a gap also existed. We compare these differences before and after the shock. Our results reveal that firms in the top 25% of Bonacich power centrality in the director dispatch network experienced a significant decline in profit levels. These findings suggest that while corporate networks support firms either positively or negatively under normal conditions, they also serve as channels that amplify economic shocks during crises, as seen in the director dispatch network.

This paper analyses how post-2008 financial regulatory measures potentially influence the shareholder networks linking Finance, Insurance, and Real Estate (FIRE) sectors to industrial sectors across distinct Varieties of Capitalism (Liberal, Coordinated, Mixed, State, and Hierarchical Market Economies). Using inferential network models on data covering the period from 2007 to 2019 (and in some cases to 2022), I explore whether regulatory tightening shapes the formation and dissolution of financial-to-industrial shareholder ties. Initial findings (Marcano Rivera, 2023) timidly suggested the presence of differentiated regulatory effects on network structures, but computational capacity limited conclusive results. By incorporating modelling changes and narrowing down the analysis to key sectors of the economy in lieu of the prior all-encompassing approach, the paper proposes to reveal cross-national variation that policymakers must consider in designing sector-specific regulatory strategies.

The primary objective of this research is to empirically verify whether corporate ownership structure can serve as an instrument for family succession, based on data from Korean chaebols. Chaebols are big business groups in Korea, characterized by a hierarchical-pyramidal ownership structure with numerous affiliated firms. A well-known feature of this structure is that family members of the group’s founder (or president) typically occupy the highest positions of the pyramid. The primary rationale behind adopting this hierarchical-pyramidal structure is believed to be its effectiveness in amplifying the controlling family’s influence over the entire affiliations.

Existing studies on chaebol ownership structures have predominantly focused on their isomorphic characteristics. However, these studies have overlooked the possibility of variation and differentiation within these ownership structures. Since chaebols are family-controlled business groups, there can be considerable diversity in the principle of family and the logic of successions. For instance, in some chaebols, multiple candidates may compete for succession, whereas in others, a successor may have been designated at an early stage. If all family-controlled business groups are treated as a single homogeneous category, such nuanced dynamics may be overlooked.

Thus, unlike previous studies that have treated the family as a single node in ownership structure network analyses, this study categorizes families into several subgroup categories to examine the structure in greater detail. The analysis employs stochastic block modeling (SBM), which allows for consideration of interchangeability or substitutability among family members. The findings suggest that chaebol ownership structures exhibit "isomorphic differentiation." Specifically, while ownership structures are fundamentally isomorphic, 'Owner-centered hierarchy', they can be classified into five distinct types. These five types appear to be closely related to different succession strategies.

This study contributes to research on family business and corporate sociology by empirically demonstrating that ownership structure and succession strategies can be interconnected. Moreover, it offers valuable insights for comparative studies on corporate structures across different national contexts.

The outer space industry is rapidly evolving into a trillion-dollar sector, driven by significant investments from major tech enterpreneurs and companies. This transformation marks the beginning of the so-called ‘New Space’ era, characterized by the emergence of profit-oriented space activities such as tourism and travel, communications and surveillance, and asteroid mining. Despite its rapid growth, the key players shaping this burgeoning industry remain underexplored. This paper identifies and analyzes these key actors, focusing on the networks of lobbying and interest groups, as well as the constellation of tech and business elites that surround them. Furthermore, the study examines the discourse promoted by these actors concerning space commercialization and its connection to pressing global issues, including space colonization, the increasing dominance of technology in society, and the intertwined challenges of climate change and diminishing terrestrial resources.

Sociological research on corporate elite mobility has often highlighted business families' social networks—especially marital ties with other controlling families—as key strategies for maintaining dominance and corporate control, as well as their collective interest as a upper class primary group. However, few studies have examined how these marital networks help sustain their organizational influence in contemporary market capitalism over time, largely due to the challenge of systematically collecting kinship data. This study analyzes data from over 800 publicly listed Taiwanese companies over four decades of observations (1981–2020, N=21,089) and maps the kinship and marital networks of controlling families. Using statistical analyses of chairperson tenure and turnover, it tests core arguments of social class perspective. Taiwan, with its market continuously dominated by family businesses and its rapid post-war economic development, serves as a suitable empirical case for testing these theoretical propositions. The analysis focuses on the effects of intermarriage among business families and whether elite status and corporate profitability moderate these effects. By illuminating the role of informal networks such as marriage in affecting elite mobility, this study advances contemporary research on corporate elites.

While traditional research emphasizes interlocking directorates, this presentation explores the socialization of business elites outside of the corporate world, using the Moroccan bourgeoisie in Casablanca as a case study. We argue that upper-class neighborhoods and exclusive private clubs are essential locations for building networks and accumulating social capital. Our analysis, based on a socio-historical approach incorporating ethnographic fieldwork and over twenty interviews, examines the residential patterns of Moroccan’s bourgeoisie and the changing dynamics of the Club des Clubs de Casablanca, Morocco's most prestigious and exclusive club.

The inner circle has been recently subject to debates in the literature on US corporate networks. Some scholars argue that the fragmentation of corporate elites leads to the disappearance of the inner circle (Chu and Davis 2016), while others suggest that it remains influential but requires new approaches to be detected in the context of the decline of interlocking directorates (Murray and Jordan 2019).

Beyond these debates, little attention has been paid to how the composition of inner circles evolves over time. While research has highlighted the growing importance of foreigners and women among national corporate elites in different EU countries, the mechanisms that shape inner circles remain largely underexplored. Importantly, the inner circle does not simply reflect corporate elites at large; its selection processes vary across national contexts.

My previous research has shown that in France, the inner circle is primarily composed of corporate leaders who graduated from École Polytechnique and/or ENA, often belonging to prestigious state grands corps such as Mines, Ponts et Chaussées, and Inspection des Finances. I investigate whether these selection mechanisms have remained stable since the 2000s, or whether they have evolved to include other fractions of corporate elites with alternative visions of public action.

I assess these changes by analysing several waves of data on the directors of large companies and think tanks. I first examine alternative indicators for identifying members of the inner circle and propose an approach based on the social mechanism of regulation. Secondly, I apply this approach to several waves of data in order to measure changes in the inner circle. Finally, I consider the consequences of these possible changes on policy-making processes.

Who represents the corporate elite in democratic governance? In his seminal work on the “corporate inner circle”, Useem (1986) identifies the emergence of three network-related mechanisms that shaped the composition and political organization of American and British corporate elites in the postwar era: organizational brokerage, elite-level social cohesion and network centrality. Subsequent research has found similar dynamics at play across a variety of capitalist societies but all studies on corporate political representation rest on network analyses of a highly select sample of leaders from the top ranks of very large publicly listed firms. We cast a wider net. Analyzing new population data on all members of corporate boards in the Danish economy (~200,000 directors in ~120,000 boards), we locate ~1,500 directors that operate as brokers between local corporate networks and measure their network coreness using k-core detection. We find a highly connected network core of ~275 directors, half of which are affiliated with smaller firms or subsidiaries. Statistical analyses show a strong positive association between director coreness and the likelihood of joining one of the 650 government committees epitomizing Denmark’s social-corporatist model of governance (net of firm and director characteristics). The political network premium is largest for directors of smaller firms or subsidiaries, indicating that network coreness is a key driver of business political representation, especially for directors without claims to market power or weight in formal interest organizations.

The role of network centrality in firms' internationalization strategies has become increasingly relevant in understanding their choice of entry modes. While previous studies have explored the impact of network positioning on international activities, research focusing on greenfield investments and mergers and acquisitions (M&A) remains limited, particularly from the home-country perspective. Existing studies are also geographically concentrated in non-European contexts, often rely on fragmented data, and lack comprehensive measures of firm networks. Here, we examine how firms' network centrality—measured through degree, betweenness, and closeness—affects their likelihood of engaging in greenfield investments versus M&A. Using panel data on firms in 14 European Union countries between 2015 and 2022, we construct inter-firm networks based on managerial interlocks and analyze their impact on firms' entry mode choices. Our results show that firms with higher degree and betweenness centrality are more likely to undertake greenfield investments, while closeness centrality negatively influences such decisions. These findings suggest that well-connected firms leverage their network ties and intermediary positions to reduce uncertainties and identify new opportunities for autonomous international expansion. In contrast, firms with high closeness centrality, being tightly embedded within the network, may favor alternative strategies, such as collaborative ventures or acquisitions, rather than independent greenfield operations. This study offers a novel analysis of firms' network positions and internationalization strategies by integrating multiple databases and methodologies. It expands the literature by providing empirical evidence from a European home-country context and highlights the distinct roles of network centrality in shaping firms' strategic decisions for global expansion.

Many global corporations have several levels of ownership going through some cities which constitute “mandatory pathways” to reach other ones. The position of “relay”, “intermediate” or “bridge”, affords cities better access to the whole network as well as increased control over information transfers (Burt, 1992, 2007, 2015). One can interpret this cumulative process of concentration as a reinforcement of “social capital” of big cities as sociologists do at the micro level (Gould, Fernandez, 1989; Walker et al., 1997). The position of relay should be more precisely defined in urban geography, as in sociology this concept integrates the role of “tertius Gaudens” or structural hole (Simmel, 1922; Merton, 1957; Granovetter, 1973; Hannan & Freeman, 1977).

To emphasis cities of the world that have “relay” role of firms from 2010 to 2022 (every 3 years: ORBIS database), we will develop different kinds of indexes. In fact, observing the relays at the micro level of companies, but building indexes at the city level, numerous possibilities exist to reveal the size and the specialization of cities in these roles. Beyond the main results of this empirical implementation, the purpose of the presentation will be to discuss how far different indexes could reveal diverse aspects of this spatial role for cities and for their dynamics. Which ones reflect more the attractiveness and power dynamics of the cities? Which ones could reveal resilience properties or spatial balance? We will discuss these measures for better using these properties of relay in the future.