Periodic Reporting for period 4 - CORPNET (Corporate Network Governance: Power, Ownership and Control in Contemporary Global Capitalism)
Reporting period: 2020-03-01 to 2021-02-28
The character of global business networks has long fascinated but continues to divide scholars of global markets and governance. Scholars disagree on the origins of emergent patterns of corporate networks. Do they reflect institutional preferences of corporate and political elites in their search for power and control? Or are they unintended by-products of corporate conduct? And how do patterns of global corporate ownership relate to actual power in the governance of such networks. This project did to do what has so far eluded existing scholarship: to fully explore the global network of corporate ownership and control. Using cutting-edge network science methods, the project analyzed granular, firm level data on corporate ownership and board networks covering over 200 million firms. Utilizing this unique data and cutting edge data science techniques we have been able, among others, to for the first time empirically distinguish between and measure how countries perform sink and conduit roles as tax havens; expose and theorize a fundamental reorganization of stock markets and corporate governance through the rise of passive investing; map, theorize and analyse the full global geopolitical infrastructure of internationally active state owned enterprises; and to map and model patterns of global corporate elite formation. nd complexity studies on the other. As such CORPNET has exposed the new power structures in the global political economy and already inspired policy makers, public debate and scholars alike.
Inter and cross disciplinary work is at the heart of CORPNET’s ambition. This shows for instance in the publication strategy where we reach out to several disciplinary journals in computer science, network science, political science and socio economics. Nearly all of our publications are co-authored by team members with a background in computer and social science. The diversity and academic ethos of the CORPNET lab is key to its success, which reflect in uur publication record with contributions in academic journals such as Nature’s Scientific Reports; New Political Economy; RIPE; ; Social Networks; Strategic Management Journal; Business & Politics; Global Networks; Social Network Analysis and Mining; The International Spectator, but also professional contributions in outlets as the Antitrust Chronicle, Goed Bestuur en Toezicht, en Economische Statistische Berichten. Firmly rooted in an open source approach, CORPNET publishes all its code through GITHUB (https://github.com/uvacorpnet). Our award wining paper on the Big Three asset managers is the best read article in Business & Politics; our paper on States vs Corporations the best read article in the International Spectator, and our paper on Uncovering Offshore financial centers in the top 100 of Nature’s Scientific Reports, ranked 16th in terms of impact (altimetric). We have had many requested interactions with policy makers on several topics; including the Fiscal Information and Investigation Service (FIOD), Autoriteit Financiale Marktenl Infobox on Criminal and unexplainable Assets (iCOV), and the Korean Tax Authority, as well as visits from international journalists. Our work on ownership reconcentration led to question in EU parliament and our study together with two related studies triggered commissioner Vestager to closely monitor these new developments. Our outreach activities are popular as well: a short animation video we made based on this article received over 10k views on Facebook and almost 18k on Youtube
Developing new methods and applying existing methods to novel applications is at the core of CORPNETs mission. For analyzing large scale networks with partitioned structures we developed centrality measure (Takes and Heemskerk 2016). In order to analyze networks at the meso level (instead of the micro level of actors or the macro level of the entire network) we invest in new motif detection algorithms (Takes, Kosters, Witte 2017). For this work Takes received the 2017 Young eScienctist award from the Netherlands Organisation of Scientific Research (NWO). Our work on network dynamics of elite formation (project 2) has for the first time applied the recently developed relational event modelling approach to the field of board interlock studies. The work on ownership chains and offshore finance builds on a novel method, developed by corpnet, to measure and count complex paths of global corporate ownership relations (parent-subsidiary relations) and the role of particular jurisdictions in these networks (project 3).
Developing new methods and applying existing methods to novel applications is at the core of CORPNETs mission. For analyzing large scale networks with partitioned structures we developed centrality measure (Takes and Heemskerk 2016). In order to analyze networks at the meso level (instead of the micro level of actors or the macro level of the entire network) we invest in new motif detection algorithms (Takes, Kosters, Witte 2017). For this work Takes received the 2017 Young eScienctist award from the Netherlands Organisation of Scientific Research (NWO). Our work on network dynamics of elite formation (project 2) has for the first time applied the recently developed relational event modelling approach to the field of board interlock studies. The work on ownership chains and offshore finance builds on a novel method, developed by corpnet, to measure and count complex paths of global corporate ownership relations (parent-subsidiary relations) and the role of particular jurisdictions in these networks (project 3).
We lacked an accurate understanding of the core organisational form in which the global economic order manifests itself: the global network of corporate ownership and control. As a result of theoretical and empirical nationalism, scholarly work was not well able to fully grasp recent important developments. This project filled the gap by studying this subject without an a priori assumption of the appropriate level of analysis. This project’s complex social network approach generated the first coherent understanding of the properties, topology, interdependencies and generating mechanisms of the global network of ownership and control.
Empirical innovation: We conducted the first big data network analysis of the ownership ties and interlocking directorates on the largest dataset currently available, covering over 200 million firms worldwide. This endeavor required a significant investment in cleaning and organizing the data in such a way that it is suitable for social science network analysis. The data enhancing and quality assessment techniques form a first set of innovative outcomes with a large potential impact on how social scientists can deal with big data in the years to come. In addition, analyzing the properties of the web of corporate governance relations uncovered new and unexpected properties, such as hitherto unnoticed levels of concentration of corporate ownership and hence power in the hands of passive asset managers; the dominance of the UK and the Netherlands as conduits for Offshore Financial Centers.
Conceptual innovation: An important theoretical innovation comes from the conceptualisation of network formation as the interplay of strategies deployed by persons, corporations and owners. This allowed for a more precise modelling of complex network interactions and network formation. In particular, we have developed a novel conceptual and empirical approach to study the flow through networks, and the importance of particular (sets of) nodes such as countries or firms in these flows. Furthermore, the big data approach makes it possible to search for those regions and clusters that are empirically distinguishable and use these as a basis for comparative research, instead of using pre-defined categories (such as countries). We generate new classifications of global communities that go beyond fixed political-geographic lines.
Methodological innovation: A key goal was to transmute the advanced methodological toolbox from network science to social network analysis. This opened up a number of new and unexplored questions, for instance on the use of approximation algorithms, the differences between approaches in computer science and physics, and the use of the toolbox for social science applications. This has made advanced and computationally strong methods available to scholars that study political economy and socio-economics.
Empirical innovation: We conducted the first big data network analysis of the ownership ties and interlocking directorates on the largest dataset currently available, covering over 200 million firms worldwide. This endeavor required a significant investment in cleaning and organizing the data in such a way that it is suitable for social science network analysis. The data enhancing and quality assessment techniques form a first set of innovative outcomes with a large potential impact on how social scientists can deal with big data in the years to come. In addition, analyzing the properties of the web of corporate governance relations uncovered new and unexpected properties, such as hitherto unnoticed levels of concentration of corporate ownership and hence power in the hands of passive asset managers; the dominance of the UK and the Netherlands as conduits for Offshore Financial Centers.
Conceptual innovation: An important theoretical innovation comes from the conceptualisation of network formation as the interplay of strategies deployed by persons, corporations and owners. This allowed for a more precise modelling of complex network interactions and network formation. In particular, we have developed a novel conceptual and empirical approach to study the flow through networks, and the importance of particular (sets of) nodes such as countries or firms in these flows. Furthermore, the big data approach makes it possible to search for those regions and clusters that are empirically distinguishable and use these as a basis for comparative research, instead of using pre-defined categories (such as countries). We generate new classifications of global communities that go beyond fixed political-geographic lines.
Methodological innovation: A key goal was to transmute the advanced methodological toolbox from network science to social network analysis. This opened up a number of new and unexplored questions, for instance on the use of approximation algorithms, the differences between approaches in computer science and physics, and the use of the toolbox for social science applications. This has made advanced and computationally strong methods available to scholars that study political economy and socio-economics.