Response to Howard Quayle, Chief Minister of the Isle of Man

Howard Quayle, Chief Minister of the Isle of Man has recently stated that CORPNET did not list Isle of Man as a tax haven:
“I don’t think we are a tax haven, for a start. That’s not me saying that, that’s the OECD, that’s the University of Amsterdam who did a report very recently listing tax havens, the Isle of Man wasn’t one of them.”

The report: “Uncovering Offshore Financial Centers: Conduits and Sinks in the Global Corporate Ownership Network”, Javier Garcia-Bernardo, Jan Fichtner, Frank W. Takes & Eelke M. Heemskerk, Scientific Reports 7, Article number: 6246 (2017)

We clarify that we did not produce a list of tax havens, but a list of jurisdictions where the ownership of companies accumulates (sinks-OFCs) and a list where the ownership of companies goes through (conduit-OFCs). While Isle of Man may not be a large player in corporate tax avoidance, it may be a large player in other types of tax avoidance, such as VAT avoidance, and that cannot be inferred from our study. Recognized tax havens by the EU and the IMF such as Barbados, Antigua and Barbuda, Grenada, Montserrat, St. Kitts and Nevis, or Turks and Caicos Islands did not show up as sink-OFCs, for the simple reason that corporations/individuals do not organize their ownership structures from those places.

In our report, we found that owners of companies do not generally place their entities/vehicles in Isle of Man. This is similar to what we find with Guernsey, Ireland or the Netherlands (Figure A as lime-colored dot). On the contrary, according to our data, entities in Isle of Man are usually owned from other countries.

However, we found that Isle of Man is used as a conduit in corporate constructions that start from a sink-OFC, go through Isle of Man, and end up in a third country (Figure B as lime-colored dot in the bottom-right quadrant). While Isle of Man may not be a sink-OFC, our results indicate that the territory may be used to own companies in sink-OFCs.

For the complete list of sink and conduit-OFCs, please see Supplementary Information.

Comparison with several list of offshore financial centers

Two PhD positions on corporate control and network analysis

Two fully funded PhD positions are available at CORPNET, University of Amsterdam.
Closing date 1 May. Starting date (no later than) 1 september 2016.

We’re a multidisciplinary team, bringing together political science, computer science, network science, sociology, and based at the Amsterdam Insitute for Social Science Research. We’re interested in networks of corporate control. See

For the first position, we are looking for somebody with experience in longitudinal / dynamic network analysis . For the second position, we are looking for somebody with corporate governance expertise, ideally from a networked perspective.

Follow the links below for more information, or see Contact Eelke Heemskerk ( for more information.
Corporate Network Governance: Power, Ownership and Control in Contemporary Global Capitalism (CORPNET)
CORPNET seeks to do what has so far eluded existing scholarship: to fully explore the global network of corporate ownership and control as a complex system. Using cutting-edge network science methods, the project explores for the first time the largest database on ownership and control covering over 100 million firms. Exploiting the longitudinal richness of the new data in combination with state-of-the-art methods and techniques makes it possible to model and empirically test generating mechanisms that drive network formation.

PhD Project: What Drives Network Formation?

This subproject studies the generating mechanisms that drive the formation of corporate governance networks. The project builds a conceptual framework that tries to explain the formation and (dis)continuation of corporate board overlap through the strategies of the actors involved. It takes a bipartite approach and starts from the notion that the generating strategies of actors do not only take place at the firm-by-firm level but are essentially located at the firm-by-person level. The methodological approach is temporal modelling (for example, stochastic actor-oriented or temporal exponential random graph models), and the project will have to develop methods and algorithms to apply modelling on a very large scale.

PhD Project: Corporate Control in Contemporary Global Capitalism

This project engages in a fundamental discussion on the relationship between corporate control, ownership and corporate power. In a world of dispersed and fragmented ownership, corporate directors and managers are the more powerful actors. In a world of (re)concentrated ownership however, directors lose power and influence. The question, then, is the extent to which new concentrated owners such as states and large asset managers, and the corporate elite of directors are separable entities. This is an important question with regard to financial intermediaries and for venture capital, but also for the re-emergence of the state as corporate owner in the global economy. Where and how do the network ties of ownership and of managerial control reinforce each other? Are ownership ties an additional tool for the corporate elite to strengthen their grip on big business? Or can we discern opposing groups of corporate managers and directors on the one hand and the (new) owners of corporate capital on the other?

CORPNET gives plenary presentation at the CSS Winter Symposium in Cologne

We are glad to announce that the abstract “On Data Quality and Centrality Measures in Big Corporate Network Analysis” has been selected for a plenary presentation in the 2nd GESIS Computational Social Science Winter Symposium 2015.

Although often depicted as atomistic and individualistic market actors, corporations are tightly embedded in networks of power and control. They share board members (interlocking directorates), share owners and engage in direct shareholdings with one another. These so-called corporate networks have long formed distinct national business communities and are part of the organisation of national economies.

New databases and methods allow us to go beyond our current understanding of how the global economy is organised. However, with big corporate network data come big problems as well. In the CORPNET research group we analyse the networks of corporate control of more than 170 million companies and their directors (sourced from the ORBIS database). This rich data source provides both new methodological challenges when interpreting the result of applying for example centrality measures as well as challenges with respect to data quality assurance.

The abstract can be found here. See you in Cologne!