Who's the Boss? Analysing Drivers of the Shape of Business Lobby Networks in the Netherlands and UK

Contemporary theories on the structure of policy and lobby networks are surprisingly static in nature. Commonly, political scientists describe the shape of lobbying network structures of corporatist arrangements as one with a core and periphery with a prominent role for selected umbrella organizations. The structure of pluralist set-ups is rather decentralized and hollow cored with a broad variety of organized interests that compete for access and influence. Our knowledge however, of what such networks actually look like, how they develop over time and what drives the shape of them in the empirical world, is very limited and has a focus on pluralist contexts. We hardly know therefore whether country-level and/or domain-level variation factors affect the structures of lobbying networks. In this paper, I argue and aim to empirically demonstrate that a shift in business representation has occurred in corporatist contexts. Some scholars argue that due to the decline of (neo-)corporatism, the role of umbrella organizations has become less central and that they are not by default involved in the policy-making process anymore. Their members, corporations, therefore have been urged to seek complementary modes of representation such as lobbying in coalitions or alone. In a corporatist set-up I therefore expect to identify lobbying networks with less central roles for umbrella business organizations and more prominent roles for corporations over time. I do however, presume to find a difference between policy domains in which the networks are embedded as umbrella business organizations tend to traditionally still have a strong political network and knowledge on specific issues such as labor policy. For this very reason, corporations might still opt for representation through associations when it comes to such issues. The current paper tests this expectation by identifying and explaining the shape of lobbying networks through employing social network analysis, topic modelling and time series analysis with temporal data (1990-2017) of British and Dutch parliamentary hearings and political news coverage.