https://orcid.org/0000-0002-4632-9288
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Regulatory initiatives are frequently shaped by the ability of the financial industry to build alliances across the wider business community. Yet comparative and international political economy scholarship remains divided over how to explain the resulting networks of financial lobbying. Using quantitative text analysis of 1300 responses to EU financial regulatory consultations between 2010 and 2018, we map patterns of lobbying coordination based on cosigning and text re-use in consultation responses for the first time. This unique dataset is used to analyze hitherto hidden patterns of domestic and cross-border coordination by financial organizations within and between European countries. We find that while distinctive national lobbying networks persist at the country level, the internationalization of financial actors is statistically associated with the formation of coordination ties with foreign financial actors. This suggests that European financial integration has facilitated the emergence of new cross-border alliances which complement – rather than substitute for – existing domestic financial interest coalitions. We argue that the text-as-data approach employed here makes an important new contribution to scholarship on business power and the political economy of Europe.
Acknowledgments
The authors wish to thank the three anonymous reviewers for their helpful and constructive comments. The article was presented at the ECPR Joint Sessions of Workshops in Mons, April 2019, the EUSA Biennial Conference in Denver, June 2019, the EPSA in Belfast, June 2019, and at workshops at the University of Oxford, University College Dublin, and Leiden University. We thank all the participants at these events for their comments, particularly Geoffrey Underhill, Matthias Thiemann, Greg Fuller, Natascha van der Zwan, Erik Jones, Amin Samman, Nikitas Konstantinidis, and David Howarth. Finally, we are extremely grateful to Steffen Murau for the research assistance and to Waltraud Schelkle and Dorothee Bohle for inviting us to be part of the Special Issue.
Disclosure statement
No potential conflict of interest is reported by the author(s).
Notes
1 Respondent type was coded as: (1) Business organization, (2) Organizations representing the interests of workers, (3) Non-governmental organizations, indifferent of the domain in which they are active, (4) Institutions, defined as ‘semi-public organizations, which fulfill a public role and which do not have profit as their first goal, but which do not have the legal status of an NGO’, (5) Public authorities, (6) Mixed category.
2 Since this analysis is computationally demanding, we first filter those pairs of sentences in the corpus that are most likely to contain text re-use by implementing a ‘Minhashing’ technique. This involves converting text strings to shortened numeric references (known as hashes), and then filtering out those pairs of sentences below a minimum level of similarity (see Wilkerson et al., 2015).
3 We excluded all sentences that shared at least 10 consecutive words with the Commission text (or an equivalent score generated by the Smith–Waterman algorithm), and excluded all sentences where the overlap with the Commission text was at least a third of the sentence.
4 Since the algorithm is calibrated to add three points for each additional word that two documents share in the same order, a score of 30 indicates a sentence with 10 shared words. We apply a high minimum SW score of 60, equivalent to 20 consecutive shared words per sentence (see Supplementary material Table A2 for example).
6 Data from the World Bank website, based on data from the IMF, International Financial Statistics and data files, and World Bank and OECD GDP estimates. We took the average of both values, individually, for the 2008–2018 period.
7 STAN data ar based on OECD member states’ annual national accounts by activity tables, supplemented with data from other sources (e.g. national industrial surveys/censuses) to estimate missing detail. We used the ISIC Rev. 4 data to derive country-level information on value-added, both measured in current prices. Data is broken down within STAN into different industry categories: we use financial and insurance activities, and divide this value by the total for all industries, to derive a country-year level observation representing the percentage contribution of the financial industry to total value-added in the economy, in a given year (see Young, Citation2015).
8 We also generated a variable that evaluates country-level financialization by production measures (rather than value-added), but these were highly correlated (Pearson correlation coefficient = 0.9429).
9 Because of the possibility of measurement error, we replicated our analysis without the financialization indicator, and found the results to be substantively the same. For each of these variables, we took the average of all available years for the period of analysis, as yearly data was only available until 2017 for only a small number of countries.
10 This variable captures the possibility that the greater openness of a financial system to international flows could affect the opportunities for coordination available to financial groups, independently of how internationalized its activities. The ‘Internationalization’ variable measures the extent to which the banking system of a country as a whole is interconnected with other countries. We used data on Banks’ Foreign Claims by Home Nationality from the Committee on the Global Financial System (GCFS, 2018). This was divided by each country’s GDP (from the Global Financial Stability database) and averaged over the 2010–2018 period because we were not confident that across-time analysis would be useful given the uneven nature of financial policies over time.
11 We cannot rule out the possibility of measurement error in this instance. Although banking internationalization is only one measure of country-level internationalization, it was the broadest indicator that we could find for this number of countries.
Additional information
Notes on contributors
Scott James
Scott James is a Reader in Political Economy at King’s College London. His research examines the political economy of financial regulation, business power and lobbying, and the importance of economic ideas and narratives in public policy.
Stefano Pagliari
Stefano Pagliari is a Senior Lecturer in the Department of International Politics at City, University of London. His research covers a number of themes related to international and comparative political economy, with a particular focus on the political economy of finance.
Kevin L. Young
Kevin L. Young is an Associate Professor in the Department of Economics at the University of Massachusetts Amherst. His research focuses on the politics of financial regulation, transnational policy networks and the role of business in shaping global governance.