![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
This article presents a three-dimensional conceptualisation of conflict of interest (COI) regulation directed towards assuring the impartial and unbiased decision-making of parliamentarians. It distinguishes and separately measures (based on a new dataset) COI strictness, sanctions and transparency and shows that they indeed constitute empirically separate dimensions of parliamentary ethics regimes adopted in European democracies. In order to illustrate the usefulness of these indices, the article then examines the relationship between the three indices and trust in national parliaments across 25 democracies. Unlike the Sanctions and Transparency indices, the COI Strictness Index (composed of strictness of rules and enforcement) has a significant and robust negative association with trust, which highlights the importance of disentangling different elements of COI regimes. While future research has to explore the causal relationships between COI regulation and trust, capturing the complexity of COI regimes in an unbiased fashion and thereby making them comparable across European democracies is an essential step towards doing so.
Acknowledgements
This article forms part of a collaborative project on ‘Conflict of Interest Regulation in Europe’ jointly conducted with Fabrizio di Mascio and Alessandro Natalini. Many thanks go to them for their contributions to the project as well as to Nicholas Allen, André Kaiser and the two WEP referees for their detailed and helpful feedback on the article. We also thank staff at the Center of Comparative Politics, University of Cologne, at the School of Government and Public Policy, University of Strathclyde, and the participants in the panel ‘Regulating Political Actors in Old and New Democracies: Parties, Politicians, Interest Groups’ at the EPOP Conference in Cardiff, for feedback on earlier versions of the article. All remaining errors are solely ours.
Notes
1. Our analysis excludes voluntary rules which are not formally enforceable (Sieberer et al. Citation2016: 63).
2. This is distinct from anti-corruption or anti-bribery legislation regulating MPs as self-interested individuals who try to gain financial benefits. Giving in to conflicts of interest in legislative decision-making is associated with bias but does not necessarily generate private financial gain.
3. Greco is the ‘Group of States Against Corruption’, an international forum which is part of the Council of Europe. Greco was established in 1999 to monitor member countries’ compliance with the organisation’s anti-corruption standards.
4. This contains two categories of regulation: those that obligate MPs to declare – regarding individual decisions – that they are affected by a conflict of interest or provisions that require MPs to excuse themselves.
5. Figure A1 in Online Appendix A shows the distribution of the rule strictness scores reflecting the combinations of legal mechanisms country by country.
6. See Figure B3 in Online Appendix B for an overview of the scores capturing enforcement in our sample.
7. The Supreme Authority for the Transparency of Public Life is coded as an independent body as it consists of six appointees from the high courts of the state (Conseil d’État, Court of Cassation and Court of Audit), two parliamentary appointees and one appointee of the President of the French Republic.
8. We assign a score ‘zero’ on one end of the spectrum to all-permissive regimes and a score ‘eight’ to the most constraining ones. A regime without an enforcement structure for preventive COI rules but an independent watchdog with broad monitoring powers for disclosing rules takes an intermediate position with a score of ‘four’.
9. In comparison to the measurement of rule strictness, we capture the constraints inherent in the COI enforcement structure overall (i.e. we do not weigh enforcement of preventive rules more heavily than enforcement of disclosing rules) to capture whether any given rules are underpinned by enforcement structures and if so by which type.
10. It may seem that combining rule strictness and enforcement can lead to the unwanted weighting of the components constituting these two parameters (OECD Citation2008: 31). Yet we avoid this problem as we systematically transfer our ratings assigned to these components into the proportions of constraints in the regulation. See Figure B4 in Online Appendix for the distribution of the COI Strictness Index in the sample.
11. See for further details on the rank assignment Online Appendix C.
12. A COI regime has a transparency rank of ‘eleven’ if all the transparency options are coded as present, which indicates the maximum possible level of transparency with regard to both preventive and disclosure rules and a ‘zero’ if none are present. A rank of ‘five’ is assigned to regimes that, for instance, have transparency requirements in relation to both preventive and disclosing rules, yet in either case the scope of the information published is limited, while no information on rule violations is released. Note this is only one possible institutional constellation that might receive a rank of ‘five’.
13. We construct an additive index of overall constraints generated by transparency requirements as linked to preventing and disclosing rules. We do not prioritise preventive rules over disclosing rules in this case as there are no theoretical reasons or empirical indication that release of information on non-compliance with one type of rule is perceived as more or less problematic by those regulated or by the public. Note that, as with the measure of rule strictness, the indicators are interdependent (e.g. if no public information release is required, the scope of such release is of no relevance). We consider the whole theoretically possible range of variation and also assure equal distances between the different levels of the index.
14. See for details on the trust data the next section and Online Appendix D.
15. We lost Norway because Eurobarometer (EB) does not include information on it.
16. Note that tests for multicollinearity of the presented models do not reveal any problems.
17. We are aware that 25 countries can be a critical sample size for a multi-level design. Note that the results of logistic regression with clustered standard errors are the same.
18. Online Appendix D provides details of additional robustness checks and tests regarding endogeneity and reverse causation.
19. Online Appendix D Table D2 provides results for additional specifications.