The hard impact of soft co-ordination: emulation, learning, and the convergence of collective labour standards in the EU

ABSTRACT

The article examines whether European countries' collective labour standards are converging and if so why. It presents new data on labour protection in 45 European countries and shows that the EU is a labour standards convergence club. It reviews the soft and hard instruments which the EU adopted in order to promote labour standards. The article models the competition, emulation and learning convergence mechanisms using a spatial econometric approach, and shows that EU countries learn from each other's successes and emulate one another. The article thus presents new evidence of the mixture of domestic and cross-national factors which shape labour standards in Europe. It adds to the growing evidence of countries' spatial dependencies in various policy areas. Most importantly, it provides evidence showing that soft co-ordination is an effective instrument which has had a positive impact on labour standards, and has led to horizontal convergence in the EU.

KEY WORDS:

• Convergence
• emulation
• europeanization
• labour standards
• learning
• spatial dependency

ACKNOWLEDGEMENT

The author is thankful to the Journal of European Public Policy's four anonymous reviewers for their helpful advice and suggestions.

SUPPLEMENTAL DATA AND RESEARCH MATERIALS

Supplemental data for this article can be accessed on the Taylor & Francis website, doi: 10.1080/13501763.2015.1022205.

Notes

1 I am thankful to an anonymous JEPP referee for bringing up this important point.

2 This specification is built on the highly plausible assumption that labour rights reforms are aimed, inter alia, at attracting foreign investment, which improves employment rates and supports growth.

3 Additional robustness checks of these results can be accessed in the online supplementary material.

4 The curiously high coefficient of EU membership may seem alarming, but as Kam and Franzese explain the coefficients of variables involved in interaction terms represent the effect they would have on y, when the other variables involved in the interaction are set to zero. ‘It may fall outside the range of what appears in the sample, or it could even be logically impossible!’ (Kam and Franzese 2007: 20).

5 Because the dependent variable is differenced, so too are the state-level variables. The veto-points score is included both in differenced and in levelled form

Additional information

Biographical note

Sara Kahn-Nisser is a lecturer at the department of Sociology Political Science and Communication, The Open University, Israel.