Individual Voice in Employment Relationships: A Comparison under Different Forms of Workplace Representation


  • David Marsden

    1. Department of Management and Centre for Economic Performance, London School of Economics
    Search for more papers by this author
    • The author thanks the DARES of the French Ministry of Labour for access to the REPONSE data, and the sponsors of WERS (Department for Business, Industry, and Skills [BIS], the Advisory Conciliation and Arbitration Service, the Economic and Social Research Council [ESRC] and the Policy Studies Institute), and the ESRC Data Archive for access to the WERS data. Thanks are due to Jörn-Steffen Pischke and Barbara Petrongolo, members of the Centre for Economic Performance Labor Market Workshop, Thomas Amossé and Loup Wolff, and two anonymous referees for their much appreciated advice.


This article considers the role of individual employee voice in regulating the “zone of acceptance” within the employment relationship and examines the extent to which different models of collective voice inhibit or foster the operation of individual voice. It focuses especially on the role of representatives who deal with job-level grievances and who operate within contrasted frameworks of collective voice. In one, representation is negotiated with the employer, and in the other, it is based on rights established in employment law. The former is commonly associated with shop stewards and unions, and the latter with employee delegates and works councils. It is argued that whereas in the negotiated model, individual and collective voice are substitutes, in the rights-based one, they are complements. The article also considers how this may alter under dual-channel representation based on both unions and councils, which is very common in European workplaces. Britain provides an example of the negotiated model and France of both the rights-based and dual-channel models. These ideas are tested using data from the 2004 British and French workplace employment relations surveys and confirmed using data from the 1998 surveys.