Action research literature promotes broad participation in order to gain better insights into prevailing issues and cope with both present and future challenges in organizations. For good reasons, action researchers view participation as desirable and even necessary. However, emphasizing participation also creates a blind spot: researchers tend to assume that employees are willing or even eager to take part in organizational change projects. A group of action researchers involved in a large-scale organizational development project in a Norwegian public organization initially also had this optimistic assumption. Over time, they realized that many employees were reluctant to participate and often kept silent, so they conducted follow-up research to explore the reasons for employees’ non-participation. The findings show that the employees had rational reasons not to participate, and that employee participation should not be taken for granted. The paper outlines eight different rationales for non-participation, and discusses implications for action research.

Across industries, union density is under great pressure from different forms of organisations and, in many ways, a more individualised working life. Employee relations within the public sector have undergone a transition due to privatisation, decentralisation, and the adoption of quality management approaches. Employee relations in Nordic countries are strongly embedded in national regulations and agreements. However, research on workplace development within the public sector rarely includes discussions of the union role. The Nordic model perspective acknowledges that the different social parties share interests and visions, and it promotes a collective effort when workplace development is sought. This paper poses the question of how public organisations can change the “boxing and dancing” behaviour in union–management relationships through the establishment of a parallel organisation (PO). The PO serves as a different organisational mode when the operating organisation is unable to successfully deal with certain prevailing issues, where knowledge rather than authority should determine decisions. The findings show that the PO creates a “dancefloor”, less confined by bureaucratic barriers, where unions and managers co-create new relations. In addition, participants experience more enhancement of their roles, and their focus towards developing their workplace collectively is more prominent. Our findings contribute to the industrial relations literature by proposing POs as a tool for building relations between unions and managers in a public organisation. Our paper also contributes to the PO literature by proposing that the inclusion of unions in a PO can be crucial when attempting to transfer outcomes into the operating organisation.