What has happened to union recognition in Britain?

This paper examines the determinants of establishment-level union recognition status using data from the three Workplace Industrial Relations Surveys of 1980, 1984 and 1990. Our theoretical approach argues that product and labour market variables dated at (or around) the establishment's setup should be considered the crucial determinants of recognition status. Confirmation of these predictions is found for private-sector manufacturing establishments in the econometric analysis. Of particular interest are: (i) the nonlinear relationship between historic industry quasi-rents and recognition status and the stability of this aspect of the model over the three data-sets; and (ii) a marked downward shift in recognition probabilities that occurred during the 1980s, which seems to be largely a consequence of the failure to organize new establishments set up in the 1980s.