Abstract
It has recently been argued in an influential literature that unions typically improve productivity by providing workers with an internal 'voice' at the workplace. The alleged efficiency gains accrue through a reduction in quits as unions collect information about the preferences of all workers, and through a union governance or job-regulatory apparatus that better monitors labour's effort input while encouraging co-operation between workers. The assumptions underpinning this new view of unionism and the methodology employed to isolate the union productivity differential are critically evaluated in this paper. It is further argued that the unions and efficiency question is fundamentally one pertaining to the long run. Empirical evidence showing a deleterious effect of unions on productivity growth and profitability, coupled with material assembled from the economics of politics literature, is deployed to substantiate this argument. It is concluded that the productivity gains attributed to unionism are, at best, inconsequential.