Media Briefings

UNIONS: New analysis of their impact on efficiency and employment

  • Published Date: September 2014

Unions have only a moderate impact on aggregate unemployment, according to research by Christian Bauer and Joerg Lingens, published in the September 2014 issue of the Economic Journal. But their theoretical analysis of the effects of collective bargaining also suggests that overall efficiency is lower: this is because higher wages deter new firms from entry while existing firms’ incentives to hire too many workers are not reduced.

Over the last three decades, almost all OECD economies have witnessed a decrease in union density, measured as the fraction of employees who are members of a union. Even in former union strongholds such as Denmark, Germany and Sweden, density declined by 6% to 9% over the last decade (with an OECD-wide average decline of 2.5%). Understanding the effects of this trend requires an analysis of how unionisation affects efficiency and employment.

That is the point of departure for this new study. The researchers show that unionisation results in a welfare loss that is rooted on the one hand in firms hiring too many workers and on the other hand in insufficient firm entry. But the effects of unions on aggregate unemployment are moderate.

Starting from standard economic modelling, this inefficiency result of unionisation seems to be straightforward. The threat of stopping production enables the union to bargain for a higher wage than the one that would prevail in the competitive situation. Consequently, the firm would hire fewer workers. The reason for the decrease in efficiency is then the decrease in employment.

An important implicit assumption for this argument to be true is that the hiring decisions of firms cannot affect the outcome of the wage bargain. But when hiring workers is a time-consuming process, firms gain a strategic advantage. The firm is tempted to hire too many workers.

With a large workforce, the firm finds it easy to replace any worker. This erodes the worker’s bargaining position and hence decreases the wage. Hiring is then partly driven by this strategic effect, putting too little weight on its effect on productivity. This is an additional source of inefficiency that is not taken into account by standard economic models.

Within this more realistic framework, the efficiency effect of unions is less clear. Unionisation (understood as collective bargaining) then might act as a ‘counterbalance’ to the firm’s strategic advantage, thereby neutralising the inefficient overhiring incentive.

Analysing a model that leaves room for efficiency increasing unionisation, Bauer and Lingens show that in this case too, unionisation in general decreases welfare (compared with an efficient benchmark situation). But the decrease is not primarily driven by its effect on aggregate employment: instead, it comes from distorting the entry decision of firms without neutralising the overhiring incentive.

The reason for this is that the bargained wage in the union case reflects the (per workers) opportunity costs of not being able to produce output and the search costs that would have to be spent to find a new workforce. But then the inefficiency effects add up. The overhiring incentive for the firm is still valid because the wage reflects the average value of production (which is decreasing in the number of workers). But in addition, the increased bargaining power of the workforce results in too high a wage, which discourages firm entry.

The researchers conclude that their results lead to a more thorough understanding of the effects of and reasons for efficiency losses caused by unionisation. First, because of the countervailing effects on firm-level employment and on the number of firms, the union has only a minor impact on unemployment.

Second, the efficiency loss is smaller in industries in which firms find it fairly easy to replace their workforce both in terms of hiring costs and value of lost production.

From a policy point of view, these results imply that policies aiming at ‘de-unionisation’ and policies that decrease the cost of replacing workers are substitutes rather than complements. This may help in the implementation of such policies.


Notes for editors: ‘Does Collective Wage Bargaining Restore Efficiency in a Search Model with Large Firms?’ by Christian Bauer and Joerg Lingens is published in the September 2014 issue of the Economic Journal.

Christian Bauer is at the University of Munich. Joerg Lingens is at the University of Muenster.

For further information: contact Joerg Lingens on +49 251 83 22 9 23 (email:; or Romesh Vaitilingam on +44-7768-661095 (email:; Twitter: @econromesh).