A Model of Union Behaviour and Benefits under Uncertainty

for press This paper offers some results on the relationship between unemployment benefits and employment in a stochastic framework. The intention is not to provide an exhaustive analysis of this relationship but, rather, to investigate the effects of uncertain unemployment insurance in a unionised economy. Uncertainty about the system of unemployment benefits may be important for those entering unemployment. There are indeed a number of features of the model of benefit incentive effects that may be discussed, but we have looked at only one aspect of the problem: the effect of compensation uncertainty on employment and union power in a unionised economy. Consider an economy with a monopoly union (a monopoly model). The union with identical individuals-members maximises an objective function taking into account the effects of its wage-setting on the employed union members. In this context, literature shows that an increase in wages reduces employment in the unionised sector, and each worker therefore faces a greater chance of being unemployed whereas the members which remain employed will achieve a higher wage. This model, where membership is endogenous and, therefore, included in the union objective function, produces, in the long run (steady state), a higher employment level: as union membership is conditional upon employment, this implies that the union prefers higher employment. In any case, the exogenous impact of a higher reservation wage (benefit in our case), affects union preferences increasing union wage. Unemployment benefits reduce the cost of becoming unemployed, leading the union to take greater risks over its wage claims. Is this theoretical result robust in a stochastic environment? In a context where there is uncertainty about unemployment benefits, the monopoly union model shows that the union moderates its wage claim and this moderation increases both employment and, in a closedshop mechanism, membership. This theoretical result has been tested for the U.K. labour market during the 1980s. In particular, we estimate a VAR model for employment, real wage and membership (that is the variable defined by the theoretical model) and carry out dynamic simulations by increasing the volatility of the benefit variable. We find that an increasing volatility (uncertainty) in the benefit variable yields, in the long-run, a higher employment level (a result which is consistent with our theoretical prediction) and a drop in union membership (a result which contrasts with the theoretical prediction). Although the employment result seems coherent with the theoretical model (and the stylised facts), the membership effect moves in a perverse direction, since the closed-shop arrangement used to model the membership dynamics yields higher union power. The model’s result is, therefore, conditioned by the degree of closed-shop arrangement. In fact, one of the key determinants of union membership used by literature as an explanation of the fall in union members during the 1980s is the government behaviour which significantly influenced workers’ choice to become union member or to renew membership.