(What) Do Unions Maximise? Evidence from Survey Data

Highly formalized models of union behaviour have been developed and are widely used within neo-classical economics. This approach is justified by the argument that most goals pursued by unions can be reduced to a wage (or wage-equivalent) - employment trade-off. However, the strong assumptions made about union goals and preferences, how these are determined and the range of issues over which a union will seek to bargain with a firm are not supported by extensive empirical evidence. This paper explores these issues through a survey of union leaders to ascertain the range of union goals and their preferences, and the types of issues over which unions bargain with employers. The results of the survey present convincing evidence that standard neo-classical models do not adequately capture union bargaining behavior, or union goals and preferences. The findings presented here also throw open, but do not resolve, the question of whether unions are 'rational maximisers'. The most interesting result - not widely found within the economics or industrial relations literature - concerns the nature of union bargaining preferences. Union bargaining over employment appears to be asymmetric around current levels of employment in that union appear to care more about the employment implications of their bargaining strategies, when they are likely to result in employment decreases