Industrial conflict in Britain
The economic analysis of conflict in Britain has previously concentrated on examining aggregate strike frequency. The thesis recognises the limitations of this approach and argues for the investigation of a broader definition of conflict and at a more disaggregated level. While weakly encompassing previous theoretical work, the principal objective is to establish the patterns and trends pertaining to wider set of measures of conflict in post-war Britain. The empirical investigation of these disaggregated dimensions of conflict and their inter-relationships appears to have previously received only very limited attention. Following a critique of the extant theoretical and empirical literature, the first substantive chapter examines the traditional aggregate- econometric models of strike frequency. These are shown to be unsatisfactory in a number of ways. The chapter then turns to the central issue of the procyclicality of strikes. It is shown that while the total number of strikes is only very loosely related to the cycle, strikes arising over the level of remuneration bear a much closer correspondence with the level of economic activity and this finding accords with many of the theoretical models that have been proposed for strike. activity. The chapter concludes with an examination of a cyclical-political model of strikes within which the impact of the recent reforms in labour legislation is also investigated. One of the central arguments of the thesis is that the emphasis on strike frequency is inappropriate. This is most clearly illustrated by the fact that while strike frequency fell by almost one quarter between 1980 and 1984, the incidence of strikes at the establishment level actually increased by 45%. An examination of the determinants of the incidence of conflict activity forms the basis of the second substantive chapter of the thesis. As a subsidiary theme, the complementary nature of strike and non-strike action is also explored. The next chapter investigates the ceteris paribus differences in strike probabilities between the public and private sectors. While the levels of strike incidence and frequency appear to be much higher in the public sector, much of the divergence is found to be a consequence of differences in the characteristics of the two sectors. Additionally, when weighted by employment and/or union coverage, strike frequency is found to be lower in the public sector and, moreover, each of these strikes tends to be shorter and involve fewer workers. The final substantive chapter looks at the impact of strikes on industry output and efficiency. The structure of the model is novel in that a production frontier is estimated without having recourse to an explicit functional form for the inefficiency component. This is due to the availability of a panel of data in which the fixed effects can be viewed as capturing both the inefficiency term as well as the industry fixed effect. A second stage estimation is then used to identify each industry's level of efficiency. While strikes do not appear to reduce output in aggregate, there is some evidence to suggest that those industries which incur a large number of short strikes do have their output significantly disrupted. This loss of output also serves to make these industries less efficient in general. Thus a major conclusion is that a disaggregated approach is necessary in order that the multi-dimensional nature of conflict and the sectoral diversity in the incidence of industrial action can be investigated in a satisfactory manner. Any new theories of conflict will need to encompass the empirical findings of the thesis.