Wage determination and strike activity in the New South Wales coal industry : trade union and employer bargaining

Abstract

The New South Wales (NSW) coal industry has been one of the most strike torn industries in the world, with violent and bitter battles between trade union and employer. This thesis seeks a greater understanding of the determination of wages and strike activity in this industry. The thesis is presented in two parts. Part one considers difficulties encountered when modelling the objectives of a trade union, and the outcome of bargaining between a trade union and firm. Part two applies models of the objectives of trade unions, and bargaining, to issues of wage determination, and strike activity in the NSW coal industry. The process of wage determination in the NSW coal industry is investigated using the reduced form of Svejnar's (1986) model. The major prediction of Svejnar's model, that there is a positive relationship between the industry surplus and the industry wage, is strongly supported. The results are improved upon by adapting Svejnar's model. A first-order dynamic adjustment model is used to allow for a more complicated dynamic structure than that assumed in the theoretical model. Furthermore, by taking into account some of the aspects of the bi-sectoral (open cut and underground) structure of the coal industry and the heterogeneity of its labour force, Svejnar's model is generalised from its specific reduced form. The thesis also provides a survey of the major, and more influential, models of strike activity. Some of these models are then applied to data from the NSW coal industry, resulting in three major conclusions: (i) with the exception of the Hayes' (1984) model, the theories considered do not provide acceptable explanations of strike activity in the NSW coal industry; (ii) there are common empirical relationships that are predicted by authors of very different theories; and (iii) there is a need to treat measures of strike activity as potentially different, rather than alternative, indicators of strike. An eclectic model of strikes is then developed. This eclectic model is based on Tracy's (1986) world-wise approach and was greatly influenced by: the theoretical survey of strike models; the empirical application of these models to the NSW coal industry; and familiarity with the NSW coal industry. Each of the different measures of strike activity is modelled using a common set of explanatory variables in an attempt to ascertain the commonality, or uniqueness, of the relationships determining strike activity. The results suggested that the explanations for strike frequency, the size of strike, strike intensity, and the average duration of strikes are not the same. Indeed, the only variable which was found to have a significant effect on all four of these measures of strike activity is the level of market concentration in the NSW coal industry. (This variable also had the highest elasticity, at the sample mean, of all the significant variables in the regressions for strike frequency, average duration, or strike incidence in the NSW coal industry.) The explanation for this dominance of market concentration on the results is unclear. From both empirical and theoretical perspectives, there appears to be a need to consider the role of industrial relations more fully. Despite this common theme that runs through the regressions for strike frequency, average duration, and strike intensity, it should be stressed that each of these regressions also exhibits combinations of significant empirical relationships which are unique to itself. This result rejects the use of these measures of strike activity as simple alternatives for each other. Show more