Over time, a body of literature concerning the magnitude and the determinants of skill differentials has been developed by Australian and overseas authors. The literature on skill differentials, however, has been limited. The studies of skill differentials in Australia have almost invariably been concerned with the criteria used by Commonwealth authorities in determining award skill differentials.
Although there has been much theorizing as to the determinants of skill differentials, little empirical work, in the sense of the statistical testing of hypotheses, has been undertaken. Many authors appear to have simply presented, and commented on, the possible factors which could be held to determine skill differentials.
This thesis is a modest step towards filling this gap. It aims, wherever possible, to statistically test the various postulated relationships contained in the body of literature concerning skill differentials. In particular, the thesis aims to empirically examine the determinants of changes in award skill differentials in Australia for the period 1953 to 1973. It also examines the question, whether or not changes in actual skill differentials are a function of market forces. The study is confined to the manual workforce. The skill differential, in this thesis, is defined as the wage rate of the skilled expressed as a percentage of the wage rate of the unskilled.
The empirical quantification of the various hypotheses on skill differentials is important since it could be used as a basis for policy decision making. The policy decisions that are associated with skill differentials are numerous and touch on the considerations of: unemployment; earnings drift; efficiency in the allocation and use of skilled and unskilled labour; the incentive to acquire skills; and wage indexation criteria.
The methodology used in this thesis i s the application of the ordinary least squares regression technique. Linear as well as nonlinear relationships have been used. Due to the existence of structural shifts, and observations which could be considered atypical, dummy variables were used where applicable. The data consisted of published as well as unpublished data which were supplied by the Department of Labour and Immigration. The analysis was carried out on an aggregate as well as on an industry basis, and, wherever possible, these results are compared.
The thesis is divided into five chapters. Chapter Two outlines the various theories on skill differentials and also examines the criteria used by Commonwealth wage fixing authorities in determining award skill differentials in Australia.
Chapter Three tests some of the theories outlined in Chapter Two. The hypotheses examined include: whether changes in award skill differentials are a function of the level of excess demand for skilled relative to unskilled labour; the level of excess demand for labour; changes in the consumer price index; and whether changes in award skill differentials are interdependent.
Chapter Four addresses itself to the correspondence between award skill differentials and actual skill differentials, and examines whether changes in actual skill differentials are a function of the excess demand for skilled relative to unskilled labour. However, due to the lack of data, an examination of the above relationship was undertaken using earnings growth and drift data by industry.
Chapter Five presents a summary and outline of the main conclusions that were derived from the thesis.
It is to be hoped that the main findings of the thesis, and the method and approach adopted here, would be of some use to writers on Labour Economics and to theorists in modelling future contributions.