부산대학교 도서관

This thesis is comprised of three essays studying the high and low wage ends of the labour market. Historically there has been little theoretical analysis of low-paid, non-standard work and so in the first and third essays I develop theoretical models incorporating two types of non-standard work: casual and part-time. The second, co-authored, essay is empirical, using data on football players to study the wage determination of very highly paid workers. In the first essay I study low wage, casual jobs. These jobs provide flexibility for firms to change the size of their workforce cheaply and quickly and for workers to choose whether to supply labour in every period. This flexibility comes at the expense of certainty for both firms and workers. I develop a search and matching model incorporating casual jobs, which I use to evaluate the effect of labour market policies on aggregate outcomes. The equilibrium of the model features the concentration of casual jobs at the bottom of the wage distribution. I find that a ban on casual jobs increases unemployment, but that the average wage of those employed actually increases. In addition, in the model with casual jobs, workers in low wage casual jobs continue to search for a higher quality match that will offer work more frequently. In a model with search frictions, this makes it harder for unemployed workers to match with a firm. This crowding out effect offsets some of the negative effects of a ban. I also consider the effect of a higher minimum wage for casual jobs. I find that the effects are limited. These results are due to an offsetting mechanism: although higher wages lead to higher unemployment, as firms offer more regular jobs, the number of workers called-up to work in any one period increases. In the second essay we turn to the other end of the wage distribution, to study the determinants of superstar wage effects, asking whether productivity or popularity-based explanations are more appropriate. We use longitudinal wage and performance data for workers (players) and firms (teams) from a particular market for sports talent: Major League Soccer in the United States. We find evidence that the top earners, whose annual salaries are mostly not accounted for by their past MLS performances, when compared alongside other footballers, are paid more because they attract significantly higher stadium attendances and thus revenues. There is no evidence that higher residual salary spending by the teams affects their relative performance in football terms, or that the amounts the teams spend on actual talent affect attendances. Taken together, these results suggest that a popularity-based explanation of superstar wage effects is appropriate among the top earners in this labour market. In the final essay I study long-term trends in part-time work. I show that there has been an increase in the part-time share over the past 30 years. There has also been an increase in the part-time work on the extensive margin; part-time workers on average work longer. Despite this, the difference in average hourly pay for part-time and full-time workers (the part-time pay penalty) has steadily decreased. I develop a flexible neoclassical model of the labour market which can explain firms' and workers' preferences for part- and full-time work. The equilibrium of the model matches key features of the labour market: full- and part-time workers undertake different tasks; there is bunching of workers at full-time hours; and full-time workers earn higher hourly wages than part-time. The model can be used to disentangle the effects of changes in workers' preferences and on firms' production technologies on the relative quantities and prices of part- and full-time labour. I provide an extension of the model which incorporates heterogeneity in workers' preferences, and will enable the study of gender differences in part-time work.