The Economics of Human Capital Investment, Rents and Wage Inequality
thesisposted on 05.02.2020, 12:31 by Akbar Ullah
This thesis examines three different, self-contained topics in the field of labour economics. The first chapter presents evidence which shows that firms are charging rent on training, capital and R&D investment, but they do not charge such a rent on working hours. It shows that under standard production functions, this evidence is inconsistent with models of human capital where wage setting takes place through bargaining and/or the worker’s quit decision is exogenous. Then, it develops a model where the worker has no bargaining power and makes optimal quit decision to better explain the evidence. In such a model, the firm optimal wage strategy is to charge rent on only those factors which are under firm’s direct control such as workers training. The model also contributes to the literature that tries to explain firms’ investment in workers transferable skills. The second chapter addresses the findings that workers who have high pre-job schooling also get more training during jobs. It develops a model of investment in human capital which shows that training can increase in pre-job schooling only if schooling increases efficiency of training in human capital production or affects worker’s preferences. But except the positive efficiency effects, all the possible preferences effects of schooling are on their own not enough to generate increasing wage function in schooling. When schooling has a positive role in human capital production, then human capital, consumption and wages are strictly increasing in pre-job schooling. The model generated net of training cost wage distribution can at least partly explain the US workers’ earnings distribution by education categories. Unlike standard labour-leisure choice models a novel and empirically testable prediction of the model is that, although highly qualified workers work for less time, they spend this time in learning rather than in leisure. Chapter three addresses the conflicting evidence about the wage return of training. Longitudinal panel data based studies during the 1990s and the 2000s give very high wage return of training. On the other hand, studies in the 2010s that are based on randomised experiments estimate insignificant return from training. Using the recently developed heterogeneity robust techniques, this chapter shows that training’s immediate wage effects are low on average but it has high and significant long term wage effects. The chapter concludes that the experiment based studies suffer from small sample size and short time span as in these studies only immediate post-training period data is considered for the analysis.