Abstract
This thesis examines how skill premium movements can be explained by productivity shocks and how these can be explored within macroeconomic models. Chapter 1 uses structural VAR models to identify shocks model skill premium responses in US data, while chapter 2 estimates a Dynamic Stochastic General Equilibrium model in the same data to explain the skill premium movements. Chapter 3 expands the model to include household education decisions, estimating a Dynamic Stochastic General Equilibrium model in UK data and simulating the impact of fiscal policy on welfare and inequality. Existing literature modelling the skill premium has relied on the assumption that patterns are caused by skilled labour being a relatively stronger complement to capital inputs than unskilled labour; the Capital-Skill Complementarity Hypothesis. However, recent studies have found that this assumption is not supported in data. In chapter 1, structural VARs are used to test whether Capital-Skill Complementarity can be rejected by the data, finding that the results cannot reject the hypothesis. While Dynamic Stochastic General Equilibrium models including the skill premium have been used in the existing literature, these models do not estimate parameters from data. Chapter 2 estimates a model in US data to find parameter values and compare the performance of alternative specifications. The results suggest that the current calibrations used in the literature are too conservative and that the standard model is outperformed by an alternative specifications, although still retaining the Capital-Skill Complementarity Hypothesis. In chapter 3, the model is expanded introduce household education spending decisions and is estimated using UK data. The findings here reject the Capital-Skill Complementarity Hypothesis. Using this model, fiscal policy changes are simulated, finding that distortionary taxes reduce overall welfare but increases in educational spending subsides and reductions in unskilled wage taxes can be used to reduce inequality.