Economic inequality over the life cycle in Australia

Abstract

This thesis investigates the key factors that govern the economic behaviour of Australian households and contributes to optimal income taxation, all within an overarching life cycle framework. Throughout the thesis there is also a focus on modelling individual heterogeneity, consistent with the growing efforts in macroeconomics to go beyond the representative agent model in examining household behavior and optimal fiscal policy.

Chapter 1 (published) explores the life cycle pattern of wage inequality in Australia. Using panel data on male wages, it explores the relative importance of unobserved worker heterogeneity versus random wage shocks in explaining the upward trend in inequality with age. It finds significant heterogeneity in wage levels but no evidence of systematic heterogeneity in wages growth. Instead, highly persistent wage shocks are found to account entirely for the rise in wage inequality with age.

Chapter 2 explores the statistical properties of the GMM estimator that underpins Chapter 1. This simulation study explores whether GMM estimation is more accurate at recovering key parameters when specified in levels or growth rates in the context of panel survey data.

Chapter 3 (review and re-submit) explores risk aversion among Australian households using HILDA panel data. Exploiting data on households' share of risky assets, risk aversion is found to be constant in wealth once measurement error is accounted for. A consumption Euler equation that adjusts for measurement error in consumption data, is used to estimate the coefficient of relative risk aversion. The preferred non-linear models suggest a moderate degree of risk aversion for the typical Australian household, with values ranging from 1.2 to 1.4. These findings can provide guidance for calibrating household preferences in macroeconomic models of the Australian economy.

Chapter 4 (under review) provides estimates of the Frisch (or wealth-compensated) elasticity of labour supply for Australia using 17 waves of HILDA data. Critically, the chapter uses an estimation approach that incorporates the extensive as well as the intensive labour supply response, and finds that the resulting aggregate elasticity is large -- ranging from 1.5 to 2.4. This is well above the 0 to 0.5 range typically found in international studies of prime-age men that ignore the extensive margin. The chapter's sub-aggregate results confirm a relatively large wage elasticity for women and older age groups, especially along the extensive margin of labour supply.

Chapter 5 reexamines optimal taxation of capital in the long run. Recent studies using OLG models cast doubt on the classical result that the optimal tax is zero. The chapter explores whether these findings carry over to a more realistic model that includes dual-earner households and life cycle marital risk. Analytical results are first derived from a simple two period model. For plausible parameter values, the saving of married households is higher and, conversely, the saving of single households is lower than in a world without divorce and marriage risk. The net impact of these competing effects is to reduce saving in the most empirically relevant cases. The chapter then examines optimal tax in a fully-fledged life cycle model with dual earners and marital risk (and other standard features) calibrated to the United States economy. The optimal steady state tax on capital is found to be zero. The main reasons are twofold. First, a relatively low tax on capital mitigates the tendency for lower saving induced by marital risk demonstrated in the two period model. Second, the dual labour supply margin acts as a substitute for precautionary saving and thereby increases the elasticity of saving with respect to the (after tax) interest rate. This chapter therefore highlights the relationship between optimal taxation and the presence of other insurance mechanisms such as dual labour supply. Show more