Tax, transfers, and income inequality in New Zealand during 1988-2013

Abstract

This dissertation investigates the role tax and transfer policy changes played in the evolution of New Zealand disposable income inequality between 1988 and 2013. Across five papers, the key changes in tax and transfer policies are identified, the labour supply response of individuals to the changes are estimated, and the impact of these changes on the income distribution is quantified. Overall, nearly 40% of the increase in income inequality during this period is attributable to changes in the tax-transfer system.

The tax and transfer payment changes investigated in this dissertation cover the gradual flattening of the tax scale over the 1980/90s, the reduction in benefit payments following the 1991 Mother of All Budgets, the introduction of Working for Families in 2005, and the erosion of transfer payments relative to the average wage throughout the period.

Given these changes, the efficacy of the tax transfer system for meeting vertical and horizontal equity goals is evaluated using data from the Household Economic Survey (HES). The redistributive effect of tax-transfer policy fell from 22.6 Gini points to 18.2 Gini points between 1988/91 and 2011/13, with a corresponding decline in the amount of vertical equity in the tax-transfer system. Between the same periods the degree of horizontal inequity rose,although this was predominantly the result of greater targeting in the tax-transfer system.

The adjustment in the structure of the tax-transfer system not only leads to a change in tax liabilities and transfer payments, but also generates a behavioural change by individuals with regards to the number of hours they would be willing to work. Preference parameter estimates over hours of work and income are generated for individuals in the sample, with imputed wages estimated for those who are out of work.

A tax-transfer microsimulation model, that utilises wage and preference parameter estimates, is then used to construct counterfactual scenarios where the tax-transfer system of a given year is applied to the population of other years. For example, the tax-transfer system of 1988-1991 is applied to the population in 2010-2013 in order to create a scenario representing what the disposable income distribution in 2010-2013 would look like with the 1988-1991 tax-transfer system. Estimates from this process suggest that nearly 40% of the increase in disposable income inequality between the 1988/91 and 2010/13 periods was due to the change in payments and labour supply behaviour associated with tax-transfer policy adjustments.

Other potential drivers of income inequality change were investigated by reweighting the HES data of one period to more closely represent the population of another period. Although shifts in the share of individuals in part time work also generated an increase in income inequality, the lift in higher educational attainment over this period is estimated to have reduced income inequality more sharply (by nearly 22%). The shift in the age distribution towards prime-aged work was not associated with any change in the aggregate income inequality measure.