Empirical Essays in Taxation - Responses to the Australian Taxation System

Abstract

This thesis contributes to the international literature on targeted tax incentives and taxpayer bunching. In Chapter 2, I introduce a new Australian Taxation Office (ATO) dataset that supported the empirical analysis in this thesis. The ATO Longitudinal Information Files (ALife) is administered and maintained by the ATO to enable more public-interest research on Australia's taxation and superannuation systems. It provides almost three decades of longitudinally linked individual-level tax records. The ALife is enabling a wide variety of public-interest research projects, reducing Australia's reliance on studies conducted overseas and on conclusions that may not necessarily hold in the Australian context.

In Chapter 3, I use ALife to examine the labour supply effects of Australia's Mature Age Worker Tax Offset (MAWTO). The MAWTO was a targeted earned income tax credit that intended to incentivise the participation of workers aged 55 and over. The MAWTO was in force from 2004-05 to 2014-15. I show that the MAWTO had little effect on labour supply rates of older workers. The results imply that almost all of the MAWTO recipients would have worked anyway. It, therefore, seemed sensible to abolish the program given the estimated program cost of $4.3 billion over its 10 years of applicability.

In Chapter 4, I examine Australia's Transition to Retirement Income Streams (TRIS) program. The program aims to prolong labour force participation of older workers (age 55 to 65 years) by offering early access to a worker's compulsory retirement savings (superannuation). I show a small labour supply response which increases after the program's initial years. The size of the effects appear consistent with the program adoption profile which was low initially. I argue that my estimates should be considered a lower bound on the true response given technical difficulties in isolating the labour supply response as the program adoption matured in later years. I estimate that a least half of program participants appear to be employing strategies to minimise tax; a behavioural response which seems at odds with the program's intent. 'Tax-effective' strategies can be devised by drawing a TRIS while, in the same financial year, cycling salary-sacrificed (pre-income tax) contributions back into one's superannuation account.

In Chapter 5, I examine Australia's Division 293 tax using the `bunching' approach. Division 293 tax intends to make Australia's compulsory retirement income system (superannuation) fairer by doubling the concessional contributions tax to 30 percent for very high earners (those within the top 2 percent). Consistent with other studies, I find that taxpayers with business income are more able to bunch. Almost all of the bunching response is accounted for by the self-employed, including those who receive trust distributions. I otherwise find females are more than twice as likely to bunch than males, despite only accounting for around a quarter of tax filers in this region of the income distribution. The results suggest that the tax induces a substantial tax planning response but little labour supply response. Show more