We evaluate the employment effects of Australia’s Mature Age Worker Tax Offset (MAWTO) using administrative data from the taxation office. Our study shows how large-scale administrative data can be used to precisely estimate the effect of government policy.
Australia’s ageing population poses a demographic challenge shared by many developed countries. The elderly dependency rate is projected by the Australian Government to increase from 23 per cent in 2015 to 37 per cent by 2055.
Responding to the increasing demand for services which older citizens typically rely upon brings fiscal challenges. Big ticket items, such as pensions, public health and aged care services, are expected to place additional stress on public sector budgets over a period when there is expected to be an increasingly smaller share of people working and paying income taxes.
The Mature Age Worker Tax Offset (MAWTO) was an Australian Government earned income tax credit which intended to encourage older workers to prolong their time in the labour market. It was introduced on 1 July 2004 and was available for 10 years before it was repealed on 1 July 2014.
Similar to earned income tax credits in other countries, the MAWTO reduced the tax liability of workers who met certain eligibility requirements. Official government estimates indicate that the MAWTO cost the Australian Government just over $4.3 billion in forgone tax revenue over the 10 years it was available.
A tax filer qualified for a non-refundable MAWTO credit of up to $500 if three eligibility tests were met. First, an individual had to be classified as an Australian resident for tax purposes. Second, the person had to be at least 55 years old by the end of the tax lodgement year (30 June in Australia). Third, only people with earnings from work (excluding passive income flows) between zero and $58,000 were eligible for the tax credit. This earnings threshold was increased to $63,000 from the MAWTO’s second year.
A modest boost in employment participation, but…
We test the effects of the MAWTO on participation and income using administrative data from the Australian Taxation Office (ATO) and find a small behavioural change. We test the effect of the MAWTO at its introduction in 2004 and its cessation in 2014. On average, across men and women, we find that the MAWTO increased participation by around 0.5 percentage points.
These average results mask some heterogeneity. The effect is larger for men at the MAWTO’s introduction, while we detect no effect for women. Conversely, the opposite is true when measuring the participation effect at the cessation of the program; we observe a larger and positive effect for women and no effect for men. We find little evidence that the MAWTO had an effect on earnings with the exception of women in the final year the MAWTO was available. Here we detect the MAWTO boosted earnings by around 1.5 percentage points.
The small behavioural response implies that almost all individuals who receive the MAWTO would have worked irrespective of receiving the credit. This highlights a typical downside of earned income tax credits in that it is not possible to provide the credit only to those individuals who were induced to work longer.
For the small share of people who responded to the MAWTO, back-of-the-envelope calculations imply that the average fiscal cost (in forgone tax revenue terms) for each individual was around $80,000 per person. This suggests that the MAWTO was a relatively expensive way to achieve a modest boost in employment for older workers.
What explains the size of the response?
The behavioural response of employment to the MAWTO should be considered in the context of other stage-of-life factors which contribute to the decision about whether or not to work for people approaching retirement. These factors include an individual’s health status, job satisfaction, partner’s retirement status or desire for more leisure time. Further, older workers are more likely to have accumulated wealth which may reduce the attractiveness of a modest tax credit.
The MAWTO was administered automatically through the tax system upon an individual filing a tax return. There are administrative efficiencies from leveraging off existing systems as the reporting burden on individuals is lessened given much of the information required for means testing is already captured in the tax return.
A downside, however, of using the tax system is that the benefits are typically transferred after the financial year in which the benefit was earned. This introduces a timing delay issue which may have dampened the intended effect of the program.
Further, some MAWTO recipients may have been unaware that they received the credit given the automatic calculation within the tax return system. Studies in the United States cite survey evidence that earned income tax credit recipients are often unaware of the program, or do not take the tax credit into account when making labour supply decisions.
In summary, it appears the MAWTO had a small positive effect on labour supply. This research adds to the body of literature on EITCs in a unique way as MAWTO was targeted at mature age workers only. How to retain mature age workers in the labour market is a question of policy interest around the world. Our results suggest that tax credits would be an expensive and relatively ineffective way to do this.
A working paper of the MAWTO analysis is available at the following link:
I think there may have been another incentive to work on scheme for those who reached 70 with a one off payment of 10,000 dollars. The study finding of people working on later in life because they want to is the main reason with many retiring early, then going back to work if not on a full time bases then part time. Well done on the paper.
Is there a tax incentive for me I will be turning 70 on the 30/06/019 still in full time work Regards Helen Reynolds