Evidence-based public policy is a strong and noble ideal.

Research that goes beneath the surface and contributes to rigorous, evidence-based public policy can be immensely challenging. Particularly if researchers don’t have the best datasets available to them.

This partly explains the clamour for governments to be doing more, and being more open in allowing access to the huge and valuable troves of administrative data that could be made available for research, in the public good.

It also explains why it is kind of a big deal when public sector agencies invest significant amounts of time, resources and energy into making these datasets widely available. That is precisely what the ATO has done, in developing the ATO Longitudinal Information Files (ALife) – and researchers are celebrating.

More than 14 million Australians filled out a personal income tax return for the 2017-2018 income year. ALife brings together the vast amount of information contained within those personal income tax forms from 1990-91 to 2017-18, alongside similar superannuation fund forms. The ATO then provides to researchers a rigorously selected 10 per cent sample on a consistent, confidentialised and anonymised basis.

It is the long history – now close to 30 years – and the higher degree of statistical precision that is possible across more than 300 personal income tax and superannuation variables, which makes the dataset a gold standard for researchers.

As the ATO’s Kirsten Fish has said,

“ALife represents a quantum leap forward for Australian-focussed tax and public policy research. It is quite a remarkable achievement for Australia”

“Data tells us how people are behaving. It’s policy in action... The insights that can be gleaned from researching this data can inform robust and sound policy and decision-making.” said Ms Fish in her opening remarks at the ALife conference in March.

“Early ALife research has already produced valuable insights. The projects using this data range from evaluation of specific policy, examining income distribution and dynamics, labour market analysis, intergenerational mobility, taxpayer bunching, and retirement income analysis.

One example is the real, important and practical contribution this research has made to informing the Government’s Retirement Income Review… which will influence the tone of retirement income policy debate for decades to come.”

ALife wouldn’t be possible without a team of visionary public servants within the ATO since 2013. The dataset’s development has been supported by ANU, particularly through the Tax and Transfer Policy Institute (TTPI) and a succession of Sir Roland Wilson (SRW) PhD scholars from the Department of Treasury.

First, 2013 SRW PhD Scholar Shane Johnson developed a ‘Tax Simulator’ that played an important role in detecting and fixing problems in early versions of the data.

Then 2015 SRW scholar Nathan Deutscher took over the reins by contributing to a future ‘ALife Family Links’ file that will enable intergenerational analysis. As Dr Deutscher has remarked,

“It’s a delight now to get emails from overseas saying ‘This is fantastic data, I want to work with it, you can be my co-author’. I am able to go back to them and say ‘Look the data is there, but you don’t need to take me along for the ride’. This is not some personal monopoly, this data is a significant contribution to the public good, supported by the Sir Roland Wilson scholarship.”

“This data is a significant contribution to the public good, supported by the Sir Roland Wilson scholarship.”

In March this year the TTPI, the Department of Treasury and the ATO convened the 2021 A-LIFE conference, and brought together a growing collection of enthusiastic early adopters of the dataset.

Current SRW PhD scholar and one of the conference co-organiser’s Tris Sainsbury remarked, “What emerged from the 2021 A-LIFE conference was how strong the demand is for a regular avenue to bring academia and the Australian Public Service together, with discussion centred around the presentation of rigorous academic research that uses these Australian administrative datasets.”

He is optimistic for the future. “The goal should be to really build on this emerging community of practice in coming years. As more researchers become familiar with the datasets, more research is advanced, more datasets are developed, and data becomes increasingly linked across government. Who knows, hopefully we will see strong and unique COVID-19 research based on these administrative datasets coming from academia in the near future.”

ANU Vice-Chancellor Professor Brian Schmidt observed that the experience with ALife is revealing something that is somewhat unique to ANU and the Sir Roland Wilson scholarship program.

“It’s the intergeneration of SRW scholars from the APS who have all contributed to building this powerful data asset, and the mobility of knowledge through the cohorts such that we can do multigenerational pieces of research with scholars over time.

"This is something unexpected and not part of the original design plans. That it emerges nonetheless offers exciting prospects for evidence-based work that is on a scale that wouldn’t be possible otherwise.”

More information on the ALife conference, including the opening and keynote address, is available on the TTPI website.

Read more about the Sir Roland Wilson Foundation. Keep updated by following us on Facebook and Twitter.