ATO ramps up data analytics for real-time focus on tax cheats

The Australian Taxation Office is ramping up its data mining and analytics capabilities for a real-time focus on tax cheats.

The Australian Taxation Office (ATO) is ramping up its real-time focus on tax cheats, enhancing its data mining and analytics capabilities in a bid to be "data smarter".

Late last month, the ATO announced it was stepping up its data mining program to target offshore tax evaders, identifying individuals with undisclosed offshore income assets.

Deputy Commissioner, Michael Cranston, said at the time that the new information gleaned from the ATO's data mining operations would be used to encourage people to disclose unpaid tax under its Project DO IT – the agency's offshore disclosure initiative, which to date has claimed AU$13 million in tax liabilities.

Additionally, under its cross-agency tax evasion task force Project Wickenby, which sees the ATO share information with at least eight other government agencies, it had raised around AU$1.9 billion in tax liabilities from over 4,000 audits under to the project — as of the end of February.

Project Wickenby combines intelligence from a range of agencies, including the Tax Office, AUSTRAC, the Australian Federal Police, and the Australian Crime Commission.

"These agencies are constantly improving their information technology systems and growing their overseas intelligence exchange networks to combat serious criminal activity," said the Minister for Justice, Michael Keenan.

The ATO derives much of its data from a number of state and federal government bodies, along with a host of private sector sources, including eBay, stock exchanges, share registries, businesses in the construction industry, employers, and banks.

According to an ATO spokesperson, the data-driven focus on offshore tax evaders, along with the ongoing Project Wickenby, are just two parts of a broader strategy by the agency to be "data smarter", using analytics to develop better-targeted services based on a deeper understanding of taxpayers' behaviours.

"We are also focusing on enhancing our data mining and analytics capabilities to enable the real-time identification of risk and appropriate responses," an ATO spokesperson told ZDNet.

"That means we can tailor our response to a range of simultaneous issues according to the risks they pose," said the spokesperson. "It ensures an efficient approach to our compliance work that gets the best results for the community."

"Analytics is used to identify outliers in the information, throwing up a red flag when an anomaly presents itself. More broadly, identifying outliers also protects the tax system as a whole. It shows when there's a sudden shift in behaviour, which indicates an area we need to focus on," the spokesperson said.

According to the spokesperson, the ATO does much of its data mining in-house, while also utilising expertise from others, and is increasingly working with new state-of-the-art analytics approaches in order to analyse ever-larger data sets.

The ATO's expanding data analysis capabilities come as its new myTax tax return lodgement portal on the Australian government's myGov website comes under fire from users reporting problems with the new platform in the first few days of its launch.

Last week, the launch of the new portal was accompanied by a number of would-be users reporting difficulty in linking their myGov login details with their ATO details.

On 3 July, an ATO spokesperson told ZDNet that "less than 1 percent" of users were affected, with the rest successfully lodging online using e-tax.