Tax evasion has been a major pain point for countries across the world – with a staggering US$427 billion lost each year. This missing tax revenue has been especially harmful last year, crippling governments’ pandemic response, wrote a report by the Tax Justice Network.
Australia recently uncovered an elaborate fraud scheme that amounted to $17 million. “Not complying with tax obligations is not a victimless crime – the whole community is impacted by this behaviour,” said the Australian Taxation Office’s (ATO) Assistant Commissioner, Aislinn Walwyn.
GovInsider caught up with Marek Rucinski, ATO’s Deputy Commissioner of Smarter Data Program, to find out how the agency is tackling fraud with data.
Fighting scams and fraud
The ATO has “advanced analytics and matching techniques” to monitor transactions and identify suspicious activity, Rucinski says. The agency matches data from its systems with information from their partners to look for patterns or activities of interest.
Singapore’s Accountant-General’s Department has also turned to bots and data to help detect anomalies in agencies’ finance claims. Indonesia built a system on taxpayer behaviour to predict fraudulent behaviour.
ATO is part of the Australia’s Serious Financial Crime Taskforce, which consists of other agencies like the Federal Police and the Attorney-General’s Department. These agencies share intelligence with each other to identify risks and perpetrators, he adds.
Using AI for taxes
The agency uses automation and AI to guide taxpayers and streamline work practices to improve productivity, Rucinski says.
But ATO does not use automation and AI to make decisions requiring cognitive judgment. “Human judgement and empathy are vital for good tax administration” and cannot be automated, he emphasises.
Analytics is also transforming the way ATO assesses work-related expenses, Rucisnki says. It uses analytics to compare a taxpayer’s work-expense claim against other taxpayers with a similar income and job.
That helps the agency determine how much the claim differs from the norm, then alert taxpayers when unusually high claims are made. The taxpayer will then be prompted to double check their deduction.
Data to simplify tax paying
ATO receives, matches and pre-fills large volumes of data from third-party providers such as banks and employers. That allows the agency to share information about an individual’s tax affairs with them or their agent, before they lodge a tax return.
The agency is working with these providers to extend its data sources, Rucinski says. That will allow it to expand the range of information that can be pre-filled.
Training tax officials
“Tax authorities are fast becoming the data centres of government, with all the associated risks and responsibilities”, wrote EY’s Global Chair for Tax in the World Economic Forum’s blog.
“This increases the demand for data we hold,” Rucinski says. That higher demand was very evident through 2020 and the Covid-19 response, as ATO had to support various stimulus measures.
ATO will focus on implementing self-service data tools. That will ensure the agency can continue to meet the demand for data in the future, he adds.
It aims to develop data tools that require little to no technical training, he says. It’s currently developing a data catalogue, for instance, that will help staff understand what the data means, and where it comes from.
The organisation will also build up its staff’s data literacy skills, notes Rucinski. It will equip staff to access, interpret and analyse data via e-learning and workshops.
Tax evasion has been a major problem plaguing governments, but data may just bring the criminals one step closer to justice.
Correction: A previous version of this article referred to the Australia’s Accountant-General’s Department. It has been updated to reflect the correct name.