The Government is turning to data matching to claw back some $2 billion in welfare overpayments, signalling a further expansion of its analytics capabilities.

Unveiling the Mid-Year Economic and Fiscal Outlook (MYEFO), Finance Minister Senator Mathias Cormann expected “nearly $2 billion in savings” to come “from two enhanced welfare payment integrity measures, using better income data matching to ensure welfare payments are made within the rules.”

The biggest of those two measures – expected to contribute some $1.3 billion in savings over three years – involves “recovering money from a greater number of people where discrepancies have been identified” between the income declared to Centrelink and PAYG records held by the tax office.

This is effectively an extension of a measure announced in the Budget in May. Where it focused on historical discrepancies, the revised measure expands the scope to 2014 and 2015 data.

The other $700 million in welfare savings is expected to come from a data matching exercise targeting non-PAYG income declared to the tax office.

Outside of the $2 billion, the Government also hopes to recover more outstanding debts from people that have been overpaid benefits.

The focus would be on “high value debts and those individuals identified as having the capacity to pay, for example from those who no longer receive government payments and are now employed” – again suggesting data matching between welfare and tax agencies.

Treasurer Scott Morrison told ABC’s 7.30 program that IT systems would allow the Government to pursue overpayments and incorrect claims recipients “more vigilantly”.

He also said that an officer from the Australian Federal Police had been seconded to work with the Department of Human Services for the project.

“What we're talking about here is people who owe debts, who have been overpaid with benefits or may have achieved benefits to which they weren't entitled; and ensuring that those debts are being repaid,” he said.

“And even at the levels we're talking about, we're talking only about a 20 percent recovery on those, and I think we can go even further still.”

Any failure of the debt recovery component would not result in a negative impact on the budget’s bottom line.

“While these debts would otherwise not have been recovered, write-offs do not impact the fiscal balance and so improvements to debt recovery are not reflected in the expenses associated with this measure,” the MYEFO states.

The welfare measures also provide the embattled Office of the Australian Information Commissioner a financial lifeline, amounting to about $4.7 million over four years for a role in the non-PAYG data matching exercise.

Not everyone is convinced that data matching will lead to such a large projected saving.

Welfare support groups including the Australian Council of Social Service and the Welfare Rights Centre said the savings targets were “unrealistic” and cautioned against “aggressive” recovery efforts, given the complexity and scale of the welfare systems made errors possible.

The Government is looking further afield than welfare to make savings using data matching and analytics.

It wants to match data on people with outstanding Higher Education Loan Programme (HELP) debts “with international agencies to identify debtors working overseas”.

This is part of an existing crackdown on a loophole that allowed Australians working overseas to avoid repaying higher education debts.

The tax office is also in line to receive almost $62 million over four years to ramp up its “data analytics capability … to improve taxpayer compliance and reduce compliance burdens by pre-populating additional information in their returns”.

In past years, the ATO has used data matching systems and techniques to identify Australians with lost or unclaimed superannuation and other monies.