16m a year

The Government is about to use complex computer software to beat sophisticated welfare fraud, saving an additional €16m a year alone on bogus dole claims. The Department of Social Protection reviews over 1m welfare claims a year.

It expects to save €510m this year in payments to bogus claimants.

However, a pilot predictive analytics programme has already helped detect nearly 50% more non-compliant jobseeker payment recipients than all other risk-profiling methods.

The discovery of bogus claims or errors from 6.3% of cases probed as a result of the trial computerised system compares to a 4.5% rate currently being found among investigated clients.

Its use on a wider scale just to drive jobseeker investigations could deliver €50m more in welfare savings in three years, according to a 2013 consultants’ report for the department.

However, the new system also has the potential to save on the cost of detection and investigations, and to identify overpayments earlier.

It will use thousands of variables generated from existing information held by welfare officials, and other agencies, such as the Revenue Commissioners, to decide the bulk of cases which should be probed.

The system will initially pick cases for investigation among recipients of jobseekers’ allowance and benefit, the one-parent family payment and disability allowance. The four schemes account for almost €4.5bn in annual payments to more than 500,000 welfare recipients, out of nearly €20bn spent annually on supports.

Tánaiste and Social Protection Minister Joan Burton said a small amount of fraud can add up to a lot of money. “By tackling such fraud, we ensure that the welfare budget is there for those who need it most, and I make no apology for this,” she told the Irish Examiner.

“Data analytics offers another tool for combating welfare fraud and the department will be seeking to maximise its significant potential.”

It is planned to begin operating the system as soon as possible. The approach will be extended to other social welfare schemes once it is up and running.

The department said it is too early to say if its use will lead to reduced reliance on other methods of identifying claims for investigation.

“The challenges of detecting and preventing fraud are intensifying with the growing sophistication of fraudsters’ strategies,” said a spokesperson. “There is a need therefore, to use more advanced methods to combat fraud.”

Systems of detecting higher-risk cases for investigation include: reviews of income and earnings; checking data with other Government departments or agencies such as Revenue, registry offices, or Department of Education; reviews after tip-offs; reviews of payment components, such as qualified adult or child dependent entitlements; address or identity checks; medical reviews; mailshots; and reviews of assets in estate cases.