Statcounter

31 May 2017

Could ID cards address rising benefit fraud?

According to the latest figures from the Department for Work and Pensions (DWP), benefit fraud is costing the UK £2.0bn a year. From scammers posing as deceased relatives through to ex-pats failing to mention they live in Alicante, it’s seemingly only getting worse. (h/t Dave)

The welfare system hasn’t exactly made life easy for itself. An entangled network of needs-based benefits, coupled with antiquated methods of ID verification, has made it easy to exploit. But due to a wave of new technologies, this tired old framework is finally set to change. Quite how far will depend on the decisions we make collectively.

Reducing fraud at the point of application

Until recently, it’s been notoriously difficult to check someone’s circumstances across different public sector departments. This lack of cooperation has led to millions being lost through fraudulent claims for Jobseeker’s Allowance - even by those working within the public sector itself. It’s also enabled thousands of people to receive a pension in the name of a deceased relative.

Part of the problem is being addressed with the introduction of Universal Credit (UC), which is a clear acknowledgement that the benefits system has become a labyrinth of complexity. Having one single application process for a range of awards removes the risks associated with conflicting information. The DWP is already saving money - in 2013/14, it recovered an extra £100 million of fraud and error-related debt compared to 2009/10.

As a further safeguard, the verification process for a UC claim now draws upon up-to-date earnings from HM Revenue and Customs (HMRC). The technology, understandably known as “Real Time Information” lets the DWP check an applicant’s current income. HMRC now insists that employers and pension providers provide these figures immediately after making a payment, rather than annually.

But the most notable new technology is the National Fraud Initiative’s (NFI) AppCheck tool. For the first time, any department can verify an individual’s circumstances with over 1,300 other organisations including the DWP, the Home Office, local councils, police authorities and private companies. In total, around a third of a billion records are trawled through. The chances of successfully claiming a benefit fraudulently are diminishing for each new data source that’s added.

Beyond this, the next step brings with it a whole new level of controversy. Last year, the DWP began a trial using blockchain, which involves claimants downloading a mobile app to receive and spend money. Every transaction is recorded on a ledger to help them improve their budgeting. Potentially, it could allow the DWP to specify where recipients spend their cash - although this has been denied by Conservative peer Lord Henley. Ministers are currently considering extending the trial, which was originally carried out by Disc (formally GovCoin).

Monitoring changes in circumstances

The National Benefit Fraud Hotline receives over 1’000 calls a day, but according to a freedom of information request to the DWP, 85% of reported cases turn out to be false. So it’s quite plausible that the cost of investigations could easily negate the overall savings.

But the Government’s reliance on anonymous tip-offs is set to be replaced by improved data matching. In addition to its Appcheck tool, the NFI also carries out frequent data comparisons across a myriad of sources. In 2016, it identified 172,907 fraudulent claims, resulting in savings of £198.2 million. Almost half of this was derived from simply comparing pension payments with the Disclosure of Death Registration Information from the General Registrar’s Office. Even though this doesn’t seem to be particularly groundbreaking, it’s actually highly significant given the public sector’s usual reticence for collaboration.

It’s only a matter of time until machine-learning algorithms can identify fraudulent activity with even more certainty. Exposing artificial intelligence to the NFI’s records will enable it to accurately spot discrepancies and even predict where fraud could occur.

Identification checks

The current ‘approved forms of ID’ have been around for decades. Even a new claim for UC only requires a driving license, passport or EEA national identity card. But reams of forged documents are in circulation and every year around 1,000 fake passports are seized at border points. A quick search for “how to create a fake driving license” lists almost two million results.

Manchester-based VST Enterprises has created a technology, known as the VCode system, which makes it decidedly harder to claim benefits in a false name. It’s currently being trialled in India and the US where users can authenticate themselves using a variety of secure personal data. Louis-James Davis, CEO, explains:

“Currently in India and the USA, benefit claimants are provided with a magnetic striped card loaded with funds that they can redeem at grocery outlets etc. That system has caused plenty of difficulties that VCode addresses. It’s been found to be more secure and effective in ensuring benefits go to genuine claimants than the magnetic strip system it’s replacing.”

In the UK, the technology could potentially be integrated into ID cards “as a multifunctional system, securely containing biometric and demographic data”, such as “ID photographs, fingerprints and iris scans.”

Moral dilemmas

As a society we need to think carefully about our next step. Do ID cards encroach on our civil liberties? Should we dictate where benefits are spent? How far are we prepared to move towards one centralised record which is accessible by private companies?

Technology is evolving to such an extent that benefit fraud could be eradicated within the next ten years. Perhaps the only potential stumbling block is our inbuilt resistance to change.

Source

No comments: