Future of e-government: prevention is better than the cure

30 Jul 2013

Sean Shine, senior managing director of Accenture's Health & Public Service practice in Europe and Latin America

When people hear about analytics and big data they may think of it as something that belongs in the world of corporate strategy, where marketing boffins are studying every social media stream, database, weather pattern and news report, trying to figure out where to squeeze the next cent out of consumers.

Indeed, bookies firms like Paddy Power employ platoons of statisticians and analysts to figure out the perfect odds for upcoming races, boxing matches and soccer games. In Las Vegas, the house always wins because casinos invest heavily in intelligence systems to ensure every revenue-generating opportunity is optimised and every risk minimised from every individual that crosses the threshold.

It will, however, come as a surprise to most citizens of Ireland that some Irish Government departments are already making use of this kind of technology to analyse risks and behaviours of tax-paying citizens in real-time in order to minimise losses to the exchequer.

Last month, Richard Bruton, Ireland’s Minister for Jobs, Enterprise and Innovation, appointed Liam Casey, CEO of PCH International, John Herlihy, head of Google Ireland, Heather Reynolds, owner/director of Eishtec, and Bill McCluggage, the new Irish State chief information officer, to join senior department officials in a new big-data taskforce, which will include the piloting of two public service big-data projects.

The Real Time Risk Framework

One system already in operation is the Real Time Risk Framework. The Revenue Commissioners introduced the system in 2011 and it has been credited with saving the exchequer between €2.5m and €3m in erroneous tax refunds or credit requests.

The system, introduced to the PAYE sector, will eventually roll out to include VAT payments and taxes of self-employed small business owners. It uses predictive and behavioural analysis to prevent incorrect claims and detect under-declared income.

The Real Time Risk Framework is one of a number of new systems in use by the Revenue Commissioners. There’s also REAP (Risk Evaluation, Analysis and Profiling), which interrogates more than 50 separate data sources from the public and private sectors, and a Social Network Analysis tool that is being piloted to identify possible fraudulent activity.

Declan Rigney, assistant secretary general in the planning division of the Revenue Commissioners, said the purpose is to cut down on errors and wasting time.

“Regardless of what channel someone makes a claim – be it online or inputted by a staff member here – it is risk evaluated in real-time.”

Rigney said the analytics system is designed to create a smoother flow of claims processing and at the same time if errors are spotted, to ensure prompt action.

“Most of the time it throws up ‘yes or no’ scenarios that require a second look or further evaluation but there are behavioural elements to it, too,” Rigney said.

“In the past, there have been instances where there have been problems with a claim or refund and that has often come down to innocence or where people have over claimed or put in the wrong figure. We are able to, based on the entire population, predict from the characteristics of a claim if it is likely to be fraudulent.

Things pass through seamlessly and quickly when there is no problem so there is a benefit to the compliant tax payer, Rigney said. The system also helps staff move quickly and deploy resources to where the risk is the greatest.

Rigney said that in the past if mistakes slipped through due to human error, it meant money that went out the gate would be harder to recover.

“Now, before money goes out the door vis-à-vis an instruction to an employer or a direct refund, we’ll stop it.

“We save the hassle of trying to recover money after the fact and this has saved us over €2m in the past year alone.”

Rigney said the real-time risk analysis system has been integrated into the Revenue Commissioners’ public facing systems and back-office apps and is now part of the tax body’s day-to-day modus operandi.

“The question then is whether to roll it out further into other areas like VAT to make sure we are stopping money going out rather than trying to recover it after the fact.”

Analytics team behind Revenue Commissioners’ system

A team of 30 workers, including PhDs, mathematicians and statisticians, based at Accenture’s Analytics Innovation Centre in Dublin, built the system deployed at the Revenue Commissioners.

The centre, announced two years ago, will employ 100 people at full scale.

Sean Shine, senior managing director of Accenture’s Health & Public Service practice in Europe and Latin America, said while analytics can be used to achieve cost savings, it can also help government agencies get by with fewer staff and free up resources to focus on the core business and provide better services for citizens.

“What people don’t realise is that there is more usage of analytics capabilities in Ireland than in many other countries, particularly some of the Government departments who are using it today,” he said.

For example, the Department of Social Protection has been working on some projects in this area and has identified savings that it can make through advanced analytics, said Shine.

“A lot of people see analytics as examining something that happened in the past, but predictive analysis is aimed at finding out what might happen in the future based on all the bits of data you can pull together,” he said.

“The idea of finding out about issues or errors before they happen or predicatively analysing where things might occur is very, very powerful.”

Shine said he is working with medical insurers around the world to reduce fraud and error. For instance, there is a pilot project in Valencia, Spain, that is focusing on people with chronic illnesses – a situation where a small number of people take up a lot of medical resources.

In Spain, about 4pc of people are responsible for 60pc of the costs in the health system. The reality is that people who have illnesses, such as heart disease, have a higher probability of other illnesses, such as diabetes, Shine said.

“The idea is to use analytic technology to identify patterns and enable people to take corrective action before they fall into the trap of chronic illnesses. This involves certain procedures and tests at home to keep them from becoming chronically ill and therefore reduce the usage of scarce hospital capacity,” said Shine.

“We are doing more of this kind of work with the health system in the UK – a lot of private care insurance and hospital groups are working together on similar systems so they can reduce the cost of medical delivery by preventative maintenance to improve outcomes.”

Like the old adage goes, prevention is better than cure.

A version of this article appeared in the Sunday Times on 28 July

John Kennedy is a journalist who served as editor of Silicon Republic for 17 years