Making big data corruption's worst enemy
text size

Making big data corruption's worst enemy

Anti-graft activist Veera Somkwamkid files corruption complaints with the Office of the National Anti-Corruption Commission in this photo from October last year. (Photo by Tanaphon Ongarttrakul)
Anti-graft activist Veera Somkwamkid files corruption complaints with the Office of the National Anti-Corruption Commission in this photo from October last year. (Photo by Tanaphon Ongarttrakul)

In the last decade or so, the world has witnessed an unprecedented explosion in the quantity of information available as our ability to generate and store data rocketed through the stratosphere. This availability of a seemingly infinite amount of information, known as "big data", has fundamentally altered the traditional way of doing business in many industries.

Today, companies routinely employ sophisticated techniques to meticulously analyse big data to gain a competitive edge. Facebook's ability to customise your news feed and tailor advertisements to you, often with surgical precision, is surely not accidental.

Nevertheless, big data is not merely confined within the realm of the private sector. Governments across the world are also actively employing big data to aid their various endeavours. Big data analytics allow them to better understand the underlying themes and patterns of various national issues and tackle problems in a more targeted and efficient manner.

Combatting corruption is one avenue where big data can prove to be an indispensable asset. In a country where graft is rampant like Thailand, it can be difficult to determine where and how to begin ferreting out fraudulent activities. Big data can be that guiding light in the face of overwhelming ambiguity.

Natchapol Praditpetchara is a researcher at Thailand Development Research Institute (TDRI). Policy analyses from the TDRI appear in the Bangkok Post on alternate Wednesdays.

The US government's use of big data to crack down on Medicare fraud is a prime case in point that our anti-corruption bodies would do well to study from. Medicare is a single-payer health insurance system for the elderly and certain disabled citizens administered by the US Department of Health and Human Services (HHS) since 1966. It currently serves around 55 million Americans.

Albeit one of the nation's foremost healthcare schemes, Medicare has suffered from a chronic problem of fraudulent benefit claims by doctors, pharmacists and other medical providers alike who, for example, claim for procedures that never took place or for medications at an amount far greater than what was actually prescribed. HHS estimated that in 2014 alone, the government lost over US$60 billion (2.1 trillion baht) to Medicare fraud.

In 2007, HHS and the US Department of Justice (DoJ) jointly established a task force, namely the Medicare Fraud Strike Force, to tackling the problem. The task force also enlisted the help of federal, state and local law enforcement agencies nationwide.

From the start, the task force made big data a cornerstone of its investigative efforts. It relied on centralised real-time information of Medicare benefit claims throughout the country. This amounts to millions of claims per day and constitutes the heart and soul of the task force.

The task force meticulously analyses this data for suspicious patterns. If big data analytics show an area with several suspicious billings, the task force will delve deeper into that area. It will analyse to see what types of claims and from which medical providers that are particularly reoccurring and irregular. It will then assign investigators to check out on-site suspicious activities. These investigators will interview doctors, nurses, staff and patients in addition to analysing the providers' billing records, bank accounts and business ties among other relevant evidence.

The availability of real-time information nationwide helped facilitate and speed up investigations. In the past, if the DoJ wanted information on benefit claims, it would have to contact individual Medicare Administrative Centres (MACs), state-level offices that administer Medicare, which would often take weeks, if not months, to produce and provide the requested information. This significantly delayed and hindered investigations.

In addition, as claims are generally not reimbursed immediately, the DoJ has a window period to suspend reimbursements of suspicious claims and thus withholds payments if the claims are found fraudulent. With big data, investigators can be more targeted and proactive in their investigations, which is essential given the sheer size of the Medicare programme.

For nearly a decade, the task force has discovered and recovered billions of dollars in fraudulent claims. It has also indicted over two thousand individuals connected to Medicare fraud. It has expanded considerably and currently operates in nine cities across the US that are considered potential "hot spots" for fraud.

As for Thailand, the government certainly has a large amount of data on corruption cases. The government's 1111 hotline receives and handles over 100,000 complaints a year. Anti-corruption agencies, particularly from the Office of the Auditor-General, the Office of the National Anti-Corruption Commission and the Office of the Public Sector Anti-Corruption Commission, together receive over 10,000 complaints annually. The Administrative Court also handles around 5,000 to 7,000 cases a year.

Nevertheless, the government needs to streamline its various information sources in order to make the best use of the available data.

It is imperative that those government agencies compile all complaints that they receive and enter them into a centralised database system that is constantly and updated on a timely basis. Having access to the shared database, all of them can have a clearer, broader and more holistic picture of the complaints they are investigating. I myself observe there has not been sufficient data sharing among the anti-graft bodies.

Another means to create a centralised database is to simply have one common system for all government agencies to handle complaints on corruption. This would mean that all complaints will be funnelled into the same database directly.

South Korea, for instance, has implemented such a common complaint system for national and local state agencies. This has helped its anti-corruption body to conveniently analyse information to parse out common themes and patterns to make operations more targeted.

Moreover, it is also useful for the government to make such a centralised database system available for free to the public in a readable format. This will engage individuals in monitoring, investigating and reporting on corrupt activities. Everyone can take the role of anti-graft watchdog.

Many non-governmental anti-corruption organisations currently struggle to investigate cases as they lack readily available data at their disposal. Nevertheless, it is imperative that personal information is withheld to ensure that privacy is not compromised.

Do you like the content of this article?
COMMENT (5)