Closing government budget deficits by eliminating the insight deficit

By Carl Hammersburg, Manager, Government and Healthcare Fraud and Risk at SAS

The Canadian economy and budget are taking hits from all sides.  The falling price of oil and the lower dollar are taking a toll, with deficit estimates for next year ranging from $18 billion from the government to $30 billion from private economists.  With those challenges looming, the time has passed for continuing to do the same things over and over again. Expecting different results from that approach is the definition of insanity.

There is money being left on the table every day, and very cost effective-approaches to gaining access to that money.  The underlying issues cut across benefits programs, in the form of fraud, waste and abuse, as well as tax evasion and the underground economy.  Utilizing data at hand and layering analytics on top to detect and prevent these problems can save millions from a problem standing in the billions. 

It’s important to understand the size and scope of the problem.  In the U.S., the Internal Revenue Service (IRS), counterpart to the CRA, estimates the tax gap every 5 years, and last calculated it at $385 billion USD ($518 billion CDN).  Similar work is done in the U.K. and elsewhere, but Canadian authorities haven’t followed suit, leaving it to economists to estimate. If the problem is similar in percentage terms here in Canada, that could mean as much as $45 billion slipping through the cracks every year.

None of these numbers even begin to address the second threat to national taxing systems in the form of a dramatic increase of data breaches aimed at identity theft. Service improvements from the CRA and counterparts magnify the problem.  In years past, refunds could take months.  Now filing is online, and refund times have dropped dramatically, with the CRA estimating an average of eight business days this year.  This creates a system that hands out money first, then audits later.  That approach opens the door for identity thieves, many of whom quickly transfer the funds out of country and out of reach.

In the U.K., the HMRC, which collects taxes at the federal level, invested deeply in their ability to analyze billions of records quickly, finding fraudulent returns and optimizing their auditing and collecting activities, saving an estimated return of ₤7 billion ($13.2 billion CDN). 

Belgium, like all of Europe, was facing fraud in the form of falsified transactions and requests for refunds of VAT (GST).  They estimated the problem stood at €1.1 billion ($1.61 billion CDN) annually.  Tackling this type of tax fraud involves advanced analytics, the understanding of networks of many companies, and data flowing within and across borders.  Upon implementing a comprehensive solution, they erased 98% of the problem, saving more than $1.4 billion Euros every year.

On the benefits side, the Canadian Health Care Anti-Fraud Association estimates that two to 10% of all health care spend is abusive or fraudulent, an estimate that is low compared to international estimates, which typically range between 10 to 20%.  Similar gaps exist in cash, food and other aid and subsidy programs, unemployment and workers’ compensation.

Here too, there are examples of government programs that have taken significant steps to address abuse and organized fraud.  In Los Angeles County (population 10.1 million), a project aimed at breaking down silos of data and gaining an organized view of subsidized child care fraud showed amazing results.  The county’s ability to detect organized fraud rings jumped by 3200%, and a pilot program alone was valued at $6.8 million USD ($9.15 million CDN). Some estimates show it may be saving as much as $31 million USD ($41.7 million CDN) every year. 

Employment insurance, which is handled at the state level in the U.S., has also shown ripe for tax evasion, as well as fraud through identity theft, including manipulation through forming fake businesses to show employment and create eligibility for claims.  In North Carolina, analytics helped uncover more than $1.5 million USD ($2 million CDN) from the first 12 tax cases alone, out of more than 600 initial leads, a number which has since climbed into the thousands.  Detection of a ring of 107 fake businesses helped stop $5.2 million USD ($6.98 million CDN) in employment insurance benefits from being paid, while identifying an additional $2.7 million USD ($3.62 million CDN) in fraud.

So, are analytics being utilized today at the federal or provincial level tax and benefits agencies in Canada?  Of course.  But the majority of those operations combine rules developed using tools, not a comprehensive approach.  I followed the same path in my time in government.  Tools and development of rules and outlier analysis helped the problem, but fell far short of the huge increase we saw with a comprehensive, multi-year, multi-filer analysis, which developed a return of $30 for every dollar spent.  The same has proven true many other jurisdictions.

The huge budget gap is looming today.  The solution is proven, and has worked across multiple countries, with many different programs and taxing systems.  It’s time to stop identity thieves and fraudsters from hitting the government ATM, time to shut down the underground economy, and put those billions back to work.  No more insanity, just effective government.

Carl Hammersburg is the Manager or Government and Healthcare Fraud and Risk at SAS, the leader in analytics. In his role Carl leads the team assisting government organizations in the use of analytics, process change and data sharing to eliminate tax evasion, the underground economy and abuse of social service programs.  Prior to joining SAS, he spent 20 years in tax and healthcare audit, collection and anti-fraud activities in Washington State. Carl can be reached at

Back to Top