To Catch a Thief: Building a Better Case Against Sales Suppression
Oct 03, 2019
Last year we posted an article titled Zapping the Fun Out of Tax Fraud, where we explored how businesses are using sales suppression software to steal tax dollars collected during point of sale transactions. Since then, we have seen states increase efforts to crack down on the illegal use of these “zappers”.However, the outcomes of these efforts are varied, and states are struggling to build strong cases for prosecution.
Recently in Washington, charges were brought against a restaurant owner which included six counts of first-degree theft and three counts of use of sales suppression software, resulting in an estimated $5.6 million in unpaid taxes. However, the case was dropped, and the owner only paid an $800 penalty. So, where did the breakdown between charges and prosecution occur? Prosecutors were simply unable to substantiate the discrepancies which led to the supposed unpaid debt of $5.6 million. The owner contended that a former manager may have made a mistake when balancing the books. A similar case out of Washington was considered one of the first-of-its-kind when criminal charges were brought against another restaurant owner who was accused of withholding over $400,000 in sales tax. The case began as a routine audit by the Washington State DOR. Auditors noted an unusual change in cash receipts, as compared to historical data on file, and determined that the restaurant was using sales suppression software.Upon further investigation, auditors and the DOR uncovered the use of zapper software. In August 2016, Wong pleaded guilty to first-degree theft and unlawful use of sales suppression software. The court ordered Wong to pay $300,000 in restitution to the Washington DOR. In addition, Wong’s business entered a corporate guilty plea to first-degree theft, unlawful use of sales suppression software, and filing a false or fraudulent tax return.
These two cases present differing results but similar crimes, yet one was prosecuted, and one ended with a plea agreement. What can revenue agencies do to build better cases for prosecution? Obtain stronger provable evidence with the help of data analytics. Sales suppression detection is a two-step analytical process. The first step is audit selection, which involves the use of predictive models to select businesses that have indicators of sales suppression, based on many dimensions. One indicator is derived from the distribution of cash vs non-cash transactions over time. A machine learning algorithm can then identify statistically significant deviations from the distributions of similar businesses. Once a business is selected for audit, the second step involves Point of Sale (POS) data analytics, which then uses models to determine the likelihood of zapper technology being used in tax evasion.
There are many methods to determine if zapper technology is being used. For example, an algorithm can search for missing transaction ID numbers in the POS data, indicating the potential presence of zappers. You could also compare the distribution of cash vs non-cash transactions, for a given year, against other years of POS data. Statistically significant changes in this metric, year-over-year, could indicate potential non-compliance. Regardless of the methods used, this two-step process will lead to increased auditor efficiency and accuracy. However, using analytics requires more than one dimension for selection. Using several techniques, like cluster analysis, decision trees, and regression creates an “ensemble” approach, which we have seen demonstrable improvements for audit selection over business rules.
Identification of sales suppression should take a higher priority over whether a tax zapper is on a machine, in order to prevent the loss of much needed revenue. With data analytics, states can recover revenue more rapidly and boost long-term compliance. Sales suppression may seem like a victimless crime, but the lost revenue has real consequences.