Big data analytics could fill the annual £15bn VAT and excise duty gap, according to statistical software company SAS.
HM Revenue & Customs (HMRC) said the supplier estimates the 2012-13 unpaid tax toll to be £32bn – nearly 7% of the total money due in. More than £12bn comes from failure to pay VAT and £2.9bn is lost on excise duties. Criminal activity, tax evasion and the black market account for nearly 45% of the total tax gap.
A total of $6.9tn worth of goods are imported into the EU as a whole each year. SAS said customs officers have an increasingly difficult time detecting fraud across high-volume freight traffic, as smugglers continue to outfox officials.
Smugglers use creative methods to avoid taxation, said the software firm. One cunning plan is paying for a certificate of duty for one case of alcohol and then using that same certificate for every other case in a very large shipment.
SAS UK and Ireland director for central government Simon Dennis said customs officials need to look beyond individual shipments to search for extra information that might warn of suspicious activities. However, he said doing this manually is “time consuming and the chance for human-error is high."
Data analytics enable custom officials to rank import and export transactions and manage information to create risk scores in real time, said Dennis.
“Using high-performance, predictive analytics, governments can efficiently detect and prevent fraudulent shipments, and crack down on organised crime. Most importantly, they can significantly reduce the tax gap and funnel that much-needed money to programmes and services for British citizens”, he added.
According to SAS, other EU member states have used data analytics to reduce fraud. Belgium, for example, has cut losses in carousel fraud – a type of VAT fraud – by 98%, saving almost €1bn per year. As a result, the Belgian government has gone from €1.1bn in VAT fraud in 2002 to just €18.5m in 2012.
HMRC is, however, using Hadoop to reduce corporate tax avoidance. Chief digital officer at the department Mark Dearnley told delegates at an open-source data standards conference in London in April 2014 his team has built an enterprise data hub using the Hadoop framework to ingest UK corporation data.