Gold coins on island with palm trees
rafaelbenari/123RF

The Canada Revenue Agency (CRA) is stepping up its use of “big data” as part of its broader efforts to combat offshore tax evasion, Diane Lebouthillier, minister of national revenue, said in a statement published on Monday.

“The CRA continues to prioritize obtaining better data, improving its use of data to target its compliance actions, and achieving results in its fight against offshore tax evasion and aggressive tax avoidance,” she said.

Lebouthillier also said the tax agency “fully supports” a series of recommendations in a recent letter from the Offshore Compliance Advisory Committee (OCAC) on “gathering and analyzing large amounts of raw tax data and using it as a tool to combat offshore tax evasion.”

The committee recommends the CRA: considers extending its data analysis efforts beyond high net- worth individuals to other types of taxpayers; work with other countries to keep its data mining techniques up to date; and develop metrics for assessing the success of those techniques.

In response, the CRA says it is developing predictive analytics using machine learning “to identify potential areas of non-compliance by discovering unseen patterns in data.”

The agency also says it is developing new data models to identify high-risk taxpayers, and using social network analysis to automate the identification of links between individuals and businesses. Last year, the CRA began using data about households in wealthy neighbourhoods “in a more systematic way to conduct in-depth risk assessments and initiate audits,” the agency says.

Starting this year, the CRA will also start receiving international banking data through an information exchange program between countries that will provide access to “additional information to better identify tax cheats and deal severely with them,” Lebouthillier said.

Additionally, the CRA indicates that it intends to raise public awareness of these efforts to help bolster confidence in the tax system and encourage compliance.