Canada Revenue Agency monitoring Facebook, Twitter posts of some Canadians

The Canada Revenue Agency is scrutinizing the Facebook pages and other social media posts of people it believes are at "high risk" of cheating on their taxes. It's also rapidly expanding its use of cutting-edge big data techniques for everything from improving customer service to deciding who to audit.

Agency is increasingly turning to cutting-edge data analysis techniques to improve service and 'compliance'

The Canada Revenue Agency is monitoring the social media posts of Canadians it suspects are cheating on their taxes. (Paul Sakuma/Associated Press)

The Canada Revenue Agency is scrutinizing the Facebook pages, Twitter feeds and other social media posts of Canadians it suspects could be cheating on their taxes.

That's just one example of the agency's increasing focus on what it can learn by collecting and analyzing many kinds of data — both its own internally generated information and what it calls "publicly available information."

"The CRA does practice risk-based compliance, so for taxpayers identified as high risk, any relevant, publicly available information relating to the specific risk-based factors for the taxpayer may be consulted as part of our fact-gathering processes," said spokesperson David Walters.

Among those considered high risk are wealthy Canadians with offshore bank accounts, said Jean-François Ruel, director of CRA's Strategy and Integration Branch.

"If we go with high-risk, high-wealth individuals that do offshore [banking], then we would look at all information that is public for compliance action."

Tobi Cohen, spokesperson for the privacy commissioner, said CRA notified it of its plan to collect publicly available information from social media in connection with "tax fraud and non-compliance risk analysis, audits and investigations."

However, David Christopher, of the advocacy group Open Media, said his organization opposes government agencies monitoring what Canadians are saying on social media.

"When Canadians post something on Facebook, they believe that they are sharing that with their friends and with their family. They don't believe that they are sharing that with some government bureaucrat in Ottawa," he said.

"Unfortunately, Facebook's privacy settings are notoriously complex and many people might think that they are posting something to their friends and it ends up getting shared with the whole world."

The revelation that the Canada Revenue Agency is checking social media posts comes as the agency is also expanding its use of cutting-edge technology and data analysis to better catch tax cheats, to target people for audits and to improve its service for Canadians.

Big data

Business intelligence, also known as big data, is a rapidly growing area within CRA. In 2016 alone, the agency posted three separate privacy impact assessments centred on its plans to use business intelligence techniques in its operations.

Andrew Treusch, who retired as CRA commissioner in June, previously wrote about how technology is changing the way the tax agency operates. (CRA)
Former CRA commissioner Andrew Treusch says technology is changing the way the agency operates.

"Evolving technology is having a significant impact on our approach to compliance," he wrote in the agency's 2016-17 Report on Planning and Priorities.

"Data analysis and business intelligence are providing us with better insight into taxpayer behaviours, allowing us to spend less time and effort on lower-risk groups of taxpayers and focus our resources on dealing with deliberate non-compliance."

In the report, the agency says business intelligence, which includes "mining accessible data," is a key area.

"This is because of the far-reaching opportunities it presents to enhance strategies for compliance, services and debt collection. Business intelligence encompasses big data to improve non-compliance detection (predictive analytics) and behavioural economics (such as 'nudging') to improve compliance."

Future techniques

An internal CRA document, which CBC News first found on the website of the U.S Internal Revenue Service, outlines CRA's plans to use business intelligence techniques in the future.

The document, prepared in 2014, describes plans to move into such areas as predictive analytics, which can use data and algorithms to help officials decide whether someone who hasn't paid their taxes should get a gentle reminder, a phone call or an audit.

When Canadians post something on Facebook, they believe that they are sharing that with their friends … They don't believe that they are sharing that with some government bureaucrat in Ottawa.- David Christopher, OpenMedia spokesperson

A chart in the document indicates CRA also planned to get data to analyze from "web and social media."

Ruel said while the agency was looking into the prospect at the time, it is not currently planning to include social media analysis as part of the business intelligence side of its operations.

"A lot of steps and a lot of work in terms of privacy would have to be done first and it's not one of the priorities right now."

However, he said the business intelligence unit does analyze data from CRA's interactions with Canadians through its own Twitter and YouTube accounts.

"When they communicate with us, we analyze this information to make sure that we understand what is happening," Ruel said. "If there is concern around scams, for example, we will take that information and then create a communications strategy to make sure that people have the information that they need."

Identifying offshore tax evasion

Ruel said the business intelligence section has a dual focus: improving service for Canadians while at the same time identifying cases that should be audited.

On the audit side, it is working extensively with data it has been receiving from Canadian financial institutions since January 2015 whenever someone makes an electronic transfer worth $10,000 or more.

The result has been a lot of information about countries of concern for offshore tax evasion.

"If we look at the case and we see there are a lot of funds transfers toward a country where we know there could be some offshore activities, then we link that information with all the other information that we have and we prioritize that way," Ruel said.

Privacy Commissioner Daniel Therrien's office says social media posts may be publicly available — but government agencies should only collect information directly related to its activities. (Sean Kilpatrick/Canadian Press)

The agency has also begun text analytics, also known as text mining. For example, Ruel said the business intelligence section is analyzing auditors' notes to detect new tax schemes or techniques for avoiding taxes.

But as CRA and other federal departments increase their use of data, privacy advocates say they should also be asking tough questions.

In a speech prepared for a conference in December, Patricia Kosseim, senior general counsel at the Office of the Privacy Commissioner, warned that just because something is publicly available online, doesn't mean individuals have waived their rights to privacy.

"Government's objective of being more nimble and responsive to the needs and expectations of Canadians is playing out in a data-rich context nurtured by individuals who are increasingly exposing their personal lives on social media, discussion forums and online networks. Never before has it been so easy to data-mine these 'non-traditional' sources of information.

Government use of big data techniques is causing concern among privacy advocates, University of Victoria professor Colin Bennett says. (CBC )

"While it may be tempting to cull data from these sites for recruitment, performance management, law enforcement or security related purposes — not to mention out of sheer curiosity — these are not lawless zones."

Colin Bennett, a University of Victoria professor who specializes in privacy, said government use of big data techniques is causing concern among privacy advocates.

When Canadians provide information to the government, they provide it for a specific purpose, not for algorithms and predictive analytics, he said.

"You are getting beyond the initial reason why the data were collected in the first place, which is to administer our tax system, and it is becoming far more hypothetical, far more speculative and concerning."

Mobile users: View the document
(Text KB)
CBC is not responsible for 3rd party content


Elizabeth Thompson

Senior reporter

Award-winning reporter Elizabeth Thompson covers Parliament Hill. A veteran of the Montreal Gazette, Sun Media and iPolitics, she currently works with the CBC's Ottawa bureau, specializing in investigative reporting and data journalism. She can be reached at: elizabeth.thompson@cbc.ca.


To encourage thoughtful and respectful conversations, first and last names will appear with each submission to CBC/Radio-Canada's online communities (except in children and youth-oriented communities). Pseudonyms will no longer be permitted.

By submitting a comment, you accept that CBC has the right to reproduce and publish that comment in whole or in part, in any manner CBC chooses. Please note that CBC does not endorse the opinions expressed in comments. Comments on this story are moderated according to our Submission Guidelines. Comments are welcome while open. We reserve the right to close comments at any time.

Become a CBC Member

Join the conversation  Create account

Already have an account?