Major institutions, including the Bank of Canada, are increasingly looking to the private sector to provide them with debit-card purchase data, search engine results and other economic information – a trend that has been accelerated by the pandemic.
The data has the potential to help policy makers respond to challenges more quickly than they could by relying on official government statistics, which are often months or years behind because of data collection techniques that, in some cases, rely on businesses to fill out forms. The private-sector data can also drill down to precise locations, which can help officials get a better read on how small and medium-sized businesses are faring.
But privacy experts warn the process of disclosing this data has to be handled carefully, because customers may have no idea their information is being used this way.
“During the pandemic, there’s been this enormous increase in research using these real-time datasets,” said Etienne Lalé, an economics professor at the Université du Quebec à Montreal, who has met in recent weeks with teams at the Bank of Canada and the Canadian Chamber of Commerce that are working with private-sector data. “It is really hot.”
The private-sector information can come in many forms.
Google has long saved information about what users of its search engine are looking for. It uses the data to serve up relevant search results and to sell targeted advertisements. The company publishes some of the information through Google Trends, a free service that lets people type in phrases and see how frequently users have entered those specific words in Google’s search bar.
Economists have begun to experiment with trying to draw insights from this data, for example by looking at the rise and fall in people’s use of search terms such as “unemployment benefits” or “mortgage.”
Earlier this year, the Organisation for Economic Co-operation and Development (OECD) began publishing weekly estimates of gross domestic product for 46 countries, including Canada. Its model, based on Google Trends data, has generally been in line with official GDP statistics, but with slightly stronger swings up and down.
The model projects that, for the week of Nov. 21 to 27, Canada’s GDP was running 1.6 per cent below prepandemic levels. A special-projects team at Statistics Canada is evaluating the use of a similar model based on Google data.
Another treasure trove of customer data is the GPS devices in cellphones. Telus launched an initiative called Data for Good in the early days of the pandemic. It shares anonymized location data with government and public-health agencies, so they can use information on how people are moving and where crowds are gathering to inform decisions on COVID-19 restrictions.
In October, the Toronto Region Board of Trade launched its Recovery Tracker project, which aims to look at economic activity in the Greater Toronto Area and parts of southern Ontario as the region emerges from the pandemic. The project combines data from Statistics Canada with real-time movement data from Telus to zero in on activity within certain areas. In October, for example, it estimated that only a quarter of workers in Toronto’s financial district had returned to the office.
Jan De Silva, president of the Board of Trade, said her team hopes to share its techniques with other parts of the country to help them track local economies.
“We’re really excited about how it can solve pain points in our largest economic zone, but also how it can create a proven tool for other orders of government to get behind,” Ms. De Silva said.
The Recovery Tracker also uses data from Moneris, a payment processor that tracks debit and credit card transactions that flow through its networks. The company boasts that its market penetration gives it a window into a third of all transactions in Canada.
“We’re in a position to capture all that information,” said Peter Goldsztajn, director of corporate data analytics at Moneris Solutions.
The timeliness of private-sector data makes for a stark contrast with official government statistics, which often take months to prepare.
For example, Statistics Canada began to publish estimates of business openings and closings in the summer of 2020 as a key indicator of economic health during the pandemic.
The estimates are derived from monthly payroll forms that employers submit to the Canada Revenue Agency. The forms are matched against the list of companies in the national business registry. If a business submits payroll data each month, Statistics Canada considers it to be active. If a business stops submitting, it is considered to have closed, according to Amélie Lafrance-Cooke, chief of the agency’s economic analysis division.
But the CRA collects those forms over five-week periods, and then Statistics Canada analysts take three and a half weeks to analyze the information. Ultimately, Statistics Canada releases its estimates on business openings and closings three months behind the reference periods. For example, in November, it released information about business activity in August.
Prof. Lalé has been working with two American economists to develop an alternative, faster measure of business closings. Their method uses anonymized data provided by Homebase, a software tool used by more than 100,000 American businesses – mostly small ones in the service sector – to schedule workers. If a business does not schedule workers, Prof. Lalé's group checks its activity on Google and Facebook to see if it has temporarily or permanently closed.
Prof. Lalé said his model indicated that small businesses in the United States did not close at a higher rate than larger businesses. His group has not yet conducted similar research on Canadian companies.
Using private-sector data in this way can raise concerns about protecting the privacy of individuals.
In an annual report tabled in Parliament last week, federal Privacy Commissioner Daniel Therrien said he was concerned about what he called a rise in the surveillance economy.
“Personal data has emerged as a dominant and valuable asset and no one has leveraged it better than the tech giants behind our web searches and social media accounts,” the privacy commissioner said in his report.
In 2019, Statistics Canada halted a pilot project that would have involved the harvesting of personal financial information from banks, after the commissioner warned it was an invasion of privacy.
In his report, Mr. Therrien said the government had a role to play in ensuring corporations were being responsible with their customers’ data. He urged the government to once again table a bill to regulate digital privacy. The Liberal government introduced such a bill last year, C-11, which died when Prime Minister Justin Trudeau triggered an election in August.
John Wunderlich, a privacy consultant who has worked with the Ontario privacy commissioner and other organizations, said the federal government must step in to ensure companies have a fiduciary responsibility to their customers’ data. This would make companies as responsible for sensitive information as banks are for monetary deposits. Otherwise, he said, corporations may default to thinking only of how to monetize the troves of data they are collecting.
“Structurally, do we want the use of our data for public good to be determined by private actors?” he asked.
Spokespeople for the Bank of Canada and the Canadian Chamber of Commerce confirmed the organizations are studying the use of real-time private-sector data for making economic projections. They declined to specify which data sources their organizations are using, but said findings would be made public in coming months.
Prof. Lalé said he does not see the appetite for private-sector data going away.
“There’s really a demand to track the activity of Main Street businesses,” he said. “It will be part of the political legacy of the pandemic.”
Our Morning Update and Evening Update newsletters are written by Globe editors, giving you a concise summary of the day’s most important headlines. Sign up today.