France Uses New Measure to Calculate 35% of Economy Is Shut

(Bloomberg) -- As France’s statistics agency prepared its routine economic analysis in early March, Director Jean-Luc Tavernier decided it was time to attempt something more radical.

Trying to make sense of the impact of the coronavirus, restrictions on movement and school closures, he instructed colleagues at Insee to give up on the report they were due to publish and reinvent how they measure the economy.

The breakthrough for the economists in their new task was getting hold of credit card data that allowed them to track consumer behavior in a lockdown -- such as panic buying -- and how different businesses were faring.

“We said to ourselves: we are changing worlds and the exercise no longer makes any sense,” Tavernier, 58, said in a video interview from the confinement of his home. “From that point on, it was something completely new and in no way the same methods as following such a plunge is unprecedented in the history of a statistics institute.”

The results of the frantic effort to cross check previously untapped information with card transactions were alarming: 35% of French economy activity had disappeared.

Published on March 26, the efforts struck a chord with private sector economists, and the 35% estimate has been cited in multiple research reports. Bank of America used as a guide when it made dramatic downgrades to its euro-area outlook last week, and ING economist Julien Manceaux described it as an “extraordinary light in the thick economic fog in which we must have the honesty to admit we are all immersed.”

What Bloomberg Economists Say...

“French statisticians were the first to provide some insight into the economic impacts of strict containment measures in Europe – and the costs are eye-watering. We think at least 80 billion euros of income will be lost over March and April and the French government appears committed to offsetting a substantial part of that. It’s a big cost, but one worth paying to help the economy rebound swiftly in the second half of 2020.”

-Maeva Cousin, euro area economist. Click BECO for more.

Tavernier has also received calls from other European statistics agencies asking for details on how Insee worked and saying they may try similar methods.

The rush to reinvent the analytic toolbox reflects the cataclysmic impact of the virus on economies. With huge recessions forecast, the worry is that companies go bankrupt and unemployment surges. Establishing accurate measures is crucial for governments and central banks struggling to work out how to deploy hundreds of billions of dollars of stimulus.

Insee’s approach resembles the practice known as “nowcasting,” when various types of high frequency data, from GPS records to words in press articles are used to forecast the present in real time -- a kind of weather report for the economy.

Particularly useful in the confinement situation was timely credit card data, which Insee got from France’s interbank network Groupement Cartes Bancaire CB. The two had been in negotiations for a long time, but the crisis proved a catalyst to overcome legal and institutional hurdles.

The card data was used to cross check other novel data sources, including electricity use and surveys of business panels used by research institutes, notably France’s Rexecode.

“My hope is that we decide this data has proven a real service so let’s make it easier for public statisticians to use it,” said Tavernier, a longstanding civil servant who was head of cabinet at the budget ministry during the financial crisis.

When Insee publishes its next study on Thursday, it plans to give an even more detailed breakdown of its findings.

Tavernier said his greatest surprise was how quickly his team was able to put the study together and how coherent the different data streams were. Still, he acknowledges there may be blind spots. While measures of industry are relatively reliable and tested, business services are much more opaque, especially with many working from home.

“It’s a scientist’s illusion to think that collecting data gives you a perfect, real-time knowledge of the economy,” Tavernier said. “Things are much more complicated and we are not going to chuck out our old tools right away.”

“Today’s truth isn’t necessarily tomorrow’s,” he said.

©2020 Bloomberg L.P.