When France and its European partners find themselves confined during the first epidemic wave, the almost complete shutdown of entire sections of the economy raises many questions.
How long will this situation last?
What impact on GDP?
On unemployment?
On debt and deficits?
Beyond the large economic aggregates, other questions appear quickly, in all fields and sectors regularly auscultated through quantitative studies, from demography to sport.
At the statistics directorate of the Organization for Economic Co-operation and Development (OECD), Jorrit Zwijnenburg recalls: responses to the methodological and conceptual challenges posed by the pandemic ”.
Instructions and hotlines reserved for statisticians
As of March 26, 2020, Eurostat will begin to publish additional guidelines in order to support national statistical services in their way of dealing with the new difficulties they may encounter.
The body responsible for collecting and harmonizing the statistical series of the States of the European Union and its closest partners is even setting up a
hotline
of excellence in charge of finding solutions to the thorniest problems of European statisticians.
Among the questions raised: how to present the trend of unemployment in OECD countries, when France has set up a partial unemployment system while the United States and Canada have chosen a model of temporary layoff?
INSEE then speaks of a "trompe-l'oeil drop" in unemployment when the New York Times headlines its front page on "the worst unemployment rate since the Great Depression".
Hitherto coherent, the group "OECD countries" temporarily becomes a completely unusable aggregate.
Another puzzle emerges then, with a name that can give chills to the uninitiated: the correction of seasonal variations.
In short: the techniques which make it possible to compare the variations between a month of February with 28 days and a month of October with 31 days.
This is explained by Stéphane Ducatez, Deputy Director General of the Pole Emploi Network, in charge of studies and performance: “Normally, to understand a variation over several months, we
seasonally adjust
the data, that is, say that we correct according to the number of equal working days and the month, in particular.
We will have to exclude the extreme points of 2020 to achieve these adjustments.
Nothing impossible, of course, but precautions to be taken.
No panic for the collections, but many precautions on the analyzes
Same impression among sports statisticians.
Within the analysis department of a professional football club, Pierre (the first name has been changed) had to deal with the sudden disappearance of his raw material: no more matches, no more data.
But “we have the chance to work mainly in minutes played rather than per month or per quarter,” he explains.
We just risk having to take precautions on how we will have to judge the effectiveness of the substitutes, because since the resumption of the championship, the clubs have the right to 5 changes of players and no longer to 3, it will be necessary to find a solution, but that's not the main part of our job.
"
But the coronavirus crisis did not only have disadvantages for Pierre: “During the cessation of the championship, we were able to take the time to develop and improve our statistical models, to better monitor the performance of our players and teams. other clubs.
"
Newsletter It pays me
The newsletter that improves your purchasing power
Subscribe to the newsletterAll newsletters
Another football professional, Hugo Bordigoni is the founding partner of SkillCorner, a start-up specializing in real-time data acquisition and processing.
For him the pausing of the championships and the continuation of these without an audience is even a godsend.
“Normally, club recruiters travel a lot to discover new players or observe players they have already spotted.
Since this has not been possible since the start of the pandemic, recruiters are relying much more on video and data.
"
Everyone makes their revolution
Even the most official institutions took advantage of this period.
At the OECD, “the growing public demand for data on a very local scale and available very quickly, to compensate for the absence or long delays of the usual publications, has enabled national statistical services to explore new sources of data and improve the processing of traditional sources, but also to use them to develop new indicators, ”explains Jorrit Zwijnenburg.
At Pole emploi, studies were even more numerous than normal: “the statistics published on the number of job seekers had ceased to appear every month to better describe the tendency of the labor market.
With DARES, we have chosen to resume this monthly publication during the crisis to enlighten decision-makers ”confirms Stéphane Ducatez.
Ditto for the survey on the workforce needs of employers, conducted by Pole emploi at the end of 2019 but published at the heart of the first epidemic wave, a very bad timing which aroused some criticism: "We therefore proceeded to an update results after the first confinement, explains Stéphane Ducatez.
Understanding the variations that have taken place allows us to better guide job seekers, and perhaps direct them towards more promising sectors.
"
Finally, 2020 has pushed statisticians to go a little further, in detail, in regularity, in granularity, showing a complexity that the usual aggregates sometimes synthesize too much.
Jean-Luc Tavernier, Director General of INSEE, notes in particular that “on complex socio-economic phenomena such as purchasing power or the situation of the labor market, INSEE has always recalled, and continues to do so, that they cannot be reduced to a single indicator ”.
2020 will have proved him right.
