Evry has since last summer handled half of the data collection that takes the form of telephone interviews to Statistics Sweden's Labor Force Survey, LFS. But after finding deficiencies in the material, Statistics Sweden now chooses to end the collaboration.

Today, Aftonbladet also revealed that the prevailing working conditions for the people who collect data at the company are so bad that it stimulates cheating. This in turn gives incorrect data.

The tasks in the LFS are an important basis for decision-makers and the incident has led to Statistics Sweden's Director General Joakim Stymne being called to the Government for talks on the situation.

Statistics Sweden has been using Evry since 2015

- In 2015, Evry took over 20 percent of LFS and since last summer it has managed 50 percent, says Johannes Cleris, press spokesman at Statistics Sweden, to SVT News.

However, he points out that in the beginning Evry showed good results and that it was necessary to buy the service as it became more and more difficult to get people to participate in telephone surveys. Now, Statistics Sweden simply ignores the data collected by Evry and will release new revised figures using only the material that Statistics Sweden itself collected.

"The result is that statistics will not be as well distributed, but the big numbers still work," says Johannes Cleris.

Statistics Sweden warned about the shortcomings last month

It was last month that Statistics Sweden assumed that they had identified deficiencies in the data collection for their LFS. The authority had seen a lack of quality for some time and was able to trace the errors back to summer 2018.

- The sudden rise in unemployment we have seen in recent months is probably misleading. The increase started as early as last spring. The current level of unemployment is also estimated to be overestimated, said John Kling, head of unit for LFS, in a press release.

The bad working conditions at Evry that Aftonbladet reports on are low wages and strict supervision of the work. Several sources also tell the newspaper that they have regularly been involved in events that indicate that material from false telephone interviews has been submitted to Statistics Sweden.

Statistics Sweden's revised figures will be released on November 14.