Statistics Finland could give up unreliable GDP data, says the director general

In an article published on Linkedin and Facebook, Sovala took a stand on what was published in HS to the news Kalevi Sorsa foundation from the research reportwhich investigated the methods and historical relevance of research institutes’ growth forecasts.

The Kalevi Sorsa Foundation is a social democratic think tank.

The forecasts had generally gone badly wrong and usually in the same direction.

As one the reason was the strong reliance of the forecasts on the GDP advance data of Statistics Finland.

The preliminary data had deviated greatly from the actual development of the economy, between 2011 and 2025 by an average of 0.7 percentage points.

Sovala reminded that in his previous job at the Ministry of Finance he was responsible for its forecasts in 2013–2016.

“And on that basis, I dare to say that among the group responsible for the forecasts, their uncertainty is understood very well. The users of the forecasts, including the media, seem to be a bigger problem,” he wrote.

It would probably be better to base life on less forecasts, Sovala admits, but reminds that decisions regarding budgets have to be made years before the final information is available.

“Some common calculation basis for decisions is needed – even a bad forecast is better than nothing.”

Statistics Finland reductions in statistical production will be decided in the next few months on the basis of the information and positions collected in the spring opinion round.

“In theory, giving up forecasting GDP data could be a very smart savings target, but EU regulation sets limits on this. We have to think about it,” Sovala wrote.

To Helsingin Sanomat, Sovala elaborates on the phone from his vacation that, as far as he remembers, the removal of preliminary GDP data during the statement round has not been in the actual removals.

“It is a theoretical option, but not in active consideration to do so.”

According to him, GDP typically fluctuates more in small countries and is more difficult to predict.

“In larger countries, the law of large numbers ensures that predictions are more stable.”

“When the growth has been close to zero, the margins of error are very large in relation to the growth. If the economy grew by five percent, a one percent forecast or advance information refinement would not catch the eye.”