On June 30, the INE gave the advance data for the second quarter of 2021: GDP had grown by 2.8% in said quarter, a very interesting figure. We still had not recovered the pre-pandemic level but the path was clear. However today, when the INE gave the definitive data, places growth at 1.1%, a much worse figure.
Normally the advanced data of the INE usually coincide enough with the definitive ones and the corrections are minor, one tenth at the most. But this time the disaster has been huge and puts us in a much worse situation than expected. What happened?
The problem is that the advanced estimates provided by the INE are made with incomplete or even estimated data. According to the INE’s own note of the final data, there are three factors that explain the error:
- The first is that the data with which INE produces this leading indicator are only those for the first two months of the quarter (April and May).
- The second is that, to compensate for the lack of definitive data for June, advance data or estimates thereof.
- And the problem was with these early estimates of the June data, which were of poor quality. Specifically, the information on the sale of large companies and corporate SMEs, which revised its leading indicators also down.
The INE is following a good methodology, you just have to see the high quality of the advanced data in a normal situation. But Since the Covid-19 pandemic emerged, making estimates is becoming complicated. The situation changes every month and the trends are not firm.
The INE is not the only one that has a problem with estimates
The problem is not exclusive to the INE. Last year, AIREF had to stop its GDP predictor, a leading indicator of growth in real time, as the quarter progresses, since it gave very bad data because it takes a while to collect the abrupt collapse of the economy. (By the way, the leading indicator of GDP for the second quarter by AIREF was better than that of the INE, it indicated a 1.6% growth for the second quarter and was given a day before that of the INE, on July 29)
And although now there are no subsidences, we are on a roller coaster: when the incidence of Covid-19 rises there are confinements (forced and not forced, out of fear) and less economic activity, which picks up quickly when cases drop.
Therefore, it is advisable while the situation lasts not to take the advanced data reliably, as in its day we recommend not paying much attention to the forecasts. This is not the best time to forecast.