The best remains – Newspaper Kommersant No. 51 (7496) dated 03/27/2023

In 2022–2023, the peculiarities of Rosstat’s method of calculating inflation led to some underestimation of its estimates — the discrepancy between the dynamics of average consumer prices and the consumer price index (CPI) has increased in the last two years, the group of economists who studied the issue believes. In February 2023, without this effect, the CPI would have grown by 12.5% ​​year-on-year against the official 11%. The source of the effect is a sharp change in the range of goods. However, it was at this time that the situation was unique. If during a standard crisis, more expensive and high-quality goods were the first to be washed out of the assortment, then in 2022-2023 the situation was reversed.

The economists of the Solid Numbers telegram channel Rodion Latypov, Elena Akhmetova and Yegor Postolit published a preprint of the work “Price puzzle a la russe”, showing, apparently, the first real effect in recent years in the calculation of inflation by Rosstat, leading to its systematic underestimation. We are talking about “an increased discrepancy between the dynamics of average consumer prices and price indices for similar goods and services.” Theoretically, such a discrepancy should be near zero, but the constant change in the assortment (primarily its improvement) creates a difference between the change in average prices and indices: Latypov, Akhmetova and Postolit pay attention to the fact that in 2022–2023 it became very noticeable.

The attention of researchers was attracted by the “almost monotonous” decrease in prices for electronics since the second quarter of 2022 in the CPI components of Rosstat, with an increase in average prices for it against the backdrop of a weakening ruble. A review of the components of the CPI for which this discrepancy was largest in 2022 showed that it relates to many goods and services (maximum – upholstery, the difference between the price and the index is 39.3%). The Rosstat methodology assumes the replacement of the observed product in the CPI in January, associated with an improvement in its quality: the same wallpaper became the leader here (price growth in January 2023 by 129.2%). January jumps, due to the Rosstat methodology, are not reflected directly in the CPI (if the product previously estimated in it is no longer there, and the similar product replacing it is new and better, this does not change the price of the product in the base period), but the accumulated difference can lead to a price divergence in CPI components and in average prices. The researchers tested this assumption on one of the most popular products – smartphones: the change in the average price for them in January 2023 was 30% compared to the previous month, while the consumer price index for it decreased by 0.09%. The explanation looks like this: a smartphone, in the understanding of Rosstat, has ceased to be a device with a built-in memory of 32-128 GB and has become a device with a memory of 64-256 GB, 30% of the difference between them in Rosstat is higher than in the price lists (7-19% ), which may be due to the narrowing of the range – Rosstat counters had to fix prices for those smartphones that were on sale.

The paper provides estimates of the extent to which this effect (this is not a “mistake” by Rosstat; the index is considered correct, but consumers usually pay attention to absolute prices, and not to CPI) was large-scale in different years, based on the “historical” deviation of average prices from indexes. The average spread across the basket in 2014-2022 was 0.3%. In 2015-2016, at the peak of economic activity, it was negative – at the level of minus 0.5% (and overestimated inflation in the understanding of Rosstat). In 2022, on the contrary, the “expansion” was high (about 0.9%), in 2023 – up to 1.4%: according to the calculations of Latypov, Akhmetova and Postolit, in February 2023, without taking into account this effect, all inflation could be 12.5% ​​year-on-year vs. 11%. The authors suggest an obvious reason: due to a non-standard crisis, instead of expensive goods in 2022-2023, cheap ones were washed out of the assortment faster. The “caught” effect, which the authors do not claim, was most likely purely “sanctioned”: with a high probability, the underestimation of the CPI of Rosstat in this case is a reflection of logistics interruptions, since the turnover of expensive goods in retail is usually less than for cheap ones, and the replenishment of the assortment was slow or absent.

Dmitry Butrin