On the Statistical Reflection of Long-Term Growth in Agricultural Production

Abstract

The article presents the results of a study aimed at proving the existence of systematic error in traditional calculations of long-term growth rates of agricultural production based on chain indices of agricultural production. According to the authors, the article also introduces a more accurate assessment of its dynamics with the account to the structure of the relationship between prices and the volume of agricultural production. The paper describes a theoretical model that is a methodological basis for the study and explains the discrepancy in assessing the dynamics of agricultural production using chain indices and indexes at constant prices. It allows establishing differences in the ratios of the Laspeyres, Paasche, and Lowe indices for crop and livestock production, due to factors in the formation of demand and the complex structure of the relationship between the price level and the volume of production. The adequacy of the constructed theoretical model is proved based on aggregated data that eliminated the influence of incompleteness of the initial information. As a result, it was established that livestock production is characterized by time-distributed changes in prices and quantity of products, which makes it possible to assess its dynamics using both chain indices and symmetric indices. It is proved that the dynamics of crop production cannot be adequately described using chain indices, since a positive correlation of prices of the previous period and production volumes of the current period causes an overstatement of the index in comparable prices of the previous year. Based on calculations within the proposed aggregated model, it is proved that the use of constant prices as the Lowe index weights, updated every five years, is an acceptable approximation of the Fisher symmetric index. Application of the indicated methodology for calculating the index to the data on Russian agricultural production by main types of products in 1990-2018 allowed to establish an overstatement of dynamics by 11.9%. The main difference falls on crop production (+ 19.6%), while for livestock - the differences are insignificant (-0.7%).