Porównanie cen i wskaźników cen konsumpcyjnych: tradycyjna metoda uzyskiwania danych a źródła alternatywne
Comparison of prices and consumer price indices: traditional data collection and alter-native data sources

Summary/Abstract: One of the major challenges official statistics is faced with in the 21st century is the use of alternative sources of price data in order to modernise consumer price statistics and, as a result, to improve the accuracy and reliability of inflation data. Data collecting based on the traditional method encountered numerous difficulties caused by COVID-19 (distance-keeping restrictions limiting price collectors’ fieldwork, closures of points of sale). As a consequence, the work on alternative data sources intensified. The article presents the results of an experimental study involving the use of prices collected by means of the traditional method (by price collec-tors), and scanner and web scraped data from one of the retail chains operating in Poland. The aim of the study was to investigate the occurrence of differences in prices and price indices of selected food products and to estimate them, using the traditional method and alternative data sources, i.e. scanner and web scraped data. An additional goal was set to identify source-based reasons for these differences. The empirical study covered the period of February and March 2021. The results based on data from different sources were compared using both graphical methods (histograms, box plots) and the calculation of elementary price indices according to the Dutot, Carli and Jevons formulas. The findings revealed certain, sometimes serious discrepancies in the distributions of prices obtained from various data sources, which suggests that the application of scanner and web scraped data may lead to the over- and understating of price indices obtained via the traditional method. The article also discusses the main methodological aspects of obtaining and applying data from alternative sources, and indicates the probable causes of the differences observed both in distributions of product prices and in monthly price indices calculated using data from various sources.

• Details
• Contents