Peter Kn\u00ed\u017eat, Statistical Office of the Slovak Republic, Faculty of Management, Comenius University Bratislava, Slovak Republic<\/p>\n
\nType of article: scientific article
\nPages: 27 \u2013 44<\/p>\n
The digitalisation of retail trade is changing consumer purchasing behaviour while creating new opportunities for the modernisation of official statistics. While traditional price collection is based on selected representative products, new data sources enable statistical institutions to work with large databases of prices and transactions obtained directly from online shops. One of the most promising data sources is transaction data from online retailers, which, in addition to prices also contain information on the quantities of individual products sold and their sales value. Such data provide a more accurate representation of actual consumer purchasing behaviour and create the potential for compiling more economically accurate consumer price indices. The objective of this paper is to present the potential use of transaction data from online retailers for the compilation of consumer price indices at the Statistical Office of the Slovak Republic. The paper describes the automated data collection process, presents the methodological framework for bilateral and multilateral consumer price indices, and analyses the differences between weighted and unweighted approaches. The empirical analysis is based on experimental transaction data obtained from the Heureka online platform for selected product categories on the Slovak market. The results indicate that transaction data enable the construction of economically more representative weighted price indices while multilateral methods substantially reduce the volatility of price index series. The findings also confirm the importance of filtering extreme price changes before index compilation. The paper contributes to the ongoing research into the implementation of alternative data sources in official price statistics and demonstrates the potential of transaction data from online retailers for future consumer price index production.<\/p>\n