Using the Main Sources of Big Data to Generate Economic Statistics

Abstract

The article examines the experience of National Statistical Offices (NSOs) of the Asia-Pacific region in the use of big data for the production of economic statistics, as well as the prospects and challenges of integrating such information sources into the system of official statistics. The purpose of the study is to substantiate the feasibility of integrating big data into the system of economic statistics, and to identify ways to improve its quality, relevance, and analytical value in the context of modern global digitalization. The study demonstrates the application of big data in estimating key macroeconomic indicators, including consumer price indices, residential property price indices, tourism statistics, and others. It is proven that the most effective sources of big data are online price data, retail scanner data, mobile phone information, administrative tax databases, financial transactions, and other digital traces of economic activity. The advantages of using big data are highlighted, including increased timeliness of statistical information, the possibility of generating more detailed indicators, reduced costs of conducting surveys, and expanded analytical capabilities of official statistics. At the same time, the article identifies challenges associated with their implementation, such as access to information, ensuring the confidentiality of personal data, harmonizing new data sources with traditional methodologies, and the need to develop technical infrastructure. The study shows that the NSOs of the Asia-Pacific region are implementing numerous projects related to the use of big data in the field of economic statistics, but most of them remain at the experimental stage. An exception is price statistics, where several NSOs have already integrated scanner data and/or online data into the calculation of the consumer price index for certain product categories. In some cases, these non-traditional data sources can fully or partially replace traditional methods of data collection. However, big data are most often used in addition to traditional sources, providing more timely and detailed information as well as new analytical opportunities.