How experts evaluate plans to reform statistics

Victoria Pavlyushina, head of the socio-economic research group at the Agency for Transformation and Economic Development (ATRE), notes that now statistics often lag behind economic development: for example, the classifier of products by type of activity does not yet include “new” types of goods, in particular digital ones. However, she notes, some successes in terms of statistical accounting can already be noted – for example, the collection of data from many different departmental databases to assess the values ​​of indicators of national goals. “Statistics reform resource-intensive in every sense,” says the expert. In addition, the process will require significant analytical and methodological work, additional adjustment of departmental systems, including increasing the technical capabilities of data storage and processing.

As the ex-head of Rosstat, director of statistical research at the National Research University Higher School of Economics Alexander Surinov notes, federal bodies collect a fairly large amount of information, and “we need to use this wealth rationally, not waste money on collecting information that has already been collected, and not torment respondents with the same things.” same issues” – for this, national statistics must have access to such data. “For example, in the Scandinavian countries, little information is collected from respondents – the bulk of information for statistics comes from administrative registers. There are others where the situation is the opposite. Russia is somewhere in the middle; there is already good interaction with many departments,” he says.

The problem itself, adds Alexander Surinov, is quite difficult to solve for various reasons – in particular, there is a methodological problem associated with the fact that departments collect data to perform their functions and the information often requires processing in accordance with the methodology (for example, Rosstat and the tax service have some difference in understanding the pay data). The use of big data, which is generated spontaneously by private businesses, without the use of official classifiers, also requires solving methodological problems – statistics must “digest” them for qualified use, determining which data is suitable for statistical purposes and processing them in accordance with the methodology. “If big data can be harnessed and made a serious source of primary information, that will be great – it solves the problem of speed and can represent socio-economic phenomena in detail,” he says.

Antonina Levashenko, head of the Russian Center for Competence and Analysis of OECD Standards at RANEPA, explains that the modernization of statistics “is a process that requires consistent reform of methodological approaches and significant resources to increase the coverage of surveys and high-quality collection of statistics.” In particular, she believes, it is important to remove barriers to data exchange both between various government bodies and between the state and business – for example, today, to measure data on the income of the population, obtaining data from Sberbank on the income and payments of the population may be crucial, but for This requires more systematic work at the federal level.

Evgenia Kryuchkova