The mathematical-statistical methodology of the contemporary Soviet family budget survey

Abstract

The study describes and assesses the mathematical-statistical methodology of the contemporary Soviet Family Budget Survey, both in regard to the sample design and in regard to the processing and analysis of the survey data. A wide range of methodological deficiencies are identified, accounting for the widely recognised unreliability of the data produced. The problems of using the survey data in various fields of policy-making, planning and research are explored. It is shown that Soviet data-users where possible avoid relying on data from this survey. The historical and social factors influencing the methodology of the Family Budget Survey are discussed. The most important causes of the deficient methodology are found to be the neglect of mathematical statistics and sampling theory in Soviet socio-economic statistics, originating in the Stalin period, and the bureaucratic inertia of the Central Statistical Administration.