Combining Statistical and Rule-Based Expert Knowledge to Measure Employment Precarity

Abstract

- The measurement of precarity and the identification of a set of indicators that can be used for its assessment has been established as a key issue in Europe, central to the entire discipline of labour statistics, social policy, and sociology of work. Most recent studies agree upon the basic characteristics that a worker should have to be considered as precarious: insecurity, vulnerability, and no or limited entitlements. The present paper offers an innovative method that combines statistical analysis regarding the measurement of nine key indicators that are linked with precarity to a lesser or greater extend, with a rule-based expert system to rate each worker’s precarity. Raw data are drawn from the EU-Labour Force Survey (EU-LFS) for the case of Greece. However, the suggested method can be applied with minor modifications to the remainder thirty-four participating in the EU-LFS countries since a common questionnaire is used for all countries. The estimated indicators refer to three domains that are linked with precarity: labour market conditions and job insecurity, limited entitlements, and insufficient resources. Having estimated a precarious score for each worker, the socio-demographic characteristics of precarious workers are identified, extracting valuable knowledge on their profile.