Use of Panel Surveys to Measure Employment Precarity in a Cross-National Framework: An Integrated Approach to Harmonize Research Concepts and Longitudinal Data

• Published in: Survey research methods Vol. 17; no. 3
• Main Authors: Katarzyna Kopycka, Anna Kiersztyn, Zbigniew Sawiński, Stefan Bieńkowski, Viktoriia Sovpenchuk
• Format: Journal Article
• Language: English
• Published: European Survey Research Association 13.10.2023
• Subjects: cross-national comparability; employment careers; panel surveys; precarious work; precarity index; sequence-based measures
• ISSN: 1864-3361
• DOI: 10.18148/srm/2023.v17i3.7989

In this article, we introduce a methodology to measure employment precarity in cross-country research based on individual career data from national panel surveys. First, we propose a measure of employment precarity, which is comparable across countries differing in their institutions, legal regulations and practices concerning the organization of labor relations. To address the comparability issues raised by using fixed-term employment indicators in the study of employment disadvantage, we conceptualize employment precarity in terms of sequences of labor market experiences, which carry a universal meaning as indicators of labor market attachment: non-employment, low income from work, and job separations. Drawing on recent developments in the field of sequence-based indices, we develop a Cross-National Precarity Index (CNPI) and test its performance using data on employment biographies from the German Socio-Economic Panel. We confirm good construct validity of the proposed measure by comparing its distribution across subpopulations as well as by assessing the statistical association between the index and typical correlates of precarious employment identified in the literature: employment status and life satisfaction. Second, we outline a methodological framework for the ex-post harmonization of career data gathered in various types of individual panel studies. Acknowledging the methodological differences between these studies, we propose adopting a common, calendar-based format for longitudinal data organization and constructing harmonization control variables to account for the recall and measurement bias resulting from the specific methodological solutions adopted in domestic surveys.