Liberal and trade-unionist concepts of flexicurity: Modelling in application to 16 European countries

Abstract

The notion of flexicurity was introduced in the late 1990s. It promotes the idea of compensation of deregulation of labour markets (= flexibilization) by advantages in employment and social security, in particularly for flexibly employed (other than permanent full-time, called also atypically employed). This paper suggests an operational definition of flexicurity, taking into account different views of liberals and trade unions. The corresponding flexicurity indices are derived from (a) data on the dynamics of employment types, (b) scores of the strictness of employment protection legislation provided by the OECD, and (c) qualitative juridical data on social security. To convert the latter into numbers, eight employment types (permanent full-time, fixed- term part-time, etc.) in 16 European countries are ranked with respect to their eligibility to five social security benefits (unemployment insurance, public pensions, etc). To avoid known shortages of ranking, the Method of Total Ranks is proposed. The ranks replace continuous variables as index entries, and a dedicated model estimates the total index error which results from such an 'ordinal rounding' of the index input. The flexicurity indices are calculated for 16 European countries for the years 1994-2003. Contrary to theoretical opinions, the current deregulation of European labour markets is not compensated by improvements in the social security. If the flexicurity advantages/disadvantages are accounted proportionally to the size of affected groups then the factual trends are negative even from the viewpoint of liberals, to say nothing of trade unions. The reciprocity of the advantages/disadvantages turns out to be illusory, because gains are smaller than losses and winners are fewer than losers. Thereby the study warns against promoting flexicurity policies with no operational control and empirical feedback.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.