The heterogeneity of informal employment and segmentation in the Turkish labour market

Abstract

This paper aims at investigating the heterogeneity of informal employment on the Turkish labour market. To circumvent the constraints imposed by the traditional parametric methods, finite mixture models are estimated in order to identify the optimal number of segments within the informal employment and their respective returns to individual characteristics. In particular, it sheds light on the potential voluntary nature of informal employment by comparing the estimated probabilities of segment membership with the theoretical probabilities that would result from a competitive labour market under the hypothesis of income maximization by workers. Results show that the classical self-employed versus informal wage-workers divide is not the best split of informal employment. Furthermore, the two estimated informal employment segments are both less desirable than formal employment. Thus, the hypothesis of labour-market segmentation, even after taking informal-sector heterogeneity into account, seems to hold, supporting the traditional dualistic view of informal employment in the Turkish context.

Keywords:

• informal employment
• labour market segmentation
• Turkey
• finite mixture model

JEL classifications:

• J31
• O17
• C14
• N35

Notes

26 outliers have been deleted: those whose hourly wage exceeded 50 Turkish liras, the mean hourly wage being 4.14 Turkish liras (see Table 1).

Casual wage-earners do not have to be registered with the social security institution in Turkey, but the 17th International Conference of Labour Statisticians stated that jobs not subject to the social insurance system as a result of their occasional nature have to be considered as informal jobs.

Other usual instruments such as the number of children under 14 years old, the number of dependents over 14 years old in the household and the size of the household do not pass the tests.

This figure is very similar to that found by the household labour-force survey (65%), which covers all those in employment, including non-salaried family workers (see Ben Salem et al. 2011a and 2011b).

Assuming that individuals live and work in the same area.

The column with the formal sector wage equation results only appears once in this table: the results of the division of the informal sector into two known segments were very similar to those turned up by the mixture model.