External labour market flexibility and social inequality

Abstract

In this paper we examine the impact of temporary work on two dimensions of social inequality: income and career mobility. Additionally, we are taking a comparative perspective on this subject by comparing Germany and the UK. To investigate the effects of temporary work we use data from the German Socio-Economic Panel and the British Household Panel Study on non-self-employed respondents. The results show that temporary work does influence the system of social inequality: we found wage penalties and an increased probability of severe negative effects on the working careers of temporarily employed persons in both countries (net of education, age, and a variety of other covariates). Thus we can conclude that temporary employment represents a substantial socio-economic risk for employees. Most importantly, this holds true for both the German and the British case, two quite distinct labour market regimes.

Keywords:

• Labour market flexibility
• Social inequality
• Germany
• UK
• Temporary employment

Acknowledgments

Financial support from the Deutsche Forschungsgemeinschaft (DFG) (Project ‘Die Auswirkung atypischer Beschäftigung auf das System sozialer Ungleichheit’) is gratefully acknowledged. We thank the UK Data Archive and the Deutsche Institut für Wirtschaftsforschung for providing The British Household Panel Survey and the German Socioeconomic Panel, respectively.

Notes

Atypical employment refers to employment relationships that differ from the so called ‘standard employment relationship’ (Mückenberger 1985; Kress 1998; Hoffmann & Walwei 2000), which empirically is mainly described by being full-time, permanent, and non-self employment.

Dr. Martin Groß is senior researcher at the Institute of Social Science at Humboldt University Berlin. He has carried out a number of empirical studies on educational systems, social inequality and labour markets. Recent publications include (with J. Giesecke) ‘Temporary Employment: Chance or Risk?’ European Sociological Review 19 (2): 161–77.Johannes Giesecke is junior researcher at the Institute of Social Science at Humboldt University Berlin. His current interests include research on social inequality, labour market flexibility, and quantitative methods.

Temporary employment refers to employment relations with fixed-term contracts or to seasonal/casual employment.

Here, the ‘ceteris paribus’-clause is very important: we strongly believe that the temporary character itself has an income effect. People in temporary employment may also earn less than people in permanent employment because they are less educated, younger and so on. But our hypothesis suggests that there will be an effect of the temporary character of the position beyond the effects of the characteristics of the people occupying the position.

The theory of ‘implicit contracts’ (Baily 1974; Gordon 1974; Azariadis 1975) deepens this line of argumentation. Implicit contracts between employers and employees govern the exchange of job security for a certain ‘insurance premium’. However, this insurance is not available to all workers – only those with valuable qualification, with considerable investments in (specific) human capital, are protected from being laid off (Diekmann 1982: 242ff.). Employees with low qualifications and a low productivity, who form the ‘periphery’ of the work force, find themselves in jobs that are less secure. These workers represent a buffer to save the core. When demand for goods and services is low, workers in the periphery are dismissed first, i.e. costs resulting from adjustments to production and demand have to be borne by the periphery in order to protect the human capital investments of the core.

In the approach of Doeringer and Piore the third segment, the ‘upper primary labour market’, consists of well-paid, but rather unstable jobs.

It is important to note that the mobility chances of people with the same characteristics differ, i.e. the constraints for mobility stem from the characteristics of the labor market and cannot solely be explained by individual traits.

These institutional differences are to a large extent reflected in what Soskice (1999) calls ‘(industry-)coordinated market economy’ (Germany) and ‘uncoordinated or liberal market economy’ (UK).

This is the main reason for calling Germany a credentialist society (cf., e.g. Groß 1998).

In Germany, the Dismissal Protection Act does not apply to companies with less than six employees (excluding apprentices). In these enterprises the only statutory dismissal protection is given by notice periods. In 1996 the corresponding number of employees was increased to eleven. This number was reset to six employees by the Schröder government at the end of 1998.

One example for the concept of just cause is that terminations have to be socially justified. Thus, even when terminations are based on economic reasons, a procedure of social selection sets in. For a detailed discussion of the concept of just cause see Birk (1993).

For example the qualification period of employment for a claim of unfair dismissal or redundancy payment was raised from 6 months to 1 year (1979) and finally to 2 years (1985) by the Thatcher government (Rogowski & Schömann 1996). In 1999 this number was reduced to one year by the New Labour government. For a complete list of the changes in legislation made by the conservative government see, e.g., Deakin & Reed (2000).

For the role of temporary work in the British public sector see, e.g., Conley (2002).

For a list of acknowledged factual reasons cf., e.g., Walwei (1990).

Importantly, the Employment Promotion Act (EPA) did not replace the old law. Thus, temporary employment could be based on the EPA or on factual reasons. In 2001 the two old laws were combined into one.

cf. Projektgruppe Sozio-oekonomisches Panel (SOEP) (1995)

cf. Taylor et al. (1999)

In the cross-sectional income analysis we additionally use the sample D, which consists of immigrants, who moved to West Germany after 1984 (first conducted in 1994/95), and sample E, which is a refreshment sample (first conducted in 1998). Furthermore, we include respondents from sample C, who moved to West Germany.

The number of these cases is very low. In 1998, only 0.5 per cent and 0.1 per cent of the respondents of the Sample A and B, respectively, lived in East Germany. Furthermore, the labour market chances of these respondents can be assumed to be more similar to West German than to East German patterns.

cf. Halpin (1997).

In respect to deriving individual working careers from the single waves, the BHPS is somewhat ‘superior’ to the GSOEP as in the BHPS information on all job shifts are reported, while in the GSOEP only information on the last job are given. The problems resulting from this shortage of information when using the GSOEP are discussed in Giesecke & Groß (2003).

For the German data, the so-called ‘Arbeitsbeschaffungsmaßnahmen’ (ABM) are not excluded from the analyses. They are rather marginal for the West German labour market. Rudolph (2000) shows that the overall rate of temporary employment is only slightly increased by ABM, while in East Germany ABM plays a significant role in explaining the overall rate of temporary employment.

Sometimes higher numbers of temporary employment are reported for Germany (e.g., OECD 2002). This is due to the inclusion of apprenticeships, which are almost always on temporary basis.

During the 1990s the gender gap in the UK became much narrower compared to the situation in the 1980s (see, for example, Robinson 1999).

The categories of the original Casmin classification are related to the new Casmin variable in the following way: categories 1a and 1b are combined into Casmin low, categories 1c, 2a, and 2c_voc into Casmin middle vocational, categories 2b and 2c_gen into Casmin middle general, and categories 3a and 3b are combined into Casmin high.

In case of the mobility analyses age refers to age at the beginning of a particular spell.

In case of the mobility analyses we use a variable that distinguishes marginal from part-/fulltime employment (GSOEP) and part-time from full-time employment (BHPS), respectively. This is because in both the GSEOP and the BHPS there are no information about the actual amount of hours worked in the last job(s), but there are the classifications mentioned above, which are made by the respondents themselves.

For this variable we used information about the periods of unemployment, which was collected retrospectively, next to information stemming from the single waves.

Of course, the full models can be obtained from the authors directly.

However, the coefficients of the interaction terms for vocational educational degrees are substantially (about 10 per cent) though not significant. The lack of significance could be caused by the fact that only few employees are to be found in these categories while at the same time a lot of variables are included in the model. At any rate, the model clearly suggests that the income loss in temporary employment rises with education.

For a wider discussion of the differences between fixed-term contracts and seasonal work for the UK see Booth et al. (2002). Also, Gallie et al. (1998) provide hints that there may be two types of temporary workers (those with short contracts and those with medium term contracts) that should be distinguished in the analyses. Unfortunately, the distinction between seasonal/casual work and fixed-term contracts is not possible in the GSEOP. However, in contrast to the result in Britain we would expect a significant negative effect of fixed-term contracts in Germany.

To interpret this result correctly, it is important to keep in mind that in the mobility analysis, age refers to the age at the beginning of a particular spell, and not to age in terms of the position in the individual life course.

In another model (results are not presented) we analysed the probability of getting a temporary contract in dependence of the employment status of the previous spell (unemployed versus not unemployed). The results are similar to those of model I: unemployment increases the risk of receiving a temporary contract.

Since in the GSOEP it is technically difficult to distinguish between the single categories, we only distinguish between the categories ‘unemployed’ and ‘not unemployed’.