Predrag Lazetic

A cross-national analysis of 17 countries

Predrag Lazetic, University of Bath

This paper investigates the diversity in job quality of university graduates in 17 European countries using multilevel regression modelling, based on combined REFLEX and HEGESCO graduate survey data. The focus of the research is on aspects of graduate jobs that affect quality, especially the analytically neglected aspects of skill utilisation and work autonomy, as well as income, job security and work life balance.

The paper analyses variance in graduate job quality across 258 sectors of economic activity in the 17 countries studied, and identifies a number of factors that are correlated with overall job quality and its dimensions.

The main research focus, however, is on contextual factors in the wider society and economy that help explain both diversity in job quality and differences between different sectors of the economy and different occupational groups. In particular: 1) the adoption of new computer technologies; 2) exposure to globalisation, and 3) high educational attainment in the labour force.

The study tests two broadly contrasting theoretical approaches to differences in graduate job quality: skill-biased technological change theory (Acemoglu, 2002) and the new institutionalism (Baker, 2014) on the one hand, and the conflict theory of global knowledge capitalism (Brown et al., 2012) on the other, and in empirical terms finds more support for the latter of two theoretical accounts.

The findings point to one fact, neglected in mostly cross-country comparison-focused comparative research on job quality, which is that differences between countries in terms of job quality are not significant on any dimension apart from income, at least not in the case of the graduate worker segment of the labour force. With income level, the country of employment matters much more than for instance the sector of employment in which graduates work or the type of job they do. The differences in the macro level are however significant, and in some cases substantial, at the level of sectors of the economy, and these should be considered more in comparative studies. Only by taking into account sector features can one understand why, for instance, the primary and secondary education sectors across Europe provide the best graduate jobs, and the sectors of wholesale and retail or transport and communications the worst graduate jobs. A multilevel perspective on graduate job quality also reveals the important point that sectoral comparisons have limits, as most variance in graduate job quality is located at individual level (e.g. there is greater difference between individual graduate jobs within one sector of the economy than between graduate jobs in different sectors of the economy in Europe).

Here the key findings with regard to variables related to the main theoretical framework and the main research question asks:

Use of computer technologies at work: There are differences here at both the individual and sectoral level. Above average use of computers and the internet compared to other graduates in the same sector was not associated with any premium or penalty in terms of job quality in any of its dimensions apart from job security, where it brings some rewards, and is indicative of a more permanent position. On the other hand, graduates working in sectors characterised by high utilisation of computers and internet technologies are on average slightly better paid than graduates in other economic sectors (β = 1.265). But higher than average use of computers and the internet in a sector of the economy, contrary to the theoretical assumptions of skill-bias technology theory, is associated with lower than average skill utilisation (β = -2.082), which potentially indicates a deskilling effect of high computer use in a sector. Such sectors apparently do not provide jobs in which graduates can use the skills acquired through higher education. This study shows that graduates in Europe primarily associate skill utilisation with the extent to which their work calls on their disciplinary knowledge and analytical skills. Softer skills and abilities come only in second place.

Globalisation exposure: Similar to technological growth and level of technological development, globalisation exposure in the sector seems to have not so straightforward negative or positive effects on graduate job quality. High globalisation exposure in the sector of graduate employment is positively correlated with skill utilisation (β = 0.119) and job security (β =0.421), while it has a small but negative effect on earnings (β= -0.077). This seems to support the view of skill-bias theorists about the impact of globalisation on high skilled workers (Goldin and Katz, 2007). While globalisation, with a related increase in competition, outsourcing and offshoring might have a more negative effect on low-skilled workers lacking higher education, highly skilled graduates in Europe do not seem to be as affected by outsourcing and offshoring activities as their low-skilled co-workers.

Tertiary education attainment of the labour force: the findings indicate a significant, yet relatively small, positive correlation between skill utilisation (β = 0.052) and work autonomy (β = 0.052) and the proportion of workers with higher education degrees in a sector. Higher educational attainment in some sectors (usually education, professional services, ICT or finance) is associated with slightly higher skill utilisation and work autonomy for every graduate worker regardless of the occupational group he or she belongs to, however it has no associations with income. This partly supports the views of new institutionalism ( Baker, 2014) on the emergence of a schooled society. However, the past rate of tertiary attainment growth across sectors of the economy is negatively associated with skill utilisation (β = -0.090), which also potentially supports the arguments of new institutionalism. This negative effect is theorised as a short term effect until the occupational system adjusts to this new type of workers. Nevertheless, this finding also indicates that (hypothetically) in sectors where the proportion of highly educated workers grows fast, skill utilisation levels might fall, due to the filtering of new graduate workers down occupational hierarchies.