Abstract
Is the middle class shrinking? This article contributes to the debate on job polarisation in Europe. Based on data from the European Union Labour Force Survey (EU-LFS) and looking at 12 European countries, it shows that there is an evident trend towards job polarisation. While this polarisation takes various forms, it is clearly the highest- and lowest-skilled jobs that have increased most rapidly among the active population over the past 20 years, to the detriment of middle-skilled jobs. The article then goes on to demonstrate that polarisation also exists when it comes to working conditions insofar as the lowest-skilled jobs are also where the most precarious employment conditions are found. Conversely, the remaining middle-skilled jobs are relatively shielded from this decline in working conditions.
Since the early 2000s, a substantial body of literature has focused on the issue of the polarisation of social structures in western countries. While the proportion of highest-skilled jobs has risen considerably over the last few decades, several empirical investigations have shown that this is not the only pattern in play and that the proportion of the lowest-skilled jobs also seems to have risen. This result was first shown in the USA (Autor et al., 2003, 2008; Wright and Dwyer, 2003) and the UK, before being extended to the rest of continental Europe (Goos et al., 2009). Insofar as these studies also highlight a decline in jobs in the middle of the skills structure, they conclude that employment structure in the western world is becoming increasingly polarised, essentially in connection with the effects of technological change (Goos et al., 2014).
However, this conclusion has not met with overall consensus. With the same data, Fernandez-Macias obtains very different results by using more detailed job classifications. He distinguishes between three patterns of change in employment structure in Europe (Fernandez-Macias, 2012). The main issue that has sparked controversy and led to different conclusions has been the changes affecting middle-skilled employment: to what extent has its proportion declined? Put differently, is the core of the middle class being eroded? This question of the shrinking middle class has been subject to much debate since the mid-2000s (Alderson et al., 2005; Atkinson and Brandolini, 2013; Pressman, 2007).
This article aims to contribute to this debate by analysing the evolution of job structure in 12 European countries over the last 20 years and asking two questions. First, to what extent are we witnessing a significant decline in the proportion of middle-skilled jobs and a rise in low-skilled jobs? Second, beyond quantitative changes in the proportion of workers in different categories of employment, what changes can we see in job quality among the fast-spreading jobs in the lower fractions of the social structure? In other words, are we also witnessing a general polarisation of working conditions? This question is important insofar as it sheds light on the trajectory of the lower fractions of the middle classes, destabilised by the erosion of middle-skilled employment, and raises the question as to whether they are doomed to being downgraded to precarious and badly paid jobs.
Analysing the polarisation of job structure by taking employment conditions into account breaks with prevailing methodological choices in this field. Irrespective of their results, the studies cited above analysing changes in employment structure all take an approach that measures job quality solely by wage.
Wage is an important indicator of the quality of a job and its ranking in the social structure. However, two criticisms can still be levied against these studies. First, they are based on a questionable methodological choice insofar as they take into account hourly wage, which fails to account for the expansion of the most unskilled and precarious employment. In such jobs, insufficient working hours can be a central feature of employees’ working conditions. Consequently, hourly wage fails to account for the reality of changes in social structure: due to the fragmentation of labour, many employees in the most unskilled jobs are poor workers. It is therefore essential to take into account working hours as a parameter.
Furthermore, in order properly to describe job quality, factors other than wage must also be taken into consideration. In countries that have set a minimum wage, this salary can apply to a large range of jobs that nevertheless belong to very different sectors and, once again, involve very different working conditions.
In order to address these two criticisms levied against studies taking hourly wage as sole indicator of job quality, I suggest introducing two new factors. First, an approach in terms of socio-professional categories aimed at showing changes in the proportion of jobs in different employment categories. Socio-professional classifications are imperfect tools, but they do afford a finer-grained approach to job quality than one solely based on salary. The European Socio-economic Groups Classification (ESeG) distinguishes between jobs according to level of skill and sector of activity, thereby making it possible to better describe the types of job that are in expansion. Second, in order to take into account insufficient working hours in certain jobs, I also use the proportion of involuntary part-time employment to better describe job quality in the lower fractions of the social structure.
After outlining the data and methodology, I will show three forms of job polarisation in Europe that go hand-in-hand with a fairly evident decline in middle-skilled employment. I will then, however, show that the remaining middle-skilled jobs are relatively protected from the instability and fragmentation of employment that particularly affects the least-skilled fractions of the service sector.
Data and methodology
The present study draws on data from the European Union Labour Force Survey (EU-LFS). This data set makes it possible to describe the main changes that have affected job structure in around 12 countries between 1993 and 2013.
Several arguments suggest that there has been a certain homogenisation of social structures among western countries over recent decades (global trade and the concomitant alignment of economic policies, technological change). However, we also know that these overall trends should not prevent us from observing potential national specificities linked, in particular, to the welfare regime in place (Esping-Andersen, 1990). The latter influences the rate of growth in the service sector but also employment structure and the ethno-racial and gendered division of labour. Where the service sector is concerned, recent figures provided by Oesch show that expansion of the lowest-skilled jobs in the service sector is connected to the type of welfare regime in place (Oesch, 2015).
In this study, I recreate the groups of the ESeG classification based on the labels of professions coded in the International Standard Classification of Occupations (ISCO). 1 This classification is the result of two decades of debate at a European level (Filhon et al., 2013). At its most aggregated level, the ESeG project posits seven groups for the actively employed: managers, professionals, technicians and associated professional employees, clerks and skilled service employees, industrial skilled employees, and less-skilled employees (in agriculture, industry, or the service sector). In this study, I use a more detailed level of the ESeG classification that distinguishes between clerks and skilled service employees (Amar et al., 2014).
In the second section of this article, I use the proportion of workers in involuntary part-time jobs as an indicator to describe working conditions at the bottom of the social structure. The EU-LFS data set provides information about working time as well as the reasons for part-time work (person still undergoing training, for personal health reasons, need/choice to look after children or incapacitated adult, person could not find a full-time job, other personal reasons). It is therefore possible to calculate the proportion of part-time workers in each group of occupations who state they were unable to find a full-time job or contract.
The erosion of middle-skilled employment: Patterns of change in European social structures
Here I present changes in share of the workforce for four groups of the ESeG classification: 2 the high-skilled jobs of managers and professionals, the less-skilled jobs, and two middle-skilled categories, namely less-skilled industrial employees and clerical workers in administration and the service sector (Figure 1).
Three patterns of change in European social structures.
The results allow three groups of countries to be distinguished. In France, Sweden and Austria, polarisation is particularly strong. Four trends can be seen in these countries: (a) an increase in the proportion of highest-skilled jobs (managers and professionals); (b) an increase in the proportion of less-skilled employees; (c) a decline in the proportion of industrial skilled employees; and (d) a decline in middle-skilled clerks’ jobs. In these three countries, polarisation is not just about the concomitant rise in the proportion of the highest- and lowest-skilled jobs, it has also resulted in a significant decline in the proportion of middle-skilled jobs over the past two decades.
A second group of countries comprising Spain, Italy, Greece and Germany has also seen a form of polarisation insofar as there has been an increase in the proportion of the highest- and lowest-skilled jobs. However, while these countries have also seen the fraction of industrial skilled employees decrease, the same is not true of clerks. In other words, in absolute terms, there has been no significant erosion of the share of middle-skilled jobs across all sectors of activity. In relative terms, however, the latter is indisputable.
Finland, Denmark, Portugal, the UK and the Netherlands form a third group of countries in which polarisation takes a less obvious form, insofar as the proportion of less-skilled employees seems relatively stable. However, these countries are facing a decline in the proportion of middle-skilled jobs. For these countries, it all therefore depends on how one chooses to define polarisation: the fact that middle-skilled jobs in industry and administration are in decline, whereas the proportion of less-skilled employees remains stable, could be seen as indicating a more moderate form of polarisation or, put differently, ‘relative polarisation’.
What conclusions can be drawn from this quick overview? Seven of the 12 countries in question are facing a very clear trend toward polarisation of social structure and particularly an increase in the proportion of the lowest-skilled and most precarious jobs, largely in the service sector. Moreover, my results show a fairly clear trend towards a decline in middle-skilled employment. In industry, this has decreased sharply in all countries. In administration and bureaucracies, it has declined in eight countries. These results are congruent with those obtained recently by Cirillo. Based on different data but taking an approach through socio-professional categories, she demonstrates a fairly clear trend towards job polarisation while also emphasising the need to take into account sector of activity (Cirillo, 2018).
Is this enough to allow us to talk about a shrinking middle class, more generally? These middle-skilled jobs in industry or the service sector in fact correspond to the lower fractions of the middle class. Conversely, technicians and associated professional employees – who constitute the core, or perhaps even the higher fractions, of the middle classes, with higher levels of income and qualifications – are increasing in proportion across all countries. With the decline in jobs at the interface of the upper working classes and lower middle classes, the middle classes are being undermined from the bottom up. Nevertheless, there are variations between national contexts. Some countries are not experiencing any clear decrease in administrative jobs and others, where the latter are in decline, have not experienced any real polarisation insofar as there has been no significant increase in the least-skilled jobs.
In order to account for changes in social structures and look beyond common trends (i.e. the decreased proportion of industrial workers and increased proportion of highest-skilled employees), several factors must be taken into consideration, and first and foremost public policies that shape the effects of technological change. The unequal decline in the proportion of administrative employees is partly linked to differences in structure according to sector of activity: in Germany, 25% of administrative workers are employed in industry as opposed to 14% in France, according to the EU-LFS data set. Differing rates in the penetration of new technologies is probably also a factor. However, where public administration is concerned, it is important to also look to the effects of differences in the time frames of administrative reform. Similarly, differences in the increase rate for the proportion of ‘lowest-skilled’ occupations can be linked to a certain number of public policies that affect job expansion in certain types of services, for example, in the sector of interpersonal services, where states can either encourage the creation of these jobs or not, particularly on a fiscal level (Carbonnier and Morel, 2015; Morel, 2015).
Approaching this issue through broad occupational categories makes it possible to describe an initial form of polarisation, highlighting an increase in the proportion of less-skilled employees in seven countries and a consistent decline in middle-skilled job across most countries. It is nevertheless necessary to go beyond simply describing general changes in different job categories. To complete this study of the dynamics of employment structure in Europe, it is important to focus on the issue of working conditions. Are we witnessing a general trend towards the polarisation of working conditions? Above and beyond the erosion of a large number of middle-skilled jobs, is there also a trend towards more precarious conditions for workers who manage to retain these kinds of jobs?
The polarisation of working conditions: Are the remaining middle classes being spared?
By using a more detailed level of the ESeG classification, it is possible to take involuntary part-time employment as an indicator of job quality (Table 1).
Proportion of involuntary part-time jobs in 2013.
First, in 2013, in all countries, the highest proportion of involuntary part-time work is found among the less-skilled employees, ranging from 11% in Austria to 40% in Italy. Second, and in comparison, the proportion of involuntary part-time work is much lower among clerks (ranging from 3% in Portugal to 13% in Sweden) and among skilled industrial employees (ranging from 1% in Austria, Denmark and Portugal to 6% in Greece). This result underscores the fact that the middle-skilled jobs that do remain are relatively well protected when it comes to working conditions and are fairly comparable, in this regard, to professionals and managers, as well as technicians and associated professional employees. It is clear that the deterioration of working conditions first and foremost affects the least-skilled employees in the service sector.
In this regard, the second observation is that there is a division in most countries between the more-skilled employees in the service sector and their less-skilled counterparts, 3 as the proportion of involuntary part-time work is significantly higher among the latter. Two Scandinavian countries present a notable exception: in Denmark and in Sweden, the two proportions are comparable. In the Netherlands, the discrepancy between the two is not considerable and the same is true, to a lesser extent, of Finland. Conversely, it should be noted that this divide is particularly significant in France and countries in southern Europe where the discrepancy between skilled and unskilled work in the service sector is particularly pronounced.
The third observation is that the proportion of involuntary part-time employment is at its highest among the lowest-skilled service employees in Italy, Spain and France: 40%, 36% and 31% respectively. The proportion of service sector employees within the active population is particularly high in Spain and France and increased sharply in Spain between 1993 and 2013 (Bernardi and Garrido, 2007).
Fourth observation, it would seem that the proportion of involuntary part-time jobs is higher in southern European countries and lower in northern European countries. However, this link is not automatic, as Sweden has a high proportion (24%) and Portugal a relatively low proportion (18%). Within these results, France clearly falls among the southern European countries, whereas Germany (20%) and the UK (18%) are in an intermediate position. These results should be considered in light of public employment policy. In France, the 2005 Borloo law relating to the interpersonal service sector aimed to double the number of employees through a policy reducing employer costs and making employment conditions more flexible. Conversely, in Finland the sector has historically been run by local authorities: most workers are state-employed workers covered by collective agreements that afford them much greater protection (Kröger, 2011).
Finally, while the proportion of involuntary part-time employment in the service sector differs quite considerably according to country, these employees’ working conditions are far more similar than those of managers or experts. Whatever the country, the proportion of managers or experts facing involuntary part-time work is lower than 5%. This means that the polarisation of working conditions is particularly stark in countries where the proportion of involuntary part-time work is high among low-skilled service employees. In Spain and France, for example, the greatest rise in employment share has been, on the one hand, in managerial and professional jobs and, on the other, among low-skilled service sector employees – in other words, the jobs that are the most protected from labour fragmentation and the jobs that have been most affected by this.
Conclusions
By describing three patterns of change in employment structure in Europe, my results underline the decline in the proportion of middle-skilled jobs in most of the countries. The proportion of skilled industrial workers has decreased in all countries, whereas the proportion of administrative employees has seen considerable erosion in eight of the 12 countries. These changes are a sign of the undermining of the lower fractions of the middle classes. However, the remaining jobs in these sectors are relatively protected from the deteriorating working conditions and increasingly precarious employment affecting the least-skilled jobs in the service sector. The jobs that remain in this area of middle-skilled employment continue to offer relatively good employment market participation. One question does remain, however, that would require individual and longitudinal data to answer: what has happened to the millions of employees in Europe who have lost their jobs as skilled labourers or clerical staff? Have they increased the competition for less-skilled and less-stable jobs, therefore fuelling a massive flow of occupational downgrading with potentially major social and political consequences?
These results also make progress in analysing the factors that explain European employment patterns. The fact that change has not followed one single pattern shows that technological change is far from the only factor to take into account when it comes to explaining the evolution of social structures; institutions and public policy should also be taken into account (Fernandez-Macias and Hurley, 2017). In each country, patterns of change in employment sectors and working conditions are also the result of political choices and power relations between social actors.
This also means that type of welfare regime should be taken into account, even though traditional classifications do not provide a perfectly adjusted interpretative grid for this. The differences observed between northern and southern European countries are particularly enlightening in this respect. With the exception of Sweden, countries in the North seem to have contrived to contain the expansion of the lowest-skilled jobs and to avoid the decline in working conditions of people in such employment. This is not the case for countries in the South, where the increase in the proportion of the lowest-skilled jobs has gone hand-in-hand with the fragmentation of labour and ever more precarious working conditions.
Supplemental Material
Appendix – Supplemental material for The decline in middle-skilled employment in 12 European countries: New evidence for job polarisation
Supplemental material, Appendix for The decline in middle-skilled employment in 12 European countries: New evidence for job polarisation by Camille Peugny in Research & Politics
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental materials
Notes
Carnegie Corporation of New York Grant
This publication was made possible (in part) by a grant from the Carnegie Corporation of New York. The statements made and views expressed are solely the responsibility of the author.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
