An Increasingly Educated Workforce

While the recent development of atypical work has been massive, the most spectacular phenomenon in the composition of the European workforce over the last twenty years has been its educational improvement. Figure I.6 reports the number of persons in employment holding a tertiary diploma in France, the UK, Germany, and Spain. Despite breaks in the series, it appears that their number has risen dramatically over the past twenty-five years. This evolution is the result of the widely documented increasing demand for educated workers, driven by technological and organizational progress. This demand met an increased supply of higher educated workers, boosted by the democratization of tertiary education or, in some countries, by immigration.

Figure I.6. Employment of workers with tertiary education, 1992-2013 (in thousands) France, Germany, Spain, United Kingdom

Source: For Germany, Spain, and the United Kingdom, Eurostat-LFS, workers aged 15-74; breaks in the series. For France, INSEE estimations adjusted for breaks, released in February in 2015

Even during the Great Recession, the number of highly educated workers in employment did not decline in Spain,^[1] while the increasing trend was unaffected in the three other countries. Differences in their level and evolution in Spain, as compared to the three largest European economies, may be related to the distance to the technological frontier. Indeed, Spain has the lowest R&D spending of the four countries, and statistics for STEM jobs (Science, Technology, Engineering, and Mathematics) reveal important differences. According to our estimations using the EU-Labour Force Survey,^[2] the share of STEM workers in the workforce in 2012 was 7.4 per cent in Great-Britain, 7.8 per cent in France, and 8.2 per cent in Germany, but only 4.7 per cent in Spain.

These trends in education are consistent with an ongoing technological revolution and corroborated by Nick Crafts' analysis in Chapter 2, as well as by the micro-evidence presented country by country in Chapters 4 to 9. Since young women are more educated than young men, this steady demand has been favourable to women in recent years. It accounts for part of the reduction in the gender gap in employment rates. According to Eurostat-LFS, this gap fell by 2.3 per cent in the EU for workers aged twenty to sixty-four;by 1.7 per cent in the UK, 2.1 per cent in France, 3.5 per cent in Germany, and by 8.7 per cent in Spain.^{[3] [4]}

In the long run, educational upgrading of the workforce should enhance productivity. However, it may also dramatically affect the cyclicality of productivity if firms exhibit different behaviour vis-a-vis highly educated workers and the others. Figure I.6 suggests that at least in France, Germany, and the UK, the demand for highly educated workers is acyclical, which contributes to the under-adjustment of employment and to lower productivity growth. As Boeri comments in Chapter 9, the productivity puzzle in Germany may be viewed as the incidence of a low Okun's coefficient. To be more precise and to decompose that effect, one should consider that the elasticity of (unemployment to variation in GDP is an aggregate of different educational attainments. Table I.3 reports the short-run adjustment of employment for the tertiary and non-tertiary educated workers in the EU15 and our four selected countries between 2005 and 2013.¹¹ In France, Germany,

Table I.3. Quarterly sensitivity of employment (aged twenty-five to seventy-four) to changes in GDP by educational attainment

Note: GDP seasonally adjusted, quarterly data; Eurostat-LFS quarterly employment, unadjusted twenty-five to seventy-fouryears. Authors' estimations: equations in first differences, fixed effect per quarter, clustered by quarter.

Thirty-five observations. ***statistically significant at 1 per cent level; ** at 5 per cent. Interpretation: in the EU15, a 1 per cent increase in GDP is associated with 0.35 per cent growth of jobs held by workers who have a secondary education or less, just as a 1 per cent decline in GDP is associated with a drop of0.35 percent in jobs they hold.

Source: Eurostat

and the UK (and on the average in the EU15), the quarterly evolution of tertiary educated employment is negatively, but non-significantly, correlated to GDP changes. In Spain, the coefficient is positive but is slashed by half in comparison to the non-tertiary educated.

The chapters on national economies acquaint readers with workplace- level evidence showing that firms hoarded highly skilled workers during the recession. Numerous arguments are presented in the various chapters, rationalizing why highly educated/high-skilled workers are hoarded or hired. In Chapter 6, Andrews reminds us of the Beckerian mechanism: firms invest more in the human capital of skilled workers and are thus more reluctant to lay them off. Also, in Chapter 4 on France it is argued that highly educated workers are more mobile, which may improve the matching process, especially in the case of a macroeconomic shock; if highly educated workers are involved in long-run projects, the opportunity cost of hiring such workers is counter-cyclical.^[5] In Chapter 7,

Bellmann, Gemer, and Laible stress that German firms faced great difficulties at the end of 1990s in finding highly skilled workers after many had been dismissed during the post-reunification recession. As a consequence, firms now try to avoid repeating such errors by hoarding their core workforce, which is one of the goals of concluding pacts for competitiveness. The demographic decline should theoretically have magnified this behaviour.

Therefore the hoarding of skilled workers and the dramatic educational amelioration of the workforce, combined with labour market reforms and labour market policy reactions to the recession, constitute important hypotheses to explain a lesser adjustment of the aggregated workforce in the three largest European economies during the Great Recession and the apparent productivity slowdown. These hypotheses may be combined with other explanations based on technology that are developed in both parts of the book.

• [1] And the share of high-skilled workers on temporary contacts increased during the crisisin that country, showing that skilled workers were relatively protected.
• [2] STEM jobs are ISCO-08 211, 212, 213, 214, 215, 216, 251, 252, 311, and 314.
• [3] In Spain the decrease in the gender gap is also related to the fact that job destructions inthe construction sector concerned essentially males.
• [4] Eurostat provides quite consistent data for this period with no (or small) breaks in theseries.
• [5] A companion paper (Askenazy, Chevalier, and Erhel, 2015) formalizes this idea. Themodel shows that if firms are not facing harsh credit constraints, they may even hire more