The role of human capital in the productivity slowdown in France

What are the main structural factors explaining the productivity growth slowdown? How much is human capital contributing to growth? Did the financial crisis of 2008-2009 play a role in the productivity slowdown? As the world is going through a COVID-19 recession, this last question is even more important today.

Our note uses a time series analysis of growth using national accounts data, to identify structural breaks trend of productivity growth between 1976 and 2018. France has experienced three structural breaks in labour productivity growth since the mid-1970s. The first two occurred in 1986 and 1993, and the last in 2004, well before the 2008 crisis. Although no structural break was identified at the time of the 2008 crisis, it may have temporarily amplified the earlier slowdown identified in 2004, due to its impact on the financing of the economy, the negative shock for businesses and the public policy responses.

We decomposed the sources of growth using standard macroeconomics growth models. Our results suggest that human capital contributes between two thirds to three quarters to the trend of productivity growth over the whole period. In fact, between 1976 and 1986 when productivity growth averaged 3.4% per year, human capital contributed 2.2 points to the productivity growth. This contribution gradually and progressively decreased to bottom out at 0.6 points since 2004. Concomitantly, productivity growth only progressed by 0.7% per year in the last period.

Over the past fifty years, each successive new cohort entered the labour market with a higher level of years of schooling. However, the growth of human capital has significantly decreased over the period. Between 1975 and 2000 the share of secondary school graduates in the French working population rose from 37% to 68%, that of higher education graduates from 10% to 25%. The average annual growth rate of human capital was 2.5% between 1975 and 2000 for secondary education and 3.5% for tertiary education. This high rate of growth in the beginning of the period decelerated significantly subsequently. Over the following twenty years, in 2000-2020, it decreased respectively by 1.5 and 1 point. The growth in years of schooling is reaching a natural limit as we get closer to an entirety of a cohort completing secondary schooling.

Our note does not aim at identifying a causal link between human capital to productivity growth. However, our macro-econometric growth decompositions point out that the deceleration in human capital growth explains between 60% to 75% of the productivity growth slowdown.

The note ends with a reflection on the needs to develop non-cognitive skills. At the firm-level, the quality of human capital has a direct impact on the determinants of growth highlighted in empirical micro-econometric studies. Both cognitive and non-cognitive skills are linked to better management quality, innovation, the adoption of new technologies and innovative forms of work organisation.

The limits to the expansion years of schooling suggest that the policy focus should switch to increasing the quality of schooling, and on fostering life-long learning systems that increase human capital throughout a worker’s life.