The surprising disparity between industrial areas

This note therefore studies the mainsprings of performance in the French provinces, in particular the most industrial ones, over the post-crisis period from 2009 to 2015.

The surprising gap between industrial areas: understanding economic performance and decline


Since the late 1990s, the gap has widened between dynamic regions, which are creating jobs, and those in difficulty, which are losing them. This observation feeds into a simplistic notion of “two Frances”, split between urban areas presumed to be the “winners of globalization” and outlying areas seen as the victims of inevitable deindustrialization.
This way of seeing things, which frequently came up during the “yellow vests” crisis, does not however stand up to analysis: economic and social differences do not necessarily correspond with geographic ones. French towns like Vitré, Figeac, Issoire, Cholet, Saint-Nazaire and many other employment-rich zones remind us that areas of average size and population can turn out to have a robust economy, especially in industrial sectors. Yet this heterogeneity is still largely unexplained by the economic literature. This note therefore studies the mainsprings of performance in the French provinces, in particular the most industrial ones, over the post-crisis period from 2009 to 2015.

It first confirms that industry generally – but not systematically – has a strong knock-on effect on employment in the other sectors in the region, while the opposite does not seem to be true. This is why it is important to take a very close look at what drives local industrial sectors.

Next, the analysis looks at the wide variety of regional trajectories. For example, some employment zones have benefited from a growth in their industrial base, even though French industry is known to have lost jobs overall. In addition, positive and negative variations in employment zones do not appear to be related to the size of their industrial base: “major industrial areas” are not doomed to failure any more than major clusters and other metropolitan areas are destined for success. Last and not least, our calculations show that the specialization of areas in more or less buoyant sectors is a fairly weak – and even increasingly weak – “predictor” of their performance. In other words, the key to understanding regional trajectories is the “local effect”, i.e. the residual performance that cannot be explained by their activity portfolio. As it turns out, a region’s local impact could be very positive in a fairly weak sector of activity, and rather low or even negative in another.

After taking stock of this wide variety, the note presents an econometric analysis that identifies some determining factors of this local effect and rules out others. In a very significant way, a region’s industrial employment is boosted or dragged down by employment in the surrounding areas. Thus, the western and southern coastlines of France, the northern Alps and the heart of the Paris region harbour vast clusters of positively driven areas; in contrast, the northeast quarter of the country and a broad diagonal cutting down to the southwest are dominated by groups of negatively driven areas.
These clusters with common dynamics bring a reminder that regions are not islands. On the contrary, a prosperous region can stimulate both local services and industry (suppliers, sub-contractors) in neighbouring areas, especially when partner-based systems like competitiveness clusters play a part. These synergy situations turn out to be a lot more numerous than cases of eviction or rivalry, in which the development of one region works to the disadvantage of its neighbours.

We also observe that the concentration of employment in a few major industrial firms, although sometimes a factor of vulnerability (e.g. when they are mostly in the same sector on an economic downturn), generally has a positive influence on employment growth.

Lastly, the note points out that not all activity sectors react in the same way to different employment determinants: cluster effects for example, sometimes hastily presented as a universal parameter of regional performance, are mostly relevant for knowledge-intensive industries. As a consequence, depending on the specific combinations of their dominant activities, regions give different results on the parameters studied: some suffer from an initial high concentration in industrial jobs, or on the contrary from an economy too exclusively focused on residential activities, while others do not. The note thus presents a typology of territories in four categories.

In conclusion, if the structural characteristics of territories do not provide a simple explanation for their performance in terms of industrial employment and total employment, the reason is probably because these phenomena are increasingly multi-parametric, all the more so because their “social capital”, and in particular the effectiveness of institutions and cooperation between diverse actors, play a crucial role. The main reason is that companies’ momentum and their capacity to generate growth, which do not exclusively depend on the local area, remain the principal driving force behind local development.
All local actors can contribute to creating an attractive, accommodating environment, in particular companies themselves. To understand how to encourage the most dynamic companies and transform individual success stories into collective performance, a good solution seems to be to carry out more qualitative case studies.


Couverture Note 30 - L’étonnante disparité des territoires industriels
Learn more
Téléchargez la Note au format PDF
Learn more
Téléchargez la Synthèse au format PDF
Learn more
Consultez le communiqué de presse
Learn more