SUMMARY
Chloé Duvivier, Stéphanie Truchet, Nicolas Mauhé and Marouene Mbarek
La Documentation française | " Économie & prévision "
2018/2 n° 214 | pages 97 to 139
For almost 15 years, public authorities have been deploying the Internet in rural areas that have been abandoned by private operators for lack of profitability. These public investments are motivated by a desire to ensure territorial equity, but they also aim to boost the economic attractiveness of these areas.
In order to assess the effectiveness of these policies on this second objective, this research analyzes the effect of very-high-speed internet (VHS) deployment on business start-ups in low-density communities. To this end, we evaluate the impact of the first years of the Auvergne Très Haut Débit program, launched in 2013, which aims to eliminate all "white zones" not served by THD in Auvergne by 2025. To achieve this goal, the program is deploying several technologies: fiber-to-the-home (FttH), whose deployment is being prioritized, but also alternative technologies (DSL and Wifimax speed upgrades) and professional fiber (FttO).
In order to provide a detailed analysis of the program's impact, we proceed in two stages. First, we assess the average impact of the program, i.e. its effect on all beneficiary communes and business start-ups. Then, we study whether the program's impact is heterogeneous, by assessing whether it varies according to the type of beneficiary communes and businesses. Furthermore, in each case, we examine whether the effect of deploying fiber optics is greater than that of deploying all technologies combined.
The article's originality lies in its assessment of the impact of access to THD on the economic attractiveness of rural areas, particularly in France, since few studies have yet tackled this issue, and most of them focus on North America.
The analysis is carried out on the 1,175 Auvergne communes located in the public initiative zone, over the period 2010-2017. Three sets of data are mobilized for this purpose.
On the one hand, data on the rollout of the program, provided by Régie Auvergne Numérique, indicating for each commune the date of deployment of each technology. Secondly, annual data on the number of establishments created per commune, taken from the Insee Répertoire des entreprises et des établissements (REE/Sirene). Finally, we mobilize a variety of information on the characteristics of communes from different databases (Census of Population, Permanent Equipment Base, etc.). To measure the causal impact (or own effect) of the Auvergne Très Haut Débit program, we use impact assessment methods. More specifically, we use a double-difference method with matching, which consists in comparing the evolution of business start-ups in communes benefiting from the program with those in similar communes not yet benefiting from THD deployment. This is particularly appropriate in our case, as certain communes are given priority
based on their observable characteristics. Compared with standard matching, which assumes selection on the basis of observables, this method also makes it possible to control for unobserved heterogeneity constant over time and for temporal effects common to treated and untreated communes, thus eliminating the main sources of potential bias.
According to our estimates, the deployment of both THD and Fiber has not stimulated business creation when all beneficiary communes and businesses are considered. Estimates by type of commune and sector of activity indicate that the program has affected business creation in four sectors in particular communes. Deployment stimulated business creation in three sectors: catering and accommodation; public administration, education, human health and social action; and other service activities. On the other hand, negative effects on real estate have been identified. These effects are most noticeable in communities with a high level of natural amenities, or those that have benefited from
significant THD deployment. Moreover, it appears that the impact of Fiber deployment is greater than that of HSBB for all technologies combined, and that the program's effects are reinforced over time. Finally, it should be remembered that the effects measured here are short-term effects, produced over the first four years of the program.
As a result, it appears that while the deployment of HSBB is not sufficient on its own to boost rural economic development, it can be a lever for development when certain conditions, relating to the territorial economic context and the design of the digital policy, are met.
Keywords : territorial attractiveness, Business, internet, technology, specific territories, rural territories, rural areas
