Ilyes Boumahdi (p), Nouzha Zaoujal

The disruptive effect of automation affects both the economic structure of countries and their labor markets. We have highlighted the impact of automation on the labor market on the national and territorial level of Morocco and the 27 countries of the European Union (EU27) (12 Moroccan regions and 81 territories of the Nomenclature of territorial units for statistics (NUTS 1) from the EU27), drawing on the work of Frey and Osborne (2013).
The geographic choice of this study is dictated by the increasingly important link between the productive ecosystems and, consequently, the employment basins of Morocco and those of the EU27 taking into account the fragmentation of the chains of value of industrial production and services between the two shores of the Mediterranean.
We found that, in addition to Morocco (63%), fifteen countries have a prevalence of automation higher than the EU average of 27 (47%). The lowest rate is that of Luxembourg (38%) whose tertiary sector occupies 77% of total employment in the Duchy against only 46% for Romania which has the highest prevalence of automation (58%). The automation rate is spread from the center of Europe to the east and slightly to the south, roughly following the historical trend of the enlargement of the European Union.
This spread of the automation rate is noticeable at the subnational level (example of the Comunidad de Madrid (44%) and the Île de France (42%)) relative to the peripheral regions of the countries (Respectively the Sur region (49%) and Bourgogne - Franche-Comté (47%)).
All EU27 regions have a lower prevalence of automation than the Moroccan regions except the Romanian regions of Macroregiunea patru (60%) and Macroregiunea doi (63%) which have a higher prevalence respectively than Guelmim- Oued Noun (58%), Rabat-Salé-Kénitra (59%), Oriental (59%) and Casablanca-Settat (60%).
The automation rate also differs according to the differentiated specializations of the regions with a variability which decreases from countries with the highest national rate to those with the lowest rates. Thus, the automation rate in Morocco varies from 52% (Laâyoune-Sakia El Hamra) to 69% (Drâa-Tafilalet) against a national average of 63%, or 17 points difference when this difference is only 1.7 points between Oost-Nederland (41.7%) and West-Nederland (40%) against an average of 41% in the Netherlands.

Xiumin Li (p), Mi Guo

This paper aims to analyzes the impact of information and communication technologies (ICT) on regional economic growth in China during the period 1995-2017 from provincial-level perspectives. We estimate the ICT capital stocks of 30 provinces in China (not include Tibet because of the restriction of data acquisition) using perpetual inventory method firstly; then we analyze the relationships between ICT capital per capita and GDP per capita for all provinces; finally, we divided the economic growth of 30 provinces into different sources, the growths of inputs include labor, ICT capital and non-ICT capital and growth in total factor productivity, according to the framework of growth accounting formed by Jorgensen & Griliches (1967) and Schreyer (2001).

The primary data sources for the analysis are China Statistical Yearbooks, China Electronics Industry Yearbooks, China Electronic Information Industry Statistical Yearbooks, Statistics of China's electronic information industry from 1949 to 2009, as well as China’s Regional Input-Output Tables of 1997, 2002, 2007, 2012. In the estimation of the ICT capital stocks, the ICT investments of the interval years between the two years the input-output tables set up are calculated by the method proposed by Akihiko(1996, 1998, 2003) . The growths of total factor productivity are calculated according to the framework formed by Jorgensen & Griliches (1967).

It is found that during the period from 1995-2017，the ICT investment increased rapidly in all provinces, there are positive relationships between ICT capital per capita and GDP per capita for all provinces, the contribution of ICT capital to regional economic growth rose as the ICT investment increased, the contributions of ICT capital to regional economic growth are different among the provinces, which resulted in the obvious regional disparities. Therefore, increasing ICT investment in underdeveloped provinces will be helpful for narrow the regional disparities.

Frank van der Wouden (p), Hyejin Youn

Individuals who collaborate locally are more likely to learn from one another than those collaborating non-locally. We examine the publication records and knowledge portfolios of almost 1.7 million scholars who published a co-authored paper and subsequently published a single-authored paper. We investigate to what extent the geographical distance between the collaborators impacts the probability of them learning through collaboration. We find evidence that local collaboration is associated with a learning premium of between 40% and 85%. Surprisingly, this learning premium increased over time, despite advances in communication and transportation technologies that supposedly abolish the friction of geographical distance. In addition, we find that the learning premium from local collaboration is greater in individuals in their early career stage than in mid or late career stages. Furthermore, individuals from lower ranked institutions are most likely to learn from collaborations. Our findings also indicate that increasing geographical distance between collaborators hits the collaborative learning of those in STEM-type fields the hardest. We observe highly significant positive effects from local collaboration on learning even after controlling for confounding factors and using matched data in the statistical models. These results suggest that geography plays a key role in mitigating the impact of learning through collaboration. This has important implications on innovation policy, the structure of effective research teams, and the broader processes of knowledge production and diffusion.

Chloe Duvivier, Claire Bussière (p) , Emma Cazou

As broadband access is widely assumed to foster economic development, a number of large-scale broadband deployment programs have been initiated in many countries (e.g., the “National Broadband Plan” in Ireland, “The French Very High-Speed Broadband Plan” in France, “Connecting Canadians” in Canada). One objective is to foster economic development in less densely populated areas, which are often ignored by private telecommunication operators. Although massive investments are being made by public authorities, the genuine broadband effects remain uncertain however.

On the one hand, the emergence of the internet has led to considerable hopes for rural areas, with the idea that it will compensate for their geographical remoteness. On the other hand, however, broadband economic effects in rural areas are largely uncertain for a number of reasons, including average low rates of broadband adoption, high costs of broadband access, and an unfavorable local economic context.

We investigate whether the expansion of broadband fosters economic development in rural areas. To do that, we evaluate whether the French Very High-Speed Broadband Plan significantly affects municipal-level economic outcomes (firm location; employment level and structure). We also investigate whether broadband impacts vary across industries and rural areas, and whether ultra-fast broadband (optical fiber) generates additional benefits compared to first-generation technologies.

The study is carried out on the 31 285 French rural municipalities, from 2010 to 2017. Different sources of data are used, including detailed information on broadband coverage at the municipal level (share of buildings with internet access, by technologies and speed).

To identify the causal impact of broadband on rural economic development, we use spatial differences-in-differences with pre-matching. The methodology allows controlling for two potential sources of bias: selection effects and spatial interactions.

According to the first statistical analyses, broadband effects are not significant when considering all rural areas and all economic sectors. On the other hand, positive broadband effects are found for specific economic sectors and rural areas. In addition, ultra-fast broadband generates higher effects compared to first-generation technologies.