Unfolding the innovation system for the development of countries: co-evolution of Science, Technology and Production (original) (raw)

We show that the space in which scientific, technological and economic developments interplay with each other can be mathematically shaped using pioneering multilayer network and complexity techniques. We build the tri-layered network of human activities (scientific production, patenting, and industrial production) and study the interactions among them, also taking into account the possible time delays. Within this construction we can identify which capabilities and prerequisites are needed to be competitive in a given activity, and even measure how much time is needed to transform, for instance, the technological know-how into economic wealth and scientific innovation, being able to make predictions with a very long time horizon. Quite unexpectedly, we find empirical evidence that the naive knowledge flow from science, to patents, to products is not supported by data, being instead technology the best predictor for industrial and scientific production for the next decades. Knowledge production and organization represents the main activity of modern societies – " learning economies " [1] in which most of the wealth of a country is intangible, and the organization of the national innovation system [2], and of diffused creativity [3] are the crucial capabilities for success. Therefore, in the last thirty years the relationships between science, technology and economic competitiveness has become an important focus for social sciences in general and economics in particular [4, 5]. Even though the standard narrative links science, technology and economic productivity in a direct flow [6], actual interactions are likely multi-directional [7] and emerging from the non-trivial interplay among their individual components: the footprint of a complex system. The new literature of Economic Complexity uses techniques which, differently from traditional social science approach , do not try to average out the complexity of the system, but embraces it by explicitly building on the het-erogeneity of individual actors, activities and interactions to extract relevant parameters to characterize the system. Trying to recover the qualitative insights [8] and the few quantitative attempts [9, 10] of the heterodox economists and social scientists, researchers used this approach to study unobservable characteristics and capabilities of countries [11–13], and to unearth unexpected synergies among different activities [14, 15]. Following this line [9, 14–16], here we create the network of interactions between the different human activities. We build on the assumption that if two activities co-occur significantly more often than randomly (in terms of appropriate null models) in the same countries at the same time, then there is an overlap between the capabilities required to achieve proficient level (i.e., competitive advantage) in both. For the first time our network encompasses activities in different realms (or layers): scientific fields, technological sectors, and economic production. In such a comprehensive multi-layer network [17], i.e., a system where entities belong to different sets and several categories of connections exist among them, we particularly focus on interactions among the different layers. As detailed in the Supplementary Information, to task we build the adjacency matrices M L c,a (y) connecting country c with the activity a belonging to layer L if, in year y, country c was expressing a competitive advantage in activity a with respect to the world average. L stands for the layer of analysis, consisting in the set of all activities related to either Science, T echnology or P roducts export. Note that each of these layers has an intrinsic hierarchical structure: for instance, in the science layer we can consider activities like Physics and Astronomy or corresponding sub-activities (like Statistical Physics, Condensed Matter Physics, Nuclear and High Energy Physics). Thus, our matrices do depend on the resolution used for activities classification (even if not explicitly reported in the notation). We use different established databases to construct the multi-layer space: for S cience, we take bibliometric data on papers in the period in the various scientific fields from Scopus (www.scopus.com); for T echnology, we consider the number of patents in different technological sectors extracted from Patstat (www.epo.org/searching-for-patents/business/patstat); and for P roducts export, we use the export data collected by UN COMTRADE (https://comtrade.un.org/). Using these matrices we compute the probability of having a comparative advantage in activity a 2 ∈ L 2 in the year y 2 , conditional to having a comparative advantage in activity a 1 ∈ L 1 in the year y 1 (Figure 1), that we define as the