National Innovation Systems Archetypal Analysis

Abstract

The national innovation system (NIS) determines the innovation capability of a country, and its economic development. However, recently, very little is known regarding the determinants of NIS functioning in various countries. Probably the easiest way to obtain such an understanding is to begin with the structural representation of the NIS. Particularly, it is quite natural to assume that there exists several ‘cornerstone type NIS’ or ‘archetypal NIS’, and all the other types can be considered a mixture of them. The aim of this paper is to somewhat study the advances in the structural understanding of the NIS. For this purpose we conducted our study based on the data set from the Global Innovation Indexes’ (GII) seven pillars and using archetypal analysis. It is also important to note that the concept of entropy was also naturally determined under archetypal analysis. We demonstrate that each NIS can be considered a mixture of three archetypical NISs, which are as follows: The first one is a prototype of a highly developed NIS (with a high level GII score and a low level of entropy); the second one is a prototype of an underdeveloped NIS (with a low level GII score and a low level of entropy); and the third one is an intermediate form of NIS (with a medium level GII score and a high level of entropy). Hence, we establish that such a multidimensional phenomenon, such as the NIS (described in this study as the 7-dimensional vector – GII pillars), with an acceptable level of the accuracy, essentially can be considered a 2-dimensional object; and the corresponding barycentric coordinates are a convenient means of describing NISs. We also introduce an important indicator – the NIS entropy – which characterises the level of the disorder or randomness in the NIS.