How the use of knowledge complexity and relatedness measures can lead to smarter bottom-up economic policies

The idea for this work emerges from those instances of economic and geographic literature that aim to provide with tools to better understand and study regional economies by concentrating on local, place-based capabilities. This kind of approach tries to focus on development opportunities and limits that are tailored to individual territories, and rejects the one-size-fits-all type of policy. In recent years some measures of economic diversity have been introduced as a proxy to study the complexity of economies. The underlying idea of this field of research is that the more complex an economy is, the less ubiquitous its capabilities and products are, and the more potential gains it can aim to. Since the very introduction of well-implemented measures of economic complexity, it has been shown that they strongly correlate with some of the most important indicators of economic wealth, and that they can be predictive of future economic paths. The attention to local peculiarities has also been deepened with the introduction of measures that account for the relatedness among regional capabilities. These measures can act as summary indicators to minimise the risk of new investments when studying economic policies for regions. This work's approach tries to recombine some complexity and relatedness measures into a framework for smart specialisation, as suggested by previous literature. The framework aims to position the regions' different policy options into a two-dimensional space, in which one dimension represents the policies' expected gains (with complexity measures as a proxy), and the other dimension represents the levels of risk associated to those policies (as summarised by relatedness measures). This structure will then be used to examine a case study of Piedmont and the city of Turin, to analyse the economic status of this region for recent years, and to see if this approach can help to interpret the past fluctuations in the local industrial history. For the purposes listed above, we will compute the main measures of complexity and relatedness included in the framework on a patent database, assuming that patents can be considered as a summary of the technological capabilities a region owns. In some parts of this work, we will also use macroeconomic data on regions' population and GDP. Through the analyses carried out for this work, we find confirmation that complexity and relatedness can effectively be used to explain some aspects of technological development in regions, both in general and when applied to our specific case study. We then use the framework environment to assess a few examples of local economies at different points in time, by comparing Piedmont and some similar regions at different points in time, and observing the evolution of available bottom-up economic policies for specific place and time-based capabilities.