The p-Innovation ecosystems model

TL;DR

This paper introduces the p-Innovation Ecosystems model, a spatial clustering approach to identify economic regions that preserve industry interactions, aiding regional development and policy planning.

Contribution

It formulates a novel mixed-integer programming model for clustering regions based on industry interactions and provides a heuristic solution approach.

Findings

Model successfully clusters Colombian municipalities into innovation ecosystems.

The approach preserves key industry interactions within regions.

Potential policy applications demonstrated through case study.

Abstract

In this paper, we propose a spatially constrained clustering problem belonging to the family of "p-regions" problems. Our formulation is motivated by the recent developments of economic complexity on the evolution of the economic output through key interactions among industries within economic regions. The objective of this model consists in aggregating a set of geographic areas into a prescribed number of regions (so-called innovation ecosystems) such that the resulting regions preserve the most relevant interactions among industries. We formulate the p-Innovation Ecosystems model as a mixed-integer programming (MIP) problem and propose a heuristic solution approach. We explore a case involving the municipalities of Colombia to illustrate how such a model can be applied and used for policy and regional development.