Policy Priority Inference: A Computational Method for the Analysis of Sustainable Development

Abstract

We develop a computational framework to support the planning and evaluation of development strategies towards the 2030 Agenda. The methodology takes into account the complexities of the political economy underpinning the policymaking process, for example, the multidimensionality of development, the interlinkages between these dimensions, the inefficiencies of implementing policy interventions, as well as the institutional factors that promote or disencourage these inefficiencies. The framework is scalable and usable with publicly-available development-indicator data, and it can be further refined as more data becomes available, for example, on public expenditure. We demonstrate its usage through an application for the Mexican federal government. For this, we infer historical policy priorities, i.e. non-observable allocations of transformative resources that generate changes in development indicators. We also show how to use the tool to assess the feasibility of development goals, to measure policy coherence, and to identify accelerators. Overall, the tool provides a systemic framework that allows policymakers and other stakeholders to embrace a complexity view to tackle the challenges of the Sustainable Development Goals.