Evaluation of public procurement efficiency of the EU countries using preference learning topsis method

Abstract

Holding governments accountable for public procurement efficiency has been high on the agenda of public finance practitioners in the last few decades. The European Commission has developed a set of value-for-money indicators for public procurements within the Single Market Scoreboard. Although this matrix is actively used to rank countries, a number of downsides have been hitherto reported. This paper proposes preference learning (a machine learning method) for criteria weight estimation in combination with Technique for Order Performance by Similarity to Ideal Solution (as a multi criteria decision making technique) to re-evaluate the public procurement performance of the EU countries. This approach can be used for unbiased ex-post evaluations and focus of efforts and resources on critically important public procurement policies.