Competitive impact of scoring design in public sector invitations to tender

Abstract

Public procurement represents a field where significant public funds are expended to provide services to the general populace. Improving public procurement methods has been identified as a potential means of reducing public spending. While public procurement has long been in the interest of academics, empirical research literature about tender scoring practices remains scarce, particularly in terms of its competitive impact.

This thesis examines scoring design and its relation to competition in public procurement invitations to tender (ITT). The study was conducted as a sequential mixed methods case study of HUS medical devices and supplies (CPV-33) procurement. The specific research questions posed are: 1) What kind of scoring design and -criteria are used in Finnish healthcare ITTs, and 2) How are scoring criteria associated with competition in such ITTs? These questions are addressed in two sequential research stages.

The first stage of research is an exploratory content analysis of HUS ITT documents. The content analysis aims to collect, categorize, and analyze the scoring criteria in single lot sealed bid auction ITTs from 2019-2022. This thesis discovers that the scoring has been conducted using either the Lowest Price or the Economically Most Advantageous Tender (EMAT) scoring methods. In the case of EMAT, the weight of price criteria has ranged from 40% to 100%, with an average of 3.7 quality criteria per ITT. The first stage results also indicate potential instances of human bias in the design of scoring rules.

The second research phase begins by testing the difference in the number of bidders between Lowest Price and EMAT scoring. It appears that EMAT scoring tends to correlate with fewer bidders. This phase then presents a regression analysis of scoring criteria data gathered in the first research phase. The outcome variable in this analysis is the number of bidders, which corresponds to the number of bids, as the final sample considers only single lot ITTs. The results of the regression analysis are somewhat limited, largely due to the restricted data sample.

The empirical findings of this thesis suggest that real-world scoring practices in public procurement have potential for improvement. Theoretical perspectives on different scoring practices diverge, warranting further research to create a more comprehensive understanding of the actual implications of various scoring details on competition, as well as other procurement outcome measures.