Twice the price : An empirical study in procurement of homogenousgoods for Swedish municipalities

Background: Studies of complex and big-ticket tenders such as construction contracts have been the main research focus of public procurement. The efficiency of these low frequency transactionsis inherently hard to study, since the real world offers a wide range of variables that could explain the discrepancies. What is not studied to large extents are high frequency transactions, with low transaction volume, of common goods. These purchases are mostly overlooked due to the lack of documentation that exists, primarily explained by the low degree of digitalization and transparency on smaller contracts/transaction that fall below the EU reporting thresholds.Purpose: The purpose of this research is to examine the existence of price discrepancies for homogenous goods in public procurement between Swedish municipalities. Furthermore, provide insight and possible explanations to this phenomenon in accordance with economic theory.Method: This study opted to use a quantitative method of research to look at different municipalities and their procurement invoices. The quantitative data has been compiled and adjusted to comparable numbers, so that differences can be identified with clarity.Conclusions: The study concludes that there are significant price discrepancies for homogenous goods in the public procurement sector. With price variations ranging from being negligible to paying twice the price. The selected theoretical framework points towards opportunism and asymmetric information being the main sources of the price discrepancies. Improvement could possibly be made by shifting policies towards further digitalization of the procurement process and horizontally integrating the public sphere. In turn reducing asymmetric information and transaction costs, making less room for opportunistic behavior due to improvements in ex ante and ex post costs. From the results of the empirics, a speculative cost saving potential of 7.90 % is estimated