The tradeoff between rules and discretion has been a central topic of research in public procurement. Kelman’s (1990) early work stressed the costs of rigid regulation in the U.S. government procurement, and made the case for discretion. Since then, research on the benefits of discretion has progressed rapidly in Europe (for example, Spagnolo 2012, Coviello et al. 2018, Bandiera et al. 2020, Decarolis et al. 2020). New research confirms that politicians do not trust the bureaucracy, even in countries with high human capital and efficient institutions.
My new paper, joint with Simeon Djankov at the London School of Economics and Professors Ed Glaeser and Andrei Shleifer at Harvard, adds to these studies with a broader geographic and theoretical focus. It is background research for the Contracting with the Government indicator in Doing Business. We cover 187 countries and the complete path of the procurement process. The theory delivers a basic prediction, namely that procurement regulation is more valuable when the incentives of the bureaucrats are not closely aligned with support for social welfare. Properly motivated bureaucrats require fewer rules. Countries with weak bureaucracies need strict laws to regulate them; countries with strong bureaucracies can allow the regulator more discretion.
We use the sample of countries with arguably the strongest bureaucracies – the OECD countries – to check the validity of the tradeoff point. The prediction of the previous research and our theoretical model is that these countries would manifest the greatest discretion, allowing regulators to use different rules for different projects – either based on the size of the procurement project, or its nature – delivery of goods, services or works. The result is surprising: rules are nearly uniformly implemented for small and large projects alike; for buying pencils, delivering nursing care or resurfacing roads alike.
In the study, we have selected a works project worth $2.5 million as representative of the general state of public procurement globally. We first ask the question whether more discretion is given to regulators for smaller-size projects. The answer is inconclusive. Of the 37 OECD members, 18 (Austria, Belgium, Colombia, Estonia, France, Greece, Iceland, Israel, Italy, Japan, Korea, Latvia, Mexico, Poland, Portugal, Spain, Switzerland and Turkey) have no thresholds at any level, that is, the same rules apply irrespective of project size. Four economies (Canada, Lithuania, New Zealand and Slovenia) have thresholds for much smaller projects: C$750,000, €145,000, NZ$100,000 and €100,000, respectively), but beyond this size regulation is uniform. The remaining 10 European Union countries have a threshold of €5,350,000, while Australia, Chile, Norway, the United Kingdom and the United States have thresholds that vary between $3.5 million (Chile) and $7.8 million (United States).
Analysis of the rules that apply above the $2.5 million threshold shows three differences: 1. limited tenders or negotiated proceedings are prohibited or must satisfy very strict conditions; 2. tenders require more extensive publication, including in media outside of the home country (this applies to the European Union); and 3. standstill periods during which unsuccessful bidders or other parties can file complaints are mandated by law in more detail. Finland, for example, requires that tender notices above the EU threshold be published in both the national e-procurement portal HILMA and at EU level e-Notices portal. Legislation in Luxembourg establishes procurement remedies applicable above EU thresholds, and provides a 15-day standstill period between the notification of award and the contract signing, during which an interesting party can initiate a suspending procedure before the Administrative Court. With the evidence that the law allows some more discretion to the regulator below these thresholds, we can conclude that half of the OECD utilizes the tradeoff between rules and discretion, while the other half is fully in the regulation corner.
We next ask whether rules change for different types of procurement: goods, services or works. Here the answer is unambiguously no. Only 4 of 37 OECD members (Canada, Chile, Germany, Ireland) use different regulation for different projects. In Chile, for example, the procurements of goods and services below a certain amount can be contracted directly by public bodies. In Ireland, contracts for goods and services with an estimated value of less than €5,000 can be purchased on the basis of verbal quotes from three competitive suppliers. In Canada, government agencies can procure services up to C$25,000 directly from a single supplier, avoiding complex processes. In the remaining 33 OECD economies, the rules are uniform.
The conclusion is clear: in the tradeoff between rules and discretion, rules rule.