Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - REPCOR (The Role of Reputation and Corruption in Procurement)

Teaser

Public sector procurement accounts for approximately 20 percent of GDP in developing countries and 10-15 percent in developed countries. Nearly all activities in which the public sector is involved, from defense to transportation, from education to healthcare, require the...

Summary

Public sector procurement accounts for approximately 20 percent of GDP in developing countries and 10-15 percent in developed countries. Nearly all activities in which the public sector is involved, from defense to transportation, from education to healthcare, require the public sector to procure construction works, goods and services from private contractors. Moreover, during the recent financial crisis, public procurement has often been invoked as an instrument to sustain demand or even promote growth by fostering innovation.
The proper design of procedures through which procurement takes place is crucial to avoid waste and enhance social welfare. The appropriate selection of contractors clearly plays a prominent role among these procedures and, indeed, a large body of academic and policy research has been devoted to the design of tender awarding rules (Laffont and Tirole, 1993). Preventing corruption and ensuring that the selected contractor will comply with the obligations in its bid are often seen as major goals of a successful awarding procedure. Nevertheless, there is surprisingly limited evidence as to how different awarding methods are susceptible to corruption. Similarly, there are staggering differences across countries in the extent to which contractors’ past reputation is used to award new tenders.
The roles of corruption and reputation within a procurement system are distinct, but closely connected. Fundamentally, the procurement problem is that of achieving the best value for money. Both in private and in public procurement, however, the individuals in charge of the procurement decisions are often distinct from the individuals who will benefit from the object procured. This generates a basic agency problem that is typically addressed by specifying rigid regulations and transparent awarding rules, like auctions. Price-based auctions are indeed a transparent mechanism that fosters competition between contractors and, under certain conditions, can be the “optimal” procurement mechanism (Myerson, 1981). Nevertheless, the lack of flexibility in the choice of contractors might imply that a low price at the time of the auction results in a low delivered quality. Past performance of the contractor is thus one of the prominent tools through which contractors can be given the right incentives. The choice of how to integrate reputation within a transparent procedure is, however, not trivial and a poor design can foster corruption and inefficiency. In the US, a major reform in the 1990s, intended to reduce the rigidity of procurement procedures, introduced into the Federal Acquisition Regulations more flexible purchasing methods, similar to private sector practices, among which was the placement of a stronger weight on suppliers\' past performance (Kelman, 1990).
The issues of corruption and reputation are central to the debate behind the recently released European Union Directives on Procurement (2014/24/EU), the body of rules that regulates public procurement practices in the EU. The major innovation of these Directives is their greater emphasis on more flexible awarding rules with price-based auctions being replaced by scoring rule auctions as the default awarding method. The latter mechanism entails weighting multiple criteria to select winners. Nevertheless, reputation is not an admissible criterion and can, at most, be used to exclude some potential participants. The effects for society of correctly designing these regulations are fundamental and a vibrant academic literature, that I will discuss below, has informed the reforms in both the EU and the US. Furthermore, the goal of curbing corruption through procurement design is now prominent in the work of development economists (Olken and Pande, 2012).
This research proposal describes three projects that will advance the frontier of our understanding of the roles of corruption and reputation in procurement. It is motivated by the consideration that the lack o

Work performed

In the first thirty months of the project, I have worked on all of its three components and make progress in each of them. At the moment of writing this report, the advancement on the three components can be summarized as follows.

For Component 1, I completed the first core paper and began the analysis of a more recent wave of data to be used for a second core paper. The completed paper (“Past Performance and Procurement Outcomes”, joint with dr. Riccardo Pacini and prof. Giancarlo Spagnolo) is under submission at a leading economics journal, the American Economic Review. We see it as an important contribution to the economic analysis of public procurement as it provides strong, new evidence on the positive effects of introducing reputational elements in the selection of contractors. This is a hotly debated policy issue in Europe and indeed the research has appeared in the Italian press and we have been invited to present it to both Italian (Anticorruption Agency) and European (European Space Agency) institutions, as well as in an ample list of academic seminars and conferences, including: the Barcelona Graduate School of Economics, The University of California Berkeley, Stanford University, Stockholm University, University of Bologna, University of Paris I-Sorbonne, Mannheim University, the University of Oslo and the ZEW Mannheim.
In the last months (from April 2018 onward), I have also been granted access to a more recent wave of data regarding the large scale experiment object of component 1. In my original ERC proposal, I had explained that obtaining these additional data would have been very important to understand the long term impacts (10 years after the beginning of the experiment) of the introduction of reputational concerns. Although the analysis is still in its preliminary stage, the initial results are offering interesting insights on how a series of modifications in the procurement rules interacted with the reputational system. I am therefore optimistic that a second core paper with novel and important results will soon be ready.

For Component 2, I completed most of the quantitative analysis pertaining to two papers, both of which are core relative to this component. One paper is essentially the one extensively discussed in my original proposal: an analysis of how discretion is associated with corruption in the procurement of public works. This first paper, tentatively titled “Some new facts (and a little new theory) on the corruption of government procurement” (joint with Raymond Fisman, Paolo Pinotti, and Silvia Vannutelli) presents evidence of a nuanced behavior for which discretion is less used by public buyers more at risk of corruption, but when they use it the results are detrimental in terms of the likelihood of the contract going to firms potentially involved in corruption activities. We offer a simple model and a vast amount of new evidence on corruption in procurement that we believe will be influential for future academic and policy studies. We are in the drafting stage of the paper, but we have already started to present it (or committed to present it in the next few months) at academic seminars, including those at: Northwestern University - Kellogg School of Business, Boston University and Bocconi University. Earlier mentions of some preliminary results from this project where also presented at seminars at the University of Pisa and State University of Milan (Transcrime group), where I was invited to present the other paper emanating from Component 2 that I now describe. This other paper, titled “Corruption Indicators in Awarding Procedures: an Analysis of Italian Contract Notices” (joint with Cristina Giorgiantonio) is based on an in depth analysis of about 3,500 call for tender and awarding notices concerning public works for road constructions awarded by Italian municipalities. We identified a series of systematic vulnerabilities in how the calls are written and quantified how thes

Final results

For both components 1 and 2, the novel results that I obtained in the various papers described above already represent advancements relative to the state of the art. In terms of both understanding the role of reputational incentives in public contracvting (Component 1) and measuring the association between different types of procurements and the risk of corruption (Component 2), the results obtained are novel and fill gaps in the existing litwerature. My main goals for these two components is twofold. On the one hand, I will use the next months to promote and disseminate the results achieeved. In this respect, in 2019 I am planning to host one of the two conference events for which I requested founding. On the other, I will keep on working on the data associated with Components 1 and 2 with the aim of producing additional papers. Additional papers for both components are likely to represent novel and interesting findings as well. For instance, with respect to Component 1, the availability of data until 2017 allows me the to complement the earlier analysis performed over the period 2007-2010 with a long term perspective on the program. In terms of Component 2, instead, the richness of the data has not yet been fully exploited as there are various aspects deserving dfurther investigation, like the interaction between firms\'s technical and financial characteristics and their propensity to engage in corruption.
Finally, as regards Component 3, this is the component on which I will likely work the most in the following months. The difficult task of obtaining all the required data has been successfully accomplished with regard to the Italian case. In the following months, I expect to colose a first paper analysizing the evidence from the Italian data. At the same time, I will pursue a data collection effort to obtain similar data for other European countires.