Curbing corruption: what can be done?

Anti-corruption, Uganda, economics

At a time when the aid sector is battling to maintain its reputation for spending public money well, corruption in public procurement – the purchasing of public services such as health, education or infrastructure – particularly when funded by development aid, is a delicate topic.

Liz David-Barrett (Senior Lecturer in Politics, University of Sussex) and Mihály Fazekas (Post-Doctoral Research Fellow, University of Cambridge), researchers on a British Academy Anti-Corruption Evidence programme-funded project, are working on an innovative methodology for analysing big data from major donors to help identify indicators of corruption in aid-funded procurement. In conversation with Antonio Andreoni (Senior Lecturer in Economics, SOAS) and Luca Tasciotti (Lecturer in Economics, SOAS), they presented their findings at a seminar event at SOAS University of London, and sparked a debate over how feasible it is to use top-down controls to prevent or punish this kind of corruption.

Antonio: In both developed and developing countries public procurement plays an important role and significant economic resources are involved. How does corruption manifest in public procurement?”

Liz: “Procurement involves the state (the buyer) allocating state resources (public money) to external actors (usually private companies, but sometimes NGOs and SOEs). The ideal scenario is that state actors make those decisions about how to spend public money, and to whom to award contracts, on the basis of setting out clear criteria for what is needed and then choosing the company which can supply the goods, works or services in the best way and at a reasonable price. But those decisions involve a lot of discretion and often individuals on both sides of the transaction have an interest in distorting or manipulating the process so as to control it for their own private benefit.

“The procurement process is complex and there are many ways in which it can be manipulated.

“It can happen right at the start, when public officials are assessing the needs of the public authority and working out what to spend money on. It can happen when they write the tender specification: if they specify it too narrowly, they can fix it so that only one company (which they favour, for some corrupt reason) will find it worthwhile to bid or is likely to win. Alternatively, the selection panel that chooses the winning bidder might be subject to corrupt influence. And sometimes the contract seems to have been awarded fairly, but the winning bidder then inflates the costs or the authority refuses to pay unless the company pays a bribe… And these are just a fraction of the ways that the process can be corrupted!”

Luca: Public procurement is known to be high-risk for corruption, but evidence is frequently anecdotal rather than accurate. Data on this issue in Europe is scarce and not strong enough to be representative – what strategy did you adopt to capture accurate data on this phenomenon?”

Mihály: “Thanks to extensive transparency laws in this field, and recent advances in computer science, we were able to collect data on over 17 million publicly announced contracts and the tendering processes they were awarded in. Such an enormous dataset allowed us to carry out elaborate text mining and to identify high risk constellations in the market, such as only one bid being submitted in an otherwise competitive market, or where the tendering terms seem like being tailored to a specific company. Using this and building on anecdotes and court records we were able to identify corruption risks at a large scale.”

Antonio: “The statistical analysis you undertook suggests that a couple of factors, such as short advertisement periods and single bidder tenders, where only one company submit a bid for the public tender, are good predictors of potential irregularities in the procurement process: the shorter the advertisement periods and the fewer the bidders, the higher the chance of corruption.

“However, a challenge is that in developing countries industrial sectors are pretty underdeveloped and often only very few firms have the capability required to bid for large procurement projects. In these countries, cases of single bidding tenders might be the result of this lack of a sufficient number of competitive firms, more than the result of irregularities. Moreover, the few firms operating in these countries might decide to collude and manipulate the bidding process by giving an appearance of a multi-bidding tender process. How is your research addressing some of these dynamics?”

Liz: “Our research has two parts. Our big data analysis looks at overall patterns in more than 100 countries and over almost two decades. But we have also conducted qualitative analysis in some countries, to try to learn more about how to interpret this big picture.

“One of our aims is to understand how the context – including the nature of local markets and the types of corrupt practices – might mean that some red flags are more useful than others.”

“So, in Tanzania, for example, our research on the ground suggests that ‘signature period’ and ‘cost overruns’ might be better able to capture corruption patterns than ‘single bidding’, since tender manipulation appears to be opportunistic and public-sector-led. In Ghana, our research suggests that procedure type is important for capturing corruption risks in procurement, since there is extensive use of sole-sourcing on spurious grounds to facilitate favouritism in the allocation of contracts. ‘Single bidder’ is also a relevant red flag in Ghana, while local corruption practices are also reflected in ‘cost overruns’, ‘transparency/publication of tenders’, and ‘tax haven registration’.”

Mihály: “We often narrow down the analysis to tendering methods which are competitive by intent such as international competitive bidding procedures run according to World Bank rules. In such cases, even if local supplier markets are weak, the very design of the process aims to draw on global suppliers to make the process competitive.”

Luca: “Your most recent work looks at corruptions patterns in aid projects of 500 million US dollars and more, funded by the World Bank, Inter-American Development bank and EuropeAid. How do changes in the donor rules and context-specific political factors affect corruption in aid projects?”

Mihály: “Our findings confirm that widening access to public tenders through electronic procurement and advertisement and increasing donor controls through more extensive audit rights work very much in line with what theory would predict. As a result of these reforms, single bidding in public procurement has decreased from about 22% to about 18% globally. And this effect is entirely produced in countries with low to medium state capacity, whereas the effect of these reforms is zero in high state capacity developing countries like Brazil.

“Before we get carried away, no technical fix will, on its own, alter the underlying political economy constraints on anti-corruption.”

“However, they could play a powerful role in locking reforms, as was the case in Ukraine’s revolution, where the changing political economy led to and was facilitated by measures such as the ProZorro electronic procurement platform, which allows transparency across the public procurement process.

“Thinking more broadly, our research has also shown that excessive bureaucratic controls in public procurement act as a barrier to entry in these markets, effectively making corruption, as well as collusion, easier to organise and maintain. This suggests that decreasing bureaucratic controls and increasing output/impact monitoring may lead to lower corruption – even though this sounds counterintuitive in light of many traditional anticorruption recipes.”

Liz: “Procurement is so prone to corruption because it is a major way in which public money is allocated. Tackling corruption in this area cannot be achieved without considering the wider political economy in a given context, such as how power is distributed and whether institutions have autonomy from politicians. Building a strong institutional oversight ecosystem can help to reduce the opportunities for manipulating procurement processes for private gain, and can encourage people to hold officials to account if they see evidence that only government cronies are winning contracts or that contractors are failing to provide services to the standard required.

“Greater transparency about procurement data can be an important part of that. Not just because the public can see what is going on, but because contractors themselves can put pressure on governments to clean up procurement. I met a number of construction engineers in Tanzania and Ghana who are proud of the high standards to which they work, but very frustrated about losing contracts to inexperienced contractors who win – they perceive – because of their political connections. The private sector can also be an important force for change in this area.”

What do you think? Can changes to procurement such as single bidder tenders help reduce corruption? What other anti-corruption strategies would you suggest to reduce the vulnerability in this key public procurement sector? Let the Anti-Corruption Evidence Research Programme know your views.


Want to learn more?

The Department of Economics at SOAS offer both undergraduate and postgraduate programmes in Development Economics.

BSc Development Economics offers students a thorough grounding in economic theory, analysis and policy, while studying the economics of a wide range of developing and emerging countries.

Students on the MSc Development Economics programme will complete courses on macroeconomics, microeconomics, econometrics and growth and development, plus have the chance to take a range of modules to study the economics of specific regions.


Find out more


Share this post