The Consequences of Zero Tolerance

The chart above shows what happens when policy is based on a slogan. In this case “Zero Tolerance.” Procurement rules in both Peru and Colombia require that any public contract tainted by corruption be terminated immediately. As the Brazilian investigation into construction giant Odebrecht unfolded, it became clear that many projects to build highways, power plants, and other infrastructure projects in the two countries had been corruptly awarded.  Authorities in both countries then did what the law told them they must: cancel the contracts.

Most large infrastructure contracts in Peru and Colombia are in the form of Public-Private Partnerships (PPPs), and the immediate termination of a PPP can be enormously costly.  Not only to the firms that paid bribes to secure the contract, but to lenders, suppliers, and the hundreds of other contractors on the project who had no knowledge or involvement in the bribery scheme.  The greatest costs are likely be felt by the citizens of Colombia and Peru.  For as the chart shows, the consequence of zero tolerance is a halt to new spending for roads, power, and other essential facilities as investors and project developers shy away from the risk future contracts will be terminated for the tiniest of infractions by anyone associated with the project.   

Colombians and Peruvians may today be proud their governments are so tough on corruption neither one will tolerate a speck of it in any contract for infrastructure.  Tomorrow citizens of the two countries may have a different view: when power shortages mean the lights won’t come on and the failure to build new roads and maintain old ones produces horrendous traffic jams.  

Last week the World Bank hosted a presentation by Inter-American Development Bank staff where the issue of why “zero tolerance” is a good slogan but a bad policy was examined and means for addressing infrastructure corruption without producing the results shown in the chart was discussed.  A paper the IDB presenters recently published, the source of the figure above and the basis of their presentation, is here.   A video of the session here.  

16 thoughts on “The Consequences of Zero Tolerance

  1. Thank you for sharing Rick! Mexico is currently experiencing this very problem with a half-built 18 billion Airport project. President Lopez Obrador had a public vote on whether the project should be scrapped claiming that it was spoiled with corruption. The public voted to scrap it, but this of course had major consequences. The peso slid more than three percent against the U.S. Dollar after the project cancellation was announced. It is also unclear what will be done with the foundations that were already built for the airport. Here is an article with more detailed in case you are interested in learning more:

  2. Rick,

    As you know, I agree that the “zero tolerance” slogan, though rhetorically useful in some contexts, is not very helpful as a guide to anticorruption enforcement policy, in light of the inevitable trade-offs involved. (One of my earliest posts on this blog made that point:

    That said, I respectfully think that you’re commentary here is a bit oversimplified, and a bit too dismissive of these stringent laws. Of course these laws, whenever they’re actually invoked, have significant negative consequences. But that’s almost always true of punishments that are meant to deter. Putting someone in jail costs society money and doesn’t undo the harm that the perpetrator inflicted. Fining or debarring a large corporation harms innocent third parties (shareholders, employees, customers, etc.). Oftentimes the social costs of imposing these punishments far outweighs the social benefits (mainly the benefit of deterrence). But not always. And my concern about your post is that it doesn’t mention the (possible) deterrence benefit of a strict automatic-cancellation rule at all. So will these rules, and the cancellation of these projects, lead to worse infrastructure in Peru and Colombia overall, as you suggest? Well, maybe it will, for the reasons you suggest. But what if, in seeing this drastic outcome, going forward procurement officers and/or bidding firms are scared out of rigging the bids or inflating the contract prices in exchange for kickbacks, and as a result the efficiency of infrastructure investment goes way up?

    To be clear, I’m not a fan of automatic contract cancellation, or for that matter of automatic debarment, largely for the same reasons as you: It seems to me that the collateral consequences are really large, and also a similar deterrent effect might be achieved with an alternative penalty, such as a sufficiently large fine. (See: So we might not disagree on the bottom line. But I don’t think the issue is quite as simple–or the laws in Colombia and Peru quite as foolish–as you make them sound.

    • Dear Matt and Rick
      Many thanks for posting the presentation of our paper at the World Bank. In this paper we did not advocate for dismissing from the legal framework the possibility of cancelling contracts because of an event of corruption. We tried to show that when cancelation is your only choice, the negative consequences appear to outweigh the benefits. Moreover, it seems clear – as in the case of some of the countries mentioned – that innocent parties (firms, beneficiaries of public works and citizens) suffer direly from this choice. Another issue we did not explore relates to the legal costs associated with cancelation, but anecdotal evidence suggests that countries incur in significant legal contingencies. These contingencies do not surface in full at the time of cancellation, but typically years later when cases mature, many times in a different jurisdiction from the one where the execution of the contract took place. Again, many thanks for opening this discussion in your blog.

      • Fair enough. I was reacting to Rick’s summary in the post, rather to your paper itself, or the video of the presentation, neither of which I’ve yet had the opportunity to look at. I will check them both out as soon as I can!

        Again, for what it’s worth, I’m broadly sympathetic to your bottom line conclusions.

  3. Glad to see the post caught readers’ attention. Asseret, many thanks for flagging the cancellation of the half-built Mexico City airport. Another example where a knee-jerk reaction to the discovery of corruption in a large project is likely to be quite costly.

    Matthew, as Roberto explained, the policy advice that flows from his paper is not that one should never, ever cancel a project tainted by corruption. It is that a rigid rule mandating that government always cancel it – no matter what – is bad policy. For it will result, as you say, in instances where “the social costs of imposing these punishments far outweighs the social benefits (mainly the benefit of deterrence).”

    My favorite example of a government realizing the cost and, fortunately for its citizens, quickly reversing course comes from the United States. An Air Force investigation in the early 2000s had found that Boeing Company subsidiaries and several employees were in possession proprietary documents of a competitor pertaining to satellite launch vehicles. (The employees later pled guilty to stealing trade secrets.) Accordingly, on July 24,2003, the Air Force suspended the subsidiaries from receiving any contract to develop a low- cost method for launching satellites into space. A month later the Air Force waived the suspension and awarded Boeing a $56.7 million cost-plus contract to deploy a rocket to carry the Global Positioning Satellite System into space. The waiver was granted because the Air Force realized that Boeing was the only firm capable of “actually launching the GPS satellite” into orbit.

    The flip-flop on the Boeing suspension is described in an article by then U.S. Army Judge Advocate Major Jennifer Zucker in volume 13 of the Public Procurement Law Review. Major Zucker’s paper is one of several explaining how U.S. procurement rules deal with a contract under execution when it is discovered that the contractor has committed a corrupt act, either in connection with the contract being implemented or elsewhere. The tensions between the desire to punish the contractor and deter it and others from fraud and corruption and the government’s need to see a contract completed are described from several different angles: the government’s, the public contracting community, public interest groups, and the academy. For those readers (surely very few) who do not subscribe to the Public Procurement Law Review, the papers are available at:

    All are well-written and reward study.

    • Great, thanks for the clarification and the additional examples.

      The only thing I would add at this point (and here I’m picking up on what I said in that debarment post I linked to above) is that all of this makes me think that monetary sanctions on firms that engage in corruption should be much, much larger. The impulse to automatically cancel contracts and debar firms comes, in my view, from the (likely correct, and certainly plausible) instinct that monetary penalties–at least at their current levels–are not nearly high enough to provide adequate deterrence. If that’s true, and if it’s also true that other sanctions, like contract cancellation and debarment would have much worse collateral costs for third parties than do monetary sanctions, then the best way forward would be to raise the monetary penalties–perhaps by an order of magnitude. But I’m not sure if there’s any political appetite for that.

      (Of course the other thing we could do is put some actual human beings in jail. I’m for that too, though I think the issue is a bit more complicated than sometimes portrayed, for reasons I’ve discussed in other posts.)

      • I think we have a similar rule in Lithuania, leaving both options on the table – either terminating the entire contract, or applying alternative sanctions. In my opinion, in cases where there would be more harm for the public interest if the contract was terminated, alternative sanctions should be applied. For example, if the company in such cases would be obligated to pay damages AND a fine while also finishing the project, this would most likely have a deterring effect as well. In such cases, however, I would like to see the court taking such a decision after evaluating the situation and evidence.

        (The link to the paper IDB paper is not working right now and I cannot find it online, maybe this is addressed there?)

  4. In 2015/2016 the Colombian government learned that bribes had been paid by Odebrecht to ensure the adjudication of infrastructure projects in Colombia. We knew for sure that Odebrecht had paid 6.5 million USD to get at least one of those projects, and we were unsure of what other projects had been tainted by bribes. Therefore, the challenge for the government was to decide whether the contracts should be terminated, declared void, or if they could be assigned to a different contractor.
    We concluded that the contracts had to be declared void – this is, they never existed – because their object/purpose and cause were illicit (objeto y causa ilícita). This conclusion stems from a strongly held principle of continental civil law according to which, when a contract is entered into with illicit object or cause, it is deemed to be non-existent. Such a contract cannot be enforced, and things need to go back to their original condition. My point here is that the execution of that principle has nothing to do with the zero tolerance principle. It has to do with a legal principle embedded in thousands of years of history, that can be traced back to Rome.
    Thus, if the zero-tolerance policy changed (understanding by “policy” the scope and nature of political speech, and even some regulations on procurement or anticorruption), – say, loosening it – the legal consequences would probably remain the same. In fact, I would argue that while (in 2016-2017) the Colombian executive branch was trying to find a solution to facilitate the assignment of the contract, aiming to prevent the collateral consequences of its automatic termination, the Ombudsman took a more radical approach by petitioning to an Administrative Tribunal the termination of the contract, under the illicit cause/object argument. The Tribunal ruled in favor of the Ombudsman of the Nation, and ruled that the contract should be immediately terminated. A positive consequence was that Odebrecht was fined 400 million USD. But the negative consequences were huge. The works were paralyzed, thousands of jobs were lost, and the project will have to wait for at least two or three more years to be newly adjudicated and accomplished. Further, Odebrecht sued Colombia before an international arbitration tribunal, which still has not ruled a final decision on the case.
    In sum, changing the outcome that derives from the illicit object/cause principle, would require a profound change of continental law principles, that I wouldn´t expect to occur in the short run. Considering that such a principle has been in place long before Odebrecht, I would suggest that the reduction on PPP´s in the region is not a consequence of the zero-tolerance policy. Instead, a would suggest another explanation which I find plausible. Investment stops because corporations think governments are corrupt and because they do not believe they will enjoy fair procurement processes.

    • Camilo, thanks for the fascinating explanation. I have no background in civil law so I wonder if I can ask you to provide a bit more doctrinal detail. You stated that the Odebrecht contracts in Colombia had to be voided because their “object/purpose and cause were illicit (objeto y causa ilícita).” But, it seems to me that while the *cause* of those particular contracts being entered into was illicit (bribery), the object/purpose of the contracts was legitimate: i.e. to build roads, or bridges, etc. In my mind, an illicit purpose would be something more like a contract to perform a hit on someone, rather than the contracts at issue in the bribery case. Am I misunderstanding what object/purpose encompasses in the civil law tradition? Or is the test disjunctive, so that if *either* the object/purpose *or* the cause of the contract is illicit, the contract must be void ab initio? Very curious! Thanks!

  5. Thank you, Rick, for a really interesting post that gives us a lot to think about. I agree with Professor Stephenson that the high costs of project cancellations are the key factor making the non-discretionary cancellation rule a good deterrent. If the costs were on the lower side, the deterrent effect of the rule would not be very significant. However, the associated high costs also make me question the sustainability of the non-discretionary cancellation rule. Indeed, it might be the case that the “tomorrow citizens” of Colombia and Peru will have to live with the consequences of “zero tolerance” of corruption. However, I think that it is more likely that they will live in a country in which the “zero tolerance” policy will no longer exist.

  6. Dear Camilo and Jason, your comments are generally in line with our suggestions in the paper.
    There are, at least, two separate issues. The first one is whether your legal framework allows applying either remedy (termination as well as other measures). The second one, is whether this other measures can have a deterring effect or not. I believe this two issues may be dealt with separately. I also believe Camilo is right, most of the legal frameworks in countries with a continental legal tradition only admit the “zero tolerance” response. in order for our suggestions to work, the review of said legal tradition requires review. On the other hand, it is difficult to anticipate and predict costs of cancellations. The degree of execution, the complexity of the infrastructure and the particular subsector deserve particular analysis. Renegotiate fines or termination of a contract for a highway with 10% of execution is different from reviewing the dredging of a river with 80% of execution. The moment the company stops dredging, you are facing a different scenario in terms of your cost-benefit analysis when deciding whether to cancel, terminate or continue.

  7. Pingback: Even “Tough on Corruption” Proponents Should Worry about “Zero Tolerance” Rules | GAB | The Global Anticorruption Blog

Leave a Reply to asseret10 Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.