What Does “Zero Tolerance” of Corruption Mean? A Comment on Labelle

Rick’s last couple of posts (here and here) critiqued Bill Gates’ claim that, because corruption in development aid projects as relatively small-scale (allegedly around 2%), it’s therefore a manageable “tax” on aid. Rick asserted (correctly, in my view) that the corruption problem is much bigger, and that the 2% figure is essentially a made-up number. Over at the Huffington Post, Huguette Labelle, the Chair of Transparency International, also responds to Gates. Most of what she says is pretty standard (which doesn’t mean it’s not right). But near the end of her post, she takes a striking position that’s worth thinking about a bit more critically: She argues forcefully against “[a]ccepting low levels of corruption as a pragmatic fact of life,” and instead advocates “[z]ero-tolerance for corruption.” This is quite different from Rick’s argument; Rick pointed out that corruption in development aid projects is not in fact low-level. But Ms. Labelle’s “zero tolerance” stance implies that even if Bill Gates were right about the facts, he would still be wrong in his conclusions. “Zero tolerance” of corruption certainly sounds good. But what exactly does it mean?

One possibility is that “zero tolerance” of corruption just means that we should always view corruption as undesirable. But even if we put aside for the moment the fact that certain forms of corruption may be justifiable (like bribing a Nazi official to escape a concentration camp), this doesn’t seem like a very productive understanding of “zero tolerance.” After all, Gates seems to agree that corruption is bad (he doesn’t “tolerate” it in the sense of thinking it’s not a problem at all).  And his view is probably the norm. So I don’t think that’s what Ms. Labelle means when she talks about a “zero tolerance” approach to corruption.

A second possibility—which takes the “zero tolerance” language literally—is that all feasible measures to minimize corruption must always be used. But that can’t possibly be right. At some point, the costs of fighting corruption exceed the costs of the corruption itself–perhaps by a lot. After all, the only surefire way to reduce corruption in foreign aid programs to zero would be to eliminate the programs themselves. Even stopping short of that extreme, anticorruption measures consume scarce resources that could be used for other purposes. In this, corruption is no different from any other bad thing that we would like to eliminate. We may say we have “zero tolerance” for murder, child abuse, disease, discrimination, and so forth. But in practice every society “tolerates” all of these horrible things–not in the sense that we ever view them as acceptable, but rather in the sense that we don’t devote all of our resources to fighting them, at the expense of everything else. So I don’t think this is what Ms. Labelle can mean either.

Which leaves a third possibility: perhaps “zero tolerance” isn’t meant as a (banal) normative position or as an (unrealistic) policy prescription, but as an attitude. Mr. Gates may agree with Ms. Labelle that corruption is bad. And Ms. Labelle, despite her rhetoric, may agree with Mr. Gates that just because some amount of corruption exists in an aid project, that’s not a reason to cancel the project or to spend all available resources on squeezing out the last little bit of corruption. The real source of disagreement seems to be that, from Ms. Labelle’s point of view, Mr. Gates just doesn’t take corruption seriously enough. “Zero tolerance” means, not literally zero tolerance, but rather making anticorruption a high priority. That’s my best guess.

I can sympathize—I’m more with Labelle than Gates on this one—but I do worry a bit about the “zero tolerance” language. It sounds good, but there’s a risk that strident, almost self-righteous rhetoric about “zero tolerance for corruption” can obscure the fact that there really are difficult and complex trade-offs involved in fighting corruption, and that the benefits of various anticorruption measures always need to be compared to their costs.

6 thoughts on “What Does “Zero Tolerance” of Corruption Mean? A Comment on Labelle

  1. Dear Matthew,
    just a couple of thoughts to follow up on this interesting post, as my U4 colleague Nils Taxell and I have recently published a paper on the concept of Zero Tolerance to Corruption as applied by Donors

    In short, I would say that we agree with most of your conclusions. There is definitely lack of clarity as to what it is meant by ‘zero tolerance against corruption’; in fact, this was one of the starting points of our analysis (on which those interested can also see http://tcr.sagepub.com/content/11/2/221.short).

    The working definition is that ZT policies fully investigate, prosecute, and sanction of all instances of a certain rime, no matter how minor.The question is whether this concept can be applied to corruption.

    We conclude that this expression is largely used as a ‘PR tool’, but that when rubber meets the road it is all but impossible to put in place a full fledged zero tolerance anti-corruption policy, because, among other things, the costs would be greater than the benefits (as, by the way, Klitgard noted in his classic ‘Controlling Corruption’). The consequence is that in practice the ‘zero tolerance’ commitment needs to be applied with a certain degree of flexibility and realism and depending on the context and circumstances, but that the ZT rhetoric can still be useful to symbolize strong resolve against corruption.

    However, I would disagree with you when you compare ZT anti-corruption policies to ZT designed to counter other types of crime such as murder. I think part of the confusion is exactly that the ZT narrative comes initially from violent crime. Now, in the case of violent or deadly crimes, such as murder, or violations hat result in deaths such as drunk driving, one could argue that the cost of reducing the violation to or close to zero is eclipsed by the importance of saving human lives. The situation is quite different for corruption, where other considerations come into play.

    Thoughts and comments on our paper are welcome.
    All the best,

    Francesco De Simone
    U4 Anti-corruption Resource Center

    • Francesco,

      Thanks very much for the thoughtful and detailed reply. I hadn’t seen the paper you did with Taxell. Having looked at it, I can see that you guys took what was for me a gut reaction and delved much more deeply. I definitely recommend it to GAB readers out there.

      I think we’re very much on the same page. “Zero tolerance” may accurately describe the policy of a particular overseer or enforcement institution, in that it captures the policy that all violations, no matter how small, will be punished. Maybe that’s a good policy, maybe it’s a bad policy, but it’s at least a coherent and implementable policy. But as a broader philosophy of setting policy priorities — where we’re not just talking about what to do with instances of bad conduct we uncover, but how we should allocate resources to activities that create the risk of the bad conduct, or may lead to discovery of bad conduct, ZT can’t really be taken literally, because we never do (and probably never could) try to minimize the bad conduct in the way that ZT rhetoric implies. Maybe the rhetoric is helpful, maybe not (and you and I might differ slightly on this), but I think we’d both agree we don’t mean it literally.

      On the comparison of ZT for corruption with ZT for other crimes (like murder), here I actually think the comparison is apt, and may highlight the different senses in which the ZT language might be used.

      Do we (or should we) have ZT for murder in the sense that any murder that law enforcement discovers leads to investigation, and, if the evidence is there, to prosecution? Yes.

      Do we (or should we) have ZT for murder in that we allocate 100% of our resources to preventing and prosecuting murder, and that we make ever single policy decision exclusively to reduce the risk of murder? No.

      So one sense of ZT for murder is plausible; the other isn’t.

      Maybe my original criticism of Labelle can be understood in the following terms: She seems to criticize Bill Gates for not embracing a “ZT for corruption position”–but Gates DOES seem to embrace ZT in the first sense (he says explicitly that every time we identify corruption in a project, we should act to eliminate it and punish those responsible). Hence, I interpreted Labelle as endorsing “ZT for corruption” in the second, broader sense–and that’s where I run into trouble, because I don’t think that understanding of ZT is tenable.

  2. I agree with your last reflection differentiating the two interpretations of ZT.

    Another aspect to consider: ZT policies are largely based on the so-called ‘broken windows’ theory of law enforcement, which more or less argues that even small violations should be repressed swiftly, because they can eventually lead to more serious ones (interestingly, the first proponents of this theory say it has been misunderstood and has led to policies that were not intended). This theory was famously applied in NY under Giuliani and is commonly argued to have resulted in a significant decrease of violent crime in the city (although, as you know, this is a hotly disputed claim).

    One question we had in mind when drafting the brief was whether this concept can be transferred to corruption. In other words, is there evidence that petty bribery, for example, can eventually lead to grand corruption? Because if so, then perhaps applying ZT policies against corruption may be a viable avenue. Of course this is a greatly simplified framework.

    Interestingly, we found little research and no evidence that petty bribery and grand corruption are correlated and ended up leaving this entirely out of the brief. The gut feeling is that they are not, as countries with little petty corruption, such as most OECD countries, still have numerous instances of political or grand corruption. Nonetheless, not much work has been done in this area, and this could be an interesting field of further analysis, perhaps for a PhD thesis.


  3. Among those who advocate tolerating “small-scale” corruption, has anybody described a threshold beyond which a corrupt act is no longer small-scale? For example, is a 5% “tax” on aid efforts too high? If so, should aid officials aim to bring that figure down to the 2% that Bill Gates described? Or does everyone who tolerates small-scale corruption just go with a “smell test”?

    • Raj, I think your question highlights a distinction between zero tolerance of corruption writ large and zero tolerance of identifiable corrupt actors that speaks to two different operational principles. One consequence of Matthew and Francesco’s understanding of the ZT attitude seems to be that there is no de minimis threshold below which organizations or enforcers should give a pass for individual acts of corruption – there is no ‘harmless’ corrupt act, regardless of scale and impact. But this is different from Gate’s pragmatic stance that, because resources are finite and not every act can be uncovered, corruption will inevitably comprise some portion of the broader waste and leakage in program costs.

      With this in mind, ZT may be less useful as an allocational principle for enforcement agencies than as an organizational principle for donors, contractors and public ministries setting their own institutional standards. Zero Tolerance toward individual offenders within organizations sends a message of high expectations and equal treatment, regardless of rank or connectedness, that are part of a sound ethics and compliance policy. In contrast, ZT in the selection of projects or funding decisions is likely not practical without shutting down most programs, though it may be possible to incrementally extend high expectations to outside parties through pass-through provisions and the like. Framed this way, the ZT attitude is not about irrationally devoting all efforts to extinguishing corruption: It’s about sticking to principles in the breach, regardless of whether the stakes are small or large and even where there may be a monetary or reputational cost to the donor or an effect on the success of individual projects. ZT advocates would say that there is a long-term value to boosting integrity, both real and perceived, that would justify these costs.


Leave a Reply

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

WordPress.com Logo

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

Twitter picture

You are commenting using your Twitter 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.