There’s been an interesting mini-debate over at the FCPA Blog about whether, or to what extent, corruption is partly responsible for the severity of the Ebola crisis in West Africa. Richard Cassin, the publisher and editor of the FCPA Blog, argued that it is. He made this argument initially in a post from this past August entitled “Ebola tragedy is also a story of graft.” He offered as evidence the following observations: (1) the WHO and other observers estimate that a very high percentage–perhaps up to 25%–of global spending on public health is lost to corruption; (2) the very high Ebola fatality rates in West Africa have been attributed in part to the lack of adequate intensive care facilities to administer the treatments; and (3) the countries hardest hit by the Ebola outbreak–Guinea, Liberia, Sierra Leone, and Nigeria–are reputed to be highly corrupt, as indicated by their very poor scores on Transparency International’s Corruption Perceptions Index.
Many critics who commented on Cassin’s initial post complained that the evidence offered did not in fact support the strong claim in the title that corruption has contributed significantly to the Ebola outbreak. In particular, the critics pointed out that: (1) the fact that a great deal of public health spending generally is lost to corruption does not actually tell us whether corruption was a major factor in the particular case of the Ebola outbreak, and (2) the low ranking of the affected countries on the CPI likewise–even if we concede that the CPI is a decent measure of actual corruption–does not indicate that corruption caused (in any significant way) the Ebola outbreak to be as lethal as it has been; at most it shows a correlation that might be explained by any number of other factors.
Cassin responded with a second post last month in which he rebutted the critics. He acknowledged that while one can never establish with “scientific certainty” that corruption has a causal effect on the severity of the Ebola outbreak, there is powerful circumstantial evidence that corruption is a “gateway” to this and other public health crises (as well as other problems like terrorism and crime), because it siphons off public resources. Cassin cites to a couple of research papers that purport to show that corruption in general has adverse impacts on public health, in particular because it adversely affects access to clean water and sanitation.
While I’m generally sympathetic to Cassin’s larger point, I think that the criticisms are fair ones. Here’s my take.
- First, the critics were right to take Cassin to task for emphasizing–in the initial post–the low CPI ranks of the hardest-hit countries as evidence that corruption was partly/largely to blame for the outbreak–and Cassin was right to back off of that aspect of the argument in his follow-up. Ebola is native to a part of the world that happens to have very high (perceived) corruption rates; we don’t have enough variation in corruption levels in the “hot zone” to make any inferences about how much of an impact corruption has had on the severity or lethality of the outbreak. And there’s obviously lots of room for what a statistician would call “omitted variable bias” — third factors that might affect both the disease burden and (perceived) corruption, like tropical climate or poverty. And to criticize this sort of inference isn’t to make unreasonable demands for “scientific certainty” — it’s to recognize that in a case like this, we can’t draw any meaningful inferences about the role that corruption played in the outbreak.
- Second, Cassin is on much stronger ground when he emphasizes the evidence that corruption has adverse impacts on public health by diverting money that’s supposed to be used for medicine, sanitation, clean water, and other health services and public goods. I agree with him that those problems are fairly well-established. (For more discussion of those and related points on this blog, see here, here, and here.) But I do think the critics are right to press Cassin on the fact that although there’s good evidence that corruption is generally bad for public health outcomes, there isn’t all that much in either of his posts establishing that corruption contributed specifically to the Ebola epidemic (either by creating conditions that were conducive to an Ebola outbreak, or by impeding the effectiveness of the response, or both). He does have one sentence in passing in the first post suggesting that West Africa lacks adequate intensive care facilities, but he doesn’t establish whether this is because the money for those facilities was stolen, or whether it’s because no investments in such facilities were made in the first place. And in the second post, Cassin hints that poor sanitary conditions and lack of access to clean water might have contributed to the Ebola outbreak, but he doesn’t come right out and say it — the paper he quotes appears to be about public health more generally — and I don’t know enough about Ebola to know whether that implied causal link is a plausible one.
So I guess my (tentative) position is that while there is indeed strong evidence that corruption contributes to public health problems generally, Cassin was perhaps arguing a bit too far beyond his evidence when he claimed (or strongly implied) that corruption was specifically a major contributing factor to the current Ebola crisis. It’s certainly plausible, given the general evidence he cites, that corruption may have contributed in some way, but there’s really nothing in either post that establishes any such link with any specificity.
By the way, one minor observation that tends to corroborate my take on this: One of the papers on the adverse impact of corruption on public health that Cassin showcases in his second post is by Professor Taryn Vian. Cassin’s post emphasized Professor Vian’s impressive credentials–but oddly neglected to mention that Professor Vian herself was one of the people who criticized Cassin’s first post, finding it to be “not  as factually based” as other FCPA Blog posts. I tend to agree with Professor Vian’s more nuanced assessment of the issue:
If there is any message about corruption which we can draw from the Ebola epidemic it is this: corruption is a public health risk factor. Corrupt practices in the health sector reduce the resources we have to respond to epidemics like Ebola. It also weakens public trust in government health systems, trust which we need to rely on in emergency situations. Anti-corruption strategies can strengthen public health systems and response networks so they are there when we need them.
That lesson is a fair one. But declaring, without more, that the “Ebola tragedy is also a story of graft” or that “graft [is] a gateway for Ebola” probably goes too far.