Henry Kissinger once quipped “in academia, knives are so sharp because the stakes are so small.”* This insight is largely correct. In the absence of large pecuniary rewards, academics tend to receive remuneration disproportionately in the form of prestige and reputation. One the positive side, these can be achieved by hard work and brilliance. On the negative side, by manipulation, conspiracy, and backstabbing.
In the modern academy, peer review plays a major role in the establishment of reputation. On the positive side, it can serve as an important tool in identifying errors, poor reasoning, and weak exposition. On the negative side, it can serve as a mechanism for enforcing consensus even around dubious concepts, punishing dissent, and protecting incumbent interests. As is the case with many social mechanisms, it tends to be conservative in nature, and hostile to true innovation. It tends to filter out the different, with mixed consequences. Bad work is often different, but not all that is different is bad. Moreover, since in its operation it tends to empower the avatars of prevailing views, it can be used and manipulated to punish or otherwise quash those that challenge said views.
The Hadley CRU emails provide a window onto the often shabby operation of this process. The Wegman report characterized the historical climate reconstruction community centered on Michael Mann as a clique. The members work with one another and review each others’ work. As a result of the insularity of this community, and its hostility to outsiders, the published peer reviewed work in the area all tended to reinforce the views of its members. There were many different published papers, but because they were written or reviewed by the same people, often using the same data, these papers were not independent, in the mathematical/statistical sense. A mere count of the number of papers gave a misleading view of the actual number of independent data points. In statistical parlance, they gave an exaggerated view of the “size” of the tests in favor of the AGW hypothesis.
Moreover, by exerting undue influence on the peer review process, the clique suppressed dissenting views and the publication of truly independent work. It then added insult to injury by condescendingly dismissing contrary work as not being peer reviewed, or being published in the leading journals (over which they exerted their influence). It was a scientific Catch-22.
The Hadley CRU emails show just how right the Wegman report was. Indeed, they demonstrate that if anything, “clique” is an understatement. “Cabal” or “mafia” is more appropriate. The enforcers of consensus resorted to threats, slander, and crude insult to protect their precious reputations and squelch opponents. They reveled in the death of one gadfly and expressed the desire to inflict violence on others who had the temerity to present papers that dared suggest the Emperors had no clothes.
In brief, the climate reconstruction mafia behaved exactly as Kissinger described, wielding their sharp knives.
The problem is that, unlike the run of the mill academic disputes described by Kissinger, in this matter the stakes are anything but small. Indeed, they are staggeringly large. In such matters, typical academic failings are intolerable. There needs to be a different mechanism for evaluating and appraising such policy-relevant research than reliance on a self-selected, self-reinforcing group of interested individuals.
This will present a great challenge. When the political consequences of science are so great, it will be very difficult to keep politics out of science. Where to begin? The recommendations laid out in the Wegman report (emphasized in a comment by Mike Giberson) are a good place to start, but likely additional steps are required.
Follow up (noon, 11/27/09). The normally sensible Megan McArdle states that “Sexing up a graph is at best a misdemeanor.” I think she means “at worst,” but regardless, in context it is clear that she means to minimize the importance of the data and methodological abuses employed to create the hockey stick and related graphics. I think this is incredibly wrongheaded. There is NO excuse for lying with statistics in this way. And, as the post points out, the stakes in this issue are immense. The “sexed up” graphs have been used repeatedly in order to advance an agenda that will have enormous economic, and perhaps environmental, consequences. (The economic costs are indisputable. The environmental consequences are far less certain, because (a) as the debate shows, the scientific basis for AGW is on shakier ground now, and (b) even if AGW is correct, the effects of the measures currently contemplated may be very small indeed.) Given these consequences, the expected cost of any distortion of the evidence is extremely large. What’s more, the emails provide strong evidence that the sins go far beyond tarting up an inconsequential graphic. These sins plausibly include, based on the code extracts and comments particularly, wholescale fraud.
* I have used this line from time to time in the past. For instance, I used it in response to a lawyer’s line of questioning about what people have said about my academic qualifications during what was at the time the largest fraudulent conveyance claim trial in US bankruptcy law history. The judge chuckled, and said something like “that’s true of judges councils too” and suggested that the lawyer try another line of questioning. So, you see, I’m environmentally responsible: I recycle!