Partisan Science 2
Partisan Scientists in Public Service: The Strange Case of the Interagency Grizzly Bear Study Team
Dr. David Mattson
In Part 1 of this series, I painted a picture of corrupted science behind the current push to remove Endangered Species Act (ESA) protections from Yellowstone’s grizzly bears. On the face of it, there are other possible interpretations. But what makes the case for corruption particularly compelling is the extent to which this scientific endeavor is beset by a perfect storm of influences that in combination virtually guarantee subversion of the entire undertaking—even with participating scientists who have the best of intentions.
The ingredients of corruption
Any useful explanation of my claim necessarily starts with some basic observations about the human condition. Most fundamentally, all people are captive to their subjectivity, which axiomatically debars any of us from being truly “objective.” With that, we are prey to bias. This fundamental psychological truth makes us additionally prey to a phenomenon called “groupthink,” which arises when we surround ourselves with like-minded people who reinforce our prejudices and perspectives. Most of us are also notoriously vulnerable, directly or indirectly, to the sway of money, which becomes all the more potent when we are subsumed in a system that rewards us for bringing money in the door. Finally, given all of these powerful bias-inducing forces, science only truly progresses—or produces “reliable” results—when done under circumstances of the utmost transparency and openness, including ample provision for the attempted independent replication (or refutation) of any research results.
With these basics in mind, it is pretty easy to compile a list of hugely problematic circumstances that beset the government scientists producing the information being used by the US Fish & Wildlife Service (USFWS) and regional state politicians to advance the agenda of delisting Yellowstone’s grizzly bears. Seven toxic ingredients…
1. Self-delusion: Every Interagency Grizzly Bear Study Team (IGBST) scientist that I’ve known, past or present, has at some time publicly claimed that they are “objective”—sometimes fervently and defensively so. As often as not, they also claim that they merely “let the data speak for themselves.” And I truly believe that they truly believe what they say. This certainly holds for the current IGBST Leader, Frank van Manen, as well as his predecessor, Chuck Schwartz.
But putting beliefs aside, whenever I hear anyone make such claims red flags go up. Rather than providing reassurance, I typically end up being convinced instead that the person making such assertions has little self-reflective capacity—which is a big problem. I contend that the only way any of us can attain even some measure of objectivity is by being acutely aware of the inevitable biases that arise from our immersion in subjectivity. Without such awareness, I suspect that people such as Frank and Chuck are (or were) all the more vulnerable to the biasing forces that pervade their professional practice. And, in fact, “confirmation bias” in research is such a widespread phenomenon that it has spawned its own body of scientific literature.
2. Groupthink: Put succinctly, the IGBST’s scientists and their invited collaborators are pretty much holed up in a private clubhouse with guns pointed outward ready to blast away at anyone who questions their science. A recent critique of the IGBST’s primary method for monitoring Yellowstone’s grizzly bear population by Dan Doak and Kerry Cutler was met with a putative rebuttal by the IGBST that, at least to my eye, relied more on the cumulative biomass of authors than on any particularly cogent arguments. Doak and Cutler’s rebuttal of their rebuttal was aptly entitled “Doth Protest too Much.”
The point of all this being that the IGBST and its helpmates have existed for years in a sort of echo chamber, which is perfect recipe for “groupthink” typified by conformity and an inability to imagine alternatives. In fact, sociologists who study scientists have long noted the prevalence of groupthink among research teams or lineages, most compellingly Thomas Kuhn in his work on scientific paradigms. Of relevance here, I see ample evidence of entrenched groupthink among the IGBST’s research group, made all the more problematic by the aggressively defensive response of this cabal to any perceived threat to their scientific hegemony and the deliberate exclusion of any public critics.
3. Monopoly: Under normal circumstances a research team afflicted by groupthink and populated by scientists prey to delusions of objectivity wouldn’t be a particularly big problem for the advancement of our collective understanding of the world. But this holds true only when all who might be interested have free access to the same physical system (as in physics or chemistry) or, in the absence of opportunities for exact replication, have unimpeded access to the same data, which holds for most wildlife field studies, including those of relevance to Yellowstone’s grizzly bears. Or put another way, any kind of monopoly on inquiry is antithetical to scientific integrity and the production of reliable information. Period. (The literature substantiating this basis point is beyond summary, but see Philip Kitcher's 2002 book for an interesting perspective).
In the case of Yellowstone’s grizzly bears, the IGBST maintains a stranglehold on virtually all of the relevant data (Fischman and Meretsky provide context and background on this). There is only one Yellowstone grizzly bear population, only one dataset of direct relevance to management of these bears, and only one research team privileged with access to it. And this monopoly doesn’t exist by chance. The IGBST’s scientists, aided by the USFWS, have fought tooth and nail to maintain this monopoly and, in the one instance where a Freedom-of-Information-Act (FOIA) request forced them to disgorge some data, they have since worked hard, politically, to dismiss the resulting science. Even more egregiously, the current IGBST Leader, Frank van Manen, publicly stated that he needed to perpetuate a monopoly to advance his career (follow this youtube link). Mind you, all of the data as well as Frank himself are paid for with taxpayer dollars. A monopoly is a huge problem.
4. Lack of replication: I invoked replication immediately above, and I did so because replication of research, or the opposite (lack of replication) is the foundation of any scientific progress—or reliable information. And attempted replication is best undertaken by an independent team of preferably cantankerous and critical scientists, or at least scientists who have a burning desire to genuinely understand how the world works. Peer review is sometimes invoked as THE cornerstone of science, but it isn’t (more on that immediately following). Independent attempts at replication are.
In the case of Yellowstone’s grizzly bears, where you have messy data collected under messy field conditions, this amounts to independent analysis that deploys alternative models and statistical methods given that modeling is the only means by which useful insight into environmental change—e.g., loss of whitebark pine—can be achieved. But there are numerous possible models parameterized by a number of possible statistical methods, and none are “correct.” In fact, different models will give you different results, even using the same data and asking the same question. Illustrative of this, a recent article in Science featured what happened when 29 research teams were given the same dataset and asked to determine whether a particular effect was “significant,” not unlike loss of whitebark pine. You got essentially 29 different answers, ranging from non-significant, to slightly significant, to highly significant. In application to any complex system, there are multiple possible models and multiple possible outcomes and, in any situation that is contested and consequential, as many of these models and results as possible need to be on the table for consideration.
My point? The deliberately perpetuated monopoly that I described above debars replication, diverse analytic perspectives and, with that, any reliable science.
5. Politicization of peer review: But what about peer review? This question matters because the USFWS, the IGBST, and its parent organization, the US Geological Survey (USGS), constantly assert that peer review is not only the cornerstone, but also the guarantee of quality science (for example...). In other words, invocation of peer review is a primary means by which these invested federal agencies skirt the monopoly issue in the case of Yellowstone’s grizzly bears. Don’t worry about independent inquiry; we’ve got the bases covered with peer review—as an article of faith. And so on. But do they?
As it turns out, the research on peer review shows that it guarantees nothing (the literature on this is vast, but see these two emblematic papers). The best that any half-way objective person can say for peer review is something similar to what Winston Churchill said about democracy: “Democracy is the worst form of government, except for all the others.” Hardly reassuring. One could say that peer review serves roughly four functions: censorship; improvement; maintenance of identity; and political advantage. On the censorship front, peer review works just about as good as flipping a coin for detecting and excluding egregious errors. Insofar as identity is concerned, it clearly constitutes a cleansing ritual that helps sustain the self-ascribed elite status of most scientists. It works perhaps best as a means of improving scientific endeavors and products. But in the case of Yellowstone’s grizzlies, it is clear that peer review is used primarily for political purposes by the involved bureaucracies, primarily by making inflated claims for its efficacy as a basis for maintaining control over scientific inquiry, and, with that, advancing the delisting agenda.
6. Corrupting money: Almost all scientists scramble for money to do research, which holds for IGBST scientists as well. They have their salaries more-or-less covered by agency budgets, but rely on outside money to hire consultants, recruit other outside help, and execute many field studies. But with money comes the potential for often tacit corruption, especially when the research is highly politicized, the money fronted by an entity with a vested interest, and where there is explicit provision for this party to have a seat at the table when shaping research questions and vetting research results—all of which holds true for the post-2007-litigation science produced by the IGBST. Chris Servheen of the USFWS ponied up most of the taxpayer dollars to support this research, along with a mandate to target the whitebark pine issue at the heart of the Courts’ 2009-2011 rulings, and a timeline for cranking out the science in support of a delisting rule prior to the 2016 elections (for example, see the IGBST study plan which explicitly targeted whitebark pine and was signed off on by Servheen). Moreover, email traffic obtained through a FOIA shows that Dr. Servheen was deeply involved in crafting, reviewing, and flogging this science. As I described last week, Recovery Coordinator Servheen is hardly a neutral civil servant. At a minimum, he pretends to play legislative politics.
7. Business models of science: And if this weren’t bad enough, the IGBST’s federal scientists are subsumed in an agency—the USGS—that has adopted a business model of science. Which means what? For one, those (such as the USFWS’s Chris Servheen) who pony up money for USGS scientists are referred to as “customers.” This implies that the IGBST is producing a product to satisfy the customer, which is, in fact, a highly prioritized outcome in the USGS. Amazingly enough, the USGS sends out questionnaires to “customers” asking whether they are satisfied, which is to say, asking whether the science served their purposes? (For example, follow this link). In a case such as Yellowstone’s grizzly bears, all of this predictably amplifies the corruptive potential attached to the money that the IGBST received in recent years from the USFWS, which, like all government bureaus, has its special interest agendas.
Perhaps for some readers who are sold out heart and soul to the capitalist model of everything, the notion of a business model is prima facie a good thing. But it is actually a bad thing when, in fact, the “customer” of a federal science agency is (as it should be) that amorphous blob of people called “the public,” and where the ultimate purpose being served is fulfillment of the public trust, not the expectations of a privileged few (see this recent insightful paper by Adrian Treves on the public trust doctrine with reference to predators such as grizzlies).
What to do? Some recommendations
What to do about all of this? Perhaps seven antidotes for seven toxic ingredients…
First, without intending to sound flip, I would recommend that leaders of the USGS, and most members of Congress for that matter, go back and take a civics course, and perhaps even bother to read some of the original texts authored by Adam Smith. A little Max Weber might not hurt either. And then they should root out the corruptive business model—“pay to play”—that has taken hold in the USGS and other federal government bureaus.
Second, the monopoly that plagues Yellowstone’s grizzly bear science absolutely needs to be broken; which means that all of the data, bar none, should be made freely available to scientists and citizens for multiple independent inquiries. The argument advanced by the USFWS that these data need to be sequestered to protect Yellowstone’s grizzlies from poachers should be recognized as the ruse it is.
Third, the move to delist Yellowstone’s grizzly bears should be put on hold until there is, in fact, some modicum of reliable science upon which to base deliberations. As is, I would argue that there is none to be had, for all of the reasons listed here.
Fourth, it might not hurt for the current crop of IGBST scientists and their compadres to sit through a class or two devoted to social-psychology and, better yet, the sociology and philosophy of science. There is clearly a knowledge deficit on this front.
Fifth, it also might not hurt for all of the involved scientists to dig into the research that has focused on the efficacy (or inefficacy) of peer review, the critical importance of replication, and the problems of model-based inquiry into messy data. The current naiveté and faith-based approaches to scientific practice are, quite frankly, staggering.
Sixth, it might improve matters if some of those in the USFWS who have overstepped the bounds of ethical behavior were appropriately chastised for their overreach. Clearly, people as high as Dan Ashe and as low as Chris Servheen don’t have a good grip on what is and is not appropriate behavior for a civil servant in an administrative agency. Perhaps clearing house might be in order.
Finally, seventh, I strongly recommend humility and caution on the part of those managing and studying Yellowstone’s grizzly bears. No matter how good our science, we will always be dealing with a huge amount of uncertainty rooted in the massive global changes afoot.