What Should be Disclosed…
The ongoing kerfuffle over Soon’s seeming ethical transgressions brings up a good point: the science being produced by corporate money can be perfectly good science. As a post in RealClimate notes, it is not the source of funding that matters but whether the science is any good. That post argues that Soon’s science wasn’t very good and so scientifically his ethical problems are irrelevant.
But is that a naive view?
On one level, there is a level of trust that scientists share that when a scientist reports observations, that those were indeed the observations. The interpretation could be wrong, there could be issues with how the observations were made, etc., but the basic notion is that there really were those observations. Now, is that trust violated if the scientist has lied about his or her funding sources?
On the next level, it has become apparently that the use of non-disclosure agreements in medical science has hidden results that were unfavorable towards drugs or chemicals, which has led to increased calls for the results to be made available regardless of a clinical trial’s outcome. This kind of tool has been used to block medical analysis of impacts from oil and gas development, it would seem. If the fossil fuel industry is distributing research money with NDAs like those, then it is possible that some perfectly fine research is going unpublished because the funder can block it. Is this actually happening? Maybe we’ll learn from the Soon affair if such tools are being used, or if Soon, on his own, decided to hide his funding sources. And it is happening with studies of seismicity associated with injection wells: industry-funded seismic networks are not releasing their data; what little comes out of such projects could well be cherry-picked; we don’t know because we don’t see the raw data. And, rumor has it, recently industry has used its funding of seismic monitoring to punish some who have said things industry dislikes: management of a network in the Salton Sea was transferred to a private company willing to submit to an NDA after the academic operators spoke out about how industry could be triggering seismicity. So this should be a real concern.
Long ago NDAs were a means for those in industry to share information that was simply a material advantage with academic studying something else. So a seismic profile created at a cost of a few million dollars couldn’t simply be released or other resource companies would gain access to it for free; usually these profiles could be shared if the exact location was not published or some parts of the profile were not published. This allowed researchers access to materials they otherwise couldn’t see. Now, though, it seems NDAs are used for the opposite purpose: to be able to control what researchers say in any way, shape or form–not only about materials provided by industry, but any results those researchers create that might have no dependence on competitively important materials provided to them.
What we need is a total change in industry support. It should be enough for companies to be steering the direction of many research projects simply by what they choose to fund; controlling what comes out of those studies should simply not be part of the program. All data needs to be publishable and there should be no veto power of industry on that research. Anything less means that lots of folks will continue to view industry-funded research as tainted, and so controversies like Soon’s the the resulting black eye for any researchers accepting industry money–or even taking positions that might seem to favor industry in some way–will continue to emerge and complicate scientific discourse on controversial topics.