A small number of truly horrible scandals in medical research led to the development of the human subjects protection system. No one denies that the abuses were real, or seriously suggests that medical researchers had shown themselves capable of collective self-regulation. So something like the Institutional Review Board was more or less inevitable. Yes, it’s a lot of mostly unnecessary expense, and there’s a certain amount of censorship in the name of human subjects protection (e.g., of studies of the potential benefits of controlled substances) but the biomedical world is sufficiently well-funded in general that the costs can simply get built into the grants. Independent and poorly-funded researchers are, of course, out of luck, but there’s no evidence that the people who run Big Science care. It’s a lousy system, but it’s not obvious one could design and run a better one, with one obvious exception: those who propose research should have at least minimal due process rights and a right of appeal from arbitrary and unreasonable IRB rulings.
However, to my knowledge there have been no comparable abuses in social science research, simply because asking people questions doesn’t put them at risk to the same extent as dosing them with toxins. Yet the human subjects protection racket has been able to extend its claws around social science reserach, subjecting it to the same rigamarole as demanded of the medical types. This is stupid, because almost no social science research actually poses important risks to its subjects that couldn’t be handled perfectly well by an informed-consent system audited retrospectively rather than a pre-approval system. Worse, since social-science research is often controversial, the risks of censorship are much more prominent, especially given diversity requirements demanding, for example, that a representative of prisoners be involved in clearing any study involving prisoners.
Any collection of data from an identifiable person counts as “human-subjects research,” even, for example, interviewing a group of judges about how they handle probation revocations. How answering such questions puts the judge at risk is more than I can figure out. And heaven help you if you submit a proposal saying “I intend to ask judges what happens in probation cases.” That’s far too vauge: you have to submit a questionnaire for review, as if you knew in advance what questions were going to turn out to be relevant.
Though censorship is an issue, mostly social-science IRBs don’t kill projects, they just waste both man-hours and calendar time with pointless requests and heckling and impose dumb rules such as limits on payments for inteviews. (It’s OK to offer a college student $10 for spending an hour filling out a questionnaire, but offering the $25 that would actually get students to do it reliably would be “coercive.” Go figure.)
The problem is that much social science isn’t especially well funded; we can’t afford to keep dedicated staff members around to fill out IRB paperwork. Now we’re being told the requirement even extends to completely unfunded (and typically short-deadline) student projects.
In addition, every person running a project (“principal investigator” is the term of art) now has to go through a “training session.” The one I’m doing is on-line, which I suppose is a mercy compared to having to listen to this nonsense live, but it’s written at an eighth-grade level in a literary sense and a fourth-grade level intellectually. It’s mostly a fairly dishonest ad for the necessity of human-subjects review, heavily tilted toward the problems of medical research. Nominally, it’s designed and run by my university, but in fact doing so is one of the requirments imposed by the Department of Health and Human Services, which, due to the fact that the UCLA Medical School is a huge magnet for NIH dollars, is our primary federal funder.
So here’s my proposal for President Kerry: abolish the “lead agency” system. Let HHS make rules for medical research and supervise the institutional arrangements for compliance, and let the NSF make a (much looser) set of rules about social science and manage compliance with them.
Now, if you’ll pardon me, I have to get back to reading sentences such as “It is important that investigators complete and return the continuing review application as soon as possible in order to avoid a lapse in approval.”