Scholarship and Censorship
To the Editor:
I am a recent college graduate who has conducted research that required approval from an institutional review board (IRB). I also firmly believe in the importance of free speech in academia and frequently agitated for such on my college campus during my time as an undergraduate. I would normally be predisposed to enjoy Philip Hamburger’s article “The Censorship You’ve Never Heard Of” [July/August], but I believe the piece is flawed.
While he highlights the Tuskegee syphilis study, Mr. Hamburger sells short the historical case for IRBs, which is much more robust than a single abuse of black prisoners. To the list of studies that damaged their subjects, we might also include Stanley Milgram’s electroshock study, in which subjects were ordered to “shock” an actor, who suffered no actual harm but screamed in fake agony. Many subjects were then told, to their horror, that they administered so many fake shocks that they had “killed” the recipient.
Or one might look to the infamous Stanford Prison Experiment, in which Philip Zimbardo divided 24 students into groups of “prisoners” and “guards” to observe their behavior—only to suspend the study as the mock guards became increasingly violent. Both studies are standard fare in any introductory psychology course as they make a simple point: Researchers can harm their subjects because the relationship between them is asymmetric. Indeed, in many social-science studies, subjects may not be told the entire purpose of the study or what researchers are measuring until after the study concludes. In these scenarios, the duped subjects cannot knowingly provide their full consent to participate and someone (read: the IRB) needs to look after their interests.
While on the topic of Mr. Hamburger’s examples, we might note that he cites only a single instance of research being squelched because of lack of IRB approval: a 2006 study involving doctors’ washing their hands. He would like us to believe numerous researchers fear speaking out about IRB-based censorship, and the number of squelched studies is much higher. But if this brand of censorship were truly as great as he claims, we would expect more than one example. As the wealth of IRB-approved research published daily shows (and any search with Google Scholar gives a taste of the firehose of research gushing from our universities), America still produces studies of all ideological, intellectual, and disciplinary shapes and sizes. Censorship this is not.
Other forms of censorship are truly worrisome prospects at the modern American university. Unfortunately, Mr. Hamburger neither proves censorship nor explains why we should abandon common-sense safeguards such as IRBs.
New Haven, Connecticut
To the Editor:
Philip Hamburger hyperbolically claims that IRBs violate First Amendment rights by licensing speech. These IRBs arose after some egregious research took place, only the most notable of which was the Tuskegee study, as Mr. Hamburger notes. I doubt that Mr. Hamburger would care to participate in a study involving heart catheterization, for example, without his consent and without its having been reviewed by an appropriate body to protect him from undue harm.
Having served on the IRB of my medical school for more than seven years, I know its functioning well. In order for the research to be ethical, it must be well designed, so that it has a reasonable chance of adding to new knowledge and of benefiting the larger society beyond the research population. Most important, the research subject must give informed consent to participate in the study after knowing its risks and benefits. Thus, much of the work of the IRB is to ensure that consent be obtained properly and that the subject understands what he or she is consenting to. This is hardly licensing speech.
I agree with Mr. Hamburger that the Johns Hopkins study of using checklists in the operating room to curb infections was inappropriately stopped because the Department of Health and Human Services required that not only patients but also doctors be considered human subjects who needed to give consent. But that could be fixed by developing regulations for quality-of-care research that exempts medical staff from having to give consent to being evaluated, without trying to reverse the whole system of subject protection and informed consent.
Department of Epidemiology and Population Health
Albert Einstein College of Medicine
Bronx, New York
Philip Hamburger writes:
I am grateful to Nathaniel Zelinsky and Sylvia Wassertheil-Smoller for their comments. I share their concern for human subjects, but not their confidence that IRBs really protect such persons. In fact, it is difficult to find any scientifically serious empirical data to support such a proposition. Instead, what is abundantly evident from numerous examples is that IRBs suppress scientific knowledge. Indeed, notwithstanding Mr. Zelinsky’s protests, the IRB regulations and the IRBs themselves repeatedly suppress knowledge. They thereby delay the development of medicine, and they leave many persons, even tens of thousands, to suffer and even perish for want of the suppressed information.
Mr. Zelinsky and Ms. Wassertheil-Smoller defend this suppression on the ground that human-subjects research can be harmful; it certainly can be. But there is no empirical data showing that anything is more dangerous when done in academic research than when done outside it. Indeed, it is difficult to document many deaths in American academic research, either before or after the establishment of IRBs, and it is especially difficult to document such deaths in the human-subjects research at stake here—namely, the human-subjects research not covered by FDA regulations. It thus is telling that Mr. Zelinsky alludes merely to instances of psychological harm. The harms covered by the human-subjects research regulations are relatively undocumented and small in scale compared with the harms that occur in a wide range of everyday and lawful activities.
The defense of IRBs, therefore, appears to rest on a strange proposition: In order to prevent harms that are largely undocumented and that generally have been nonlethal, the government should employ a mode of regulation that suppresses scientific knowledge and thereby annually leaves many thousand to die. If there is a logic to this, it is elusive.
The reality is that the regulation of human-subjects research expressly takes aim at speech and publication, and this is why it has such perverse consequences. Rather than focusing on harm, it targets academic attempts to develop “generalizable” or publishable knowledge. Moreover, rather than punish harms after the fact, it requires prior licensing of attempts to develop publishable knowledge. It thereby subjects vast amounts of inquiry to prior censorship even where the inquiry and speech are harmless. It almost inevitably causes much more harm than it prevents.
Of course, some human-subjects research is harmful, and some of it, very harmful. But the desire to regulate dangerous activity does not justify a system that focuses on speech and publication, let alone a system that imposes prior licensing on speech and publication.