Criminal Prosecution for Scientific Fraud
I spoke to a reporter at the end of last week about the criminal prosecution of scientific fraud. I’m not sure how coherent my end of the conversation was at the time, but I thought the topic was interesting enough to return to it briefly here.
Let’s put aside potential investigations and prosecutions by the federal Office of Research Integrity (part of DHHS). Granted, the ORI has claimed an extraordinarily broad mandate (funded and unfunded applications!), which might be worth returning to one day. But on the whole, such cases seem to me to be a fairly mundane application of the general contract fraud principles.
Instead, I’ll concentrate on a free-floating action in fraud against a scientific investigator for having misled potential patients. Thus, consider the scenario of a doctor faking an experiment to show that Drug X prevents heart attacks and has no side effects, when, in fact, it has no preventative powers, and it causes immediate hair loss. Is that doctor criminally liable? Civilly?
I’d guess that to the extent that general fraud often requires an intent-to-induce element, most scientists would be able to successfully assert that they did not intend for patients to rely on their work. In the civil context, I also assume that a consumer’s action would fail on the “justifiable reliance” end. If this weren’t true, I imagine that most scientific papers would end with a disclaimer that they are not intended to be relied upon, and that patients ought to consult their physicians (etc.)
But let’s put aside the doctrine for a moment and consider the policy arguments for attacking scientific fraud with prosecution. There are at least two reasons to think this is a bad idea (again, apart from the government-contract fraud case).
First, I worry about chilling the wide co-authorship norm that is widespread in scientific literature. Hwang Woo Suk’s now-discredited cloning article in science had 24 co-authors, including this apparently innocent scientist at Pitt. While not each and every author faces the same legal risk, the possibility that a co-author’s wrongdoing will result in a scientist facing hard time would inevitably raise the costs for collaboration across universities and borders. These costs are particularly unnecessary in the scientific field, where private enforcement is likely to be effective. Not only is the gold-standard in scientific methodology replicability – meaning that much work will be double-checked – but the reputation costs for falsifying data are severe.
Second, I wonder about proof. Few cases are going to be as spectacular as the Korean cloning fiasco, which has resulted in a nationwide self-examination . Most cases are going to be considerably closer, and the untangling of scientific fraud, like the untangling of corporate fraud, may entail unusually complex demands on jury decision making. I’m a big defender of the jury system, but that defense doesn’t require me to be blind the possibility that juries will allow scientific-fraud cases to turn into morality plays. Top scientists, like top executives, are likely to be arrogant and abrasive, and run the risk of pro-prosecution jury nullification.
Thus, at least on first glance, I’m not convinced that expanding civil or criminal liability for alleged fraud in scientific publications is a good idea. But I’m willing to be convinced that I’m wrong…