One recurring problem in economics, and the other social sciences all the more, is that researchers will accept a lot of conventional wisdom on a topic if it suits their preexisting biases, especially if it is not an area which they have researched themselves. Yet this entire question is — surprise, surprise — largely unstudied. Social scientists love to talk about themselves, but critical self-scrutiny backed by data is less popular.
Jason Briggeman just wrote a GMU dissertation to investigate these and similar questions, here is his abstract:
In the United States and most other wealthy nations, all drugs are banned unless individually permitted. This policy, called pre-market approval, is controversial among economists; the preponderance of the economics literature that offers a judgment on pre-market approval is critical of the policy, but surveys of U.S. economists show that many, perhaps a majority, support pre-market approval. Here I analyze the results of a recent survey that asked economists who support pre-market approval to justify, with reference to the economic concept of market failure, their support of the policy. I find that, while almost all the economists surveyed could point to a market failure or failures that may plausibly exist and affect the market for pharmaceuticals, none were able to make a well reasoned connection between those market failures and the particular remedy of pre-market approval. None of the economists surveyed cited in support of their position any literature specific to pre-market approval. I supplement the survey findings with a review of relevant reading material assigned in health economics courses at top universities, searching that material for discussions of what may justify pre-market approval. I find a strong argument that the prospect of overt disasters being caused by avoidable mistakes can justify some intervention in pharmaceuticals; however, I find little to justify the other interventions that are part of pre-market approval. I suggest that future inquiry into possibilities for liberalizing reform concentrate on understanding matters such as the informational effects of product bans, the distinction between safety and efficacy, the nature of demand for drugs about which little is known, and the political economy of drug substitutes.
The upshot is that economists hold a lot of views whose justifications they cannot articulate very well. I think you would find the same when it comes to the Ex-Im Bank (are you sure it fits the model of strategic trade theory?), the mortgage agencies (what was that externalities argument for home ownership again?) or all sorts of random regulations. The relatively interventionist economists will pull some justification out of a hat, and the relatively pro-market economists will be pretty skeptical.
For the pointer I thank Daniel Klein.