What if there were a financial products approval agency modeled on the Food and Drug Administration (FDA)? A recent paper
by law professor Eric Posner and economics professor E. Glen Weyl proposed such an agency to review new financial products and approve only the ones that improved social utility. The article, plus a Bloomberg op-ed summarizing the argument, touched off a lot of discussion, including an approving New York Times
column by Gretchen Morgenson.
There are two reasons to dislike this proposal: theory and practice. In theory, the goal is a bad one. In practice, the proposed agency is designed to resemble the FDA, which is the poster child for regulatory capture.
Regulatory capture is an old idea, most famously expressed by President Woodrow Wilson (a Democrat and leader of the Progressive movement) in 1913:
If the government is to tell big business men how to run their business, then don't you see that big business men have to get closer to the government even than they are now? Don't you see that they must capture the government, in order not to be restrained too much by it? Must capture the government? They have already captured it
The economic theory of regulatory capture is not controversial. Many economists believe that it is so severe that any regulatory discretion is likely to be turned against the public interest, while many others see regulatory capture as a danger that can be controlled through careful agency design. The first camp is likely to cite the FDA as a clear example of regulatory failure, while the second might point to an organization like the Centers for Disease Control and Prevention (CDC) which seems to serve the public health more than its staff or the pharmaceutical industry. That makes it ironic that anyone would want to create a new FDA.
Regulatory capture is not the idea that all regulators are incompetent or evil. It’s the claim that even the best intentioned and most qualified regulators will be unable to resist capture in the long run. Consider a new agency created for some public purpose. Assume it is staffed with qualified, energetic people whose only concern is how to serve the public best. Naturally, there will be some disagreements among these people about agency actions. Those regulators whose ideas are most costly to industry will find difficulties: evasive actions, stalling, court challenges, and political complaints. Those regulators whose ideas are least costly to industry will find their paths greased. Industry will cheerfully supply data and guidance, even commissioning papers from respected experts. It will voluntarily adopt changes, and let Congress know how much they respect the regulator. The first regulator will likely have a short career in government, and few employment options afterward. The second regulator will find lucrative post-government employment lobbying for the people he used to regulate.
Remember, all this started with qualified and energetic regulators interested only in the public good. The second regulator will believe sincerely and reasonably that his policies were the right ones. He didn’t choose them because they were good for industry, but he would have to be a saint to refuse praise and rewards for doing what he thought was right in the first place. In the real world, some regulators will be neither qualified nor energetic, and will be interested only in their personal good. Eventually, there will be a revolving door of people moving back and forth between the agency and private employment, or sometimes doing both at once by serving on “advisory panels” or getting “research grants.” These people have a huge interest in keeping regulations complex, because they’ve spent years learning them, and also keeping regulations flexible enough that any desired decision can be reached.
It’s important to distinguish regulatory capture from corruption. There may be some corruption as well, but that’s a different problem. In regulatory capture the agency acts as a broker between the industry and Congress. The industry wants profits, which means keeping prices high, costs low, and competition out. A small group of people gain a lot from the agency, no member of the public individually loses much, concentrated money trumps dispersed interests in politics. Congress doesn’t want problems, scandals, or complaints. So the agency works to enhance industry profits, while trying to make sure that nothing happens so bad, or angers anyone so much, that Congress hears about it. The agency may also employ former Congressional staffers, supply free work or other benefits to help Congresspeople, or make itself useful in other ways.
As I said, this story is not controversial. Some economists believe is that it’s possible to fight regulatory capture. You can have rules against revolving-door employment and former staffers representing private clients. You can try to instill a culture of public service. You can offer regulators high pay and prestige. You can insist rules be simple and decisions transparent. You can hold the agency to strict objective metrics. Other economists will tell you that none of this can work for long; people will eventually find ways to maximize their interests.
Let’s look at the FDA from this perspective to see why it’s about the worst possible model for a new agency. The Food, Drug and Cosmetic Act was passed in 1938 after over 100 people died from Elixir Sulfanilamide, a patent compound that contained the poison diethylene glycol. Congress gave the FDA the mandate to review the safety of all new drugs before they could be marketed. What it had in mind was that the FDA would look for ingredients known to be harmful, and require animal testing for ingredients of unknown toxicity.
Unfortunately, safety testing is difficult, because you’re looking for a negative. You might test the drug on 100 animals and miss the fact that it kills one out of every 200. Or you might miss the fact that it doesn’t hurt animals, but does kill humans. Or maybe it’s safe for most humans, but deadly to, say, pregnant women or Pacific Islanders. Some dangers appear only after many years. Testing with humans instead of animals is more accurate, but it means risking people’s lives. Also, not every unsafe drug kills people. How about a drug that sometimes causes nausea, or immune suppression, or blindness? Should it be approved or not? Moreover, almost all drugs were already safe. Manufacturers had reputations to protect and lawsuits to fear. This made the agency’s job difficult and unrewarding, while it had to take the blame for any problem that slipped through.
Therefore from the beginning the FDA started doing something else. The regulators in charge decided that safety could only be defined relative to benefit. If a drug cured a life-threatening condition for which no other treatment was available, you would tolerate a lot of danger. But if a drug was used for a minor complaint for which there were already safe remedies on the market, you would want to be very sure it was harmless before approving it.
This ingeniously shifted the task from testing safety to testing efficacy. Efficacy is much easier to test because you’re looking for a positive. If you give a drug to one animal and it gets better while an untreated animal does not, then the drug works (at least some of the time). The switch did something more important, it made the FDA very valuable to established companies. If someone had an innovation to compete with an existing product, it faced a high regulatory barrier, because the FDA said there was already a safe alternative on the market. But if a company wanted to introduce a new drug of its own, it could slide through as long as it was defined as a new market. There was tremendous regulatory discretion on the definition of a market, and essentially no oversight. Companies with good FDA relations got monopoly protection for their existing products and easy paths to market for their new products. Companies with poor FDA relations might as well go out of business.
One problem was the FDA angered a powerful interest group: doctors. The agency usurped their authority to decide what medicines their patients should take. The ingenious solution was the prescription. Although there was no provision for this in the Act passed by Congress, the FDA started designating both existing and new drugs as safe only under the supervision of a medical professional. This brought the medical profession inside the scope of regulatory capture and gave doctors a lucrative new monopoly.
The FDA allied with pharmaceutical companies and the medical profession became an overwhelming powerful force. It was able to get legislation through Congress authorizing the powers it had initially usurped decades earlier. The one thing no one seemed to care about was whether the FDA was actually doing its job of making sure safe and effective drugs reached the market, while unsafe or ineffective ones were blocked.
No one until Sam Peltzman, that is. Peltzman is an economics professor at the Booth School of Business at the University of Chicago (disclaimer, I took courses from Professor Peltzman many years ago). In 1973 he published an authoritative paper
“The main finding is that benefits forgone on effective new drugs exceed greatly the waste avoided on ineffective new drugs. The estimated net impact is equivalent to a 5-10% tax on drug purchases.”
This stimulated an large body of work. People looked at the drugs disapproved by the FDA but allowed in other countries versus the drugs disallowed in other countries but approved by the FDA and found the drugs disapproved by the FDA were significantly safer and more effective than the drugs the FDA approved (to be clear, this does not count drugs either approved or disapproved everywhere). Delays in drug approvals, often ten years or more, were shown to have cost hundreds of thousands of lives, while at most hundreds were saved by rejecting unsafe drugs.
Another approach is to study “off-label” uses of drugs. These are drugs used to treat something other than the condition for which the FDA found the drug to be effective. If the FDA were doing any good, drugs should be safer and more effective for approved uses rather than unapproved ones, after all, that’s what efficacy testing is supposed to do. In fact, drugs are significantly safer and more effective in off-label uses.
Although these findings are disputed by some people, it is only to claim that there is no proof the FDA doesn’t work. No one has ever produced evidence that the FDA does work to bring safer, cheaper or more effective drugs to the American public. And there is overwhelming evidence that the FDA acts to improve the profits of pharmaceutical companies and the income of doctors, which in turn generate sizable campaign contributions and lobbyist activity for Congress.
If we create a new FDA for financial products, we can expect a similar narrative. Congress will pass a law requiring the agency to review new financial products for safety. The agency will figure out that task is difficult and unrewarding. Staffers who try to do it will find their way blocked by courts and complaints to Congress. More clever staffers will decide that safety has to be compared to effectiveness, and the agency should discourage innovations where there is an adequate existing product while allowing established companies to enter new markets even at some risk to the public. This view will be supported enthusiastically by established companies.
Brokers, financial advisors and other professionals will object, so the new agency will decide some products are safe only under the supervision of a qualified professional. A powerful new group will emerge, with the money and lobbyists to gain after-the-fact Congressional legislation to ratify its actions. Financial products will become more expensive, less safe and less effective.
If you’re an optimist, you can hope the new agency will act more like the Centers for Disease Control and Prevention, protecting the financial health of the public through information sharing, data collection, emergency response, research, communication, and education (in finance, this is more or less what the Securities and Exchange Commission is supposed to do). Economists who have studied the CDC have mixed opinions about its effectiveness. However the argument is not about whether it does good, it clearly does a lot, but whether or not that good could be accomplished better and cheaper through private means. That question is more political than empirical.
Optimist or pessimist, a financial products approval agency is a terrible idea. But if you’re going to support it, you would be wise not to mention the FDA.