A couple of people have asked me to expand on what I said yesterday about the parallels between financial products and other sorts of new technologies.

America's regulatory structure is mostly the child of the Progressive Era, when well meaning, well educated protestants thought that they could save the world by putting bright technocrats from the right kind of families in charge of the messy, sprawling economy and make it clean and tidy and safe.  That sounds sarcastic, but it wasn't entirely unreasonable. The first great victory of the Progressive Era, the major revolutions in public health, did just that:  made life safer and nicer for everyone, with minimal inconvenience, by putting experts in charge of things like sanitation and quarantine and the water supply.  Before Hayek, we didn't have all that much reason to think that this feat couldn't be repeated elsewhere.

But now we have had Hayek, and the failure of the Soviet Union, and a hundred other ways to learn that in any sizeable economy, the information problem is simply too big.  Even leaving out the various incentive problems ably detailed by both Marxists and public choice economics, a well-intentioned bureaucrat cannot know enough about what's going on in the world to thoroughly manage even a static economy, much less one that has to cope with millions of constant changes, from hurricanes to new babies.

Yet our regulatory model still works on the assumption that the technocrats can figure out what is safe, and then let the public buy it, and nothing will ever go wrong.  Oh, maybe it's not working that way right now . . . but that's because the other party is in charge.  Or the regulators got corrupted.  Or the Moon was in Taurus while Saturn was in Capricorn, a constellation that will not recur for another seven thousand years.

It's perhaps easiest to see when you look at the FDA rather than the SEC.  The very early purpose of the FDA was to ensure that the label correctly described the contents of the package, a noble purpose entirely proper to any regulatory agency.  Under FDR, it took on the job of ensuring the safety of products and prevent false therapeutic claims.   Over time, this has grown into the agency we now know: the one that forces companies to go through three stages of trial for every drug.  Stage I is small-scale and mostly ensures that the substance isn't horribly toxic; Stage II looks at safety and efficacy; and Stage III are large-scale trials comparing the drug to either placebo, or the standard treatment.  After these three stages, the high priests of the FDA retreat into their sanctum and eventually emerge to give the thumbs up or thumbs down.

This is simultaneously too much and too little.  The FDA is notoriously risk averse when it comes to new drugs, a legacy of events like the Thalidomide horrors, when pointless foot-dragging on the approval accidentally protected American mothers from limbless babies.  So drugs that could help people don't reach the market, either because the testing is too expensive for small market drugs, or because the FDA erred on the side of caution.  At the very least, it delays possibly life-saving advances.

But once a drug is approved, the FDA doesn't require follow-up studies.  These are mostly done by the companies, which have incentive to bury the ones that show their product is horse puckey.

It seems to me that the FDA should put much less energy into preventing drugs from reaching the market, and vastly more energy into assessing them after they have.  Moreover, though it may sound un-libertarian, I'm not clear why the companies are in charge of the drug testing.  This seems like the sort of thing the government should do; the government is the one with the neutral incentives.

If we looked at the FDA more like a neutral broker, a source of authoritative and up-to-date information on the food and drugs out there, I think we'd be much better off.  But instead, we've turned them into the Drug Nannies.

Similarly, the SEC, indeed the entire financial regulatory apparatus, revolves around writing rules, and then making sure that companies live up to them.  Some of those rules, like financial statement filings, enhance transparency.  But many others jealously guard the perogatives of the SEC and its minions--the safe harbor provisions for the three big regulatory agencies, for example, which lets the SEC control the definition of what a good, safe security is.  From what I understand, one of the reasons that so many at the SEC opposed GLB was that merging investment banks with commercial banks meant handing over some of their regulatory power to the agencies that oversaw the commercial sector.

But among America's artificially fragmented grab-bag of agencies, a leftover from the 30's notion that bank panics would be prevented if we just kept the banks small and manageable, no one had both the authority and the apparatus to oversee systemic risk.  The Fed has stepped into this vacuum with Treasury close behind, and in my opinion they're doing a pretty good job on the fly.  But this isn't the sort of thing that should be done on the fly. 

I would much rather see America's regulators do what government is really good at:  transparency and coordination.  Providing as much detailed information about firms, securities, and markets as they can, particularly in assessing systemic risks.  Serving as a locus, with muscle if necessary, to help multifold private actors create institutions that reduce risk.  Delineating and carrying through the procedures for dealing with the inevitable failures.

That last is really, really important.  Most people don't seem to have noticed that commercial banks have been failing too.  They haven't noticed because the FDIC does an outstanding job of overseeing the disposition of insolvent banks.  They have a process, everyone knows what it is, and they are swift and efficient at carrying it out.   But we can't build that kind of process unless we're willing to admit that despite our best efforts, failures will happen.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.