A Defense of libertarianism (and response to Zed, etc.)

Medicare is already running in the red, by the way.

Just because the seatbelt is broken, doesn't mean we should abandon seatbelts. It means we should work harder to make better seatbelts.

An argument I see when libertarians are faced with such questions is that libertarians aren't 'against' seatbelts, but they want to explore the alternatives. And yet they don't seem to come up with any other alternatives...

What we should not do is idealize the paper seatbelt or pretend that it works when it does not, or pretend that the design of that paper seatbelt hasn't been heavily influenced by corporate lobbyists, corrupt politicians, etc.

If I walk into a bank and consider depositing my hard earned cash, perhaps I'd demand enough transparency to determine whether to trust that entity with my cash, in exactly the same way that I must trust Amazon when I place an order online, etc.

Information will always be imperfect, but it's in both parties' interest to increase transparency. It's barely even a coordination problem in today's world.

As a note to your remark about my speed limit explanation, why is it then that on an extremely icy road people go exactly the posted speed limit rather than, say, 90 miles per hour? Clearly they are taking the sign literally and at face value.

So one regulatory approach that would be an improvement over the current one might be to require a reserves "rating" to be published, much like the smoking causes cancer warning. Perhaps right under the bank's logo it would have to say "This bank has a reserve score of 71 (moderately risky)".

And yes, there would be room for 3rd party ratings to be applied to bank solvency as well. I'm not sure what the best approach would be, but consider that most securities regulations were made in the 20s before the information revolution (ironically, the "Financial Services Modernization Act" removed a few of the depression-era regs that were actually smart and prevented moral hazard and conflict of interest). Surely there would be a lot of clever approaches. One idea that intrigues me is that the government could offer its rating, but banks would be required to obtain several. I'd love to see someone like Michael Moore create his own such agency and use his genius to publicize it.

Analogously, consider the FDA drug approval process. Everyone knows that the FDA process is flawed, as people are routinely killed by side-effects of drugs. So why not have several ratings: One from a consortium of insurers, one from the AMA, and one from the FDA -- each group has a slightly different downside risk (and upside) to approving a dangerous drug or failing to approve a safe one in a timely fashion. Why not have a grid of 3 checkboxes on every drug. The absence of an AMA approval might be a warning sign to some, but not others.

If you think this idea is silly, consider what happens when the FDA gets it wrong and lots of people die because of a drug... exactly nothing but a too-late knee-jerk reaction. Usually the drug is immediately pulled even if it has other valid uses, etc. One perverse example: many doctors knew the dangers of Vioxx well before the scandal (the same risk exists in lesser form with all such drugs, even Ibuprofen) and good ones didn't use it on patients at risk for bleeding without great caution. Yet it took a few years and then bang, it's gone from shelves even though there are a few essentially identical drugs that stayed on the market and do still occasionally kill people when used without caution by docs.

I'm not claiming to have the whole idea of optimal regulation for the 21st century figured out, but I'm just brainstorming. Surely there would be a lot of great and clever ideas. The current, one-size-fits-all approach leads to massive and often ineffectual regulatory apparatus.

Why shouldn't coming up with ad-hoc regulatory / information-dissemination / transparency-reducing systems be something that startups do? Why not in the case of medical trials have data published to the web so that people can make mash-ups with all the data that is being considered by regulators? A few weeks ago the SEC approved a plan to put financials in XML format, but it's all very primitive and is unlikely to be all that useful, but the core idea of transparency through an API is a good one, I think.

The role of the "regulator" of the future, I think, is to design a good API spec (metaphorically and literally) and possibly make a list of all of the watchdog groups offering analysis. State and federal governments could fund their own watchdog groups too -- maybe I notice that Wyoming's prescription-drug watchdog group has been right a bunch of times in a row, so I begin trusting it more than the one in my own state. Maybe I decide to trust Michael Moore's group or Paul Graham's. Over time the best would rise to the top and every mistake would give new groups a chance to enter the limelight -- maybe a YC startup was the only one not to approve Vioxx, for example.

Toward the libertarian goal of efficient, small government, we should be focusing on regulation that increases transparency and also increases the number of entities that can act as private watchdogs. It's exceptionally telling that the SEC received repeated complaints about Madoff's fund from various sources and still did nothing. One might argue that the SEC is understaffed, but it found the staff to shut down prosper.com, a startup that was making humanity better off, on a technicality, not even on any remote accusation of fraud. More likely, prosper.com competed with the established finance powerhouses, so the SEC was probably just acting to keep competition out of the industry.

Most of what all of the regulatory agencies do is coordinate information, and if the modern world (and the internets) have done one thing, it's make such problems extremely easy and inexpensive to solve. My basic argument is that the antiquated structures create a false sense of security and get in the way, as well as soak up tons of industry lobbying dollars to the point where industries tend to become highly regulated and highly protected, the two biggest characteristics of libertarian dystopia. The military industrial complex and now the financial and auto industries are that sort of thing. Big agriculture is too, as well as many others.