Against Government Regulation

There is a strand of thought that goes something like: “Of course we need government to regulate industry. That way we don’t end up buying dangerous products and hurting ourselves through our own ignorance or because of dishonest corporate marketing.” I think this is deeply misguided. It relies on a concept of “government in theory” and ignores the realities of government in actual practice. We implement the regulatory state with the government we have, *not* the government we wish we had. The performance of regulation in practice is an *empirical* question; it’s performance is not something we can just presume upon or deduce from first principles.  Some examples of the regulatory state in actual practice:  

Take the recent example of the moral panic over vaping. The lung diseases that sickened a few hundred people and killed a handful were tied back to black market vaping products*, not above-board, legal nicotine based products. The legal market here was *not* the problem. But the government response was to point the finger at “vaping” generally. The response of our very wise officials was to recommend or in some cases to actually enact bans on vaping (in particular, bans on flavored vaping). This is unbelievably foolish. It doesn’t solve the problem, and it almost certainly makes it worse. Vapers don’t respond to a vaping ban by simply quitting. They respond the way all consumers respond: by finding the nearest substitute good. In this case that’s either 1) smoking actual cigarettes or 2) black market vaping products. Some vapers might outright stop using all smoking and vaping products (behaving the way our “public health” officials think a pliant public ought to behave), but you also get people substituting these more harmful behaviors for the relatively safe one that they’d rather do. The regulatory response 1) has nothing to do with the actual problem that’s supposedly being addressed and 2) makes the problem worse. This is a common theme.

Another recent moral panic is the so-called “opioid epidemic”. Our politicians and public health officials have gotten this one badly wrong and crafted a false narrative. The notion that “over-prescription of opioids created a new class of opioid addicts” is just wrong. The official government numbers for rates of opioid misuse and opioid addiction did not rise over the period when prescriptions were increasing, and according to the government’s own data the vast majority of “abusers” aren’t getting their supply from a doctor. (Please see SAMHSA and Monitoring the Future reports for the numbers.) The second leg of the crisis has nothing at all to do with legal prescriptions; the skyrocketing rates of overdose in the 2010 to present period are mostly due to heroin and illicit fentanyl. And, no, those users mostly didn’t start out as prescription opioid patients. The chain of causation for the standard narrative is broken at several links. And yet the reaction of our political institutions has been to fan the flames of this moral panic and call for *general* restrictions on opioid prescribing, something that has *already been tried* and has been making the problem worse for ten years. Tragically, the real victims in this story are the pain patients who can no longer get the only medicine that has ever worked for them, because a general reduction in opioid production/prescribing quotas is a blunt instrument. It hits everyone. This is the result of a “We have to do something/Think of the children” mindset without giving actual thought to the consequences. Once again, the government response 1) doesn’t address the actual problem and 2) makes the problem worse.

These aren’t cherry-picked examples of the regulatory state getting things badly wrong. This is business as usual. These are perfectly typical examples of how our political institutions work. An example I’m more familiar with in a professional setting is the regulation of insurance, and it’s a joke. We go through several rounds of objections with the state Department of Insurance, and at no point does the consumer end up “protected” because of the changes/explanations we provide. It’s just a huge waste of time, and it makes your premiums higher because you’re paying for a “compliance” department at the insurance company. Labor regulations tend to be counter-productive. Mandates to provide all employees with X mean that employers subtract the dollar cost of X from employee salaries. It’s a wash (worse than a wash if the employees don’t value X at the cost of the foregone wages/salary). Shallow thinking leads people to believe the mandate transfers the dollar value of X from employers to employees, but that’s plainly wrong. Employers respond by paying less in salary and more in benefits. Minimum wages cause employers to reduce hours (which, in the case of Seattle, led to a net *reduction* in pay for the affected workers), cut fringe benefits, and implement stricter work protocols. And occupational licensing (which is supposedly for the benefit of the consumer) is just bald protectionism for the licensed workers. Collect a few examples of the occupations that are licensed (interior designer? Florist?) or just note that the same profession is *not* licensed in other states and there’s no problem. There is no rational reason for this. This is just a way for incumbent workers to keep the competition out, using the machinery of the state.

There is a good case for regulating pollution, or regulating costs to third parties. Regulating automobile safety for the sake of pedestrians and *other* drivers is legitimate; if your car explodes on impact, that’s of interest to third parties who might be caught in the explosion. That said, regulating for the sake of the people inside the car isn’t legitimate. People have the right to decide what level of vehicle safety they are comfortable with. (If you push back on this point, should we ban motorcycles? Is there any standard enclosed automobile more dangerous than a motorcycle?) Granting the point about externalities, most of what the regulatory state does is attempting to “protect” the consumer directly, not to protect third parties.  And it mostly fails.

The marketplace provides most of the regulation that’s needed. Companies that provide substandard or other-than-advertised products are threatened with lawsuits or (likely more damaging) loss or reputation. Stiff a single customer, and likely many others will know about it and stop patronizing your business. Businesses that maintain liability insurance probably have to comply with the insurer’s underwriting standards; in this way a single small company might behave *as if* it were a large company with a lot of reputation to lose, because it’s being disciplined by this market mechanism.  And there are plenty of private institutions that provide information about safety and quality (Consumer Reports, Underwriter Laboratories, the Highway Loss Data Institute for automobiles, Yelp, Amazon reviews). It doesn’t matter that most consumers don’t bother to collect all this information. If only 10% of customers pay attention to this stuff, companies trip over themselves to comply with whatever safety or quality standards exist to capture this market share. Don’t take this paragraph as overstating the adequacy of market-supplied regulation; it’s possible that these constraints are somehow inadequate and the hand of government is still necessary. But at least acknowledge that these mechanisms *exists*and are quite powerful in most cases.

If your reaction is to start enumerating instances of businesses behaving badly, my response is to ask “So what happened next?” Did they lose their customers/employees? Okay, then that’s a case of natural market-supplied regulation is working. Did they *not* lose any market share? Okay, then apparently their customers/employees, the people *actually affected* by the problem, didn’t seem to care very much. Maybe this was just another overblown viral outrage story. Did they not lose market share because the harm was to third-parties, but not their own customers or employees? I acknowledged the possible need for regulation for the sake of externalities (pollution and other third-party costs) above, so this may be a legitimate example of the need for government. But please recognize that the threat of lawsuits still applies here. Our government may already be supplying the remedy by way of the court system and liability rules *without* resorting to regulation (and there is also non-government conflict resolution: arbitration and mediation). 

Even granting that the rationale for government regulation is legitimate in some specific instance, once again, it matters what government does in actual practice. Is it an effective solution to the problem described? Our regulators aren't doing disinterested cost-benefit analysis and stifling their impulse to interfere when their calculation counter-indicates it. They are part of the same public that indulges its populist outrage over viral news stories, or if they are any wiser, they are nevertheless trying to enforce the general public's "do something" mandate when it breaches the 51% threshold of popular support. Even if you can enumerate instances of government regulation actually solving a problem, these have to be weighed against all the times that it makes the problem worse. 

_______________________________________

*Black market products are dangerous *not* mainly because they aren’t subject to government regulations, but rather because they have an uncertain supply chain that is constantly being disrupted by law enforcement; reputation solves that problem better than government mandates. But reputation is hard when you're forced to operate underground and anonymously, especially when you can't count on being around very long because your profession has a high rate of arrest and murder. 

I started writing this as a facebook post. It got long, so I'm sharing here instead.