Varieties of Regulation

My weekend post on regulation prompted interesting followups from Tim Lee and Mark Kleiman and I basically agree with what they have to say. Meanwhile, it occurs to me that my original post shouldn't have just thrown around the term "regulation" since, obviously, regulations come in different sorts and some regulations I very much favor.

Most notably, pollution -- especially air pollution while the prospect for alternative policies centered around establishing defined property rights seems very dim -- has to be controlled through a regulatory framework. Similarly, I'm a believer in a healthy dose of paternalism, product safety, and public health regulations. But the midcentury effort to transform vast sectors of the economy into tightly regulated monopolies -- the era in which nobody could own a phone, you had to rent one from AT&T -- was ill-advised. Today, on a smaller scale you see a profusion of occupational licensing regulations whose function is simply to arbitrarily make it more difficult for people to start new businesses and compete with incumbents. You also frequently see politicians wanting to find regulatory solutions to what are basically distributional issues. The appeal here is that trying to make businesses behave in such-and-such a way rather than just straightforwardly spending the money necessary to get the thing done allows you to avoid tax increases.

The trouble is that the distorting effect tends to be much larger than what you could have gotten by just spending money. Or you'll have regulations at cross-purposes like in DC where there are all kinds of impediments to building more housing units (maximum lot occupancy rules, maximum height rules, restrictions on your ability to subdivide, etc.) combined with regulations designed to ensure the availability of affordable housing.

Comments

Sounds like you may be going out on the limb to argue that regulations which make sense should be supported while regulations which do not make sense or have harmful effects should not be supported.

I would agree.

Posted by El Cid | February 26, 2008 12:29 PM

The basic problem Matt needs to understand is what is sometimes described as "regulatory capture."

In theory it is all well and good to support only those regulations legitimately designed to protect public goods. The problem is that you need to give some agency the power to create and enforce those regulations. That agency then becomes an extremely tempting target for the incumbents in the regulated industry, who can turn that power toward protecting themselves rather than protecting public goods (and that is just as true for, say, incumbent property owners in the zoning context). And the public is ill-equipped to fight against the incumbents for control of these agencies for a variety of reasons (such as informational asymmetries and a relative diffusion of interests).

Unfortunately, I have no comprehensive solution to offer to this problem. But at a minimum, we should be highly aware of how difficult it is in practice to make sure agencies are only creating and enforcing the kinds of regulations Matt favors.

Posted by DTM | February 26, 2008 12:41 PM

Absolutely right.

Check out this paper from Alfred Kahn (one of the leaders of airline deregulation) about his fear of "progressive" re-regulation. One interesting part is where he describes Ralph Nader's 1970s opposition to airline deregulation.

Posted by ed | February 26, 2008 12:45 PM

Oops, forgot to provide the link:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1089038

Posted by ed | February 26, 2008 12:49 PM

Matt's argument is that deregulation is always good, because 50 years ago you had to rent telephones?

Seriously, we're all a bit sick of eating contaminated meat, and we're a bit sick of our pets mysteriously dying for absolutely no reason other than their food was tainted. We're sick of buying toys that kill our children, even after they've been certified as safe. We're really sick of our bosses breaking every labor law they can think of, and the LRB and the DoL looking the other way.

So Matt can go take his regulation philosphy and go to the Republican party. We sure as hell aren't ever going to buy into it again.

Posted by Soullite | February 26, 2008 12:55 PM

I know that's not exactly what Matt is arguing for here, but the past has made damn clear that deregulation is a slippery slope. Once people like Matt start cutting Regulations, they never stop.

Deregulation is dead. Privatization is dead. Liberal interventionism is dead. If you think tank types don't get that through your head, the left will spend the next eight years ripping itself to shreds and the Republicans will get both houses of congress again.

Posted by Soullite | February 26, 2008 12:59 PM

DTM has it right. It's very easy to think of a problem, and then think of how government might solve it. But what does your solution look like after it's gone through the meatgrinder of the legislative process--if it makes it through? And then, what happens when an angency is charged with implementing it?

Every step of the way, special interests can gain concentrated benefits for themselves by steering things their way. The general public faces diffuse costs when the solution goes astray. And you end up with one more wayward government program, created with the best of intentions.

Try skepticism when you'r offered these government solutions. You'll usually be right.

Posted by jbd | February 26, 2008 1:08 PM

But the midcentury effort to transform vast sectors of the economy into tightly regulated monopolies -- the era in which nobody could own a phone, you had to rent one from AT&T -- was ill-advised.

The reason AT&T is no longer a monopoly (though like an annihilated Terminator, it seems to be slowly reassembling itself) is because the Justice Department Antitrust Division beat it into the ground and broke up the company. I'd suggest that the Sherman Act is itself a form of regulation.

A tightly regulated monopoly may not be heaven on heaven, but its better than the usual alternative, an unregulated monopoly (AND DTM is right, "regulatory capture" is a big deal).

I seem to recall Greg Palast writing a book showing that publicly regulated utilities are more efficient and affordable than unregulated utilities, Here it is-- http://www.democracyandregulation.com/

Posted by beowulf | February 26, 2008 1:17 PM

By the way, I want to make it clear that I think regulating in order to protect public goods is absolutely necessary. In my view, there really is not some simple solution to this problem, and aside from a few relatively minor suggestions (e.g., leveraging the power of the Internet to make public oversight more effective), the best idea I can come up with is something along the lines of "be skeptical and watchful when it comes to your regulators."

Posted by DTM | February 26, 2008 1:24 PM

The difference is between regulation designed to protect jobs (bad because it hurts growth) and designed to redistribute wealth from the rich to the poor (bad because tax-and-spend does it better) and regulation designed to internalize externalities (good if the benefits outweigh the costs).

The first two types of regulation are bad in economic terms, but may be better than nothing if you don't have an adequate social insurance system. The Nordic model is essentially to have an adequate social insurance system and then get rid of protectionist regulation.

Posted by Pithlord | February 26, 2008 1:27 PM

heaven on heaven, geez--- I suppose that should be heaven on earth.

One other problem with regulating corporate power is that our version of federalism makes it so easy to game the system. When Congress is easier to bribe, then you argue in court that tighter state regulations violate the Supremacy Clause and should be overturned. IIRC, that's why ERISA has been used to abolish the right to sue your employer-provided insurance company in state court for a bad faith denial of disability insurance.

On the other hand, if a state has no usury law (that is, a cap on loan interest rates), the credit card companies argue "state rights" and the courts allow that state to override any federal or sister state regulations that attempt to cap interest rates. Though even that goes back to a corrupt Congress, Uncle Sam has the authority to step in and regulate interstate commerce, but apparently the anti-usury lobby is underfunded.

Posted by beowulf | February 26, 2008 1:32 PM

Deregulation is dead. Privatization is dead.

This comment is an illustration of why Matt Y.'s point here is so important. In political debates, a lot of very different issues are jumbled together in a way that makes for bad analysis and bad policy.

Deregulation and privatization are completely different things.

For services provided mostly or entirely by government (eduction, public safety, the military) privatization is almost always a terrible idea. It doesn't in any way reduce the demands on government -- which still has to oversee the contractors -- while reducing accountability and creating all sorts of potential for corruption and waste. Unless there is not already a well-functioning market for something (obviously the government doesn't need to make its own pencils) there's really no advantage to government hiring private contractors to do it.

The kind of deregulation Matt is talking about is another story. In a lot of cases, regulations are in intent or effect just protection for incumbetns. Here in NYC, we're currently having a huge spat over allowing more street vendors to sell fruits and vegetables, a very sensible public health measure that is violently opposed by small retailers. In other casess, regulation is just privatization by a different name -- it's an effort to get the private sector to do something that the state should do directly. Obvious case: at some point, expanding health coverage by regulating private insurance versus providing universal public insurance.

So, yes, privatization and deregulation get grouped together, but they're closer to being opposites.

Posted by lemuel pitkin | February 26, 2008 2:09 PM

I think one of the lessons you could learn from regulatory capture is that regulations are much more likely to work if their is strong public backing and effective public oversight is feasible. This is because the public has to serve as the counterbalance to the industry's lobbyist. On some issues, this is much more plausible than others. Some liberals think that anything that could theoretically be achieved by regulation should be implemented. Libertarians argue that since the government sometimes screws these issues up, no regulation should be attempted at all. In actuality, some things can be pretty well regulated and others are pretty hopeless.

Posted by mpowell | February 26, 2008 2:14 PM

Re: Privatization

Lemuel Pitkin hits on an important point about government contracting out. I wouldn't go as far as to say it is always a bad idea, but it can create bad incentives, and should not be confused with a competitive market.

Outside the US, that isn't the main focus of "prvatization". It is more a question of government divesting itself of assets it shouldn't have in the first place. You have scores of countries in the world spending precious tax money on "national champion" airlines and jute factories, instead of primary education and public health. Even in contemporary Canada, almost all natural resources are state-owned, as are liquor stores (in most provinces), energy utilities, auto insurance companies (in a number of provinces).

Posted by Pithlord | February 26, 2008 3:42 PM

Negative externalities is one very important idea here. Reduced air quality, personal bankruptcies due to predatory lending, food borne illness, impaired health due to bad work conditions, etc. are all negative externalities -- some companies or people reap the benefits, other people (or companies) pay the price.

Regulation is required to redress these problem, but it can be very clean, efficient regulation. Figure out what the costs are, make the responsible parties pay the costs. To the predictable squeals of outrage "We can't make a profit if we have to pay our costs!" we can reply "Free market, buster!"

Of course this is too simple, there will be debates about the true costs, attempts to create loopholes, claims these are stealth taxes, etc. etc. But if we can get a meme going that "Every business should pay all its costs -- it is only fair!" and over time standardize and streamline the measurement of costs and the regulatory process, we can progressively (heh) eliminate special pleading.

Non-hack economists will agree that well-managed "bill back" of negative externalities increases economic efficiency.

There are other needs for regulation I'm not addressing (for example public goods, which have positive externalities) and maybe I'll talk about them another time.

Posted by Jed Harris | February 26, 2008 3:43 PM

Negative externalities is one important case where you need regulation.

The other is is setting minimum standards when you really want 100% certainty about the outcomes -- and this applies even if the costs are all internalized.

For instance, food safety. Even one fatal foodborne illness is too many. And this is true even if the csots can be internalized, e.g. if the victim's family can reliably sue the producer.

A third case is where it's a precondition for a well-functioning amrket -- a lot of consmuer information stuff falls in here.

And a fourth is to overcome collective action problems or what econoists call coordinatiion failures. Labor regulation is justified on these grounds -- workers are collectively better off with a minimum wage, but in many cases it's impossible for them to bargain for one directly.

At the same time, I think contemporary liberals are often too quick to turn to regulation. There are many problems we try to solve through regulation that could be better addressed through realaligning market incentives, direct public provision, or empowering the affected parties themselves -- e.g. unions are much better at regulating working conditions than OSHA. (And of course there are some problems that government doesn't need to be solving at all.)

Posted by lemuel pitkin | February 26, 2008 4:04 PM

"e.g. unions are much better at regulating working conditions than OSHA"

Well, apart from the fact that there are so few unionized jobs left...

Why not have both? In fact, the unions can use OSHA to protect its workers. Given where we've been with worker safety, I wouldn't want to give an inch on this issue.

How about another example?

Posted by MattD | February 26, 2008 4:30 PM

I promise I will stop beating this drum, but I just wanted to note that I intended to include all of the various perfectly valid theoretical uses of regulation within my catchall phrase "protecting public goods".

Indeed, whether or not that phrase is perfectly appropriate for all of the theoretically valid uses of regulation (e.g., I've heard the term "public bads" sometimes used for negative externalities), the bottomline is that I'm not aware of any reason to believe regulatory capture becomes less of a problem if you vary the theoretical justifications for the regulations. The real problem, in other words, is making sure the regulators are acting in the public interest instead of the incumbents' interests, however the public's interests are being conceived.

Posted by DTM | February 26, 2008 4:33 PM

How about another example?

OK. Unions are much better at protecting workers from unjustified firing than discrimination lawsuits. How's that?

(And yes, the weakness of unions is a real probelm -- one of our biggest. But it's also a real problem with contemporary American liberalism that the usual response to an injustice in the workplace is often paternalistic regulation rather than looking for ways to empower the affected workers themselves. Among other things, a negotiated agreement between the employer and employees is likely to yield a more economically rational outcome.)

Posted by lemuel pitkin | February 26, 2008 6:35 PM

Lemuel Pitkin's examples #2 and #3 are basically asymmetric information cases. I think most economists would deny that workers as a whole would be better off with legislated minimums for labour contracts (his #4). Workers who would get at least that on the market wouldn't be affected. Workers who would do worse on the market would be better off. Workers who would be unemployed as a result would be worse off. In general, #4-type regulation is the bad kind the Nordic model avoids by encouraging flexible labour markets, full employment and state subsidies (not employer subsidies) for the poor.

Posted by Pithlord | February 26, 2008 6:41 PM

I think most economists would deny that workers as a whole would be better off with legislated minimums for labour contracts

Not true. First, the empirical evidence is very strong that at the levels typical in the US, minimum wage laws have no significant effect on unemployment. (This is what more sophisticated models of the labor market often predict, but it's the econometrics that have really shifted the consensus.)

Second, even if there were an effect on unemployment, the wage level that would maximize total wages would be somewhere above the market wage. This is, to use a often-derided term, Econ 101. If sellers can form a cartel, their optimal price will be above the market price. As long as there is reasonable circulation between unemployment and minimum wage jobs, low-wage workers as a group will be better off.

Of course, total income would be lower in this scenario, but if you assume (as nearly all economists do) a declining marginal utility of income, aggregate utility might well be higher.

In any case, this second point is sort of trumped by the first -- mandating a higher minimum wage empirically does not increase unemployment.[*] Under plausible scenarios of employer market power, actually moves you closer to the free-market outcome.

You are correct to point to assympetrical infoormation as an important general argument for regulation.

(*Yes, of course a sufficiently high minimum wage would increase unemployment. I'm talking about minimums in the range we see in practice.)

Posted by lemuel pitkin | February 26, 2008 7:12 PM

Come to think of it, the efficiency of minimum wages is arguably a case of assymetrical information too. Because employer market power generally derives from the fact that job search is costly for workers. There's a great book on this called Monopsony in Motion, if you're interested....

Posted by lemuel pitkin | February 26, 2008 7:18 PM

Well, the US federal minimum wage may have very little effect on employment, since it is so low and the US has in other respects highly flexible labour markets. That doesn't mean that regulating the employment contract is an efficient way to redistribute income or that the same is true in other countries with less flexible labour markets.

If you want to subsidise the working poor, it is better to do it by taxing and spending. Minimum wage hikes may be easier politically, but that's ultimately because of Bill Richardson-style economic ignorance of the fact that regulations are costly.

But if you are parochially concerned with the US, you are correct that minimum wages are not a big deal, while tying health benefits to employment certainly is.

Posted by Pithlord | February 26, 2008 7:36 PM

You might scoff at the AT&T now, but the benevolent telephone monopoly provided people with one of the most efficiently designed phones ever made; the Western Electric 500. It was easy to use, attractive, cheap, and incredibly durable. That phone occurred in part because the phone company's ownership of the phones gave it an incentive to make a "indestructible" phone that would infrequently need repair and/or replacement. That should be an argument for regulated industries, not against.

Posted by Michigander | February 26, 2008 8:16 PM

Cute how you linked to a post insulting your commenters and noting your tendency to piss them off. I for one appreciate that tendency of yours.

Posted by Justin | February 27, 2008 12:15 AM