Scott Sumner  

About those wacky libertarians

Nobility Defined... Concrete steps for Mr. Cochran...

Will Wilkinson has a post discussing libertarianism. He points out that many libertarians are disdainful of politics. They tend to be ideological purists who are so anti-government that they are reluctant to get involved in the rough and tumble of political horse-trading.

Non-libertarians tend to view libertarians as extremists, who are obsessed with "rights" to the exclusion of all other considerations. Progressive intellectuals are not persuaded by arguments that, "the FDA should be abolished, because no one has a right to tell me what I can or cannot put in my body".

In this post I'd like to distinguish between libertarianism as a political philosophy and libertarian public policies. I'll make two arguments:

1. There are arguments for laissez-faire that are far more powerful than most non-libertarians assume.

2. Libertarian public policies are far more effective than one might expect (using common sense.)

In other words, non-libertarians are lulled by the unimpressiveness of most real world libertarian arguments into assuming that all the arguments for libertarianism are very weak, whereas there are actually some quite powerful arguments.

A recent essay by Steve Landsburg reminded me of my years at the University of Chicago. After three years at Chicago I could offer very effective arguments for some pretty extreme libertarian positions. Over the following 35 years those debating skills atrophied, as I was no longer enmeshed in cutting edge economics. (I focused on a narrow area of monetary history.) But when I read Landsburg, and any of the George Mason bloggers, I am reminded of the skills I have lost. I see people who have kept up with the field.

Obviously it's hard to explain this to people who have not had a Chicago-style education. The best I can do at the moment is to direct your attention to this very thoughtful critique of the libertarian view of labor markets, by Scott Alexander. And then read this reply by Bryan Caplan, which is far superior to anything I could have written. But even so, Bryan's response (focusing on theory) is only one of many that could have been offered. Another strategy would have been to point to empirical evidence that public policies aimed at improving labor market outcomes are usually counterproductive. For instance, studies show that OSHA does not improve workplace safety.

Bryan argues that his rebuttal to Scott is based on mainstream neoclassical economics, not fringe libertarianism. There's a sense in which that is true, and another sense in which it is misleading. Take my previous post, which argued that there should be no taxes on capital. You can certainly find many progressive economists who have made the same argument. And there is a sense in which that view represents mainstream economic thinking. In the comment section Don Boudreaux pointed out that even John Stuart Mill made a similar observation, 170 years ago. But it's also true that Bryan's views on the efficiency of labor markets and my view of capital taxation is probably held by no more than 20% of economists. Indeed I doubt whether more than 30% of economists are even aware of these arguments. I had forgotten them until I read Bryan's post.

I would not have made this claim back in the 1990s, when people like Paul Krugman and Larry Summers offered many "conservative" sounding arguments. But they no longer make very many pro-free market arguments, and the profession as a whole has definitely moved to the left. Ideas such as minimum wage laws and fiscal stimulus and capital income taxation are now much more mainstream among the elite.

The views of average economists change much more slowly; indeed many never even heard that in the 1990s fiscal stimulus was abandoned by the elite, before being rediscovered in 2009. When I think of average economists I think of someone who believes "externality arguments" provide a justification for government regulation of smoking, whereas on closer inspection they do not.

If you are a progressive who has encountered libertarian ideas based on natural rights, you may not know that there is another set of arguments for abolishing the FDA, based on empirical studies of type one errors, type two errors, incentives of regulators, etc., etc.

Because I'm a pragmatic libertarian, I'm not reflexively opposed to all government programs. I'd be happy to see the FDA, SEC and OSHA be abolished, but would be less comfortable with abolishing the EPA and FTC. But despite my wishy-washy form of libertarianism, I actually have great respect for the more extreme forms of the ideology. I don't think that most adherents of the extreme form of libertarianism make good arguments, but I do think they might be correct, despite their weak arguments.

Consider areas where even pragmatic libertarians in the tradition of Milton Friedman might see a role for government; police and fire protection, education vouchers, etc. Then consider the following:

1. Denmark (a favorite country of progressives) has private provision of fire protection.

2. San Francisco (another progressive favorite) has private provision of police services in some areas, and the quality is apparently higher than for the government police, according to the metrics that progressives might want to use.

3. Friedman probably favored vouchers because it's "obvious" that otherwise the poor would not be able to afford an education. Except there is just one problem with this assumption. Tens of millions of the world's poorest students (who are an order of magnitude poorer than America's poor) do attend private schools without vouchers, throughout many of the poorest regions of South Asia and Sub-Saharan Africa. So maybe it's not "obvious" that the government must play a role in education.

In 2010 there were an estimated 1m private schools in the developing world. Some are run by charities and churches, or rely on state subsidies. But the fastest-growing group are small low-cost schools, run by entrepreneurs in poor areas, that cater to those living on less than $2 a day.

Private schools enroll a much bigger share of primary-school pupils in poor countries than in rich ones: a fifth, according to data compiled from official sources, up from a tenth two decades ago (see chart 1). Since they are often unregistered, this is sure to be an underestimate. A school census in Lagos in 2010-11, for example, found four times as many private schools as in government records. UNESCO, the UN agency responsible for education, estimates that half of all spending on education in poor countries comes out of parents' pockets (see chart 2). In rich countries the share is much lower.

One reason for the developing world's boom in private education is that aspirational parents are increasingly seeking alternatives to dismal state schools. In south and west Asian countries half of children who have finished four years of school cannot read at the minimum expected standard (see chart 3). In Africa the share is a third. In 2012 Kaushik Basu, now at the World Bank but then an adviser to India's government, argued that India's rapidly rising literacy rate was mostly propelled by parents spending on education to help their children get ahead. "Ordinary people realised that, in a more globalised economy, they could gain quickly if they were better educated," he said.

In other words we've seen a huge global gain in education for the poor, a prime progressive goal, propelled by a radical libertarian ideological solution---private schools for the poor. Paid for out of pocket, no vouchers. And I'd guess 90% of progressives don't even know it happened.

I'm not claiming that any of these anecdotes are slam dunk arguments for private police, fire protection or schools. Indeed I have consistently advocated voucher systems, despite the evidence that they are not needed. Rather I'd make a different argument. Most progressives I talk to are not even aware of the arguments for abolishing the FDA, or for not regulating tobacco. They've never even heard the arguments. They are not aware of Denmark's private fire service, or San Francisco's private police force, or the enormous growth in private education in the third world (motivated by the appalling incompetence of the public schools in those regions.) Or they are appalled at how the "free market" in drugs is ripping off American consumers, without being aware that this is due to a government regulation that bans the importation of drugs from cheap foreign producers.

Without being aware of the serious arguments for laissez-faire, progressives are in no position to be contemptuous of the ideology. Until they've done their homework, they haven't earned that right.

PS. To head off comments on smoking, let me explain why the externality argument doesn't call for regulation. There are two arguments. One is that smokers impose budget costs on society through spending on Medicaid/Medicare, etc. This is true, but vastly overstated. Nonsmokers live much longer, and still end up dying of something. They still need Medicare. They also collect significantly more Social Security. The net budget cost imposed by smokers would call for just a very low cigarette tax; far, far lower than the current tax. The other argument is second hand smoke. That does call for regulation, but economic theory suggests that the optimal regulator is the owner of the property where the second hand smoke occurs. The owner of the restaurant, office building, or airline, is optimally placed to regulate the activity, weighing the costs and benefits of a ban. There is no market failure argument here. If second hand smoke in open air was a problem, say in parks, then government regulation would be called for on standard Coasian "transactions costs" grounds. But there is no evidence that second hand smoke outside of enclosed spaces is a significant health problem. This is EC101 economics, but I rarely meet economists who know this stuff.

Comments and Sharing

COMMENTS (44 to date)
Wacky Person writes:

Hey Scott, what do you think about the argument that second hand smoke damages the long-term health outcomes of the children of smokers? There doesn't seem to be an optimal regulator to intervene in the homes of smokers.

Adam writes:

I don't think this post really gets at the meat of Will's argument, which is that libertarians in general don't have a good model _of politics_ which is capable of including libertarians themselves. Their line is often "get politics out of X" without seeing how that itself is an act of participation in politics.

Will of all people would not argue that there do not exist strong arguments for liberty and markets.

Scott Sumner writes:

Adam, I agree with Will's post. This post wasn't really a response to Wilkinson, but rather a response to lots of comments by progressives that I've heard over the years.

Adam writes:

Got it! My mistake.

Joshua writes:

There are three reasons to embrace liberty.

1) The moral - non-aggression principle et al.
2) The utilitarian - it results in better outcomes.
3) Humility - I can't possibly know what's best for you, so it would be pure hubris to attempt to order your life in any certain way.

The strongest libertarian arguments draw from all three of these reasons, but many proponents pick their favorite and don't bother to go any farther.

Of course, the worst libertarians sneer at anybody who doesn't rely on their favorite reason. Nobody tears down libertarians worse than other libertarians.

E. Harding writes:

What about monetary policy? Can you think of any arguments in which we would be better off without a Central Bank?

Tyler Kubik writes:

Coming from an 'extremist' libertarian, I must say I like this post. Thanks!

C. A. writes:


the property argument does not convince me. The toleration of smoking is a social phenomenon. Quoting from an abstract: "Introduction of a smoking regulation may then move society from an initial no-consideration equilibrium to an equilibrium in which a large share of smokers are considerate, even in the unregulated zone." I think this is exactly true for Germany, where I live. 75% of the population does not smoke and I think the situation now is a net benefit for society which would not have occured without regulation. I have to admit that the argument is one of those example where I regard the simple economic/libertarian argumentation as naive.

Steve J writes:

Do the arguments for adopting libertarian policies rest on believing people make rational/good decisions? Does the intelligence of the population matter when we evaluate how effective libertarian policies would be?

Joey Donuts writes:

Nice arguments about smoking and externalities. However, you must not get out much. If you did, you'd have noticed the cigarette butts (mainly filters) that litter areas where smokers congregate and where smokers go in public areas. Its ugly and it costs resources to clean up. How much, I don't know but I'm pretty sure its not 0.

bill writes:

I like this post a lot.
But I have a problem with the argument about regulating smoking. A natural experiment type problem. In theory, before smoking in restaurants or offices was regulated, we should have seen more non-smoking locations so that people could have had a choice. But we saw very little of that. Yet once the regulations were in place, I heard nothing but praise and sighs of relief that we could "finally" go out without stinking from smoke etc. The law overcame a relatively unpopular social equilibrium. If a party of 8 was going out, who wanted to be the 1 person to tell the 1 smoker that they would only go to a non-smoking restaurant? And in light of that fact, very few (none that I recall) non-smoking alternatives existed anyway.

Mr. Econotarian writes:

I also believe that government can nudge social morality in a beneficial way by adjusting the equilibrium through regulation, and that game theory can provide some support for this. Regulation of public smoking and racial discrimination have both yielded not only positive economic results but clear changes in social morality. Likely these have both moved in a fashion that is more reasonable for homo economicus, but may have seemed less reasonable to normal humans living in 1960.

But we also see clear failures (such as bans on alcohol and addictive drugs).

Scott Sumner writes:

Wacky, Perhaps my health was damaged; I breathed in an enormous amount of second hand smoke when I was young. And my lungs are pretty lousy. Interestingly, this is one area the government doesn't regulate.

It's not really a market failure, although obviously there's a paternalism argument to be made.

Adam, You're right, I should have been clearer.

Joshua, Good comment.

E. Harding, Sure, without central banks it's possible that WWII doesn't happen. (I.e. there might have been no Great Depression.) Having said that, there are risks to abolishing the Fed, and right now I'd rather improve it and constrain it and hold it accountable. But there certainly are good arguments for no central bank.

Tyler, Some of my best friends are extremist libertarians.

C.A. My hunch is that property owners would have reached the same decision, in most cases. Certainly my college would have banned smoking even without a law. But there certainly ought to be at least SOME restaurants and bars that allow smoking, and in many cases the law doesn't allow any exceptions (I don't know about Germany.)

Steve, I suppose it depends on the intelligence of the public relative to the intelligence of the government. In areas where people are uneducated, governments tend to be corrupt and incompetent. So even there government is not necessarily a solution to the problem.

Joey, I'll tell you what. Let's agree to tax cigarettes as much as we tax all the other items that people litter, such as candy wrappers. I doubt smokers would complain.

And you are right, I never get out anymore---all my time is blogging. :)

Bill, But then someone could have gotten rich opening a chain of non-smoking restaurants---so why didn't they (you?) do it? I hear those sorts of stories all the time, but I'm skeptical. People say they really hated the smokey environment, but when there is a powerful unmet desire, the markets fills that need. The profit motive is very strong. I think people tend to forget how little most people cared in the old days. I'm a nonsmoker who hung around smokers when I was young. The smoke didn't bother me, and I don't recall it bothering other people. I think it's all psychological, people convinced themselves that smoking was horrible, just like at some point our society decided that dog poop outside was dirty and disgusting and needed to be picked up, whereas a few years before it was natural and no one cared. When I first went to China spitting in public was very common, and no one cared. Now they are adopting the western practice of considering it disgusting. There was no market failure before they cracked down on it, as the people in China didn't care.

Mr Econotarian. What makes you think it was regulation that changed social morality? I have a very powerful counterargument. Almost the same sort of change in social morality occurred in the area of gay rights, with comparatively little regulation. Gay marriage was approved after the younger generation has already changed their attitude.

That's not to say the civil rights laws weren't helpful, it's just not clear to me whether they are cause or effect of the change in social morality.

And one danger of that line of argument is that it can lead to the war on drugs, the war on prostitution, etc.

E. Harding writes:

"I think it's all psychological, people convinced themselves that smoking was horrible, just like at some point our society decided that dog poop outside was dirty and disgusting and needed to be picked up, whereas a few years before it was natural and no one cared."

-Don't extrapolate your experiences to those of most nonsmokers, Scott. When I was a child, I always despised the smell of tobacco smoke without any social conditioning. I lived for a whole year in an apartment in which the neighbors smoked, thus leading the smoke to come through the kitchen sink. It was horrifying. It is from that time I acquired the habit of always closing the door to my bedroom when I go to sleep. I always despised the smell of tobacco smoke on Russian trains (my time in Russian trains was easily one of my most unpleasant experiences in that country, roughly half of that unpleasantness can be attributed to the tobacco smoke). So my thoughts on smoking are the same as those of the Islamic State and a younger Bryan Caplan: ban it, put the death penalty on it. Should be easy to do, with all this new modern particulate-detecting technology.

In any case, it's clear smoking has declined:
How much of this was due to government action, I cannot say. Cowen's second law (on the effects of tobacco bans on personal preferences and visa versa), but I don't have time to check it out.

E. Harding writes:

My longer comment on tobacco smoke is being held for approval. In any case, the war on drugs was a massive failure in terms of changing the public's mind (80+% of the population supported marijuana prohibition in the late 1960s; less than 50% do today), and the war on prostitution has simply led to the weird outcome of prostitution being legal when it's on camera and, thus, is protected speech.

ThomasH writes:

I am not "contemptuous" of libertarians, but frustrated by their apparent view that policy can be made according to general principles. They do not seem to want to grapple with the cost and benefits of a policy relative to feasible alternatives. Policy making without data -- the ultimate free lunch.

[Demand for labor curves slope down, some non-zero reduction in employment will result from an increase in a minimum wage = minimum wages are BAD. Not that aggregate income of low wage workers will fall by $Z, not that the income of low income workers who retain their jobs will go up only $X for every $Y of income lost (and other harms) by low income workers who loose/do not obtain jobs, not that the minimum wage is inferior to other ways of raising the incomes of low-wage workers; just BAD.]

As to why mainstream economists spend less time proposing libertarianish (neo-liberal) reforms today than 15 years ago, I'd say it's the environment. Neo liberal reforms generally remove a price distortion AND ether compensate low and middle-income loosers or assume that the income generated by the reforms will be fairly shared by low and middle-income people. It is becoming increasingly difficult to propose redistributive taxation or public expenditures -- one political party has as its No 1 priority reducing taxes and transfers to low-income people -- and recent data on the distribution of marginal income does not make the outlook for fair sharing very likely.

Clinton could claim that reducing trade barriers would lead to better jobs for most workers and displaced workers would receive generous trade adjustment assistance. No one could argue that with a straight face today.

B Cole writes:

Everyone is against regulations, except for the regulations they are for.

Ask Mr Libertarian if his neighborhood, zoned single family detached, should go to open zoning, apartment buildings and retail stores.

And there are zero female libertarians in the United States. Show to me a woman who supports polygamy and loads of brothels staffed by low cost imports.

E. Harding writes:

"Clinton could claim that reducing trade barriers would lead to better jobs for most workers"

-Did most of the NAFTA-displaced workers get better jobs?

ThomasH writes:

@ E. Harding

Your question supports my point. I'd say that the environment for neoliberal reforms was already not good, even in Clinton's time, but few people knew it. (At least I did not.)

A neo-liberal reform today -- say elimination of the corporate income tax -- would have to have a super sure redistributive component and that would be politically impossible.

Ross Levatter writes:

Scott, could you elaborate on "This is EC101 economics, but I rarely meet economists who know this stuff."

I'm in medicine and quite honestly would be very worried if I could correctly say "This is first year anatomy... or this is second year pathology..., but I rarely meet doctors who know this stuff."

jon writes:

Ross, Scott:

I'm interpret economists not knowing ec101 as meaning they suffer from functional fixedness. They know the ec101 material but struggle to use the tools out of the canonical context.

Johnny writes:

I agree with Scott that cigarette tax should be significantly lower if based purely on Pigovian grounds. However I rarely hear those arguments from economists researching the topic. The current stream is flowing through behavioral arguments, mainly time-inconsistency of consumers in variety of addictive goods. Thus optimal "sin taxes" should actually be even higher than currently (markedly so for cigarettes).

Scott: What is your take on this view?


Don Boudreaux writes:

Much can be said in response to ThomasH's criticism of policy advocacy being guided by general rules rather than by (as he says) "data." Richard Epstein - in his 1995 book, Simple Rules for a Complex World - explained the merits and wisdom of simple rules.

Hayek also spoke eloquently on this matter. On my and Russ Roberts's blog, Cafe Hayek, I have a feature called "Quotation of the Day." Today's quotation, coincidentally, addresses the issue of rules versus discretion in policy making.

Nathan W writes:

"Tens of millions of the world's poorest students..."

While this is accurate at the global level, I think this mis-represents the situation. The people who are poorest in THOSE countries cannot afford private education. It is those who are less poor, or alternatively, relatively affluent (compared to their impoverished peers) who can afford this private schooling.

Due to the problems in public schooling in poorer countries, I think more developing nation governments should be more open to the idea of vouchers which could be applied to private schools (along with rules to ensure that the private schools didn't offer illegal kickbacks to parents who choose their school).

Nathan W writes:

I think the fact that this is all ECON101 stuff is partly to explain for the backlash.

All the neo-liberal stuff is part of ECON101 doctrine in almost every text, which almost never goes even so far as to mention that things can get more complicated. While "too much" rent control, "too high" minimum wages or "too high" taxes will obviously have perverse outcomes, ECON101 teaches that ANY rent control, ANY minimum wage and ANY tax is unambiguously bad for (utilitarian) social welfare. The same logic extends through nearly the entire undergraduate learning of economics, and it isn't until graduate school that SOME economists will take classes which teach sufficiently complex approaches which may find that certain interventions may be beneficial.

Since the entire field has whole-scale adopted neo-liberal orthodoxy, it is only natural to expect a certain backlash where critical thinkers will seek to pick holes in the rather simplistic logic which routinely forgets to mention "ceterus parabus" when teaching about real-world problems in 2-variable 1-period (i.e., highly unrealistic) models.

ECON101/201 is very strong on basic principles which are 90% of the story in 90% of cases (making up numbers). That's useful, and it gives important mathematical building blocks for more advanced work. But it is very often taught in an ideological form, and I think the UNCRITICAL acceptance/teachings of ECON101 logic may itself explain this backlash where many economists then feel that their more important role is to question this orthodoxy. After all, tens of millions are already steeped in ECON101 neo-liberal orthodoxy, so the "real professionals" might be of best use in taking a more critical outlook.

In short, "too many" people take one or two economics courses and think they perfectly understand the economy which is now taught according to neo-liberal orthodoxy (which generally fits well with many libertarian views). In response, it becomes more important for professional economists to critically challenge these orthodoxies.

Thomas B writes:

Having identified libertarianism as my natural bias, I have spent some time looking into the philosophical basis for libertarianism, and the arguments for it.

Where I come out is this: "I am a libertarian" means, in my case, that I argue that public policy should generally adopt a strong, rebuttable presumption of liberty.

Here, "public policy should" means that the state itself is most likely to be successful in its goals in the long run if it adopts this policy bias.

A rebuttable presumption means that the state can deviate from liberty for good reason. This accepts that laws regarding violent crime, fraud, externalities, etc., are legitimate. If 95%+ of the population wants something regulated, clearly it should be. If only 55%, very likely not - although voting, per se, is likely not to be the basis for specific policy decisions (or not often).

A strong presumption is very different to the current "rational basis" philosophy, which says a law is legitimate if there could be any conceivable basis for it. A strong presumption would say the law is illegitimate unless there is a clear and strong public policy argument for it.

The philosophical basis is one of utility. There is overwhelming evidence that liberty - modified, perhaps, in some ways - advances economic welfare across the board, and a strong economy maximizes the state's ability to do whatever it wants to do. Deviations from liberty may further optimize the outcome, but as a guidestar that tends to avoid both tyranny and cronyism (whether "socialist" or "fascist" in nature), liberty is an excellent starting point almost all the time.

Notice that the philosophical basis is not "natural rights". I'm not even sure what those are. I think there is a notion of "natural rights" that emerges from looking at what social conventions - like property rights - have been successful, and what conventions not. Theft, murder and slavery are all entirely "natural", but not very libertarian.

The philosophical basis is also not the "non-initiation of force". Again, property rights make nonsense of this idea: if I forget to lock my doors and wake to find someone has moved into my living room, if I have him ejected there is no "natural" interpretation of "initiation of violence" that does not have me as the initiator.

BD writes:

I myself am one of the rare libertarians in the field of public health. While I agree with your smoking discussion as far as it goes, I don't think it addresses the best arguments for the regulation of tobacco.

Consider that:

(1) Smoking is not a completely voluntarily choice throughout one's life as a smoker. For early smokers, it IS a voluntary choice, but as time goes by various levels of psychological and physical addiction remove some/most/all of the voluntary aspect (it varies by person). Certainly, some smokers are able to successfully quit, but quitters are the clear minority, even of people who say they want to quit or who try to quit.

(2) The people making these early-smoker voluntary choices are not typically adults, but rather early teens (or even pre-teens), raising issues of how much protection we should give minors who have not yet developed full agency or decision-making abilities.

I still struggle with the right answer here on regulation, but I do know that any libertarian-based arguments can't be complete unless they address these issues.

ThomasH writes:

@Don Boudreaux

I don't have a problem with policy being guided by general rules; what else could it be guided by. But when we get own to making or changing a specific policy, general rules are not enough.

Even if we think that creating the FDA was a mistake (in the sense that the people who created the institution did not consider all of the costs and benefits adequately) What do we do now? Abolish it on general principles? Or guided by an understanding of how bureaucracies treat risks asymmetrically, try to reform the rules by which it operates and perhaps the scope over whihc it operates, but with arguments based on what ever data we can generate to design a better set of rules. And repeated reforms might asymptotically get to a virtual abolition if that's the way the data on the results of reform indicate.

And to come back to a proposed increase in the minimum wage, I cannot understand taking a position on a specific proposed increase without having an estimate of how how many people who lose/do not get jobs will lose how much income compared to the income gained by the number of people who retain jobs over whatever time scale one thinks appropriate, compared to alternative ways of raising the incomes of low income workers. The general idea that some number may lose some amount of income (and suffer other harm from being unemployed) is not enough for me.

Pithlord writes:

@Don Boudreaux

Hayek's argument is unsound. From the premise that we know little about the consequences of what we do, it cannot follow that we should use general rules rather than individualized discretion. You would need the additional premise that we would have a better idea of the consequences of general rules than we would of the consequences of discretion. But what possible grounds do we have for that?

I suppose you could argue that societies that follow general rules do better than those that allow lots of discretion. But that sounds like it requires "data", which is apparently bad.

You are left with some faith-based a priori theory from von Mises or something. Which is very convincing to true believers, but I can't see how it has any higher epistemic standing than astrology or Marxism.

Come to think of it, your blog could be Maoist or evangelical if you'd replace the quotes from Hayek with the Little Red Book or the Bible. In all three cases, we are asked to believe and not to question (because that would be scientism or something).

Don Boudreaux writes:

Pithlord: About your first point: You are correct that my argument relies upon the claim that (as you put it) "we have a better idea of the consequences of general rules than we would of the consequences of discretion." I regret not making this point explicit in my original comment. You are correct to identify it as crucial to my argument.

Explanations for why we have a better idea of the consequences of general rules can be found in the works not only of Hayek but also of (to name only a few) Adam Smith, Edmund Burke, Lord Acton, Milton Friedman, James Buchanan, Thomas Sowell, Richard Epstein, and Robert Higgs. You might dispute and disagree with these explanations, but it is an error to dismiss the proposition that rules provide greater predictability than does discretion as if it is obvious that this proposition is without merit.

As for your criticisms of my blog, I invite you to read it more carefully. I believe that if you do so you'll discover that Russ Roberts and I spend a great deal of time there offering argument and evidence rather than, as you suggest, merely quoting in a conclusory fashion the works of Hayek and other market-oriented scholars.

Pithlord writes:

@ Don Boudreaux

Thanks for the reply. It makes me feel bad about being rude. But I still think there's a bit of a contradiction in the argument you and Roberts make. If we should be radical skeptics about the possibility of empircally resolving discrete questions like "how much would a $.50 increase in the federal minimum wage reduce employment among the low-skilled" then surely we should be even more skeptical about empirically resolving bigger questions like whether discretion is always worse than rules. And if there is no *empirical* answer to this, then you are left with arguments from authority. Very learned authority it is, but then it would be equally easy for a progressive to rattle off a list of great thinkers who thought otherwise.

Once you despair of data, you are left with faith.

Pithlord writes:

I could kinda see how "nobody knows anything" could lead to status quo bias. After all the one thing we know about the status quo is that it's no worse than the status quo. What I don't see is how it leads to radical libertarianism, since the one thing we don't know about the EPA, Social Security and the requirement of emergency rooms to treat all comers is what would happen if we got rid of all those things.

James writes:


You point out the EC101 story is mostly true most of the time. This is surely correct but, sadly, a lot of people would leap from this observation to the conclusion that the right amount of government intervention in the economy is considerably more than the amount of intervention we observe right now.

From a purely utilitarian perspective, government intervention is only when it improves the situation which requires that three conditions hold:

(1) The EC101 story breaks down. (2) The deviation from the EC101 story leads to a welfare loss (Distracted driving and volunteering at a food bank both represent departures from EC101). (3) Government intervention is likely to improve on the scenario (Politicians might ban coal plants next to schools or they might use clean energy subsidies to remunerate their sources of campaign funds).

Given your numbers, the probability of all three of these conditions being true must be less than 10%. When people claim that the government should be intervening in the economy even more than it already is, I have to wonder what probability they assign to each of the conditions above and if they know how to multiply.

Scott Sumner writes:

Ross, You said:

"I'm in medicine and quite honestly would be very worried if I could correctly say "This is first year anatomy... or this is second year pathology..., but I rarely meet doctors who know this stuff.""

Well it's important for doctors to understand statistics, but based on what I've read they don't. So I think this sort of problem is pretty common in all fields. But I share your frustration.

And see Jon's comment.

Don, I'm a big fan of Epstein's work.

Nathan, You said:

"While this is accurate at the global level, I think this mis-represents the situation. The people who are poorest in THOSE countries cannot afford private education. It is those who are less poor, or alternatively, relatively affluent (compared to their impoverished peers) who can afford this private schooling."

I don't think that's accurate---did you read the article I linked to?

You said:

"ECON101 teaches that ANY rent control, ANY minimum wage and ANY tax is unambiguously bad for (utilitarian) social welfare."

That's just totally inaccurate, not even close.

Nathan W writes:

Scott, perhaps you are familiar with different ECON101/201 texts than I have encountered. In the earliest course, we were always shown a visual representation of the welfare loss to any wage controls, rent controls, taxes, tariffs or quotas. At the next level, at times we were then required to calculate the welfare loss of the intervention. It strikes me that the numerical examples are intentionally calibrated to make precisely these points. I think this is pretty normal in teaching ECON101/201.

I would be quite relieved to find that I'm wrong, but I think you are very over-optimistic in claiming that this is totally inaccurate, not even close to how ECON101/201 are being taught. It is precisely how I was taught, and the fairly wide existence of people who are obviously somewhat economically literate and who also expound precisely these conclusions suggests that they did not draw these conclusions entirely on their own.

As for the "poorest" being able to access education. I don't think the linked article provides sufficient breakdown of income, or other measures of poverty, of the attendees to draw strong conclusions about whether the "poorest" are accessing these private schools. Since the relevant countries have such expansive poverty, perhaps we can extend "poorest" to being anyone under median income, rather than the poorest 20%. But still, in many cases in Africa, median income is still in the range of $1-2 per capita per day, sometimes less. While we should welcome that students in families with daily per capita income of, say, $5 per day might "easily" afford $35 a month (an example cited in the link) for better schooling, this remains utterly inaccessible to the poorest quintile in many countries, not to mention that these schools cannot achieve sufficient scale to operate in rural areas (the poorest areas).

I was very optimistic in reading the article when it came out, and even more so in combining this with your discussion of school vouchers, but I think it is overselling it to suggest that the benefits are going to the "poorest", according to relative national measures. Of course, compared to poverty in America, this can be unambiguously praised as more and better education for a very low cost for the global poor (which in some countries includes even the top quintile).

Thomas B writes:

I agree with Nathan: basic economic texts show that any tax is harmful.

Of course, basic economic texts assume property rights and honest dealing. The cost of enforcing both is assumed to be zero. If you assume that it's not zero, then it is at least possible that a third party specializing in enforcing those rights, paid for with a tax, may be a socially better (lower deadweight loss, lower cost) solution than self-enforcement. In fact, it seems pretty likely to be true.

I am unaware of any realistic assumptions that make rent control or a minimum wage better than none. There are some theoretical ones having to do with monopsony power, but that's not remotely realistic in any of the situations where these policies get applied. Nathan, your comments suggest you do think that both could be justified - why?

Nathan W writes:


Rent control - once you move in, people don't like to move, so landlords will try to game this and demand higher rents of people who live in the apartment than they might get at the equilibrium price for a new renter. Obviously, the person can just move (in the case of "too high" rent increases), but this has a high transaction cost (some monetary cost, but especially the emotional/time cost of preparing to move and going to a new place). While landlords always want to avoid vacancy, this gives them an inherent advantage in the absence of rent control.

The only rent control that I am particularly familiar with is in Ontario. It allows landlords to increase rent each 12 months after signing the original lease (they are also disallowed from demanding that you sign a new lease), and the maximum rent increase is a few percent, and is legislatively modified each year. Generally this is perceived a protecting low income people from high rent increases in a situation of gentrification. For example, I lived in Kensington market in Toronto when going to university - this was during a period somewhat early on gentrification there. My rent would easily have increased by double digits without the rent control, but instead could only increase by a few percent. I'm not sure if the specific case passes the test of being "economically optimal", and I believe that most likely most rent controls are "too strong" to be "economically optimal" when strictly considering investment in many Western cities, but they are often perceived as "fair" because poor people are not forced to leave their neighbourhoods when market prices of rents go up. At the aggregate level, someone can always build another property somewhere else and charge market rates for new renters, and in the meantime, it is commonly observed that properties to old renters in newly gentrified neihoubourhoods are very poorly maintained, because the landlords never do more than the minimum by law, because they WANT you to leave so they can rent to new renters at the new and higher rental rate.

I would consider this to be a somewhat weak case, but between a) transaction costs of moving which effectively enable landlords to sneak in higher-than-market increases, and b) the high social cost of being forced out of ones (potentially life-long) home during periods of high rent increase due to gentrification, I think it is easy to argue that SOME level of rent control can be justified from the welfare perspective, so long as the regulating body is not politically captured by renters (leading to a "too low" rate of permissable increase) or landlords (who may be donors to political campaigns, hire good PR people, or more generally be more highly engaged in the political process).


Minimum wages:

I take two views on this. (And first of all, I would point out that I think that it is generally bad to implement any minimum wage hike which exceeds inflation during a weak job market.)

The first is the short term perspective. You have supply and demand. The most simple perspective which could legitimize a minimum wage is that the supply effect (more workers looking for work) can be stronger than the demand effect (at any given price, employers may pay less). This is like saying that the employment market is not truly at full employment AND that the supply curve has a stronger slope than the demand curve. Perhaps there are quite a lot of people who might like to earn more money, but when all the advertised positions that they feel qualified for are at $8 an hour, they might not bother to apply. (Perhaps stay at home moms spend more time with children, university students study more and work less, if they can afford to, or semi-retirees who might prefer to spend more time watching their gardens grow rather than work for low wages). To fabricate some numbers, perhaps at $8 an hour you have one million job seekers, and at $10 an hour guaranteed minimum, you have 1.1 million job seekers. After all, looking for work is hard work, but if the government has negotiated a higher minimum wage for you, more people will try. Anyone who runs a business knows that it's hard to find good talent. Out of those 100,000 additional job seekers, perhaps they can find sufficient BETTER applicants to fill their positions that they may ultimately employ more people at $10 an hour than at $8 an hour minimum wage. Fanciful thinking? I dunno ... this effect must exist, but whether it outweighs the downside is ultimately empirical and very very difficult to prove how many better candidates are drawn out of the woodworks. The positive effect only needs to outweigh the negative effect of losing all marginal positions (which produce between $8 and $10 per hour of labour productivity), and it is theoretically possible to have both more positions AND higher wages at the same time. This would depend on the dynamics of a particular job market, and clearly does not hold at all times and places.

Also, people are comfortable with the familiar, and lack information, and often will not seek better jobs when they can theoretically get them. Transaction costs are very high in a job search, and busy people may simply feel that they lack the time to search for better work, even if the actual conditions of the job market are such that they "should" look for better work. While in a perfect market, we should anticipate that no worker will ever toil at $8 an hour if some theoretical market equilibrium suggests they "should" be able to find better paid work at $10 an hour, as a matter of fact, people get stuck in routines, perhaps working the same low-paid job for their entire lives even if they COULD do better. As time passes, they are almost certainly matched with technologies which, due to complementarities between capital and labour, increase the value of their labour over time. But absent a minimum wage hike, they may easily fail to a) negotiate for their share of this increased productivity or b) move to a similar position in different firm where they can obtain a higher wage for this now-more-valuable labour.

Also consider that it is very difficult for an hourly minimum wage worker to negotiate for a higher wage. Perhaps the employer profits $20 an hour from their input, but only pays $8. (Due to competition between firms and workers, this is theoretically impossible in a perfectly competitive market, but I think this is true of quite a lot of positions.) Aka, this is not a marginal job position. Then the minimum wage goes up to $10 an hour. The employer will retain every employee because they still earn profits from the employee's labour. The actual "welfare effect" is still ambiguous, but I think due to declining marginal utility, it is generally accepted that marginal dollars produce more social welfare in the hands of the poor.

I also think that people generally work harder when you pay them more. How hard would you work for $1?

Another general short-to-medium term factor is the ability of firms to pass on minimum wage hikes. A gas station, for example, can almost certainly pass on close to 100% of the wage hike to consumers (gas demand is very inelastic). What about McDonalds? (I dunno, this will very for different markets).

The second is the long term perspective. I concede that it is likely that in some (many?) job markets there will unambiguously be some job losses, and in some job markets these losses will wipe out the gains to low income workers as a group, who, in this situation, will earn fewer total dollars as a group. Refer to the short term arguments though. This will not always be the case.

In the long term, I'm thinking of productivity growth, particularly at the bottom of the wage scale. The productivity of a labour position is basically related to two things: the inherent productivity of the worker, and the production structure within which the worker is employed (the innovator of a more productive production structure may reap some super profits for a while, but then as this spreads, it becomes a part of the general productivity inherent to the system, "available" to anyone who runs a certain type of business - consider the Toyota just in time model, for example). Here, it is critical to consider labour as a complementary input to capital - the inherent value/productivity of the labour when coupled with new technologies can rise, even without changing the properties of human capital itself. When you have a minimum wage, basically you ban low productivity positions. If you want to work, you must exceed the minimum wage in productivity. If you want to run a business, you must create positions where labour productivity is above the minimum wage. The basic argument is that, in slowly increasing the real minimum wage over time, you force both workers and employers to invest in capital, be it human capital (workers' investment) or physical capital (the firms' investment). In banning low productivity options, this can stoke productivity gains in the long run.

Consider the "threat" that fast foods joints will automate some/many positions if the minimum wage is "too high". Since when is a threat to the economy to increase the level of technology? But anyways, the debate usually revolves around whether net benefit accrues to low income workers as a group, not society as a whole. In cases of replacing labour with technology, this unambiguously leads to job losses in the shortest term. But aren't economists generally optimistic of the ability of an economy to find new uses for an idle resources? (Refer to concerns of ZMP workers, or taking automation to the next level, for some counterarguments...)

In the long term, you can also consider some other positions, like hair dressers, which almost certainly won't be replaced by robots any time soon, and don't really become more "productive", in my opinion. Well, I simply argue that since people are generally in higher paying positions, we can afford it.

Also in the long term, consider a more typically left-wing demand-side argument that better paid workers contribute to increased aggregate demand (an argument that is critically contingent on the net wage income of the group increasing as a whole). After all, they will say, where's the profit if no one can afford to buy anything?

Finally, please note that I do not argue that minimum wage hikes will in every circumstance lead to these positive outcomes. What is important is a) these arguments are entirely plausible, and b) they offer some potential explanations for the large number of failures to empirically validate the assumption that minimum wage hikes must lead to job losses.

As an aside, I would like to point out a common view of the matter in Australia, where minimum wages are among the highest in the world. Regardless of economic optima, or what have you, it is exceedingly common for people to believe that a hard day's work "deserves" good pay, regardless of how much talent it takes. Now surely it is better for the economy in the long run to match skilled bobcat operators with bobcats than to keep 100 labourers busy digging dirt with shovels. After all, 99 workers are now available for other projects (or, say, perhaps 90 workers after accounting for the labour implicitly built into the bobcat, now considered as capital). But the Australian valuation of the matter is very commonly that some who works very hard digging dirt has very much earned their keep regardless of what final price this may imply for the final market output.

Oh yeah, and these are surely incomplete explanations. It's just what I can dream up on the matter :)

Floccina writes:

When progressives assume that the lowest income cannot affords something they often forget to factor in reduced demand lowering the price.

One example is that in a market for housing with restrictions on building (adding housing units) and where housing is subsidized, if you remove the subsidy the landlords will take a somewhat of the hit.

Another example is that as far as I can tell, even adjusted for inflation, our USA schools spent 3x as much as in the 1960's and output is not much improved. Now I am sure that the experience for the student is better now but I would bet that you could cut spending hugely and get pretty much the same output. And if parents payed directly you could possibly get better output due to motivation and for some other plausible reasons.

Welfare also seems to not make as much difference compared to what most people would expect. Take SS we only need it for low savers who are to weak to work who have not family or friends to care for them. That is a small percent of the population but to reach that small percent of the population without creating very bad incentives we have to give SS to everyone. A huge cost relative to the problem. BTW I think we should give every USA citizen over 67 $200/week from SS, reducing the program cost by about 1/3rd. BTW I think absent SS most people would start to save as much as possible at about 60 years old and work as long as they could and do OK.

Also I once saw a show on PBS about a guy who built a road to Miami Beach.

Narrator: One of his greatest ambitions was to persuade the government to start building better roads. He convinced the leaders of the automotive industry to jump-start the process by financing the first paved road across the country -- the Lincoln Highway, from New York to San Francisco. It was such a success that he went on to build another road -- the Dixie Highway... leading, very conveniently, from Indianapolis to the foot of the bridge he was paying for back in Florida.

Floccina writes:

On smoking, it was already declining and had Government not banned it from public places, in a few more years a movement might have started boycotting restaurants that allowed smoking leading to a domino effect.

Robb Lutton writes:


I really like you and your blog. Unlike many libertarians, such as the crew at Marginal Revolution, I sense a human warmth and an intellect that searches for answers and not cool contrarian positions.

So I am disappointed in your attacks on Government regulation of smoking. It is a tremendous victory for government coercive action that most people are very happy for. If you don't understand that, at least in the US, the overwhelming majority of people, smokers and non smokers alike love the results, I wonder who you hang out with. Or is it just "principle"?

But enough of that. Lets talk about the civil rights movement and the armed crushing of the segregationist resistance. How do you feel about that? still looking for a market solution?

Thomas B writes:

Hi Nathan,

Gosh, you did have a lot to say!

A few thoughts jumped out.

I wonder if you've ever thought about rentals from the landlord's point of view? Vacancies are a VERY big deal, and can quickly destroy several years' worth of profits. As a tenant, I make sure to pay on time and, when renewal time comes, I point that out to the landlord as I ask for a below-market raise. Which I almost always get, because the landlord faces losing an on-time tenant, painting and vacancy cost, the cost of finding a new tenant, and the risk that the new tenant may be a problem tenant. In fact, in New York - LOTS of rent control - I once had a landlord strictly enforce the lease to keep me from leaving, even though he could raise the price right away if I did (being on time can backfire...)!

Several of your minimum wage claims amount to asserting that if employers offered higher pay, they'd find and have better, more productive employees. Of course, the employer is already incentivized to figure that out, and it's unlikely the government can come up with one optimal number that's better for all employers/employees - or even, given limited information, better for many of them.

The difficulty in measuring the effect of minimum wage lies in the fact that, most of the time, governments keep it close to the minimum market-clearing wage anyway, so only a few very low-paid people lose their jobs and it gets masked by other events; and they phase it in slowly, making it even harder to see what's happening - which they would not do if it were not for the fact that they already know (or strongly suspect) it causes unemployment, and they don't want the blame for that! Finally, even without phasing-in, it does take time for employers to adapt, but they do.

The "pay people more and boost the economy" argument suffers from Bastiat's "seen and unseen" objection: you notice the extra money given to Paul, but don't ask about the effect on Peter. When a low-wage employer pays more, its prices go up and its profit margins go down. Its customers and the retirees who own its stock become poorer, and have less to spend. The only way to boost the economy is to raise productivity: artificial wealth transfers may or may not increase fairness, but they don't boost the economy. Henry Ford didn't pay his workers more so they could buy his cars: many (most?) of them still couldn't afford his cars. He paid them more because he needed skilled workers, and they were expensive.

Now, this does bring us to your "make low-wage jobs illegal as a way to boost productivity" argument. It's an interesting one, and somewhat appealing: it conjures the country as a giant gated community where, if people have few skills, they're just not welcome. But it seems rather cruel: what ARE the poor, whose labor simply isn't worth much, to do? Is prohibiting them from helping themselves, condemning them to lifelong out-of-sight poverty (not working, where we might see them), but keeping them from starvation with welfare, really closer to a social optimum? Is destroying poor people's jobs with robots, faster than the market would have done by itself, really optimal? And, what about the optimum of the poor people themselves? Why do they have to pay the price for that nice gated community the affluent left wants?

I've never lived in Australia. Visited once, enjoyed it very much. Can't comment on how the economy there works, though!

Nathan W writes:

Thomas - thanks for your well-considered reply.

Yes, I know that vacancies are a big deal for landlords. Sorry for the omission - this obviously serves as a natural check against extreme hikes. All places I have lived in the last decade or so have experienced very high annual rental price increases - perhaps double, triple or in some cases perhaps five times inflation. This offers the classic Ricardian definition of a rent - the property value increases, offering a "free" or un"earned" gain for the owner, and in the absence of controls, the landlord would naturally (profit maximizer) try to capture the full gain from existing renters, who cannot get a better price by moving to a new apartment. Do we live in peculiar times with regard to real estate prices and rents? I suspect so. In the long run, I expect market rental prices would increase at a fairly "normal" rate. But in the markets that I have experienced in recent years, and in particular having fairly low income, I very confidently assert that rent controls have protected me from landlords hiking rents in a context of rapidly rising real estate and rental prices - and that, without perverse effects on long term investment in the broader market which is very free to invest in new properties for rental at market rates.

I unreservedly accept all of your critiques. I'm just saying that it's not an outlandish idea to say that these CAN be beneficial. It will depend on the given rental/labour market, including segmentation of these markets, the ability of groups to organize for their respective self interest, and prospects for the longer run.

The point about drawing people out of the woodworks is this. It takes a lot of effort to find a job. I might have to sift through hundreds, thousands or even tens of thousands of job postings, and might take additional effort to apply to tens, hundreds or thousands of positions before landing a job. If I am GUARATEED to get at least some higher wage (a higher minimum), some people who couldn't be bothered at $8 an hour WILL bother if they are guaranteed at least $10 an hour. You counterargue that an optimizing firm would in any case bid up prices to find talent. I suspect that a lot of minimum wage employers may suffer from being penny rich, dollar poor, and do not engage in this thinking, but my logic doesn't require any consideration of THAT problem (which wouldn't exist in a world of perfect optimizers). The logic relates more to making it worth it for people to look for work, stimulating a very broad based supply response. If just ONE or TEN businesses increases wages to $10 an hour, it's still fairly hard for the job seeker to become matched with this position - while the firm which increases entrant wages most likely attracts the better job applicants out of the EXISTING pool of job seekers, it will fail to draw out additional job seekers. However, when the entire market shifts to a higher wage, the entire pool of job seekers itself may increase - new talent comes onto the market (mothers, semi-retirees), more people find that it's worth it to look for a second job, etc. I myself was recently in this position. I freelance translate academic research, but the volume of business is low, so I was looking for a part-time job to supplement income. If I knew that I would make at least $15 an hour, I would still be in Canada. But it was so much effort to search for work for piddling pay, that I simply left the country and now work overseas. I would like to think (assuming that I have some valuable skills to offer) that this is a loss for the job market, especially if multiplied many times over.

I think the comment that "artificial wealth transfers may or may not increase fairness, but they don't boost the economy" is worth additional comment. This almost certainly must be true in the short run. I am quite optimistic about the potential that this can contribute to longer run productivity gains, but indeed, I am very concerned about these very low productivity workers who are essentially blocked from achieving the sense of self-worth that comes from paying your own way. The hair dresser can simply raise his/her prices, and no matter, we can afford it. But I concede that there is very likely to be some residual body of people who are employable at, say, $2 an hour, but not in a million years will never have the brains, motivation and/or social skills to EARN $15 an hour. This bothers me too. On the dark side, one could fear that potentially some will become completely disenchanted with the system and become tomorrow's terrorists and mass murderers. I weakly point out that they might have freedom to engage in a high amount of volunteerism, production of medium-grade art and music output, or might spend very high effort ensuring that their children get the very best of opportunities.

Anyways, I would like to emphasize that the point is to challenge the dogmatic response, not to assert the truth of my claims in some/all labour market environments. I generally think that any argument for intervention should be very strong whether for "efficiency" or "justice" reasons. Having challenged the dogmatic reflex, by no means do I suggest that this in itself constitutes a winning argument for intervention in a given socioeconomic time and place.

Thomas B writes:


It strikes me, as I read, that the problem that arises with rent, in particular, is that the shortage is created by legal land use restrictions that prohibit the market from responding other than by raising prices. The landlords (and homeowners) get an artificial gain from the price increase, and so they lobby to maintain high barriers against a market response.

What we have, here, is not a market failure, but a political one. Not that that helps the tenant at all.

Worse, at the margin, given the existence of a cabal of politicians, environmentalists and property owners to block additional development, it then appears socially optimal for the rents to go up, and for existing tenants to be "gentrified out" of their homes. Why? Because, given the political limitations on home construction, the homes that do exist should be allocated to those who value them most - most likely to those who could have high-income jobs in the area if they could find somewhere to live.

It is true that low-income owners are not compelled to move away as quickly as renters, in this situation. But the logic of opportunity cost comes to bear here: you can sell your home, move away, and "live like a king" somewhere else. In the end, owners who don't place a high value on that location will do exactly that.

What rent control does is to protect incumbent tenants at the expense of economic development in the area: high-value occupants cannot move in, and don't. The property is kept in lower-value uses, i.e., occupied by people who don't value it as much as new tenants would. Frankly, I can think of some situations where this might be a good thing: arguably, cities like New York and San Francisco would have been the "winners that took all" even more than they are now, if apartment turnover had not been blocked. Instead, high-value jobs have had to move to "second-tier" cities because new employees couldn't find anywhere to live.

Of course, if lower-paid workers are being priced out of their homes, and start to move away, it will raise the pay rates for the ones who remain.

So, rent control contributes to income inequality: it allows low-end pay to stay low, while forcing employers to pay even more than they would otherwise have had to do, for new high-end employees.

The "drawing people out of the woodwork" point is interesting, and thanks for that. So is the productivity one, which I am less confident about - even in the long run - but it's worthy of further thought.

Thanks for the conversation!

Peter Gerdes writes:

[Comments removed pending confirmation of email address. This is your final notice. We have attempted to contact you several times previously. Email the to request restoring your comments in this and other threads, and your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Comments for this entry have been closed
Return to top