Sunday, March 31, 2013

On freedom-hating and the merits thereof

Scott Alexander has written what I consider the best critique of libertarianism on the web, titled "Why I Hate Your Freedom." He's joking, of course, he only hates some of your freedom. It's good because 1) It's not full of straw men, 2) He's not an economics ignoramus, and 3) It's not full of bogus economic history that competely ignores the academic literature. I agree with the majority of it.

As someone with a couple economics degrees, I think I have a few relevant comments to make. They do not substantially question the main thrust of the FAQ. Quite a bit of this is nitpicking. Be warned!

The numbered sections in this post correspond with the same sections in his, and match their order. If some numbers are skipped, it's not because I can't count, it's because I had nothing to say about that section.

1) Externalities: The first 4 of the major sections are on economics. Section 1 is about externalities: Free trade does not guarantee optimal outcomes for everybody just because they're all making the trades that they want. Their trades affect other people, sometimes negatively and sometimes positively. Both of these can be used as arguments for government intervention: negative externalities get taxes, positive externalities get subsidies.

It's worth noting that we ignore the vast majority of externalities. If I wear an ugly shirt, and you look at it and are displeased, I and the shirt vendor have imposed a negative externality on you. And nobody gives a crap. Economists recognize, and I'm sure most would agree, that the costs of passing regulation against ugly shirts far outweigh the benefits.

Many (most?) libertarians accept the existence of externalities. The moderate ones just think that most externality regultions in place are higher-cost or lower-benefit than the typical person believes. Why? The Coase Theorem! The Coase Theorem says that, in a world of zero transaction costs and well-defined property rights, property always goes to whoever values it most. If you have a world where you can get paid $10,000 to join the army, and a near-identical one where everyone is automatically drafted and can pay $10,000 to get out, if both worlds have zero transaction costs, the same people will be in the army in both worlds: those who value $10,000 more than the disutility of being in the army. The only thing affected is their wealth; in the second world, everyone who opted out of the army is poorer than the people who never joined in the first world.

Of course, we don't live in a world with zero transaction costs, so efficient outcomes from the Coase Theorem aren't usually relevant. I'm bringing it up because part of Coase's insight was that externalities result from ill-defined property rights. All externalities are bi-directional. Getting drafted robs you of your time, but not getting drafted robs the army of a soldier. There's no prima facie case, in efficiency terms, as to where the property right should lie. Even if you really think I should get my smog out of your air, I may really think you should get your air out of my smog. Or less glibly, that my driving is more important than your mild cough.

So deciding which externalities to regulate is difficult. We have to resist the temptation to pick a default property right from a given narrative. We tend to assume polluters are bad, as a default, but people complaining about pollution are bad to polluters. Thus, some regulations might be voted in because we have underestimated the utility of one party.

Of course regulations should be on a case-by-case basis, and of course doing the real empirical work is hard. It says as much many times in the FAQ. I just think any discussion of externalities is incomplete without Coase, and I disagree that the existence of externalities alone justifies various regulations. You need a bit more argument to get to taxation and regulation.

2) Coordination Problems: In econ, we call it the tragedy of the commons. It's just an extreme version of externalities. It arises from the same problem of ill-defined property rights. If an action would hurt everyone a little bit but help herself a lot, most people will take that action. When most people do it, everyone ends up hurt a lot.

From the FAQ:
2.1.1: Even without anyone declaring himself King of the Lake, the fish farmers would voluntarily agree to abide by the pact that benefits everyone.

Empirically, no. This situation happens with wild fisheries all the time. There's some population of cod or salmon or something which will be self-sustaining as long as it's not overfished. Fishermen come in and catch as many fish as they can, overfishing it. Environmentalists warn that the fishery is going to collapse. Fishermen find this worrying, but none of them want to fish less because then their competitors will just take up the slack. Then the fishery collapses and everyone goes out of business. The most famous example is the Collapse of the Northern Cod Fishery, but there are many others in various oceans, lakes, and rivers.

If not for resistance to government regulation, the Canadian governments could have set strict fishing quotas, and companies could still be profitably fishing the area today. Other fisheries that do have government-imposed quotas are much more successful.
To which I respond: Empircally, sometimes. The late Elinor Ostrom, a political scientist who won the Nobel prize in economics in 2009, did lots of on-the-ground empirical work as to how small communities solve the tragedy of the commons. She studied what policies these communities tend to share. Here is a nice summary of her findings. Note that the commonalities do not usually involve anyone labeling themselves King of the Lake. They do, however, resemble small governments. Importantly, one of the commonalities is "Rules governing the use of collective goods are well matched to local needs and conditions." This is an argument against very broad government regulations. Essentially, it's another argument for federalism. As size of the commons scales up, government should intervene more.

So for, say, global warming, regulation might be a better idea than it would be for local grazing land.

Coordination problems also aren't quite enough to get you to some of the regulations the FAQ endorses. For instance, labor unions are a coordination problem, but there are other relevant factors besides the boss-employee relationship. Workers compete against other employees, not against employers. Their coordination puts lower-productivity employees out of work. That is, unions are a monopoly or a cartel that exist to prevent employers from offering wages that are too low. To people whose skills can only earn them low wages, the job of the union is to force the employer to take on someone higher-skilled. If you take the marginal utility of income seriously (the argument used in the FAQ to justify progressive taxation), you need to make a stronger argument for unions than simply citing coordination problems.

This is why the fact that "getting fired" makes people really sad isn't itself a great argument for regulation. Not getting hired to begin with also decreases happiness.For more info and literature reviews, I recommend What Do Unions Do? A Twenty-Year Perspective, edited by James Bennett and Bruce Kaufman.

3) Irrationality: Bias in regulators can be a big deal as well. For instance, it seems to me regulators have vastly overstated the health effects of some illegal drugs and hugely underestimated the costs of illegalizing them. For instance, non-addictive non-toxic LSD is considered just as bad as heroin in the eyes of American law, and police regularly murder people for drugs less dangerous than either of those. The drug war is an obvious example, but we could quibble about others. Also see this Robin Hanson post on regulator bias and libertarianish alternatives to paternalism.

4) Imperfect Information: Just a little nitpick. From the FAQ:
No economist literally believes consumers have perfect information, but there are still strong arguments for keeping the "perfect information" assumption. These revolve around the idea that consumers will be motivated to pursue information about things that are important to them. For example, if they care about product safety, they will fund investigations into product safety, or only buy products that have been certified safe by some credible third party. The only case in which a consumer would buy something without information on it is if the consumer had no interest in the information, or wasn't willing to pay as much for the information as it would cost, in which case the consumer doesn't care much about the information anyway, and it is a success rather than a failure of the market that it has not given it to her.
Economists don't typically keep the perfect information assumption. They assume imperfect information + decision theory. That is, consumers attach probability estimates to corporate malfeasance (or whatever) and a utility (or disutility) estimate to supporting that business or buying a lousy product. Adjusting for risk aversion, they make a decision whether they want to spend resources to pursue such information. This doesn't affect the discussionat all, but hey, give us economists some credit! At least this theory actually accords with the case that follows. Really I am only picking up on this because many people criticize perfect information and we honestly rarely use it.

As a side note, reputations are becoming easier to monitor with the rise of the internet. Consumer Reports is now mostly free, and Amazon reviews give lots of info where there previously was none. I don't think the world would be quite as bad as the FAQ imagines, though I still have no problem with many regulations on this front. I'm a vegan. I don't want to send emails to every single company to find out if X product has casein or whey in it (and then be lied to without any fear of punishment). There are so many food products, no website could possibly keep track of them all.

6) Taxation: First and foremost, I just wanna say that I went through about 7 years of econ (undergrad + grad) and not once in any textbook or class did anyone ever mention "trickle-down economics." It's a political term used to describe several policies that each have varying effects on the income distribution at different periods (for instance during a recession or a boom). That's a rant for another time, I guess. It's just frustrating that, when criticizing conservativeish economic policies, people bring up this useless term. Did not so many conservatives and libertarians themselves bring it up, I'd call it a straw man.

Second, a quibble about deficits and spending. From the FAQ:
6.5: Raising taxes would be useless for the important things like cutting the deficit. The deficit is $1.2 trillion. The most we could realistically raise from extra taxes on the rich would be maybe $200 billion. The most we could raise from insane levels of extra taxes on the rich and middle class would be about $500 billion - less than half the deficit. The real problem is spending.
Yes and no.
The deficit is, indeed, very, very large. It's so large that no politically palatable option is likely to make more than a small dent in it. This is true of tax increases. It's also true of spending cuts.
Cutting all redistributive government services for the poor including welfare, unemployment insurance, disability, food stamps, scholarships, you name it - would save about $200 billion. That's less than 20% of the deficit. Cutting all health care, including Medicaid for senior citizens, would only eliminate $400 billion or so. Even eliminating the entire military down to the last Jeep would only get us $800 billion or so. The targets for cuts that have actually been raised are rounding errors: the Republicans trumpeted an end for government aid to NPR, but this is about $4 million - all of .000003% of the problem.
So "darnit, this one thing doesn't completely solve the deficit" is not a good reason to reject a proposal. Solving the deficit will, if it's possible at all, take a lot of different methods, including some unpalatable to liberals, some unpalatable to conservatives, and yes, some unpalatable to libertarians.
In particular, we need to avoid the "bee sting" fallacy, where we have so many problems that we just stop worrying. It would be irresponsible to say that since a few billion dollars doesn't affect the deficit either way, we might as well just spend $5 billion on some random project we don't need. For the same reason, it would be irresponsible to say we might as well just renew tax cuts on the rich that cost hundreds of billions of dollars each year.
I'm not sure I can blame this one on the author at all. There are lots of deficit hawks on all sides, and libertarians are just as guilty. Before I explain my problem, we need to review some econ.

There's this idea called "Ricardian equivalence." The short story is that government debt is the same thing as taxes. I'll let Steven Landsburg explain:
Suppose the government owes $100 and pays $3 a year in interest. The alternative to paying that interest is to raise current taxes by $100 and pay down the debt. If you do that, taxpayers are going to have $100 less in assets, and will therefore earn less interest on their savings. That costs them (roughly) the same $3 a year.
In other words, the damage was done back when the government spent that $100 in the first place. (Of course, if the $100 was spent wisely, the damage might have been worth doing. Or not.) Once that $100 has been spent, the taxpayers are out $3 a year forever regardless of whether the debt is ever paid off.
That’s why I say that the government’s interest payments come right back to the pockets of American taxpayers. The government pays $3 a year as an alternative to taxing you $100 and paying down the debt. The choice to do that puts an extra $100 in your savings account, which earns you $3 a year. There’s the $3 a year coming right back to you. Notice that it comes back to you regardless of whether the government makes its interest payments to Americans, Chinese or Martians. All of the benefits come back to American taxpayers.
Of course, you might choose not to save that $100 the national debt is saving you. That’s fine. Then presumably you’re spending it on something that you value more than an interest flow of $3 a year. Congratulations. You’re a winner.
Or you might grumble that you have no savings vehicle that will pay you the same rate as the government’s paying on its debt. That’s where you’re wrong. You can save by buying government bonds. That will get you exactly the same rate the government’s paying on its debt.
 But doesn't deficit spending force our profligacy onto future generations? From another post:
I was therefore maximally sympathetic to the poor XM radio host (I think it was Pete Dominick but I’m not sure) who was stuck interviewing a man named John Sakowicz last Friday. Sakowicz, who hosts his own radio show in northern California, was there to warn about the dangers inherent in our growing national debt. He was very clear about this much: the debt and its associated dangers are massive, explosive, perhaps even apocalyptic. He was entire unclear, however, about exactly what those dangers are.
Pressed for an explanation, Mr. Sakowicz rather breathlessly announced that every child born in America today is born with a $45,000 share of the national debt. (He should have said the average child and $45,000 is probably not the right number, but those are minor quibbles). The host, bless him, asked exactly the right question, namely “What does that mean?”. To which Mr. Sakowicz attempted to clarify his meaning by repeating the $45,000 figure in a considerably more agitated tone of voice. To which the host calmly replied: “Okay, but what does that mean? Take my daughter, for example. Exactly how does this affect her life? Does it meant that she’ll pay that much more in taxes…..or what?”. To which Mr. Sakowicz replied that $45,000 is a really big pile of money.
Well, yes, $45,000 is a really big pile of money, but the host’s question was exactly on target. When we say that an American child is born owing $45,000, (or $40,000, or $350,000) what does that mean for the life of the child?
Answer: Pretty much nothing. What matters to your kids is not their share of the national debt, but their overall inheritance. There are two parts to that inheritance. First, there’s what your kids get directly from you. Second, there are the factories, machines and tools that other kids inherit, creating opportunities for your kids to earn higher wages
What we collectively leave our kids is equal to what we collectively produce minus what we collectively use up. When the government commissions someone to build, say, a tank or a highway or a bridge to nowhere, using, say, a million dollars’ worth of resources, those resources are subtracted from the next generation’s inheritance—partly from your own kids’ and partly from other people’s kids’, which affects your kids’ wages. If you like, you can (perhaps partly) compensate for your kids’ share of that burden by tightening your belt and leaving a little more in your bequest. But the moral is that it’s government spending, not government debt that has the potential to impoverish our children.
To see that debt is not the culprit, suppose for a moment that we decide to eliminate the debt tomorrow by raising taxes, so the average American forks over $40,000. What does that do for the average child who’s born tomorrow? It removes a $40,000 debt burden and simultaneously cuts his inheritance by $40,000. How is that child’s life affected? To a good first approximation, not at all. (I am glossing over some complications here, but they are of relatively minor importance.) And of course that calculation makes perfect sense because the only way you can make the next generation richer is by conserving actual resources—which you can’t accomplish with accounting tricks.
Ah! you might say, but we’ve done more than that—we’ve saved that child not just from $40,000 in debt but also from a lifetime of accumulated interest on that debt. Yes, and we’ve also robbed that child of not just $40,000 in inheritance but also of a lifetime of accumulated interest on that inheritance. It all washes out.
There might be other effects of deficits. For instance, some believe that the public doesn't see debt as equivalent to taxes so they're more likely to vote for spending if it's financed by deficits. And of course, Ricardian equivalence doesn't hold perfectly 100% of the time for all people. But these are really beside the point. Advocating spending cuts over tax increases isn't necessarily an error similar to the bee-sting fallacy. Various minor tax increases and spending decreases can only take a minor chunk out of the deficit, true. But since the deficit is taxes, and since we should only be worried about overall inheritance, the only thing that can make any dent on the future's inheritance is spending cuts. Raising taxes to cut the deficit is like moving money from one pocket to another. It doesn't address the problem at all. Only spending cuts can reduce the burden of government.

So when Nancy Pelosi said, "It is almost a false argument to say we have a spending problem. We have a budget deficit problem that we have to address," she had everything exactly backwards. We have a spending problem, and the deficit is mostly a red herring.

By the by, for those of us who aren't as radical as libertarians but think we could stand to cut a whole lot of government spending, the NYTimes lets you play with the deficit projection numbers. I found it pretty easy to reduce the projected deficit increases to zero and then some without raising taxes. But my estimation of many government programs is probably much lower than most people's. And on that note. . .

7) Competence of Government: The FAQ cites lots of examples of the government doing innovative things, of the government doing as well as private businesses on some projects, and good explanations why some government organizations might not do well but we should have them anyway. It is correct and that adequately answers the objection that the government is always worse than the market.

But it does not answer the objection that government performance is usually worse than market performance on many different measures across many different tasks. And this empirical result is so large and robust, it's about as close as we can get in a messy social science to a settled question. Here comes a photo dump:







Apologies if the quality of these scans is not wonderful; I know one is crooked. Above are 6 pages of citations* comparing various studies of public (or non-profit) vs. private good provisions. These compared organizations based on cost, efficiency, profitability, and various other measures (test scores in public schools, input substitution in hospitals, etc.). The types of organizations vary greatly: airlines, banks, insurance, fire protection, garbage collection, etc. Many different countries are surveyed as well. Something like 70+% of these studies found that the private sector outperforms the government agency doing the same task on the measured dimensions.

These studies aren't perfect, and they are dated, but I promise there is more where they came from and the trend is still similar. Some of the results are not applicable for the question at hand: for instance, it's found in one paper that students at private schools outperform those in public schools, but there are probably better explanations for that (cultural, genetic, and environmental) than the superiority of private schools. (I'll address private schools in more detail later.) One could argue that some of the agencies (like utilities) are natural monopolies, but that can only get you as far as government infrastructure with private provision. The private sector's still relevant and still, on average, superior.

Why? During one of the presidential debates, Obama gave a reason for why Medicare is more efficient than private insurance:
Every study has shown that Medicare has lower administrative costs than private insurance does, which is why seniors are generally pretty happy with it. And private insurers have to make a profit. Nothing wrong with that. That's what they do. And so you've got higher administrative costs, plus profit on top of that.
Now, "every" study has not shown any such thing, but that's beside the point. The point is the main argument here, which I would put a big ol' X by if a student wrote it on a test. The definition of profit is revenue minus cost. Obama instead argued that profits are a cost of doing business. This is wrong on a very, very basic level. Like, first day of econ 101 basic. Profits are the reward for doing business well. You can increase profits one of two ways: increase revenues, or reduce costs.

Because government agencies are non-profit, and because performance auditors can never be as informed and cutthroat as the market, government bureaucracies have less incentive to do either of those things. They typically just don't face much competition; their product accountability instead lies in voting, which is often very uninformed. Thus, bureaucracies are typically more inefficient than their private counterparts.

This does not mean that we should privatize everything, but it might mean that we should move many government operations in a more libertarian direction. Libertarians aren't wrong to criticize the government for sucking very much bigtime. For many measures, like efficiency and cost, it's basically a proven fact.

*The source is Public Choice III by Dennis Mueller, 2003, pages 374-379. It's a great book and you should read it.

8) Health Care: It's difficult to know why inter-country differences exist and are so large. If we take Robin Hanson seriously, it could be less the system and more that, culturally, we use medicine to signal more than people in other countries do. Or it could be that single-payer "death panels" are very efficient. As for health care reform, I'll just note that Robin Hanson wrote the best post that I have read on what economists can say about health care market failures, and for most of them a jump to single-payer is not the optimal policy, especially considering the above info about bureaucracy.

I'll say that I think single-payer would have been better than the Affordable Care Act. It's like we got the worst of all worlds in that one. A topic for another time, I suppose.

Also, the FAQ argues that studies show for-profit hospitals are more expensive and have poorer care than public ones. Some of the studies cited above agree with that, some give the opposite finding. If we want to claim that hospitals are an exceptional case of bureaucracy that works brilliantly, it's best to back that up with a proper summary of the literature. If anyone knows of a comprehensive and statistically competent review article, please send it my way!

11) Education: Apologies, this section is long because it's a tough issue. From the FAQ:
Compared to private schools, public schools actually do better once confounders like race, class, and income have been adjusted out of the analysis.

(Yes, without such adjustment private schools do better - but considering that private schools cater towards wealthy students - who usually do better in school - and often have selective admission policies in which they only take students who are already pretty smart - whereas public schools have to take everyone including dumb kids, kids with learning disabilities, and kids from broken families in ghettos - such unadjusted data is meaningless. It's the equivalent of noting that the doctor who specializes in acne has fewer patients die than the doctor who specializes in cancer: it's not that she's a better doctor, just that she only takes cases who are pretty healthy already.)
There are stronger research designs than simply controlling for confounders. The ideal study would randomly take some students from a public school and put them in a private school, then see how they compare to the kids back in the public school (and also control for confounders, of course).

As it turns out, this has been done a few times. I'll point the reader to some review articles. The first, by Caroline Hoxby: "School Choice and School Competition: Evidence from the United States" (Swedish Economic Policy Review 2003, pdf). If you are worried about bias, you should know that Hoxby is an advocate for school choice and most of her studies on that front support its worth. However, I've read a few dozen school choice papers, and hers are almost always quite statistically competent.

She explores several good studies and explains critiques of their methods. Here is a relevant table from the paper:


All of these cities had a voucher lottery, so they all have a randomized trial design, the gold standard for evaluating a treatment effect. National Percentile Rank points refer to placement on national standardized tests; which tests used are described in the individual studies. The general finding is that vouchers don't seem to have much of an effect on white or Hispanic achievement, but increase black student achievement significantly, both in the statistical and the normal sense.

 Another paper, "School Vouchers: Results from Randomized Experiments" (Peterson et al., Harvard University Program on Education Policy and Governance research paper, 2002 pdf), looks at some of the same studies and a few more that examine effects besides standardized tests. Specifically, they look at parental satisfaction on a number of dimensions, including the amount of homework assigned, truancy problems, cheating, etc., and find significant improvements for most of them. This is less reliable, as survey data, but it is informative. Other measures include amount of spending on various school facilities, with reductions in cost in private schools except for things like private tutors, where private schools spend more.

For a more negative review, see a working paper, "School Vouchers and Student Achievement: Recent Evidence, Remaining Questions" (Rouse and Barrow, 2008 pdf). They argue that the randomized trials are not objective enough: while the assignment of vouchers is random, the decision to use them is not, so there may still be a selection effect. They review more studies than the previous two and offer some criticisms of Hoxby's work.

If you're really interested in this issue, I suggest reading the above and any other good (preferably published) review articles you might find. Here's my general take:

There are some very strong studies that find vouchers work for black students who get them, a decent number of equally strong studies that find no effect, very few (I think only one?) that find a statistically significant negative effect, several weak studies that find a mix of positive and no effect (mostly positive), and many reasons to expect we will never be able to nail this problem down accurately no matter how totally awesome our statistical method. So if you want to, you can find a reason to dismiss any study; be wary of the tendency to dismiss those you disagree with.

I expected few gains for white students and more gains for blacks and hispanics. I was wrong about hispanics, they don't seem to gain from vouchers at all.

If you had a much lower prior for vouchers working before you read the above studies, it should go up quite a bit, unless you have an extremely high estimate for the prevalence of false positives/negatives in social science research. If you had a much higher prior for vouchers working, it should go down. This just isn't in the 90% range.


Standardized tests aren't perfect and we should be wary of "teaching to the test," but they do tell us something and some of these effect sizes are too large to be attributable to such a method.

Some other objections:

1) "Vouchers screw over the students left behind. Schools that lose students to vouchers also lose tax money, so the students who don't take vouchers get a worse education."

I consider this the most interesting finding of the school choice literature: The truth seems to be exactly opposite. Public schools suddenly facing competition can get their act together. See the above Hoxby paper for some discussion on the presence of charter schools, vouchers, and even greater choice in public school districts on traditional public schools. Typically, student scores improve. See Rouse and Barrow above for a discussion of the literature on Florida's A+ program, where schools were ranked A to F and students in F schools could get vouchers. Schools rated D showed small improvements and schools labeled F showed big improvements. Yes, it's impossible to know whether that's from competition itself or the stigma of being labeled an F school, but since schools already have reputations and since other kinds of competition also seem to increase scores, I think the former is a more likely explanation. (If anyone out there does want to start an A to F program without vouchers, just to see how large stigma effects can be, please do!)

At worst, if you trust these studies, it looks like left-behind students are not affected at all. At best, it's good for them. Protecting schools from financial competition seems to be what's most harmful. But if you really want to protect low-performing schools, taking money from public schools doesn't even have to be a part of a voucher program. It can be revenue-neutral.

2) "There are a whole bunch of separation of church and state and paternalism issues with vouchers. Kids end up going to schools where they get brainwashed by religious nonsense."

Or, as the FAQ puts it,
Not-so-clever parents would get fooled by TV commercials with sexy celebrities and send their kids to terrible schools. Super religious parents would send their kids to schools that taught only religious education and shunned math and science and history as the evil trappings of the secular world. Muslim parents would send their kids to madrassas. Immigrant parents might send their kids to Spanish-only schools so that they didn't drift too far away from their families. Parents with strong political beliefs could send their kids to schools that did their best to brainwash their kids into having the same beliefs as them.
I agree wholeheartedly! We have a good credentialing system for universities in the United States, and I see no reason why states couldn't implement something similar on a primary school level. At the very least, we should simply make vouchers provisional: If your school accepts these vouchers you agree to a certain standard of curriculum (for instance, no creationism, no overt religious instruction). If you are busted violating these standards, your school doesn't get vouchers anymore ever. Since many public schools, at least in Texas, already have religious instruction as part of the curriculum (pdf), I don't see how this experiment could hurt.

The FAQ follows up, "[I]f there were school choice, if we wanted to protect equality of opportunity and childrens' rights, we'd probably have to regulate the heck out of them, which to some degree would defeat the point."

To which I respond: If I am correct about the general findings of the school choice literature, then. . . no, it doesn't defeat the point at all! Even having more choice among highly-regulated public schools improves outcomes!

3) "Public schools in Finland are awesome, and they don't have school choice."

They also don't have a significantly and systematically marginalized and oppressed part of their population that concentrates in major urban areas and is trapped in a horrible cycle of poverty thanks to things like the drug war, absurd government punishments on their own poverty, etc. The FAQ also mentions this fact, but it bears on the discussion above: remember, many studies have failed to find solid positive competition effects for white students and other ethnicities, but black students tend to do much better. School choice is a step toward no longer marginalizing black people.

12) Moral Systems

I've been too hard on Scott, considering how good his FAQ is and how honestly minor some of my quibbles are. So let me just say that I love this section. L-o-v-e it. I endorse everything in it 100%, and similar arguments are probably the strongest thing that turned me away from libertarianism. I especially love the section about how "taxation is theft" isn't much of an argument at all. Libertarians, read this. Read the whole FAQ, but especially read this section.

As a suggestion, not a critique: If there is ever a version 3.0, it might want to also deal with moral intuitionists like Michael Huemer, who argues that we should treat the government as any other person. He starts from more reasonable premises than the harm principle, because he allows exceptions, but Huemer still ends up in radical anarcho-capitalism world by the end.

Here's a start from me: when we have conflicting intuitions about what's right, as I understand it, intuitionists argue that we should appeal to more fundamental intuitions that trump everything. I think that we frequently don't have totally fundamental intuitions. Most of us just contradict ourselves and that's that. This argument would require more digging into cognitive psychology and especially experimental philosophy.

13) Rights and Heuristics

Another section I'd like to cheer on. It was kind of a relief, way back when, after I figured out that all my babble about rights was wrong and that I was full of doody. Honestly, they're hard to argue for: exceptions are so, so easy to come up with, and I just can't seriously myself in a position of saying, "Yes, literally everyone in the world would be better off in some way if we implemented the policy you're proposing, but it would violate X right so we can't do it." That position. . . doesn't make sense to me.

As a side note, if you've read this far, you might find the field of Law and Economics interesting. Common law has already evolved to allow many of the exceptions the FAQ describes in contracts and property rights. I'll save that for another post.

14) Slippery Slopes

I fully agree that a welfare state will not lead to Stalin. That just ain't how totalitarianism comes about. However, there are other slippery slopes that are not as dangerous but are pretty bad. For example, France has, for decades, had persistent double-digit unemployment and an even higher unemployment rate among the youth. Economists mostly agree that this is due to France's stringent labor regulations, which make hiring new or unproved workers very risky. Strengthening unions might be a slippery slope to that.

 15) Strategic Activism

Another segment with which I have no significant qualms, though I would like to point the reader to the literature on Public Choice. It will enlighten you as to just how difficult it is to get decent solutions to fixing government. Not that you shouldn't try, but if you are interested in strategic, government-improving activism, you pretty much have to read Mueller's Public Choice III, Bryan Caplan's Myth of the Rational Voter, Buchanan and Tullock's everything, plus a bunch of cognitive bias literature and decision theory. It's a shame government employees are totally ignorant of this stuff.

I know this has been hefty reading. My goal was simply to point out the mainstream economic arguments and literature that the FAQ missed. If version 3.0 ever comes about, I hope this will have contributed to some improvements.

If I have missed any great or comprehensive literature reviews on the subjects discussed, please cite them in the comments. I have an upcoming post on why I almost exclusively trust reviews and have little confidence in individual studies.

No comments:

Post a Comment