David R. Henderson  

Levatter on Brennan

Holiday Benevolence: Increasin... Weight and velocity of money...

As regular readers know, I generally focus on economic issues. Often I do it as an economist; often I do it from a libertarian perspective.

But I want to point out a powerful essay on a philosophical issue, "The Brennan Thought Experiment: Further Explorations," published today on Libertarianism.org by my friend, Ross Levatter. A comment on this belongs on Econlog because the essay has some economic overtones.

Let me explain.

Jason Brennan, a highly productive political theory professor at Georgetown, has been highlighting a philosophical thought experiment that he thinks sheds some light on an interesting philosophical issue: Should libertarians support free markets because free markets are what you end up with when you support the moral principle of non-aggression (from which you should never deviate), or should libertarians support free markets only to the extent (empirically, quite a large extent) that free markets lead to the best social results? Although it is an interesting and important question, I was uncomfortable with his particular thought experiment to get at the question. Something seemed wrong, and, in a hurry with other projects, I didn't take time to figure it out. But when Levatter critiqued it in September, I commented online, saying his critique was "Fantastic." I thought he had nailed it.

It turns out that Brennan disagreed, suggesting that Ross was "widely misinformed" about the state of economics.

"Wow," I thought. "Does that mean that if I thought it was fantastic, that I'm widely misinformed about the state of economics!?" That's possible, I guess, but it seemed unlikely.

But now Levatter has extended his argument, listing a full dozen problems with the Brennan thought experiment.

Levatter's #3 is one of my favorites:

3. What Happens When We Change ONE Word?

There are other reasons to be skeptical of the claim Brennan's thought experiment has anything to do with economics. Consider this variant thought experiment:

Imagine that a bunch of moral philosophers provide compelling evidence that life in a strictly libertarian polity would be a disaster. Imagine that they show conclusively that if people everywhere were to live in a Nozickian minimal state or a Rothbardian anarcho-capitalist civil society, with everyone strictly observing property right rules, that 10% of people would starve, 80% would be near subsistence, and only 10% would prosper."

What has changed? Only that the term "economists" was replaced by the term "moral philosophers." Notice I made NO changes to the economic arguments or rationales in the thought experiment, since there ARE NO economic arguments or rationales provided in Jason's thought experiment, merely an assertion by "a bunch of economists," just as, in this variant, there is an assertion by a bunch of moral philosophers. Moral philosophy is, like economics, at least in part an empirical subject. It is not like logic, after all. So, as Brennan might put it, "[t]here's some possibility, and perhaps even some non-zero probability" that a bunch of moral philosophers might provide compelling evidence for this claim.

So the question of interest is: Is this a completely different thought experiment, or a slight variation of the original thought experiment? People can make up there own minds, but it seems obvious to me that this is but a minor variant of the original. Granted, per arguendo, it is easy for Brennan to imagine markets stink, but is it easy to imagine lots of people giving one answer to his original query, and yet a significantly different answer to my variant? Is it easy to imagine lots of libertarians saying, "Oh, well if a bunch of ECONOMISTS asserted 90% of the world would be immiserated by a minimal state, I would no longer favor it, but if a bunch of MORAL PHILOSOPHERS asserted 90% of the world would be immiserated by a minimal state, I'd still strongly favor it."?

Now, if you agree that my variant thought experiment is really not significantly different from Brennan's original, that is of course further evidence his is not a thought experiment about economics. It is a thought experiment about how one's sensitivity to consequences-adverse, unexpected, highly unlikely, no-reason-provided-to-believe-accurate, stipulated consequences-impact on one's advocacy of an otherwise-favored society.

My other favorite is argument #10, which is straight economics. Here's an excerpt:
A fair reading of this graph is that everything we know from history indicates that moving toward a market economy moves us away from the entire prior experience of mankind, much worse than 90% of the populace immiserated and only 10% prospering. As we move to a world with comprehensive markets, we move to a world where these proportions are actually viewed as a bad thing rather than a blessing, where we routinely expect and receive much more. This is why I previously described Brennan's thought experiment, where he imagines markets leading to the very result they historically eliminated as akin to a thought experiment where you imagine demand curves slope upward rather than down. Both are something no one has ever experienced. And this is why, above, I describe Brennan's thought experiment as too severe to achieve his desired goals.

Argument #10 actually gets at the title of Brennan's response to Levatter: "Imagine Markets Stink. It's Easy if You Try." It's actually not that easy.

I think it is now up to Brennan to explain why he thinks his thought experiment is about economics at all.

Comments and Sharing

CATEGORIES: Economic Philosophy

COMMENTS (19 to date)
Alex Godofsky writes:

I disagree; I think #3 is pretty silly. If a bunch of moral philosophers engage themselves in rigorous empirical and theoretical study of economics, then that would make them economists (even if only by a pretty serious avocation). If their work isn't rigorous, how would the evidence they give be compelling?

Rob writes:
...that 10% of people would starve, 80% would be near subsistence, and only 10% would prosper…
I assume he means to imply a fixed population size and no other happiness modifiers? Because hypothetically, one could theorize a world where this distribution is created, but there is either a much larger population who are glad to live (cf. repugnant conclusion) and/or a world where other factors are achieved that modulate the average happiness upwards despite this distribution. Both would be relevant for utilitarianism, some version of which seems to be implied in the thought experiment.

He did use the word "disaster", so I guess it's nitpicking. But it's still worth pointing out; there is a possibility that a hypothetical minimal/libertarian state would lead to different population size / average happiness modifiers than hypothetical alternative ideologies.

pyroseed13 writes:

In a sense, isn't it difficult to establish a deontological defense of anything without appealing to empiricism? To defend the non-aggression you would have to argue that deviating from it leads to bad outcomes. So I don't really understand why it matters how exactly you construct a defense of free markets.

Pajser writes:

Brennan's though experiment must sound to libertarians like "imagine that 2+2=5, do you still believe that 1+2=3". But I think it is simple, valid experiment and discussing it on meta-level is overkill. I frequently expose libertarians to argument that strict property rights result in poverty, because every citizen is able to ban all (including EM) pollution of his property and shut down whole world industry. They easily accept that state must restrict property rights.

Kevin Vallier writes:

David, for all Ross's efforts, I'm afraid it's a rather poor reply.

Jason Brennan writes:


I think Ross's criticism completely misses the mark, so I'm saddened that you approve of it.

The issue here is very simple. All I'm asking is what role consequences play in our thinking about the justification of markets. Even though we know markets are good, we political philosophers--we people who ask what justifies various institutions--want to know just what it would mean if markets had different consequences than they in fact had. So, I'm just asking whether people would still support markets if the empirical claims of Marxist economics (or, if you'd like, more anti-market mainstream economics) were true. You're right that the issue is not actually about economics. It's about how the facts of economics bear on the moral justification of market institutions. And the way we determine that is to ask what would follow, from a moral point of view, if the facts were different.

David R. Henderson writes:

@Kevin Vallier,
Are you willing to say why?

RexInder writes:

Congratulations! You've invented the 1985 movie "Witness".

Jameson writes:

There seems to be something evasive about the whole "general trend of history" argument. Sure, you could say that moving *generally* towards markets has always alleviated prosperity. But what kind of markets? Those governed by strict Rothbardian property rights? From what research I've done, it seems there's a lot of mainstream economics out there that supports pretty non-libertarian regulations (just take the minimum wage as an example). So libertarians need to be honest: one can't just defend markets in general and claim to be defending a particular brand of libertarianism.

Also, I think Alex Godofsky was spot on in his critique of point 3.

Tom West writes:

I call the two camps the moral Libertarians and the consequentialist Libertarians.

I've only met one moral Libertarian who felt that the disaster outcome was likely, but felt that was the only moral outcome :-).

However, this does seem a tempest in a tea-cup. Society is so far from Libertopia that the two camps are identical in action for the foreseeable future. If totally free markets somehow don't work, there'll be a long lead time to change one's opinion.

Secondly, it's pretty clear that no matter what the outcome, pro-market believers will see it as progress and anti-market believers will see it as disastrous.

For something as amorphous as "desirable social outcome", there is no objective scale to measure. You have hundreds of scales to choose from, and people will choose the metrics that matter to them.

Therefor the disaster (or ideal) scenarios can't occur. Or more accurately, one person's disaster will be another person's success (short of turning the planet into an atomic wasteland).

RPLong writes:

I know Brennan's no Ayn Rand fan, but upon reading his experiment I immediately envisioned Atlas Shrugged. Let's say most people do prosper in an anti-market society, and that in order to do so, a small minority of the world's most productive people must essentially be enslaved.

So, question to Brennan: Do consequences matter here? If slavery ensures a prosperous majority, shall we enslave each other? How did that work out in the US and elsewhere?

In short, moral virtue does indeed trump economic maximization. Fortunately for us, though, free markets do maximize social well-being, so we do not really have to make that choice.

Hazel Meade writes:

Brennan's question isn't an economics thought experiment at all. It's a philosophical thought experiment meant to tease apart moral and consequentialist arguments for libertarianism.

I think there is a link between the morality of the free market and the generally positive consequences, but it's difficult to articulate exactly why a moral system would lead to positive economic outcomes.

One might find one in economics, to the extent one can prove mathematically that free agency leads to generally better results. There's certainly plenty of evidence that free market *pricing* leads to more optimal resource distribution. There's probably also something in Hayek's ideas about local knowledge. But I don't think anyone's ever formulated a general theory that shows the majority of individuals are better off when they are allowed to make their own economic decisions.

Chris H writes:

I read through the back and forth from Brennan's first post to this essay and at first I thought Brennan had the better point. But the last essay brought up a few good points and Levatter's proposed alternative thought experiment does seem to get at the tension between consequentialist and deontological beliefs in libertarians to a much greater extent. How many people being hurt on net is a libertarian society worth? That's a question which really seems interesting while at the same time creating the potential for the exact same dividing line Brennan uses in his experiment.

I didn't like all of the essay, point 6 kind of grated on me a bit. It reminded me of people who try to dodge the point of the trolley problem. One thing I've seen people do there is try to say "well OK but what if one of the five people was Hitler and the one person was Norman Borlaug" (not those exact people but you get the point). Yes that would change the calculus but that's not the point of the thought experiment, answer the question as if they were random people. For point 6 I'd say you can simply answer the question as if the pre-libertarian position was the modern world and people are as moral/deserving in the thought experiment as they are now. The exact mechanism is not the point (indeed Levatter should realize the exact mechanism isn't the point given that he himself talks about how this isn't really about economics at all).

Ultimately I think it was the example used in point 9 that made understand the basic problem with Brennan's thought experiment. It's a bit too brute force, which might be OK when dealing with people just getting an intro into philosophical ideas, but it's also good to allow for more subtle gradations (like Levatter's alternative does).

Alexandre Padilla writes:


Jason Brennan is correct. I think he's just trying to ask economists or libertarians how do you justify free markets. He's asking whether people who support free markets or are libertarians how they justify their support for free markets. If you are a consequentialist and it was proven that a central planning economy produces a better outcome than markets, Brennan is asking: would you support central planning?

Jason's assumption is that economists tend to be consequentialist and I would agree with that in general but when you survey them and ask questions related to the war on drugs, for example, it's not as clear cut that they all still apply consequentialism in their answers. My reading of Mark Thornton's research on whether economists agree on drug prohibition does show that's not entirely the case.

I am sure Jason is aware of the research Fogel and Engerman did on the economics of slavery and his work entitled "Time on the Cross" that raised some eyebrows because people perceived it as endorsing slavery. By the way, Fogel and Engerman did NOT say slavery was ok.

I don't know if Jason read Yeager's book Ethics as Social Science but it might answer some of his questions.

Chris H writes:

The more I think about this entire exchange, the more I dislike most of Levatter's response actually. I still think his thought experiment is superior, but the idea Brennan's point is so far removed from reality as to be useless seems really bad.

For one thing, it's not like there haven't been huge conceptual shifts that have happened even in sciences which have major advantages over econ in the ability to perform experiment and deal with simpler systems. Take physics for instance which from Newton until the 20th century kept on confirming and reconfirming that the universe was run on deterministic laws. Then quantum mechanics came along and showed that was an illusion. If physics, which can run controlled experiments and doesn't have to deal with layers upon layers of added complexity that human societies represent, can make very basic errors like that in the time after the scientific method was being used, so can econ. Yes markets have been associated with improvement in the human condition for centuries, but that doesn't mean that despite our best efforts to disentangle the effects of markets from other issues there isn't some conflating factor. I personally think markets are the most plausible answer and don't think a strong movement towards even more markets would reverse this (I'm highly sympathetic to anarcho-capitalism), but to say that markets failing big time without strong governments is so far removed from reality as to not be worth considering seems far too arrogant to me. I could see a physicist in 1900 saying the same thing about the universe being run on deterministic laws, but try finding one today who would argue that.

The point is yes it's an implausible scenario, but it's not past some point of no return and the more I think about it, the more Levatter seems to be making a somewhat petulant attempt to dodge the question.

LD Bottorff writes:

Imagine a bunch of economists provide compelling evidence that....

raising the minimum wage will not result in lower employment opportunities for the unskilled workers.

It depends on what the meaning of the word 'compelling' is.

Tracy W writes:

Chris H: I think you have a good point in that theories can change drastically and the case of physics is an interesting one in that Newton's Laws are still taught by physicists and used by engineers in situations of the right scale. Actions still have equal and opposite reactions at least at the macro level, macro objects in motion still remain in motion unless acted on by another force, etc.

The new theories added to classical physics rather than displacing it. Similarly any future development in economics would have to explain somehow how market societies do function reasonably well. Such a theory would probably be something like the change in physics which would be "the old theories still hold true for nearly all day to day situations but if you have an economy under these particular set of circumstances then you get these unusual results."

Hazel Meade writes:

LD Bottoroff: Absence of evidence is not evidence of absence.

Ross Levatter writes:

Jason seems unable to grasp the possibility that what he intends to be doing and what a rational critique of his thought experiment shows it capable of doing are not always the same thing. I myself support what Jason claims he intends to do. Surely he doesn't believe it is impossible for thought experiments to be poorly constructed, or for even competent philosophers to overstate what their thought experiment accomplishes.

Comments for this entry have been closed
Return to top