Bryan Caplan  

The Wise Pluralism of David Friedman

Tyler Cowen on Stories... The 2012 Vote That Really Matt...
This month's Cato Unbound discusses Matt Zwolinski and John Tomasi's "Bleeding Heart History of Libertarianism."  David Friedman's response is good enough to make me see utilitarianism in a more favorable light.  Friedman agrees with Zwolinski and Tomasi that pre-20th-century libertarians worried more about the poor.  But the main reason was that back then most people were poor:
Our lead authors' repeated references to "the poor," in an essay written by and for moderns, badly misrepresents the 18th century world and Smith's view of it. When Smith was writing, the working class, the people Smith is referring to in that quote, represented not the lower end of the income distribution but the bulk of the population.
Friedman admits that libertarianian absolutism (and utilitaranism) lead to bizarre conclusions, but Rawlish social justice theories are similarly flawed:
[S]ubstituting "an ideal of social or distributive justice" is hardly an improvement, judged by either foundations or implications. Rawls' derivation for his idea of social justice, the version that Tomasi and Zwolinski apparently wish to incorporate in libertarian thought, starts with the claim that someone facing a set of alternatives whose probability distribution is unknown--the imaginary social chooser behind a veil of ignorance--will act on the assumption that he is certain to end up with the least attractive possible outcome. That, plus a lot of hand waving, is all the justification for his "philosophically most sophisticated" version of social justice that I have been able to find.

And one implication of that version, taken as literally as I have been taking the natural rights alternative, is that it is better to have a world where everyone is at a utility level of a hundred than a world with one person at ninety-nine and everyone else at a thousand. I have never yet been able to figure out why anyone takes either the derivation or the conclusion seriously.
Friedman then ends with a wise and thoughtful moral vision:

The version of libertarianism that seems most plausible to me is one where respecting rights is seen as a good thing, a value in itself as well as a means to other values, but not as a value that trumps all others. One reason to respect natural rights is that it is a good thing to do, another is that respecting them can be expected to produce a healthier, wealthier, and happier world than violating them.

Utilitarianism does not, in my view, fully capture the range of those other values, but it comes considerably closer than social justice. I do not have an adequate derivation for my ethical views and--unlike Rand and Rawls--I know that I do not, so can only report on my moral intuitions while trying, so far as possible, to think through their implications and interrelations.

Many people picture David Friedman as a reductio ad absurdum of Chicago economics.  But that's only because they don't read him carefully.  While Friedman ends up with many extreme conclusions, he never stacks the deck with neoclassical or Benthamite dogma to get there.

Comments and Sharing

COMMENTS (10 to date)
mark writes:

I've always been mystified by the admiration of Rawls's basic work. As a minor matter, he significantly revised his views throughout his life and who knows what would further revisions he would have made had he worked longer. So it seems somewhat arbitrary. But more importantly, his views were so clearly derived from the context in which he lived - his heavily Christian upbringing, for instance, but also the secular Western values that his social class was instructed to adopt, his career choice of the not for profit sector and the postwar failure of Marxism. All of his work seems an attempt to construct a framework to privilege his subjective experiences and preferences as opposed to something more open minded. A guess it could be seen as an intellectual achievement to develop a not explicitly Christian, not explicitly Marxist framework to get to the "from each according to his means, to each according to his needs" result he desired - which is essentially what his main work is intended to do, but to do so by saying "let's play make believe" and then rigging that game to fit his subjective preferences hardly seems worthy of the adulation he has been accorded.

daryl writes:

Libertarianism is better than government at helping society reach the goal of improving the life of poor people. Not just poor people, but pretty much everyone. It's hard to understand how Rawls couldn't see that when he wrote his theory of justice.

Kevin Dick writes:

25 years ago, we had to read Rawls in our freshman philosophy course. It was presented as this huge breakthrough.

But even a lowly, liberal-raised 19-year old like me was puzzled why Rawls thought infinite risk aversion was a reasonable assumption. Maybe that's why I went into economics and decision theory instead of philosophy.

I do remember getting a good grade on an essay where I suggested we do empirical experiments on people's risk preferences and use that to compute the allowable inequality chosen from behind the Veil.

Airman Spry Shark writes:
One reason to respect natural rights ... is that respecting them can be expected to produce a healthier, wealthier, and happier world than violating them.
This is my ethical approach, which could be described as 'deontological utilitarianism'. Essentially, act utilitarianism is prohibitively difficult to implement, leaving rule utilitarianism; the process of attempting to formulate a comprehensive, coherent, consistent set of rules appears to approach deontology at the limit.
Julien Couvreur writes:

I really appreciated David Friedman's response as well, and overall this debate.

But he did not make his point about bizarre conclusions of libertarian absolutism ("it is better to have a world where everyone is at a utility level of a hundred than a world with one person at ninety-nine and everyone else at a thousand").

First, utility cannot be measured and compared across individuals.
Even assuming that you could (which people tend to intuitively assume), then probably absolutists would prefer the second world. But they might still validly object to using aggression to get from the first to the second.

In practice, there would likely be voluntary agreement that would help achieve the better outcome.
Also, if you did resort to force then things might appear better in the short-term but you'd incur a risk that aggression becomes an acceptable tool whenever a case of "sub-optimality" is presented. Because "sub-optimality" cannot be measured, you have a convenient slope for statists to take advantage of.

That said, he made a good point that property rights lack a bullet-proof justification. In a paper, he presented a interesting alternative view, which is that such norms emerge as Schelling points.

Saturos writes:

The only living ethicist better than Caplan & Huemer: David Friedman.

Nathan Smith writes:

1. The major modern meta-ethical positions, Rawls', Nozick's, and the utilitarians', all have this virtue: that they rule out the selfish protection of privilege. (Rawls' view doesn't inasmuch as he tried to avoid applying it globally, but since that omission is obviously indefensible, I just assume it away and refer to Rawlsian ethics globally applied.)

2. They therefore imply open borders, since the global apartheid of the Passport Age clearly does not serve the interests of most of mankind. (I would be interested to know if anyone makes a serious attempt to contest this. It is a devastatingly strong claim on behalf of open borders, so anyone who believes in the legitimacy of migration restrictions needs to take it on. But I wouldn't be surprised if no one has actually tried, because anyone clear-headed enough to ask whether comprehensive, discretionary migration restrictions can be justified in terms of any acceptable ethical theory may be honest enough to accept that the answer is no.)

3. Since libertarians tend to be more favorable to open borders than any other political grouping, libertarianism is de facto the most bleeding-heart major ideology today. But I think in many cases reluctantly so. Sometimes you get the sense that the combination of resentment about big government with the desire for intellectual consistency forces libertarians grudgingly to favor the right to migrate.

4. A global system of rich-to-poor transfers, some form of international socialism or moderate international socialism, would be a political ideology for which a real ethical case could be made. I think I would still reject it, but if that were actually on offer, advocated by someone, arguing the libertarian view would be more challenging. As it is, liberal-left parties and pundits tend to advocate a domestic welfare state combined with migration restrictions, possibly eased up slightly. That's not a view that deserves to be taken seriously. It has power, yes, but plausibility, no.

Evan writes:

Friedman's closing remarks:

The implications of my moral intuitions are not as tidy as the theories of Rand or Rawls or, for that matter, Bentham. But then, I know of no a priori reason to expect the truth, in moral philosophy or anything else, to always be simple.

That there is pure, unadulterated awesome. I think the bias towards simplicity in formulating ethical theories is a major roadblock in moral philosophy. When people do moral philosophy they like make grand sweeping categorical statements. But, as Thomas Sowell said, we live in an incremental world, not a categorical one.

@Kevin Dick

But even a lowly, liberal-raised 19-year old like me was puzzled why Rawls thought infinite risk aversion was a reasonable assumption.
I've noticed this problem too. In particular, it seems to me that one possible consequence of an excessively strict Rawlsianism is that it would lead to banning all automobiles except for ambulances, since an infinitely risk-adverse creature wouldn't want to risk being the one who died in a car accident, even if it meant it was likely to have the low-utility life that being car-less entails.

@Julien Covreneur

First, utility cannot be measured and compared across individuals.
It can't be precisely measured, but getting "good enough" approximations isn't impossible. The fact that humans are all one species that evolved from a common ancestor, and evolved to be social animals, makes it easier. Think about it. How could humans have possibly been able to evolve to work together if we hadn't evolved some way to sense the utility of others?

Brandon Berg writes:

Of course, there are also qualitative differences between modern and 18th-century poor. The 18th-century poor were poor because pretty much everyone was poor back then. You had to be lucky and/or extraordinary not to be poor. The modern poor are poor because they make really bad choices.

Julien Couvreur writes:


I'd like to see you try. Handwaving that it's not impossible if you just approximated is not addressing some key issues: what is the unit? how do you measure? why would units be equal (marginal utility) for a given individual? why would units be equal across individuals (interpersonal comparison)?

You suggest that the last one must be solvable somehow because of evolution. That argument only tells me that people can guess other people's preferences to varying degrees of success (maybe communication is what enabled such social behavior, rather than mind-reading). The fact that so many startups fail tell me that such skill is far from accurate.

Btw, there was a good essay just on the topic of measuring utility earlier this week:

Comments for this entry have been closed
Return to top