Arnold Kling  

Libertarianism...and its Discontents

Lifted from the Comments... Nice Sentence...

Three items crossed my computer screen recently:

--the launch of by the folks at Cato. It looks like a nice web site, one that invites exploration.

--Tyler Cowen's link to a piece by Will Wilkinson.

--two older pieces by Jeffrey Friedman, namely What's Wrong with Libertarianism (WW) and The Libertarian Straddle (LS).

Wilkinson and Friedman are partially-lapsed libertarians. More on their views below the fold.

Wilkinson writes,

organized groups can be effective agents of real change. (It takes a village.) And groups are easiest to organize along lines of common material interest. That's why, for example, "right-to-work" laws look to progressives like unilateral disarmament of the working classes, next to which the right of an individual to opt out of collective bargaining seems trivial. What's the value of having the right to individually negotiate your terms of employment when all that gets you is screwed over? A politics of nothing but individual rights in a world dominated by social forces is a recipe for domination by those sufficiently powerful or organized to shape those forces.

Earlier, he writes,

I've come to accept, for example, that diffuse cultural forces, such as racism or sexism or nationalism or intergenerational poverty, can deprive an individual of her rightful liberty without any single person doing anything to violate her basic rights. This takes me a long way toward standard liberalism. But I find that my gut nevertheless leans right on issues of personal responsibility.

I think a lot depends on whether you think that people who are "screwed over" will benefit from collective, coercive action or not. I worry that it is too easy to assume that collective action works and to forget how often it fails.

Labor unions would be a prime example. In the gauzy, Whig history of the left, unions did wonderful things, and their decline is to blame for our problems today. But if you wanted to unionize 30 percent of the private sector work force today, how would you do it? By industry, so that everyone in health care, from the brain surgeon to the janitor, belongs to the same union? By trade, so that you have a union of adjunct professors of European history, another union of aspiring Asian-fusion chefs, another union of experts on sustainable development, and so on?

Turning to Friedman, I did not expect to find these sentences, in LS:

Attempting to understand not only what someone is doing, but why he may go wrong in doing it--even with the best of motives--is an inherent task not only of the social sciences but of daily life...we have to go a little deeper than the overt intentions of those we are trying to understand. Otherwise, we would never be able to explain (as opposed merely to identifying) their errors...

...if we are to understand those with whom we disagree, we must go beyond their conscious motives and examine the subconscious impulses to which their
ideas and assumptions may commit them. the goal of understanding. If we fail to go deeper than motives, we will have to attribute error to mendacity

He is the last person I would have expected to make the case for type M arguments. The problem is drawing the line between doing intellectual history--understanding someone's views in terms of where they came from--and doing reductionism. For an example of the latter, consider the famous piece by Nozick on intellectuals, featured on He wrote,

The intellectual wants the whole society to be a school writ large, to be like the environment where he did so well and was so well appreciated...Those at the top of the school's hierarchy will feel entitled to a top position, not only in that micro-society but in the wider one, a society whose system they will resent when it fails to treat them according to their self-prescribed wants and entitlements. The school system thereby produces anti-capitalist feeling among intellectuals. Rather, it produces anti-capitalist feeling among verbal intellectuals.

Does this analysis substitute for rebutting an intellectual's anti-capitalist argument? I would think not. The argument could be right or wrong on its merits, regardless of how it relates to the status needs of the arguer. (And, yes, I am sure that you could find plenty of instances in which I have been guilty of similar reductionism.)

Perhaps Friedman would say that he is trying to explain why libertarians avoid certain issues or consistently make certain unwise intellectual moves. But the line between what he thinks of as intellectual history (good) and what he would regard as uncharitable reductionism (bad) appears blurry to me at this point.

The point of WW and LS is that libertarians should choose between a consequentialist defense of liberty and an a priori defense of liberty. I think that LS makes the point well enough, but it makes some sense to read WW first.

I have two issues with WW, one minor and the other major. The minor issue is that Friedman gives a simplistic rendition of Public Choice theory as saying that government officials are self-interested. But at least one important element of Public Choice theory, rent-seeking, does not require self-interested politicians. It simply says that government intervention invites actions by interest groups to influence the direction of intervention. They do not have to corrupt the government officials. They can use pure powers of persuasion. Failing that, they can use the electoral process to get the officials they want.

The major issue is that Friedman argues that consequentialist libertarianism (the side that I would put myself) is threatened by the possibliity that some government interventions might have good consequences. This makes libertariansim a tenuous proposition, at best requiring case-by-case analysis.

I reject the case-by-case approach. Instead, I want to focus on process. Sure, government can do something right every now and then. In some instances (urban water sanitation), government has so much experience with solving the problem that it would be nuts to go full-Monty anarcho-capitalism with no thought about transition issues. But when it comes to institutional processes, I think that exit works better than voice.

Set up two societies. Society C relies on competition, choice, and the consumer's right of exit. Society P relies on planners and policy wonks who are given monopoly power backed by the coercion of the state. Which society will, on average, arrive at the technological, economic, and institutional outcomes that make for the greatest happiness for the greatest number? I think that one can make a strong empirical case for Society C. Moreover, I think one can make a case that movements in the direction of Society C tend to be welfare-improving, and movements in the direction of society P tend to be welfare-destroying.

Focus on dynamics. Focus on process. What are the conditions under which a monopoly of policy wonks will function better than a competitive, trial-and-error system at groping for solutions in a world that is too complex for any one person to understand? I think those conditions are highly unlikely to be observed.

Case-by-case thinking reverts to neoclassical economics, in which economics is about the "allocation of resources" in a static sense. The neoclassical economist uses mathematically elegant formulations to determine when markets will and will not achieve an optimal allocation. Then, pretending that we have a policy wonk with perfect information, we proceed to prescribe interventions, such as central boards to influence medical practice in order to reduce spending.

The focus on process instead looks at dynamics. It underlies the slogan "Markets fail. Use markets." The fixes for market failure are more likely to emerge from competitive evolutionary processes than from top-down planning.

The focus on process allows radical ignorance to be the justification for liberty. What we know is trivial compared to what we have yet to learn. We can support limited government on the grounds that larger government does more to impede learning.

Comments and Sharing

CATEGORIES: Economic Philosophy

COMMENTS (9 to date)
Lee Waaks writes:

Arnold won't go "full-Monty" for anarcho-capitalism, but I think his policy conclusions push him in that direction. The "limited" state he envisions will be subject to the same public choice dymamic as the large state we have now. And now to indulge a "type M" argument: Is Arnold afraid to go full-Monty because of the (undeserved) ridicule it might bring? As for Will Wilkinson and his concern about minorities in an unbridled libertarian world, is it really the case that they would be that much worse off than under the alleged "protection" of the state? Why does he trust the state? Eg., blacks are protected against racism from whites but black on black crime is rampant. So much for protection. I think his use of terms like "social forces" that might be injurious to minorities under a libertarian society is a bit vague for me. I agree, pace Friedman, that consequences matter, but I think Jan Lester is correct that we will get more want satisfaction under anarcho-capitalism (see Lester's _Escape From Leviathan_). But having said that, Friedman's new book, _Engineering the Crisis_, is fantastic and Wilkinson's Cato essay on inequality is also fabulous.

Cal writes:

Yep, Jeff Friedman says “[l]ibertarian conclusions require not only extensive evidence of government failure, but an empirically substantiated reason to think that such failure is always more likely than the failure of civil society.” (p. 410)

And as Jan Lester pointed out in his strictly Popperian critique of Friedman's long-winded essay, Friedman is demanding the logically, epistemically, and pragmatically impossible by requiring that libertarians "empirically" substantiate that government failure is "always" more likely than civil society failure.

That libertarianism is lacking empirically is the lynchpin of his entire article and it is left completely unargued; he doesnt engage in any empirical substantiation of a non-libertarian position and he completely ignores prominent libertarians who don't fit his "libertarian straddle" and avoid ethical advocacy altogether, like David Friedman and Anthony de Jasay.

Incidentally, de Jasay is a fine exemplar of someone who makes arguments based on process and institutions.

david writes:
What are the conditions under which a monopoly of policy wonks will function better than a competitive, trial-and-error system at groping for solutions in a world that is too complex for any one person to understand?

I submit, Mr. Kling, that you need to elaborate on this. Once you have given up on the neoclassics, you have also given up on any assurance that competition leads you in necessarily desirable directions.

Dave writes:

"I reject the case-by-case approach. Instead, I want to focus on process."

This is something I've realized in the last year or so: Many libertarians (myself included) care about process more than outcomes. I sometimes hear left leaning economists declare that their preferred government intervention should be embraced by libertarians because it gives them their desired outcome. But this misses the point that libertarians don't try to steer outcomes precisely, even if they believe less government also has better outcomes for most of society.

I have the same attitude when it comes to income inequality: I don't necessarily think it matters, but I care about the causes of the inequality when making a judgment. Did it arise because a banker made millions taking on systemically important tail risk for years, implicitly covered by taxpayers? Was it a result of a credit bubble aided by artificially low interest rates from the Fed? Then yes, we should be concerned with it. But if it is simply because cheaper national or global distribution allows larger returns to the most talented people in their industries (think movie stars, software entrepreneurs), then that seems like a fair process. And I think we can infer that liberals agree to some extent since there is no Occupy Hollywood or Occupy Silicon Valley movements.

Sorry if I went a bit off topic...

Daniel Klein writes:

Jeff Friedman's work has helped libertarianism to evolve in good directions.

I invite you to disdain the consequentialist v. deontological distinction. Doesn't violating deons (duties) have consequences? That is, can't you translate your scruples about violating deons into consequences, including moral and cultural consequences? (If you can't, whose fault is that?)

And from concern about consequences we develop by-and-large principles, or duties.

Consequences and duties are a yin-yang cycle, something that libertarians grow increasingly comfortable with, I think and hope.

And the dichotomy between case-by-case and being strictly principled, a dichotomy that Arnold seems to be implying, is inadequate. A third attitude is that of presumption, as in the presumption of innocence, which places the burden of proof on the prosecution.

As for "Type M" argument, I think that Arnold is being philosophically naive.

I think that the most important theory about why people do liberalism less than justice is still Hayek's atavism interpretation of modern statism. See Hayek 1967, 120; 1976, esp. ch. 11; 1978; 1979: 153-176; 1988, esp. ch. 1.

Lee Waaks writes:

Lester's argument employs what he terms "deontological consequentialism." I am in agreement with that and it seems to mirror Dan Klein's views. In my interpretation, consequentialism and libertarian rights are on a parallel track, one reinforcing the other as they move across the political landscape together. Lester's response to Friedman is here:

Mark V Anderson writes:

I was not at all impressed by Friedman's critique of Consquentialism. His case seemed to that using empirical data to prove one's case that government intervention usually does not work means that one's work is never done. One will always have more cases to prove in the future.

Well duh. Does he seem to think that he will find the ultimate proof of libertarianism that will prove the case beyond all doubt? Leave such utopian thinking to the socialists. There are and always will be cases where government intervention actually improves the welfare of society. The point of arguing Consequentialism is to show that such cases are far and few between.

And yes, arguing trend lines are often the best way to argue this. One needs a lot of trend lines in a lot of different circumstances to make a good case, but that is the best way to convince those that otherwise don't believe in the free market. I believe that to be the case mostly because it works on me in the opposite direction. If government interventionists could come up with a number of trends that show that free markets don't work, I will likely adjust my free market beliefs. It's just that almost always those lines have not shown what they claim to show.

On the other hand, I see absolutely no future in trying to convince leftists of the natural rights argument. Of course I am biased in that regard because I don't believe in it myself. But maybe that's why I can see its fallibility. I have read dozens of arguments for such rights, and every one of them uses self serving "logic" to arrive at its conclusions.

Vipul Naik writes:

This is actually a comment on the post on "Type C" versus "Type M" arguments that you linked to (but that's really old, so I thought I'd comment here). I think there is a third type of argument that is used all too frequently, and lies somewhere in between. I'll call it "Type B" where B stands for bias/blindness.

This proceeds by trying to identify some cognitive bias, or lack of maturity or lack of experience or lack of understanding of some sort in the person making the argument. It doesn't question the other person's motives/intentions, but avoids reasoning against the specifics of the argument. Type B is roughly "you're stupid" whereas Type M is "you're evil."

Examples of Type B arguments:

* dismissing arguments against affirmative action by saying that the person making the argument has an "unacknowledged privilege" and "doesn't really know what it feels like to be in the other person's shoes"

* dismissing a critique of the market by saying, "It's really sad that you never took a good Econ class in school/college" or "it's clear that you suffer from anti-market bias. I know you mean well, but you don't understand the principle of unintended consequences or the basics of supply and demand" (without addressing the actual critique)

* dismissing a critique of redistribution by saying "it's really unfortunate that you were never exposed to the cutting edge ideas that support redistribution."

* quoting behavioral/cognitive/psychological biases to explain away an argument.

I am very wary of Type B arguments offered in isolation, because they try to get the same mileage as a Type M argument while pretending to be otherwise. On the other hand, I _think_ that a Type B argument may be offered as a speculative addendum to a Type C argument and serves well in that context, i.e., first explain why I think the other side is wrong using a Type C argument, then identify why I think I and the other side have come to different conclusions using a Type B argument.

Type B arguments proliferate. I find a large part of Thomas Sowell's writings to comprise Type B arguments, though he does make some Type C arguments.

Bob Layson writes:

The case is not one of either human rights or a cosequentialist justification of institutions.

I insist on certain rights for myself and others because I desire the consequences that naturally flow from the recognition and exercise of such rights - given the human character that evolution and culture has created. The philosopher does not discover these rights but rather grasps the consequences for human opportunity of respecting these rights to person and property and knows from theory and history the ruin that comes of state organised violation of them.

Comments for this entry have been closed
Return to top