#ReadWithMe: Power Without Knowledge Part 9: What Friedman Gets Right
Jeffrey Friedman’s Power Without Knowledge: A Critique of Technocracy is impressive in its scope and its argumentative strength. The book is a culmination of themes Friedman had been examining for years. As such, many of the arguments in the book were familiar to me, nonetheless, I still learned a lot by reading it. Some of what I learned reinforced parts of my outlook, while other arguments challenged and shifted my views to various degrees. Here I focus on what I see as the book’s strengths.
A key strength of the book – and one that’s been largely missing from my posts so far – is the frequency with which Friedman stops his arguments to consider and steelman possible objections or counterarguments to his claims. This is the mark of a writer who has carefully considered his arguments and has made an active effort to highlight all the possible weaknesses he can identify in his case.
At first glance, some readers might be tempted to dismiss some of Friedman’s insights as merely parasitic on claims Austrian economists have made for years. For example, a lynchpin in Friedman’s argument is ideational heterogeneity – the idea that each mind operates differently from every other mind, motivated by a unique web of beliefs, and that this internal mental process is fundamentally inaccessible not merely to anonymous technocrats but even to people who know each other well. (Think of all the times you find yourself surprised when a friend or family member interprets or reacts to a situation very differently than you expected, or in conversation someone suggests a thought that never occurred to you.) The basic phenomenon Friedman is describing, and its implications for the construction of social policy, was described decades ago by Ludwig von Mises in his book Theory and History:
There are also clear parallels between Friedman’s argument and the famous chess analogy made by Adam Smith, who speaks of how the “man of system” (or technocrat, in Friedman’s work) is “wise in his own conceit” and believes “he can arrange the different members of a great society with as much ease as the hand arranges the different pieces upon a chess-board.” Smith goes on to argue:
But I think Friedman takes these insights further. Mises takes the behavioral variability of people simply as given – he observes it but doesn’t really explain what drives it. Friedman explains what drives this variation, provides arguments in favor of that explanation, and draws implications about whether or not this problem is surmountable. Friedman also emphasizes, to a greater degree than Smith, that the “principles of motion” driving people’s behavior (their unique webs of beliefs and subjective interpretations derived from those webs, in Friedman’s verbiage) aren’t simply different from those of the technocratic man of system – they are also different from every other proverbial piece on the board, as well as unknowable to both the other pieces and the would-be planner. These are real and important contributions.
I also think many of Friedman’s criticisms of the economics profession are fair and justifiable. However, for reasons I just touched on, I also think his criticisms fail to land on economists who operate in a more Austrian or Smithian tradition, as well as those who operate in the tradition Arnold Kling once called Masonomics. However, given that these economists are unlikely to harbor technocratic aspirations, he gives them fairly little attention in a book dedicated to examining and critiquing technocracy.
Friedman’s criticisms of rational ignorance held some force for me as well. While I think he takes his criticism too far (for reasons I’ll discuss in the next post), he did convince me that rational ignorance is often overplayed as an explanation for political behavior, and while I haven’t jettisoned the concept entirely, I think it’s less applicable than I once believed. Naïve realism, for example, saps rational ignorance of much of its explanatory vigor. Naïve realists aren’t declining to gather more information because it’s not worth the effort, as rational ignorance would imply – they believe they already have all the relevant information. As Friedman puts it, in “the radical-ignorance view, they think their scant knowledge is adequate.”
One idea Friedman articulates better than almost anyone else is how advocacy of the market mechanism is (or at least can and should be) rooted in epistemic humility. It’s not just society that is unfathomably complex – the individuals who make up society are also complex and multifaceted ways technocratic policy can never hope to reflect. His advocacy of markets can be paraphrased as “Look, this social problem is extraordinarily complicated. I don’t know what the best solution is. In fact, even talking about a ‘best’ solution may be senseless, because different solutions will work better for different people and different circumstances. The best approach is to let a thousand flowers bloom and give people the space to work out for themselves how to solve their issues in a way best suited to their own needs and desires.” To be a technocrat is to deny that a problem is too complex for you to understand, to believe that there is a “correct” solution, that you in particular know what that solution is, and that you can effectively use policy to implement that solution by altering the behavior of people you’ve never met in ways you can reliably predict. If you reject the simple-society ontology of a naive realist, you see these kinds of claims as incredibly hubristic. But these are the claims one has to make to advocate a technocratic policy.
These are some of what I see as the key strengths of this book, though the list is by no means exhaustive. But no book is perfect, and no arguments are without weak points, so my next (and final) post in this series will be on what I see as where the arguments fall short.
Kevin Corcoran is a Marine Corps veteran and a consultant in healthcare economics and analytics and holds a Bachelor of Science in Economics from George Mason University.
Feb 10 2023 at 2:38am
Are all problems or issues amenable to market solution or ate there problems that can only be resolved politically?
Won’t many cultural matters need to be resolved one way or another.
Also the market process in many instances may require political resolution.
Feb 10 2023 at 7:23pm
I have not been reading the comments, so perhaps this point has already been dealt with.
Isn’t Friedman’s skepticism so comprehensive that it precludes any judgment about the consequences of public policies, whether they increase or decrease the amount of “statism”? (One might still form judgments based on non-consequential ideals such as fairness or decency.)
Consider that perennial libertarian whipping-boy, the war on drugs. Suppose we legalize all drugs, then what happens? Do we have a society of stoners? How will that work out? What about all those unintended consequences that we’ve read so much about? Who could presume to know the innumerable and multifarious ramifications of such an action?
Feb 10 2023 at 10:43pm
Friedman’s skepticism is indeed very comprehensive, but there is an asymmetry in its application when it comes to policy intervention or the lack thereof. Specifically, Friedman argues that it is active intervention, or technocratic policy, that bears the burden of proof in showing that the perceived problem is real, is socially significant, that they have identified the true cause of the problem, developed an solution for it, and that the solution will be implemented in an effective way that will not cause other, unanticipated problems.
To put it another way – suppose you and I see some situation unfolding. Upon observing it, I say “I don’t fully understand what is going on here, and if I try intervening I have no idea what will actually happen, so I’m going to leave things alone and not interfere.” In saying this, I haven’t assumed any burden of proof. If, by contrast, you claim “I do understand what is going on here, and and I know why it’s happening, and I know how to fix it, so I’m going to intervene,” you have given yourself a burden of proof, and a very strong one, that my position wouldn’t require clearing.
The philosopher Michael Huemer described this sort of asymmetry in a paper he published called In Praise of Passivity, where he applies it to the subject you raise – drugs and the war on drugs:
Feb 11 2023 at 11:23am
You say that Friedman “argues” that intervention bears the burden of proof, but I don’t see an argument. Your second paragraph reasserts the same thing without, as far as I can see, providing any reason to believe it. (I think that almost all arguments that depend on placing the burden of proof on the other side are bad ones, but that is another matter.)
Huemer makes two points. The first is to claim that there is a moral presumption against “coercive” intervention. This is a completely different line of reasoning from Friedman’s (as you have presented that–I haven’t read the book). Getting into what moral presumptions there may or may not be would be a long discussion and a distraction from the main topic, so I will skip this.
The second argues for a presumption against intervention based on our intuitions about a situation in ordinary life. I’m not sure whether this argument is really about our ignorance. In any case, in ordinary life we think we know a lot about other people–what they want, how they think, what is good for them, what they know and don’t know–and based on this knowledge we can gauge the unlikelihood that the woman is about to swallow poison. I don’t think Friedman is entitled to have recourse to the intuitions of people who think they know a lot more than he thinks they do.
And a presumption against intervention would mean that we don’t know enough to challenge the status quo, whatever that may be, i.e. however statist it may be. If we are in the state of nature, we don’t know enough to justify agitating to establish government, and if we have been conducting a war on drugs we don’t know enough to justify agitating to end it.
Feb 11 2023 at 3:29pm
Well, sure – because I was simply telling you what he argues in the book, not claiming to have reproduced the arguments in that comment. But to sum up very briefly and very incompletely, it starts with one of the standard rules of discourse. When you make a claim, you assume the burden of proof for that claim. If that seems unreasonable to you, then so be it, but its pretty standard procedure in intellectual discourse. Technocracy is defined as a system where interventions in society are justified by specific claims of specific knowledge – the knowledge of the existence, severity, causes, and cures of social problems, along with the net costs associated with implementing those cures. Because technocrats, by definition, claim their interventions are justified by their possession of this knowledge, they give themselves the burden of proof by virtue of having made those claims. This is what Friedman means when he says (as I’ve mentioned a few different times over this series) that his project is to put forth an internal critique of technocracy, that is, whether technocracy is legitimate according to its own standards – the standards it sets for itself by the claims it makes. If Friedman is correct that technocrats don’t, and can’t, reliably possess this knowledge, then technocracy is illegitimate according to its own standards, because technocrats would lack the knowledge that they themselves claims justifies their policies.
As far as how the burden of proof relates to my second paragraph – I thought it was obvious, but it also follows directly from the rather standard point of discourse I mentioned before. When you make a claim, the burden to justify that claim lies on you. If I was to say “I’m not going to get involved because I don’t know what is happening or why, and I don’t know what the results of my actions would be” – what, exactly, is it that I would bear the burden of proving? That’s an easy question to answer for the person who says “I do know what’s happening, why it’s happening, and how to fix it!” The person who makes that claim bears the burden of proof to demonstrate they have that knowledge, particularly when they claim that knowledge is what justifies their intervention. What would you suggest is the alternative? For someone to say “I know what’s happening, why it’s happening, and how to fix it, but I also insist that I should bear no burden of proof in justifying any of these claims, or in how I can know what the results of my intervention will be!” Someone could say that, sure, but nobody would have any reason to take them very seriously.
Your final paragraph is puzzling to me. You seem to be moving from “if technocrats can’t reliably possess the knowledge they claim justifies their policies, and technocratic intervention is therefore illegitimate according to its own standards, we must therefore always accept every status quo we find ourselves in, no matter what.” This just seems like a non sequitur of gigantic proportions. There is a wide chasm between the conclusions Friedman reaches, and the implications you seem to be claiming follows from those conclusions. It’s not as if the only possible options in the world are “technocratic intervention” and “total acceptance of any status quo you happen to occupy.” Right?
Feb 13 2023 at 6:47pm
The question is how someone who proposes to “do something” (as opposed to total acceptance), whether or not that something qualifies as “technocratic intervention”, can justify the proposal, assuming the level of ignorance that Friedman imputes to all of us. Again, for example, if you are in the state of nature, how can you know enough about consequences to justify a proposal to establish a government? (I assume you are in favor of having a government.) In doing that you would not be reporting your own reactions to things that happen to you personally, but speculating about, and trying to compare and weigh against one another, the feelings, thoughts and interests of millions of strangers.
Well, one of us has a blind spot about this, and there’s no point in beating it to death. In any case, I think that the sort of thing you are doing–the systematic discussion and analysis of a significant book–is something the political blogosphere could use a lot more of, so I hope you keep it up.
Feb 14 2023 at 10:03pm
At the risk of beating things to death (in a past life, I was robbed by a dead horse and therefore carry a terrible grudge against them), I want to take one more stab at answering your question. And, in fairness, I didn’t do a very good job before because I missed what you were getting at.
You ask we can justify any particular social policy from a given starting point, in any direction, if we are as ignorant about the specifics as Friedman argues. Friedman does address this in the book, although I didn’t highlight it very much in my various posts. But roughly, and incompletely, he argues that this is only a problem if you assume that the only way to argue for or against policies, or to analyze institutions, is by using technocratic reasoning or claiming the four kinds of technocratic knowledge he describes. But that’s not the only way to analyze or advocate for policies or institutions.
First, he distinguishes between what he calls first-order arguments, and second-order arguments. A first order argument would be what I think you have in mind when you talk about how we would need to justify a policy, and also matches Friedman identifies as technocratic reasoning. A first order argument about a policy would be to say “this policy is good because it will have specific ABC effects which exceed the specific costs of XYZ.”
But this technocratic, first-order reasoning isn’t the only way to evaluate or argue for policies. Friedman also talks about how you can use second-order arguments and evaluations for social institutions. Second order arguments are like meta-examinations of a topic. He cites Hirschman as providing examples of both, even as he criticizes the reasoning Hirschman uses. Hirschman defends technocracy from the charges of unintended consequences, first with a first-order argument, and next by a second-order argument, by claiming, as Friedman describes it,
But, Friedman says, that second claim at least had the potential to serve as a second-order argument – if Hirschman had actually done more than make an assertion. As Friedman goes on to say:
Indeed, Friedman says of Power Without Knowledge that “the book will be a second-order evaluation” about technocracy. Friedman doesn’t examine or critique the system of technocracy using evaluative methods of technocracy because, as he points out, doing that just begs the epistemic question about the reliability of technocracy in the first place. But clearly Friedman believes we can still meaningfully evaluate social systems and compare them against each other – his position doesn’t entail that we embrace total skepticism. If that was implied by his arguments, he would have been unable to raise arguments for why exitocracy would be superior to technocracy. And if you look back at the kinds of arguments he raises for exitocracy, you’ll notice he’s arguing in second order terms. That is, he’s making arguments about second-order, systemic factors that would make exitocracy more effective than technocracy, rather than trying to raise first order arguments claiming specific knowledge of specific outcomes of exitocratic and technocratic policies.
Or, recall this argument from Friedman I quote in the fourth part of the series, about why a market for political information wouldn’t produce reliable results in the way a consumer goods market produces reliable goods:
Here, instead of raising a first-order, technocratic style argument along the lines of “consumer markets produce satisfactory outcomes 93% of the time but markets for enlightenment only work 34% of the time”, Friedman is again making second-order arguments about why one particular arrangement has factors making it systemically more likely to produce a desired outcome than the other.
So even with Friedman’s strong skepticism about the reliability of technocratic reasoning and first order arguments, there are still other, better ways to argue for and evaluate policies and institutions.
(If brevity is the soul of wit, as they say, then I am officially witless!)
Comments are closed.