Value-Free Economics
I am saving the meat of this for a Sunday “religious” post, because it will get into theological issues. But to give the background, I met a guy for lunch the other day who is working on a video series to teach basic economics from a Christian perspective.
The guy was asking me about the claim that economics was “value-free.” I explained what this meant in the Misesian system.
The guy challenged me and said he didn’t think economics was value-free, or that it could be. I won’t go into it now–saving it for a Sunday–but I felt very squeamish in giving the guy my pat answer. In other words, I went through the motions of explaining “wertfrei” in standard Austrian economics, but I hardly believed it myself.
(I also told the guy the ominous–but totally true–anecdote that when I was filling out my exit survey at NYU, the 3-digit department code for Economics was 666.)
Anyway, I thought of all this when reading Steve Levitt’s discussion of one of our recent Nobel Memorial winners, economist Peter Diamond (HT2 MR):
Early in my graduate career, I wanted to be an economic theorist. My advisor at the time told me I should give a copy of the paper I was working on to Peter Diamond for comments. I was terrified, but I did it. It was a paper on crime, and a key assumption one always has to make in theory papers about crime is whether one wants to include the utility the criminal gets from committing the crime when adding up social welfare. Early in the paper, I noted that I would not be including the criminal’s utility in the social welfare calculation. Diamond highlighted that sentence, wrote that he thought I should include the criminal’s utility, and read no further in the paper. When I met with him, he simply said there was no point in reading any more once a bad assumption has been made. That’s not the way I think of the world, so that simple offhand comment he made twenty years ago has always stuck with me.
BTW, even though I know this is an exercise in futility, let me assure that I UNDERSTAND WHY DIAMOND SAID THAT. That’s the point, though: Economic science forbids you from assuming that the joy from crime shouldn’t be included in something called “social welfare,” a quantity which everyone assumes should then be maximized. And then we tell ourselves there are no value judgments being smuggled in.
I’m old-fashioned, but I sincerely believe that Betham’s utilitarian calculus is a good guide for questions like this. I believe one of his criteria for utility promoting actions was that
an action is the sort that breeds future utility promoting actions, and not one which tends to diminish future utility promoting actions. By that calculus, a criminal act is of such minimal current utility to the criminal that it is dominated by other second-order welfare effects. Why can’t you assume it away?
The way to think about, I think, is that economics is value free in that the exploration of economic relationships is value-independent. For example, if government policy X truly maximized utility Y, independent of value. Then, once that relationship is established, the economist can apply his values, arguing for what he believes to be right (“government policy X maximizes utility U, therefore we should do it because that’s a good thing to do). I’m just using government spending as an example. I’m just saying that whether a relationship holds true or not is independent of value; the value is applied once the relationship is established.
Once we get into so-called welfare economics, the normativity issue emerges. Whose welfare is maximized? In whose welfare should we be interested? Economics cannot tell us whose welfare should be of interest. Lionel Robbins made this point in the 1930s. So Diamond might have with equal justification stopped reading if the other choice had been made (that is, include the criminals’ utility). Now if our interest were strictly positive, the choice would be determined by what was necessary to understand the behavior of the agents in question. (I am not sure ehat Diamond’s model was designed to do.) There is a positive use of “welfae economics” as well as a normative use.
I meant to say: I am not sure what Levitt’s model was designed to do.
So Diamond is kind of a jerk then?
Jonathan, of course, that is the pat answer I gave to the guy at lunch. And I think that’s a nice story we tell outsiders, but I don’t think it actually describes what we as economists do. Or at least, it’s a very incomplete picture.
For one thing, the very choice to focus on people’s “utilities” in the first place influences the analysis. Suppose you detest utilitarianism and think it’s a monstrous perversion of justice and ethics. Well, how are you supposed to use the “neutral” tools of modern economic science in that context?
Mario, it’s similar with your comment: It’s not merely that we have to know “whose welfare” to maximize, but that we are even starting from that premise. (Maybe this is part of your point; I’m not sure.)
AC, no, that’s not what I’m saying. I’m saying he is being more consistent than Levitt. Yes, if you are going to use modern welfare economics in its “value-free” sense, you have no business assuming from the start that the utility of criminals doesn’t count. You might as well “prove” the case for free trade by assuming producers shouldn’t be in the calculation.
And yet, as Levitt’s common sense told him, that’s nuts.
Here’s part of what I’m struggling with: Economics started out as an offshoot of moral philosophy, but now it prides itself on being a completely independent “scientific” discipline with no “irrational” reliance on “arbitrary” value judgments. (Sorry for all the quotation marks.) And look at what economics has turned into. Not only is it now harmful as opposed to being helpful–compared to the analysis of Adam Smith and Bastiat–but it actually justifies things that are crimes in my book.
This is not a coincidence.
I’ve long believed that Keynesianism is simply a jargon-filled hoax meant to obscure through a complex mechanism the immoral looting of average people who have been convinced that the entire process is simultaneously scientific, morally neutral AND good for “society” and “the economy”. And too complicated for them to understand. Critics of the looting are then shot down if they cannot keep up with the zigs and zags of the various schools of jargon (“oh, you really can’t call THAT Keynesian!”).
At the same time, I generally would explain things just like Jonathan did.
Following up on what Mario wrote, the problem is not economics, but welfare economics. Seems to me that Mises opposed that kind of utility economics because utility is not measurable. I don’t think welfare economics has any place in economics. In classes I teach, students are able to see how stupid it is from the very first example. I get nothing but laughs and hoots from them. Of course, it could be the way I present it because I think it is stupid. Welfare econ is metaphysics and should be left to metaphysics. In the metaphysics of econ, we have to deal with morality in general and the morality of property.
I like metaphysics and think it is even more important than economics, but the two should not be mixed. Without welfare econ, economics can be value free and just point people to the policies that will lead to the desired outcomes.
I must be missing the boat on this one. If the root of economics is human action and every human action is based on a subjective value, then how can economics be “free” of value judgements?
I think the science of economics has been perverted to justify forcing “subjective” from human action. It is important that economists claim they aren’t concerned with value judgements as a Pontius hand washing.
As Mises wrote, econ can tell you how to achieve certain objectives, or if your goals are even attainable, but it can’t tell you which goals are the most important or moral.
I agree, to a great extent value is intermixed with economics. For example, economists of certain schools only develop relationships which seem to support their particular view of the world. In any case, regardless of how economists operate, I still think of economics as value-free, because the relationships developed are true or false regardless of value. How an economist goes about developing economic science is not a comment on economic science itself, rather on the economist. Similarly, biology is also value free, even if some scientists decide not to research stem cell technology because it is against their values.
In any case, I’d also be as bold as to say that you can tell a great economist from a good economist by the way he approaches the study of economics.
I know what you mean, but it’s very strange to me to say economics is the value free judgment study of value judgements.
I think Mario is on target here: economics can be value-free only in that it avoids making policy recommendations.
But Gene, suppose I go to grad school and learn what the “social planner’s solution” is, one that moves society towards a “Pareto optimum.” And then I learn about the “deadweight loss triangle” from a sales tax or a tariff.
It’s fine to then say, “Oh, but don’t misunderstand, maybe you WANT a Pareto inferior outcome, or maybe you LIKE deadweight loss. No pressure.”
(And yes, “Pareto optimum” is part of welfare economics, but take something “neutral” from Austrian economics. It’s similarly loaded.)
I’m not renouncing my faith in economics per se, I’m just saying I think we are kidding ourselves if we really think we’re not smuggling value judgments in when we “do” economics. That doesn’t mean the economic tools themselves are “tainted” with our value judgments, I’m just saying I don’t think we practice what we preach.
Part of the point I’m making below is that there can be no such pareto superior planner’s solution in the real world. If there were, the trades necessary to make it happen would happen.
Sure, maybe in some other world with a different cost structure the social planner’s plan would be better, and therefore happen, but that’s irrelevant.
Utility functions are supposed to predict behavior. When observed behavior differs from what’s predicted, the model is wrong, not the people.
Welfare economics is hopelessly confused to an extent that even many of its fiercest critics don’t appreciate.
Given the way social welfare is usually defined, any model with utility-maximizing agents will result in an equilibrium that maximizes social welfare relative to the given constraints. Thus, the minimal assumptions that are included in almost any economic model guarantee if the real world is any kind of economic model, then inefficiency can never be observed: social welfare is always maximized. Hence thinking that economic science can recommend social-welfare improvements is deeply wrong. It’s results from a misunderstanding of economic models, which are meant to predict behavior. When people’s behavior differs from what’s predicted, the model is wrong, not the people.
The confusion is introduced to students early in their economic education. The prisoner’s dilemma, for instance, is often presented as an example of an “inefficient” equilibrium, but it is only inefficient in an irrelevant sense. This is because the “efficient” outcome where neither prisoner talks is impossible given the assumptions of the model. One might as well argue that the observed equilibrium in a prisoner’s dilemma is inefficient relative to peace on earth and good will towards men.
I’ve heard this idea that economics is “value-free” stated a lot before, but it’s hard to take seriously. First, as you point out, it’s hard to see why we should be focusing on “utility” in the first place. And if we do, then the model is counterfactual anyway since utility can’t be measured. For example, how would Diamond deal with the problem of showing a rapist’s utility is less than the victim? The obvious answer is that he can’t, that an economic theory based on utility is going to say (duh!) that any given action whether right or wrong could maximize “utility” in some circumstances.
Second, even if we grant that economics is “value-free” in the sense of endorsing no value statements in particular it still doesn’t follow the discipline is thereby value-free. Consider physics, a science as value-free as you can get. Everyone working in the Manhattan project was working in a “value-free” setting of particles, waves, and atoms. But when you recognize that all of their work was dedicated to making a destructive bomb, a result which obviously would only serve the interests of a government, then it’s ridiculous to characterize the Manhattan project as a “value-free” endeavor. Similarly, if an economist like Krugman spends all his time studying methods of using a central bank to save an economy, something obviously only useful to an intrusive government, what does that say about Krugman’s priorities?
Third, socialists have been saying for a long time that economics is not scientific and “value-free” at all:
http://www.infoshop.org/page/AnarchistFAQSectionC1