The essay by Edward Glaeser to which Matt Kahn points is, in my view, astoundingly wrong-headed. And, as Glaeser notes but doesn’t reflect on, the Adam Smith of The Theory of Moral Sentiments would have agreed with me, and not with Glaeser. The fact that selfishness constrained by law and the market can generate socially useful outcomes doesn’t make selfishness, or the freedom to pursue selfish ends, good things in themselves.
But what’s most disturbing about Glaeser’s essay is not its conclusion, but its reasoning: in particular, the attempt to put a preference for freedom at the heart of economics. That’s a category mistake: freedom as a moral or political principle is neither an assumption nor a conclusion of economic reasoning.
Economic analysis tells us that, under certain conditions, free choice will lead to greater preference satisfaction than constrained choice, or no choice. It also tells us - via the principle of diminishing returns - that, other things equal, a more equal distribution of income generates more preference satisfaction than a less equal distribution. Neither of those observations constitutes a moral principle. And neither is more fundamental, more “economic,” than the other.
If an economist wants to believe in freedom, or for that matter equality, as a fundamental principle, he should do so on his own time. Neither commitment follows from the science of economics. If it’s really true that “most economists” acknowledge no duty to others beyond what is legally mandated, that reflects a flaw in their character or their understanding, not a logical entailment of anything about their discipline. But I doubt that it is true in fact,
The commitment to “efficiency” in its economic sense - not forgoing the opportunity to secure more of one good, or more benefit for one person, unless doing so provides at least a compensating gain in terms of some other good or the well-being of some other person - may fairly be called a moral commitment, and also an entailment of the economic way of thinking.
That commitment is, it should be noted, contrary - in its insistence that more of a good thing is always better, in contrast with the view that enough is enough - to most of the world’s received philosophic and spiritual wisdom, going back at least as far as Aristotle. When I teach economics, I point both to the power of its analysis and the dubious assumption on which it rests. But at least the principle of efficiency is properly a part of economic reasoning. That is not the case for freedom, or for selfishness.
(Mark): “: “When I teach economics…
Heaven help us.
(Mark): “…I point both to the power of its analysis and the dubious assumption on which it rests.”
To you apply the same level of skepticism to the dubious assumptions upon which competiting theories rest? I mean, any system which uses basic arithmetic makes huge (and likely false) assumptions about continuity and infinity. Closer to home, do you point to the “public choice” critiques of State activism?
(Mark): “Yes, in principle libertarianism and conservatism are different. In practice, both of those labels apply in U.S. politics to people who want rich people to get richer and poor people to get poorer.”
Name even one such person.
Sorry. That’s “Do you apply…”
In MR. SAMMLER’S PLANET, Saul Bellow writes: “One had to learn to distinguish. To distinguish and distinguish and distinguish. It was distinguishing, not explanation, that mattered. Explanation was for the mental masses. Adult education. The upswing of general consciousness. A mental level comparable with, say, that of the economic level of the proletariat in 1848. But distinguishing? A higher activity.”
Apologies, Mark, but I don’t understand what you’re getting at here. Are you saying efficiency is the basic value of economic “science,” rather than freedom as Glaeser says? Or are you saying that it’s neither, and econ is the study of how to make the most widgets? I’m not trying to be snotty here, I really don’t quite follow.
What bothered me most about Glaeser’s piece was this idea that free trade is about freedom, when we are free-trading with people who have very little, and I see not much evidence that it gives them more. A good friend pointed out to me the other day that it was a bad idea to give China MFN status, since we lost a big chunk of leverage there.
I for one have my doubts about what we could do to move China on human rights, and also about our own commitment here, but I have to say, I was abashed that we don’t even talk about this much anymore. I suppose our ginormous debt and cr*p trade balance have something to do with that. And I rather admire some things about China’s industrial policies - at least they have one that’s not, let’s give corporations everything they want and who cares if it benefits ordinary Americans. (Anyone remember them?)
Other than that, I see economists mostly as people who deal in issues so abstracted from life that their ideas don’t have much value. And I think economic theory is in fact cold at its heart. “Freedom” is just the nice label for it. It is a perfect system for explaining why giving a hooey about the other guy is really bad for society. I almost think it was designed that way, it fits so well.
And the other side knows the value of everything and the cost of nothing.
Mark, I think you may underestimate the degree to which economists have come to value Pareto efficiency over “greater preference satisfaction” in some larger sense. Because Pareto thought it was incoherent to compare two people’s satisfaction, he proposed voluntary exchange (which by definition is the preferred outcome of both parties compared to the status quo ante) as a matter of intrinsic worth, not just as instrumentally useful to maximize utility. (From Pareto’s perspective, there *is* no such thing as utility or happiness outside Pareto effiency or, somewhat loosely as a descriptor of a whole system, Pareto optimality.) The whole point was to rule out what you think is fundamental: arguments from the decreasing marginal utility of income, which then as now seemed excellent arguments for redistribution.
Pareto was fairly open about his political and social views, which were to a first approximation Social Darwinist and contemptuous of democracy. (While he didn’t live long enough to judge fascism in practice, many fascists loved his work.) Whether modern economic analysis requires one to hold those views seems doubtful. But that the microeconomist’s standard “ordinal” view of utility entails placing what could be called a certain kind of free choice over a certain understanding of happiness or utility-not out of love of freedom so much as out of scepticism towards social measures of utility-is not an absurd claim but to my mind a fairly standard one.
Andrew Sabl’s comment is very well put. It is not necessarily clear, however, why Pareto Optimality should be privileged over, say, Foley Fairness, as a standard criteria of policy comparisons, other than Pareto’s greater congeniality to the interests of the pre-emptively rich.
In practice, economists often use Pareto efficiency criteria to justify policies — including, as an example, a broad range of “free trade” policies — which will make the rich, richer at the expense of the relatively poor or merely middle-class, on the grounds that certain counterfactual policy adjustments, which will never be enacted, could institute the kind of re-distribution that would be make the policy pareto efficient. The expected gains from free trade in stasis are large enough that a portion could be re-allocated from the small class of big winners from the policy, to the mass of losers, so the policy is deemed pareto efficient. Could be, not will be. In practice, the pure efficiency gains are modest, relative to the collateral effects on economic rent claims, which has the effect of dynamically changing the distribution of income. In other words, there is a gain from increased efficiency, but it is swamped by the distributional effects of changing the relative power of, say, employers and employees, or banks and their customers, or manufacturers and consumers. But, once those changes in the distribution of income are accomplished facts, and dubbed “natural” market outcomes, then Pareto efficiency is an excuse to oppose remedial action.
You can see that in policies of “deregulation” and “globalization” which promised increased competition and efficiency, but also demolished sources of economic rent that formed the institutional foundations for, say, unions. So, we got lower prices at Wal-Mart, and, also, the lower wages at Wal-Mart, and a politics without an effective union movement. Or, we’re promised increased efficiency and innovation from financial sector reform, and we get ATMs and a few years of high-interest CDs, but also Big Banks, who have the political power to extract bailouts, void usury laws, and enact harsh and punitive bankruptcy laws. Growth in the median wage stagnated, while all economic growth flowed to the few, who now helpfully inform us that “we” cannot afford Social Security and Medicare.
That’s not a philosophical “skepticism towards social measures of utility”. That’s an intellectual shell game, played to serve the plutonomy.
Andrew Sabi, its true economists use pareto efficiency to get out of the diminishing marginal returns to wealth or more generally the comparisons of utility.
But what do economists reach for to analyze insurance markets?-diminishing marginal utility of wealth. Finance analyses in general rely on it. Pareto efficiency in the sense that its used in economics -that winners could compensate losers- is immediately abandoned when it comes to actually analyzing choice under uncertainty. That should tell that it is not a bedrock principle of economics.
When challenged, economists tend to insist that we do not really think that people are swine intrested only in consumption and leisure. Most economists argue that we are simply considering the economic aspect of things and that, of course, people are also altruistic, curious, and well all sorts of things.
Glaeser is plainly not like most economists. Even if we could be wealthy, leisured and selfish, this would only be Pareto efficient if we were swine. Real people want to be surrounded by not completely selfish other people, not as a means to some other end, but because we are social animals.
Glaeser has not read the statement of the welfare theorems carefully. Arrow noted that the results only heald if people only cared about their own consumption. Arrow couln’t have imagined that real people were like that (Arrow is by all accounts very kind and concerned about other people — oh and a socialist). The simple theorems illustrate what would be true if people cared only about consumption. The person who proved them did not think they were any big deal.
In fact (he can check with Mankiw down the street) the free market outcome is not Pareto efficient if people care even a little bit about strangers (caring one millionth as much about a stranger’s pleasure minus pain as about my own is plenty). All, including the rich, rationally choose redistribution. I don’t want my money to go to the poorest, but I do want your money to go to the poorest (by assumption I don’t care much about you either). So taxing and redistribution (at least a little) satisfies everyone’s preferences.
I think Andrew Sabl got it largely right - but doesn’t it usually get even worse. Since (almost) no actual change is Pareto efficient economist (very) often take an intermediate step through the Kaldor-Hicks criteria – followed by some lip service to the notion that the outcome would have to be redistributed to the losers. That they usually don’t (read “never”) have any idée how this would be accomplished and most of the time, if you ask them, is fully aware that it would be impossible – don’t seem to be of any big concern.
Furthermore, since your willingness to pay, possibility to profit from, etc - will increase with income (ceteris paribus) – the game is usually rigged to the point of absurdity.
(Robert): “…taxing and redistribution (at least a little) satisfies everyone’s preferences.”
Does this not depend on assumptions about the motives of State actors? If I prefer taxation to direct contributions, why should State actors surrender resources that they control? Who taxes the tax distributors?
Malcolm Kirkpatrick
Well the policy is modelled the way economists model policy — as if the state is a machine that applies rules with no actors, no agents and no principal agent problem.
I see some traffic to my blog from this post, so I will try to give some useful links. The link to my blog isn’t really useful.
Michael Froomkin mentiones the result (with link to me attempting a proof)
http://bit.ly/hcKiIo
Greg Mankiw notes the result in a recent article arguing aginst utilitarianism
http://bit.ly/agX42J
Mankiw cites Thurow 1971 “income Distribution as a Pure Public Good” Quarterly Journal of Economics vol 85 pp 327-336. for the basic argument.
The claim is that it is possible that no one will choose to give to people poorer than he or she is, but everyone will unanimously support a policy in which wealth (or income) is taxed and the proceeds given to the poor. Consider an example with two types of people rich and poor. Assume the marginal utility of income is 1.1 times higher for the poor than for the rich (say utility is logarithmic and the poor have consumption 10 11th of that of the rich — this ratio is not very important it Assume there are a 100 million of poor people and of rich people (this number is very important). Assume that people enjoy consumption which gives them animal pleasure and that their utility (objective, happiness, what they want to maximize) depends on their own animal pleasure and on other peoples’ animal pleasure but a million times more on their own (so they are a tiny bit altruistic but just a tiny bit).
Ah but with 200,000,000 people I care about the average level of animal pleasure 200 times more than I care about my own 200 = 200,000,000/1,000,000.
Rich won’t give anything to the poor as 1,000,000 > 1.1. A policy which takes one small unit from the rich and gives it to the poor increases animal pleasure of each poor person by 1.1 and reduces animal pleasure of each rich person 1 so it increases average animal pleasure by 0.05.
If I am rich, the policies effect on my happiness is
-1 - (100,000,000)1/1,000,000 + (100,000,000)1.1/1,000,000 = 9 >0. I like it.
If I am poor the effect on my happiness is
1 - (100,000,000)1/1,000,000 + (100,000,000)1.1/1,000,000 = 11 >0. I like it.
Everyone likes it. People unanimously support a tax and transfer program. This means rich people too. Recall these are rich people who choose to give nothing to the poor.
The point is that rich person A doesn’t want to give rich person A’s money to the poor but he does what to give the money of rich people B,C,D and etc. All the rich people can agree on a policy which takes from all of them and gives to the poor even though none of them would give voluntarily.
There is no trick. The standar result that the free market outcome is Pareto efficient is based on many assumptions and one of them is that people care only about their own consumption an are perfectly entirely absolutely selfish. With only 99.9999 % selfishness, the first welfare theorem doesn’t hold.
Thanks. I’ll have to think on that. In the meantime…
(Robert): “Well the policy is modelled the way economists model policy — as if the state is a machine that applies rules with no actors, no agents and no principal agent problem.”
Those economists probably would not include James Buchannan, Gordon Tullock, Mancur Olson, and Young and Marcoulier, the authors of “The Black Hole of Graft: The Corrupt State and the Informal Economy” (AER).
Nor…
Randall G. Holcombe
Government: Unnecessary but Inevitable
__The Independent Review__ Volume 8 Number 3 Winter 2004
Eduardo Zambrano
“Formal Models of Authority: Introduction and Political Economy
Applications”
__Rationality and Society__, May 1999.
I suspect that charity is one of those “collective action problems”.
(Robert): “Mankiw cites Thurow 1971 “income Distribution as a Pure Public Good” Quarterly Journal of Economics vol 85 pp 327-336. for the basic argument.”
I accept charity as a public good. I do not accept the case for State provision of public goods. The case for State provision of public goods contains a simple flaw: corporate oversight is a public good, and the State itself is a corporation. Therefore, oversight of State functions is a public good which the State itself cannot provide. State assumption of responsibility for the provision of public goods transforms the free rider problem at the root of public goods analysis but does not solve it.
Hi, Robert. I think I would go further and claim that many rich people (but perhaps not all rich people) would want to give their money to the poor provided all other people who are as rich (and richer) do the same. Similarly, I think a lot of business managers/owners want to run their business on a set of ethical rules (e.g., without cheating or misleading customers, without harming the environment) but cannot afford to do so if their competitors do not play by the same set of rules. I suspect you feel the same way.
I might take this even further and claim that people’s preferences aren’t independent of each other. If I join a group of people who are other-regarding, I will find myself increasingly inclined to be other-regarding. If I join a group of selfish people, I will find myself increasingly inclined to selfishness.
Hi, Mark. Are you saying that the moral commitment of economists is limited to efficiency? I’m not sure if that is right. In “The Idea of Justice”, Amartya Sen identifies two contrasting approaches to justice. The social contract approach identified with Rawls, Hobbes and Rousseau is said to focus on justice in terms of right institutions. In contrast, the “comparative approach” identified with Adam Smith and Karl Marx is said to focus on justice in terms of achievable outcomes. Economists generally come under the latter tradition although perhaps some modern economists (especially those associated with the Chicago School) seem to prioritise institutions over outcomes.
Thus when comparing alternative states of the world, economists regard the well-being of individuals in terms of health, education, social connection, etc as intrinsically valuable. The attention paid to individuals is the “moral heart” of economics (I think). Economic efficiency has only instrumental value in that greater efficiency makes it possible to achieve improvements in well-being.
There is a question as to whether economists focus unduly on what Sen refers to as “culmination outcomes” as opposed to “comprehensive outcomes”. The latter would take into account procedural fairness when evaluating outcomes on the basis that individual well-being depends not only food, shelter and other such things, but also on procedural fairness. That is, we care not just about having a full stomach; we also care about being fair to each other. Most modern economists seem to focus only on culmination outcomes. Perhaps that is why you do not think that concern for individual freedom is part of modern economics.
If Edward Glaeser thinks that individual freedom is the only thing that matters morally, then he seems to fall under the social contract tradition focused on achieving perfectly just institutions. On the other hand, if Edward Glaeser thinks that individual freedom is just one of the many things that matter morally, he might be part of the tradition that cares about comprehensive outcomes.
I feel a rustic at this particular High Table, but let me throw in a couple of new points on the maxim of non-satiation.
First, this is a tautology if you define economics, as is often done, as the analysis of scarcity. So conceived, the would-be science has nothing to say about plenty. Should we aim for plenty and taking goods out of the rationing market? It’s no longer a satisfactory answer to Marx and others who have held this up as an ideal to say that it’s infeasible. This is no longer true for many consumption goods; nor for those with zero marginal cost, like information. A free bar is not always surrounded by drunks.
I’ve heard it argued, at the OECD no less, that a zero price leads to infinite demand. This is obvious nonsense; consuming things takes time, and time is necessarily scarce. More practically, most of my supermarket trolley is filled with things for which my demand is quite inelastic over a wide range of plausible prices. I would not eat more smoked salmon if the price were half, nor less if it were double. The price of steak is half in Brazil what it is in Spain, and I buy about the same amount. I would not buy any caviar at a tenth the price because I don’t like it much. My desire for supermarket-trolley goods is pretty much satiated.
The only reason I pay attention to their price is opportunity cost. So long as I have some unsatisfied consumption desires, then money has a positive marginal utility, and I will look for and buy the cheaper smoked salmon or diesel fuel. And I do: I’d like to fly business class, travel more, change the car for a hybrid, live in a bigger house, buy pictures, provide better for my family and charities. If I had the money to do these, my tastes would probably change and I’d hanker after a yacht or a jet. Aristotle, Veblen, and Robert Frank have this one nailed. But the market economy does not require universal dissatisfaction and greed; just moderate greed at the margin; and even there, not necessarily by everybody.
There are I suspect quite a few satisfied, Aristotle-compliant people around. It sems a reasonable objective to increase their number.
(James): “I would not eat more smoked salmon if the price were half…”
…But you might feed it to your cat. If the price went low enough, you’d use it for fertilizer. If the price of bread went low enough, hog farmers would use bread as pig food. This happened.
(Mark): “ When I teach economics, I point both to the power of its analysis and the dubious assumption on which it rests.
Milton Friedman addresses the issue of unrealistic assumptions in a chapter of __Essays in Positive Economics__. Theories are tools. Defects don’t disqualify theories until a better tool comes along.
Libertarians tell me “People always act in their self interest.” Why then do people catch the flu or fall downstairs? Why did Taimanov play 46. Rf6 in his 5th round game against Fischer in their Candidates’ Match?
Standard economic theory supposes that humans have preferences. Becker discusses preferences (tastes) in his very interesting essay “De Gustibus No Est Disputandem”, and argues for a paradoxical view that preferences do not differ between people (why then would anyone trade?). One implicit assumption about preferences (an ordering of world-states) is that preferences are stabile over time. If mutual exchange raises satisfaction, two individuals on opposite sides of the Earth at 45 N and 45 S could trade houses every six months (one prefers winter and one prefers summer) and ascend a scale of ever-increasing satisfaction. Since human preferences are not stabile and appetites can be satiated, three people could trade fruit trees in a circle and ascend an ever-increasing scale of satisfaction. Two psychologists, Solomon and Corbit, discuss “the temporal dynamics of affect” in “An Opponent Process Theory of Motivation” (AER, Dec., 1978?), which Tibor Scitovsky recommended to AER. Economists will play with the definition of “transaction” or “goods” to minimize the adjustment they have to make to theory.
Freedom matters because, whatever people prefer, imposition of restrictions on their self-interested actions will, in general, reduce their satisfaction. Thus, a cost-benefit argument for State action starts in a hole: coercion (the instrument of State action) imposes an obvious cost.
Malcolm Kirkpatrick: “… But you might feed it [smoked salmon] to your cat”. Hobbes doesn’t like smoked salmon any more than I like caviar. He only eats good-quality dried cat food, don’t ask me why, and his wants for consumption goods are satiated as well. (Non-market services like brushing are quite another matter.) I wonder: is there a Marxist millennium for cats, as many sentimental believers think there’s a heaven for pets? Some migh think they’ve got it already.
Your other examples are of intermediate goods and not relevant to my point about the satiable utility of final consumption goods.
Malcolm Kirkpatrick 2:
” a cost-benefit argument for State action starts in a hole: coercion (the instrument of State action) imposes an obvious cost.”
Not, as Robert Waldmann has pointed out, if people like or are indifferent to coercing other people for the common good. They may also like cooperating with others for the common good; and find coercion acceptable to deal with a disruptive minority of free riders and thieves.
My cat is called Hobbes out of respect not agreement with his pessimistic psychology. But he was surely right to think that if people are the selfish monsters of textbook microeconomics, a peaceful society is going to require a lot of state coercion.
(Malcolm): “a cost-benefit argument for State action starts in a hole: coercion (the instrument of State action) imposes an obvious cost.
(James): “Not, as Robert Waldmann has pointed out, if people like or are indifferent to coercing other people for the common good.”
Even then. The coercion is an up front cost. “The common good” is a benefit. Whether the real coercion delivers the promised benefit is an empirical question. Aside from the evident up-front costs, opponents of State provision of the public good of charity (Food Stamps, Social Security, education services) will include also the likelihood of insider self-dealing (see Axelrod, __The Evolution of Cooperation__)and recipient moral hazard, short and long term (i.e., the corrosive effect of dependancy).
I do not reject all State interventions, but I suspect argument that supposes that State actors differ systematically from the population at large in anything but their access to the tools of State violence.