Efficient Isn’t Moral

Efficiency isn’t morality, and it is a serious confusion to think it should be. Let me try again to explain.  I said:

Economic welfare cares not about giving people experiences but about satisfying their preferences. … If we do something a dead person would have wanted, that counts as a benefit.

Adam Ozimek responded:

But we care about satisfying people’s preferences because, unlike the dead, they can know that those preferences being satisfied. … If were going to count the preferences of the non-existent, then it would seem that the number one priority of all society would be to bring as many of them as possible from non-existence into existence. The easiest way to do this is to mandate pregnancy. … If we care about satisfying the preferences of the dead even though they won’t know their preferences are satisfied, does that mean we should not be concerned with whether or not living people know when their preferences are satisfied?

Adam reminds us of Tyler’s position:

Dead people don’t count in the social welfare function. (If they did, how many of them would prefer non-democratic or racist outcomes?  And would we count that?  We shouldn’t and we don’t.)

When our distant ancestors sat around debating if to change locations, expel a troublemaker, or attack neighbors, they were often ambiguous about whether they were choosing what they wanted or what was moral; they preferred to pretend these were the same.  We similarly prefer ambiguity when we argue policy today.

So it is important to clarify: As an analysis tool, economic efficiency is designed and well-suited to finding win-win deals that [added: tend to] get us all more of what we want. It is not well-suited to achieving moral outcomes, except when morality happens to coincide with getting people what they want.  Otherwise, win-win deals will predictably not achieve morality when many involved do not want to be moral.

Many of us want things we will never experience directly; we want our children to prosper after we are gone, for example. This is especially true of our moral wants; we want our donations to Africa to actually help real Africans. So we are understandably wary of deal-making frameworks which explicitly suggest that they seek only to achieve the appearance, not the substance, of our wants.  So yes, a deal-finding analysis tool should definitely count unseen wants!  Furthermore, observers concerned that deals might neglect morals should be especially eager for our deals to achieve unseen wants.

Frameworks for finding win-win deals should also try to include as many things as possible that can have wants and participate in deals.  This includes racists, pedophiles, slaves-owners, robots, animals, distant past and future folk, and future folk who may or may not end up existing.  Yes many may be morally offended if racists get what they want, but that offense counts in what other folks want, and therefore enough offense will ensure that win-win deals will not give racists much of what they want.

Limits on contract may distort prices and interfere with the ability of efficiency analysis to help us find useful win-win deals.  But that is a good reason to enforce more kinds of deals, not to try to distort efficiency for a task to which it is poorly suited: choosing moral acts.

Added: Bryan Caplan responds.

GD Star Rating
Tagged as: ,
Trackback URL:
  • nazgulnarsil

    This is a great post that summarizes an *absolutely massive* number of the misconceptions and pointless arguments about economics I’ve experienced. How many ways are there to say that economics is positive and not normative? Even the most renowned economists seem to spend an inordinate amount of time explaining this to people.

    *they preferred to pretend these were identical.*

    people constructing narratives under which their wants are moral seems to be a human universal. It is a roundabout way of telling others that they should value your wants, or if they already do confirming your alliance.

  • misterxroboto

    efficiency is perfectly suited for utilitarians.

    waste is a theft from utility. efficiency is a way to eliminate waste, adding more utility, making it more ethical.

    if you’re concerned with the nonsense that is natural rights, then you’re absolutely correct. however, please don’t group all ethical systems together.

  • Off topic

    Robin, do you have links to or could you write a post on the limits of nanotech.

    I saw your presentation on the potential causes of the singularity where you pooh-poohed nanotech as effecting only a small portion of our economy.

    This seems wrong to me. Nanotech could potentially effect all portions of our economy that deal with the manipulation of matter.

    So we are not just thinking about manufacturing in the traditional sense, we are thinking of transportation, retailing, wholesaling, medicine, construction, personal services, in the most extreme forms potentially even education.

    Anything where are ultimate goal is to reconfigure the matter around us. Even though there are many “service” jobs the goal is often matter-centric. For example if I eat at a restaurant the entire production chain that happens after the food comes out of the ground is with the goal of placing a particular form of matter in front of me.

    However, it is all counted as services, not ” hot meal manufacturing”

  • Michael Kirkland

    While we are alive, we want our children to prosper indefinitely. Once we are dead, we no longer have that or any other wants.

    Doing something a dead person would have wanted only counts as a benefit via the proxy of a living person wanting to respect the former person’s former wishes.

    • anon

      While we are alive, we want our children to prosper indefinitely. Once we are dead, we no longer have that or any other wants.

      Perhaps. But when a living person has made a deal which has helped her pursue her wants (such as leaving a bequest to a existential-risk reduction organization), why shouldn’t we enforce this deal after her death?

  • Chris

    Indeed, and there are increasingly more efficient ways of performing many tasks than using people. Unfortunately, those tasks are also the only ones many people are ever capable of performing. What do we do with them? Ignoring them could prove catastrophic for the entire system.

    • Chris

      The usual reply is that people will shift into what they do best, creating things. The problem is that most people CANNOT perform this function to an economically useful degree.

      Unfortunately, the illusion they can was created as a result of the cognitive sorting of the 20th century. It is an exceedingly dangerous illusion and threatens the stability of the entire system because those frozen out will not simply sit on their hands.

      • anon

        The problem is that most people CANNOT perform this function to an economically useful degree.

        This is an unproven assumption. In a world where most tasks were performed by machines, even a limited amount of creativity or other human-specific input would have enormous value.

  • misterxroboto

    i’d like to revise my previous statement:

    i’ve never pressed the topic, but i’m not sure how many people would be convinced that efficiency is an end in itself. you may be willing to say that reducing waste is always and everywhere good even though you have no idea what you’re not wasting. after all, bullets are expensive compared to other methods of genocide.

    so while you’re correct that efficiency has no intrinsic value, i’m not sure i’ve ever heard someone say it does. i see it as having near-universal instrumental value, but that seems to be beyond the scope of your argument.

  • rapscallion

    Economic efficiency as social welfare is an incoherent concept that ought to be abandoned. People use it to argue that we’d be better off if constraints taken as exogenous were removed, but as soon as one allows that those constraints themselves are endogenous to a larger framework, the possibility of inefficiency vanishes. If we’re all maximizers, we’re all maximizers; the world we see is the best we can do. Nothing unobserved can be inefficient.

    • Doug S.

      > If we’re all maximizers, we’re all maximizers; the world we see is the best we can do. Nothing unobserved can be inefficient.

      Game theory would dispute this. A Nash equilibrium is not necessarily Pareto optimal. If both participants in a Prisoner’s Dilemma defect, the outcome is inefficient even though both participants are maximizing their own welfare.

      • rapscallion

        But a prisoner’s dilemma is a prisoner’s dilemma; if there’s a way out, it’s not a prisoner’s dilemma. Hence, the observed equilibrium is the best you can get under the given constraints.

      • anon

        Hence, the observed equilibrium is the best you can get under the given constraints.

        This is clearly incorrect. For one thing, the extent to which prisoner’s dilemmas and other comparable situations are effectively addressed depends on broad, ‘institutional’ frameworks which are amenable to continuous improvement. Furthermore, none of us are perfect maximizers, so we can in fact do better even when it comes to individual actions.

      • rapscallion

        i) When you augment a prisoner’s dilemma game with institutional considerations, you just create another, larger game that may or may not be a prisoner’s dilemma. Individuals will simply choose their maximizing strategies in this larger game. Since they’ll all be maximizing, the observed equilibrium will be efficient. As long as people are maximizing utility, nothing observed can be inefficient. I fear the point is deceptively simple.

        ii) I’m not sure what you mean by “none of us are perfect maximizers.” If you mean that we maximize against constraints, of course. The point is that observed outcomes are the best we can do under those constraints. If you mean, on the other hand, that we don’t actually maximize utility, then you must abandon traditional economic welfare analysis because prices cannot be taken to be measure of willingness-to-pay.

  • rapscallion

    Crap, that should be “Nothing observed can be inefficient”

  • Pingback: The Dead Don’t Want «  Modeled Behavior()

  • I can’t help but thinking there is an inconsistency here. If you assume that preferences are well-defined and time-consistent, then we can take preferences as being reflected by what people say and feel they want (a “revealed preference” model), then your economic efficiency arguments in favor of allowing puppy-torture contracts (or whatever) can be supported by economic theory. However, dead people reveal no preferences–we can just poll them after they die and confirm that. So their pre-mortem desires should be considered only as a way of convincing the currently living that their contracts will be honored post-mortem.

    On the other hand, what if preferences are poorly defined and time-inconsistent. Then expressed and revealed preferences are not a sound basis for economic analysis. This might get you your ‘enforce contracts post-mortem’ result, but that is because you are ignoring what people say they want. That reasoning also justifies enforcing rules like “you can’t contract to take that addictive drug or engage in that sexual activity for pay” because legislators argue that they know preferences better than individual agents. Isn’t that pretty much what moralists say all of the time?

    So you can argue that we should pay attention to dead people’s prior stated preferences OR that we should let expressed/revealed preferences rule and never interfere with contracts, but I don’t see how you can have it both ways.

    p.s. Current evidence demonstrates pretty convincingly that people’s preferences are unstable in the face of trivial changes, often self-contradictory and almost always time-inconsistent.

  • I think this post could have benefited from linking to your preferred version of morality.

    rapscallion, it sounds like you are rediscovering the political economy of Demsetz. It is discussed in the first paper here.

    • rapscallion

      Thanks. This basic point has been made by myriad economists, but people keep ignoring it because efficiency as social welfare is such an attractive idea. I feel the paper that best puts forth the basic argument is the little noticed:

      Economic Inefficiency: A Failure of Economists

      # Economic Inefficiency: A Failure of Economists
      # Author(s): Michael Staten and John Umbeck
      # Source: The Journal of Economic Education, Vol. 20, No. 1 (Winter, 1989), pp. 57-72
      # Published by: Heldref Publications
      # Stable URL: http://www.jstor.org/stable/1182717

  • John

    Efficiency rests on assumptions about morality. For example, if I were to argue that the annihilation of the human race would be good for efficiency because it would improve the utility of earthworms that would benefit from the large feast, you might object that earthworms shouldn’t be included in measures of total welfare. Following your line of reasoning, I should respond that you’re confusing efficiency and morality–a world where earthworms are happy isn’t necessarily good, it’s just efficient.

    I think most would agree that that’s a silly argument. Efficiency, at its root, rests upon the assumption that we care about the desires of a particular group of entities–human beings. So there’s a moral component built into the idea of efficiency–if we don’t make a judgement about which entities we should care about (people, dead people, rats, single-cell organisms, rocks, planets, trees), efficiency is meaningless.

  • James Andrix

    I’ve come to think that we use the non-experiential nature of other people’s preferences as an excuse to discount them, even to the point of citing different levels of experience as significant.

    i.e. Some people are very concerned about what other consenting adults do in the privacy of their own bedrooms. I might say to them “What do you care? It doesn’t affect you at all.” I don’t much care to help them with their preferences, (even ignoring the required control of others) and I say their preferences are silly because they are non-experiential.

    The same people would be concerned about consenting adults displaying affection in public. I might now say “What do you care? It’s not hurting you.” I still don’t care about this preference, but I can’t deny it is experiential. I claim their preferences are silly because the experience doesn’t have some other arbitrary quality.

    On the other hand, if someone were to complain about something I also find morally objectionable, even though it doesn’t affect me, I would want to help with that preference. I would never cite it’s non-experiential nature are a reason not to care. (In fact some would claim this is a sign of noble altruism.)

  • Chris, it’s an unproven assumption that people won’t sit on their hands.

    anon, you are correct that it also an unproven assumption. Robin Hanson’s presentation at the Foresight Institute explains how the Luddites may have a point however.

    Robert Bloomfield:
    “say and feel they want (a “revealed preference” model)”
    I think revealed preference normally suggests “actions speak much louder than words”.

    “legislators argue that they know preferences better than individual agents”
    I don’t think the assumption that preferences are poorly defined and time-inconsistent is sufficient to reach that conclusion.

    There are other utilitarian arguments for the extermination of humanity and more.

    James Andrix:
    The enforcement costs for activities in the privacy of ones own home are high, and the people engaging in such activities normally have a much higher willingness-to-pay relative to those opposed to it. This efficiency is why David Friedman thinks anarchy would likely (though may not be) be libertarian. Using the “least-cost avoider” approach in Robin’s follow-up, we might argue that those offended by others behavior can more easily decide not to pay it much attention.

  • eddie

    Economic welfare cares not about giving people experiences but about satisfying their preferences. … If we do something a dead person would have wanted, that counts as a benefit.

    Close, but not quite. If while they were alive we did something a dead person wanted, that counts as a benefit. Doing something now after they are dead provides no benefit.

    Finding win-win deals between the living and the dead is of no use if you don’t find them until after they are dead. At that point it’s not win-win, it’s just win. And if there were a different deal that gave more to the living and less to the dead, then that’s just a bigger win.

    The dead do not have wants. The dead cannot participate in deals. The dead should play no part in our deal-finding framework, except to the extent that the living wish the same things the dead once wished.