Economists, Why So Consequentialist?

By On February 27, 2012 · 28 Comments · In Academic Philosophy, Consequentialism, Economics

Economists are frequently (a) subjectivists about value and (b) consequentialists.  (a) and (b) are in tension.

Brief dialectical summary:

Non-Consequentialist Philosopher: What should we do?
Typical Economist: Maximize utility!
Philosopher: What should you do?
Economist: Pursue what I value.
Philosopher: What about when pursuing what you value doesn’t maximize utility?
Economist: With good institutions it doesn’t happen that much.
Philosopher: Yeah but why should you support good institutions rather than pursuing other values?
Economist: I like good institutions!
Philosopher: But what about when what you value and maintaining good institutions conflict?
Economist: It doesn’t happen that much.
Philosopher: But what if it happened? What would you have most reason to do?
Economist: It doesn’t happen that much.
Philosopher: <frustrated> But what if it did?!
Economist: <frustrated> But it won’t!

<bar fight>

Extended version:

Philosophers typically say consequentialism is an agent-neutral moral theory. Following Doug Portmore (who denies this view), I will say that a theory is agent-neutral if it gives every agent the exact same set of aims and agent-relative otherwise. Consequentialist theories tell us that each person always has reason to maximize good and minimize bad. That is, the reasons for action it identifies are agent-neutral.

And all reasons for action are ultimately based on these. So if consequentialism says that you should follow a rule that maximizes good and the rule permits you to make honeybadger jokes because you like honeybadger jokes and (assuming a subjective theory of value) honeybadger jokes are good for you, then your reason to make honeybadger jokes, while relative to you, is only a reason because it is in accord with a rule that maximizes good.

Ok, so if you’re a consequentialist, then you should believe that reasons for action are fundamentally agent-neutral. So, most economists should believe that reasons for action are fundamentally agent-neutral.

However, economists are subjectivists about value. They think honeybadger jokes are valuable because they are valued by people. There is no such thing as “intrinsic” value because all value derives from valuers and so the value of anything could be removed if valuers valued them differently (oof).

Now it’s probably a conceptual truth that you have reason to pursue what is of value. So if you’re a subjectivist about value, then you think people have reason to pursue what they value. And of course people attach different values to different things, and so each person has reason to pursue different stuff. I have reason to make honeybadger jokes (though my fellow Alabamians have pretty much ruined them for me), but you have reason to indignantly maintain that LoLcat jokes are funnier.

Subjective theories of value (like the standard interpretation of utility theory) suggest that our fundamental reasons are agent-relative. It might be that all people share some reasons, but that’s just the contingent result of us all valuing the same thing in the same way.

So here’s the problem for economists:

(1) If consequentialism is true, then agent-neutral reasons are fundamental.
(2) If the subjective theory of value is true, then agent-neutral reasons are not fundamental.

So it is false that:

(3) Consequentialism & Subjective Theory of Value

I say ditch consequentialism.

Philosophers: If you like, I can discuss attempts to reconcile the first and third personal perspectives in another blog post.

UPDATE: After reading two responses to this post, I realized that I should emphasize that I’m talking about how economists employ the assumptions of their field in their more popular writings and in conversation. I do not think that economics as such requires specific metaethical assumptions.

 
  • Rebrebrebreb

    think of it like  a complex function consequentialism is a function of  subjectivism C(S(x)) and then they’re no longer in conflict, but an integral part of  determining value

  • Brad Potts

    I don’t see the conflict.  Consequentialists (correct me if I’m wrong about this) would say that ”good”, “utility”, and “satisfaction” have agent-neutral definitions, but agent-relative conditions.  That seems to mesh with subjective value.

    • Kevin Vallier

      But the fundamental reasons for action are agent-neutral, even if they have subjectivist content. So the reasons say, “Maximize subjectivist satisfaction.” But the subjective theory of value suggests that all of our fundamental reasons for action must derive from our own point of view, that is, our reasons begin with our present beliefs and affirmations. If that is so, how could anyone have a generic reason to maximize subjectivist satisfaction unless she valued such maximization on an agent-relative basis (which almost no one does)?

  • Jessica Flanigan

    I don’t think that this is a problem unique to consequentialism, nor is it necessarly a problem for any moral theory. 

    Here is why the problem seems to come up everywhere. We can imagine a Kantian or a contractualist, who thinks that everyone has the same moral reasons to obey the formula of humanity or not act on principles that are reasonably rejectable or whatever. Still, her on the ground reasons for action on these theories will in some ways be attitude dependent, because they will depend on what people actually do value or actually would have reason reject given their own values.

    Now say you are a preference utilitarian meaning that you think that we all have reason to maximize preference satisfaction (good) and minimize dissatisfaction (bad!) Your reasons will still in some sense depend on what people actually prefer. In this sense a preference utilitarian will affirm at once that reasons are the same for everyone and also that reasons will depend on what people actually value, but so far I don’t see why there’s any inconsistency here, or at least no more than there is in other moral theories.

    • Kevin Vallier

      The problem arises when we ask which reasons are *fundamental*. Sure, everyone thinks some of our reasons are attitude-dependent. But that’s not what I’m after. Take preference utilitarianism. Its an account of agent-neutral reasons even if those reasons involve maximizing subjective satisfaction. So it says, “Everyone has a reason to maximize subjective desire satisfaction.” But that’s an agent-neutral reason with a subjective content. But the subjective theory of value suggests that all reasons for action are fundamentally agent-relative. So in that sense it is incompatible with preference utilitarianism.

  • gcallah

    Man, how you expect me to get through your post when right in the middle I have to go spend an hour at LOLcats?!

  • William Schudlich

    I don’t see
    the conflict.  “Given” the agent-relative
    aims, we can then use agent-neutral consequentialist theories on that agent.  But these would not transfer so well when aggregating agents because the aims are individually subjective. 
    If anything, I think you’ve hit on why it is so hard (impossible?) to
    build formal macro-economic theories.

    Consequentialism : microeconomics :: Subjectivism : macroeconomics

  • Pingback: Are Economic Value Subjectivism and Consequentialism Inconsistent?

  • http://www.facebook.com/people/Rick-DiMare/100000504645309 Rick DiMare

    This is from footnote #44 of Portmore’s paper. Not sure, but it sounds like the hybrid is what we need:

    “This preoccupation with just deontology and utilitarianism has left the prospects of a promising hybrid (viz., agent-relative consequentialism) completely unexplored.”

  • Roger Koppl

    Let me see if I’ve got the argument.  Consequentialists say that we want “good” outcomes.    Subjectivists say “good” varies from person to person.  If “good” is different for each person, then “good outcomes” doesn’t mean anything.  If, instead, the phrase “good outcomes” has content, then it’s gotta be the same for everyone.  Thus, you say, we can’t be both subjectivist and consequentialist.  Have I got it?

    I don’t see where there is a real problem here.  When people espouse subjectivism of this sort, they are usually talking about the methodological principle that social science must generally take preferences and human nature as a given.  What people *imagine* to be good varies from person to person.  In thinking about what we should do, however, I may nevertheless have a definition of “good” that does not vary from person to person.  I want people to “thrive” or “have pleasure” or whatever even if I know that people do what they damn well please even. Sometimes they do what they want even when it ruins their lives or kills them.  As a social scientist I generally take preferences as given, even foolish preferences.  But as a moral actor, I may try to show you that those same preferences are bad and encourage you to mend your ways.   Where is the inconsistency in all of this?

    A few other points seem relevant.  The uniform “good” of my consequentialist ethics might be pretty high-level abstract stuff.  It’s “thriving” or “pleasure” or “self realization” or whatever.  The preferences of choosing individuals are necessarily more specific.  Yes, I want to “thrive” or whatever.  But I never choose “thriving” over “not thriving.”   I choose reading a book over watching Oprah, exercise over chocolate cake, shaking your hand over shunning you.  Those individual choices depend on details of context not imagined in my ethical theory.  No ethical theory can pack in all relevant context or anticipate all contingencies.  Thus, my consequentialist principles may tell me not to second-guess the choices other people make.  There’s a “knowledge problem” with choosing for others.  Thus there is a presumption in favor of autonomy, of individual choice, of experiments in living.  Our ignore of place and circumstance limits the ambitions of our moral judgments.

  • http://www.facebook.com/profile.php?id=743482150 David Sobel

    There is a conflict if we assume that morality necessarily provides reasons to comply with it. That seems a premise that the subjectivist should not accept (unless you get all fancy Schroedery about things, which I do not recommend). So the folks who combine these tend to deny that morality is necessarily reason providing.

    • Kevin Vallier

      Thanks for bringing that assumption out. Three things:

      (i) Who among “the folks” are the best to read on this subject?

      (ii) My bet is that economists would be naturally allergic to affirming consequentialism and admitting that it is not necessarily reason-giving. Given the often “non-ideal” and “feasibility” orientation of most economists, they’d round down their moral theory to provide us with reasons to choose between two highly feasible states of affairs.

      (iii) I think you’d have to do more than deny that morality is *necessarily* reason-giving to undermine the conflict. Wouldn’t you need to claim that in at least a non-trivial number of relevant cases of apparent conflict, morality is not reason-giving? I understand denying a necessary connection, but more seems required.

  • Roger Koppl

    David,

    You said, “That seems a premise that the subjectivist should not accept.”  I don’t think I see how that follows.  I’m not sure what you mean by “provides reasons.”  Presumably you mean that is I say X is “moral,” that’s a reason to do X.  And if I say “Y is immoral,” that’s a reason not to do Y.  Is that it?  If so, then I don’t think your inference works.  Murder is immoral, which a good reason to murder anyone.  The fact remains, however, that there are murders out there and social theory should not assume them away or deny that some people have murderous preferences.  

  • Roger Koppl

    good reason NOT to murder anyone.

  • Roger Koppl

    Kevin: Doesn’t your update create a vague target?  Economics is fine and dandy, but those economists sure start talkin’ nonsense when they hang out at the water cooler.  I don’t see how it’s a thesis that can be engaged critically.  

  • Julian Sanchez

    As another commenter suggests, this purported conflict doesn’t really have any special application to *consequentialist* ethics. It’s really just a restatement of moral internalism in an economics argot, and that particular metaethical dispute doesn’t much track the boundaries between first-order ethical theories.  An internalist is going to have to allow that WHICHEVER substantive moral view she favors is contingent on people’s holding an appropriately congenial set of subjective motivations and dispositions. An externalist of any stripe can accept that subjective preferences supply almost all of the content of the external reasons people have.  Parts 1 (vol. I) and 6 (vol. II) of Parfit’s “On What Matters” are instructive (if not completely convincing) on this stuff, FWIW.

    • Kevin Vallier

      My aim was to illustrate a tension between consequentialism and plausible implications of the subjective theory of value. I did that by illustrating a tension between what is widely taken to be a necessary condition for the truth of consequentialism (the claim that our reasons for action are fundamentally agent-neutral) and the subjective theory of value (namely, its apparent internalist implications). So what’s wrong with that strategy? I attacked a view by denying a necessary condition for the view. 

      I’m actually quite happy to extend the tension I discuss to cover all moral theories that depend on an agent-neutral conception of reasons. I was hoping to draw economists’ attention to the implausibility of an agent-neutral conception of reasons in part because I want to go after self-ownership and natural rights views as well. I think the common understanding of these views depend on an implicit agent-neutral conception of reasons. 

      • http://www.facebook.com/profile.php?id=743482150 David Sobel

        I am thinking it might be worthwhile to distinguish a subjectivist theory of well-being from a subjectivist theory of reasons. A subjectivist theory of well-being might claim that everything that is of value is of value to someone and its being valuable for them depends on its being liked or desired. This is compatible with the thought that we all have reasons to promote the aggregate of collective welfare. A subjective theory of reasons would say that X only has an intrinsic reason to do something if X had an intrinsic desire or concern for that something. The latter view could imply that we fail to have reason to obey just about any conception of morality–including say Kantian or self-ownership views as just about any sensible picture of morality says I may not always do whatever my weighted desires point me towards. The Kantian picture is agent-relative in that what I have reason to do differs from what you have reason to do, but it does not cut down my moral obligations to neatly fit my desires.

      • Julian Sanchez

         I think you’re putting a lot more weight on this idea of a certain type of reason “being fundamental” (and probably also on the agent-neutral/agent-relative distinction) than it will bear. As Sobel suggests, the average economist who is a subjectivist about WELFARE probably has no particularly deep commitments re: internalism or externalism about reasons. One could believe (like Parfit) that our desires and preferences determine what will maximize each individual’s utility, but do not independently provide reasons. A Humean about reasons, on the other hand, could believe that agent-neutral reasoning is distinctly constitutive of the moral point of view, while acknowledging that it’s a contingent matter whether people are motivated to (and have a reason to) behave morally.

        In neither case is it actually a precondition of the view that one or another type of reason be “fundamental.”  A value subjectivist can hold that agent-neutral, external reasons are fundamental, but contain a “gap” filled in by subjective preferences. A consequentialist can hold that agent-relative reasons provided by desires are fundamental, and hope that people are socialized with a desire to adopt the agent-neutral perspective (even if it takes some doing to show why this desire is actually entailed by their other desires and motivations).

  • geoih

    “Typical Economist: Maximize utility!”
    How do you maximize something that is impossible to objectively quantify? The flaw is in the economists who have deluded themselves into thinking they can.

  • Haytham Yaghi

    Good points! But on the flip side, libertarians tend to appropriate economist’s rhetoric when it serves the libertarian cause while dismissing other economic advice (probably made by the same economists quoted earlier) when it contradicts libertarian principles. Again, without painting with a broad brush, or making caricatures of ideas, a libertarian would be eager to say: 

    Libertarian: “Large government is inefficient, as pointed out by economists A, B, C, D and so on”. 
    Non-libertarian: “Fair enough, but the wide majority of economists, including A, B, C and D, agree that a limited form of government providing some services and a social net, funded by some progressive taxation can be made efficient and betters off society”.
    Libertarian: “Yeh but taxation is immoral and the betterment of society should not be made a priority over my aspirations.”In this example, I find that the use of economic arguments by Libertarian to be selective. It would be more consistent if he didn’t use any economic arguments in the first place.

    • geoih

      You don’t need economics to support libertarianism. Either you agree that every individual controls themselves, or you don’t. Everything flows from this premise.

      • http://www.sandiego.edu/~mzwolinski Matt Zwolinski

        Phase 1: Everyone controls themselves
        Phase 2: ?
        Phase 3: Absolute libertarian property rights!

        • geoih

          Phase 1: Everyone controls themselves.
          Phase 2: Humans interact voluntarily in order to further their self-interest.
          Phase 3: Property rights culture occurs spontaneously.

        • JCLester

          Phase 0: Assume that interpersonal liberty as non-invasiveness (absence/minimization of proactive impositions) is applied in a pre-propertarian, state-of-nature, thought-experiment.

          Phase 1: Everyone controls (de facto owns) themselves (as this minimizes proactive impositions).

          Phase 2: Everyone controls/owns whatever else he can acquire without proactively imposing on others.

          Phase 3: Strong prima facie libertarian property rights in persons and several
          property (where acquired as in 2). But in the event of disputed or problem cases apply the pre-propertarian libertarian theory.

  • Conchis Ness

    FWIW, for an economist following the party line on this stuff*, I think the dialog would actually go like this:

    Non-Consequentialist Philosopher: What should we do?

    Typical Economist: Achive Pareto optimality!

    Philosopher: What should you do?

    Economist: Pursue what I value.

    Philosopher: What about when pursuing what you value doesn’t
    achieve Pareto optimality?

    Economist: Er… I don’t think you understand the concept of Pareto optimality.

    *The party line in economics typically denies the possibility making the sorts of interpersonal utility comparisons necessary for ‘maximise utility’ to make any sense. (Individual utility functions are taken to be merely representations of a consistent preference ordering, and are therefore unique only up to an affine transform.) This why economists have tended to focus on Pareto optimality instead. Interestingly, while Pareto optimality is (as far as I can tell) agent-neutral, achieving it never requires anyone to do anything that isn’t in their own interests. 

    Of course, many economists (IMHO rightly) don’t follow the party line on this, but I’m not sure those ones are quite so wedded to the version of the subjective theory of value you’re trying to foist on them.

  • http://blog.monstuff.com Julien Couvreur

    See Rothbard’s “Towards a reconstruction of utility and welfare economics” which tries to reconstruct welfare economics on a firm basis of subjective value:
    http://mises.org/daily/2205 

  • http://www.facebook.com/people/Andrew-Hosek/1294479596 Andrew Hosek

    “Philosopher: What about when pursuing what you value doesn’t maximize utility?”

    All values are subjective. This presumes utility may be objectively measured.