Utilitarianism Debate with Michael Huemer

Matthew Adelstein kindly invited me & Michael Huemer to hash out our disagreements about utilitarianism over on his YouTube channel.  The resulting discussion was fun and wide-ranging.  In this post, I just want to highlight a couple of major themes that seemed fairly central to our dispute: (1) which intuitions we place the most weight on, and (2) the inferential role of wrongness.

Objections to Rule Consequentialism

Those put-off by the putative counterexamples to Act Consequentialism may consider Rule Consequentialism a more appealing alternative. Michael Huemer goes so far as to suggest that it is "not a crazy view." In this post, I'll explain why I think Rule Consequentialism is not well-supported -- and, at least as standardly formulated, may even be crazy.

There are three main motivations for Rule Consequentialism (RC).  One -- most common amongst non-specialists -- stems from the sense that it would be better (in practice) for people to be guided by generally-reliable rules than to attempt to explicitly calculate expected utilities on a case-by-case basis.  But of course this is no reason to prefer RC as a criterion of right; this consideration instead pulls one towards multi-level act utilitarianism (on which the right decision procedure is something other than constant calculation).

A better argument for RC (and the one that seems to motivate Huemer) is that it better systematizes our moral intuitions about cases.  But I think this is bad moral methodology -- matching superficial intuitions about cases is much less important than conforming to our deeper understanding of what really matters.  And RC is notoriously difficult to reconcile with the idea that promoting well-being (rather than blindly following rules) is what matters.

Utilitarianism and Reflective Equilibrium

In 'Why I Am Not a Utilitarian', Michael Huemer objects that "there are so many counter-examples, and the intuitions about these examples are strong and widespread, it’s hard to see how utilitarianism could be justified overall."  But I think it's actually much easier to bring utilitarianism (or something close to it) into reflective equilibrium with common sense intuitions than it would be for any competing deontological view.  That's because I think the clash between utilitarianism and intuition is shallow, whereas the intuitive problems with non-consequentialism are deep and irresolvable.

To fully make this case would probably require a book or three.  But let's see how far I can get sketching the rough case in a mere blog post.

Consequentialism's Central Concept

Ethical theories are typically formulated as centrally concerning the concept of right action.  Introductory ethics classes may define competing theories as offering different completions of the sentence: "An act is right iff...".  And that probably works well enough for deontological theories, which are centrally concerned with delineating the boundaries of permissibility and obligation.  But I think it's very misleading to treat consequentialist theories as seeking to answer this question.  (And I expect virtue ethicists would have similar complaints.)

If forced into a deontic mould, it's natural to default to maximizing consequentialism: An act is right iff it maximizes value (or, more precisely, produces no less value than any alternative option).  But the concept of rightness has connotations that fit poorly with consequentialism, as many (from Railton to Norcross) have pointed out.  For example:

Updates to utilitarianism.net

Back in July, I mentioned our new introduction to population ethics.  Since then, I've also added a chapter on Theories of Well-being, and -- brand new as of today -- Arguments for Utilitarianism.

I'm inclined to think the best case for utilitarianism stems from simply reflecting on what fundamentally matters (and one who doesn't find the utilitarian answer here intuitively compelling is unlikely to be much moved by any other argument in support of the view).  But I'm also pretty moved by the charge against non-consequentialist views that they are steeped in status quo bias, so I was pleased to be able to make that case here.  (I don't recall seeing the point discussed so much elsewhere -- it strikes me as unduly neglected.)

The other big news today is that we're kicking off a new series of Guest Essays with an excellent article by Jeff Sebo on 'Utilitarianism and Nonhuman Animals':

This essay advances three broad claims about utilitarianism and nonhuman animals. First, utilitarianism plausibly implies that all vertebrates and many invertebrates morally matter, but that some of these animals might matter more than others. Second, utilitarianism plausibly implies that we should attempt to both promote animal welfare and respect animal rights in practice. Third, utilitarianism plausibly implies that we should prioritize farmed and wild animals at present, and that we should work to support them in a variety of ways.

Enjoy!  (And maybe consider adding the relevant articles to your syllabi if you teach on any of these topics...)

Agency as a Force for Good

One fundamental reason for favouring consequentialism is the basic teleological intuition that the primary purpose of agency is to realize preferable outcomes.  If you have a choice between a better state of affairs and a worse one, it's very natural to think that the better state of affairs would be the better option to choose.

A slightly different way to put it is that if it would be good for something to happen, then it would be good to choose for it to happen.  Our agency is itself part of the natural world, after all, and while it is distinctive in being subject to moral evaluation -- misdirected exercises of agency may be wicked in a way that unfortunately directed lightning strikes are not -- it's far from clear why this should transform an otherwise desirable outcome into an undesirable one.  There's nothing obviously misdirected (let alone "wicked") about straightforwardly aiming at the good, after all.

Consequentialism thus fits with an appealing conception of agency as a force for good in the world. Left to its own devices, the world might just as easily drift into bad outcomes as good ones, but through our choices, we moral agents may deliberately steer it along better paths.

Discounting Illicit Benefits

In 'The Means and the Good' (Analysis, forthcoming) Matthew Oliver argues that pluralist consequentialists can accommodate intuitions against using others as a means, on the model of how they can accommodate intuitions about desert:

Just as it is bad for Emily to benefit from a stolen manuscript, it is bad for anyone to benefit from the use of another’s body or resources as a means. We can call this impersonal badness an impersonal-use-cost. As with a stolen manuscript, good results that are produced by using another person’s body or resources are heavily offset by an accompanying impersonal-use-cost.

By, in effect, discounting illicit benefits, we get the result that killing one to save five does more harm than good.  But we also get the result that killing one to prevent five others from each killing one to save five likewise does more harm than good.  (I think the most natural way to understand this is not to regard the second-order killing as in itself extra bad; the killing is just as intrinsically bad as any other death, the problem is instead that any good that would follow from it -- including the prevention of other wrongful killings -- gets massively discounted.)

It's a clever and interesting view!  But it seems really vulnerable to my argument against constraints, namely, that it unacceptably devalues the lives of the innocent victims who might be rescued.  Once an innocent person has been killed in an (even wrongful) attempt to save five, it really matters whether those five are ultimately saved or not!  So we shouldn't discount the value of their lives, no matter the illicit nature of the agent's act (however bad it may have been, that harm has already been done).  Otherwise, we would violate the moral datum that One Killing to Prevent Five >> Six Killings (Failed Prevention).

My reframing of the view in terms of "discounting illicit benefits" brings out the problem most starkly.  But I think it's just a verbal difference, and Oliver's original formulation in terms of an offsetting "use cost" (proportional to the illicit benefits) has the same implications.  Does that sound right?  Do correct me if I'm wrong...

The Cost of Constraints

[An excerpt from my paper-in-progress on the paradox of deontology, setting out my core argument...]

Let's begin with some cases.^[The first two are drawn from Setiya (2018).]  In all of these cases, the background setup involves five other agents who are each about to murder a different innocent victim. Protagonist may be in a position to prevent these five murders, by means of himself killing a sixth individual. (The precise details shouldn't matter for our purposes--feel free to fill in the story however seems most sensible to you.) Against this background, compare the following four possible outcomes:

Five Killings: Protagonist does nothing, so the five other murders proceed as expected.

One Killing to Prevent Five: Protagonist kills one as a means, thereby preventing the five other murders.

Six Killings (Failed Prevention): As above, Protagonist kills one as a means, but in this case fails to achieve his end of preventing the five other murders.  So all six victims are killed.

Six Killings: Instead of attempting to save the five, Protagonist decides to murder his victim for the sheer hell of it, just like the other five murderers.

Three Dogmas of Utilitarianism

I think that something very close to utilitarianism is the right moral theory, and most of the standard objections are bunk.  But here are what I take to be three genuine flaws in "orthodox" utilitarianism. (Two can be fixed from within utilitarianism.  One pushes us to accept a slightly different consequentialist view that is no longer strictly speaking utilitarian.)

Adams' Critique of Global Consequentialism

From p.479 of 'Motive Utilitarianism':

The moral point of view—the point of view from which moral judgments are made—cannot safely be defined as a point of view in which the test of utility is applied directly to all objects of moral evaluation. For it is doubtful that the most useful motives, and the most useful sort of conscience, are related to the most useful acts in the way that the motives, and especially the kind of conscience, regarded as right must be related to the acts regarded as right in anything that is to count as a morality. And therefore it is doubtful that direct application of the test of utility to everything results in a system that counts as a morality.

How must the right motives be related to the right acts?  Plausibly through a principle that links normative and motivating reasons: an agent acts from the "right reasons" when their motivating reasons are the very normative reasons that make the act worth doing.  Something along those lines.

Political Beliefs, Uncertainty, and the Expected Value of Paralysis

Jason Brennan argues that most people can't have the faintest clue about the expected value of voting for either candidate in an election:

They don't know the difference in the value between the candidates and they don't know they probably of being decisive. If it's not rational to buy a lottery ticket in such situations, why would it be rational to vote?

But I think his framing is misleading.  It's true that a robust, precise, and dialectically persuasive estimate would take a lot of work.  But it would also take a lot of work (much more than Brennan does here!) to show that most people have no good reason to think that one candidate would be better than the other, or that they epistemically must be indifferent.  Yet that is what Brennan really needs to show, if he is to undermine the rationality of voting.

Moral Theory and Motivational Contingency

Parfit sometimes suggested that Act Consequentialism might best be understood, not as a moral theory, but as an external rival to morality.  The implicit thought seems to be that morality involves some essentially social mode of thought (perhaps concerned with public codes, or norms of praise and blame, or what principles you'd want everyone to accept), AC does not depend upon any such mode of thought, but it nonetheless offers an account of what you really ought (rather than just "morally ought") to do.

I'm not a fan of that conception of morality, largely because it seems to devalue it, depriving morality of its interest and significance -- if it is not what you really ought to do, then who cares what you "morally ought" to do?  Putting that worry aside, though, it does seem interesting that (only?) consequentialists are apt to regard their normative commitments as not contingent upon being the true moral theory.  That is, if I learned that Kantianism was true, my response (besides incredulity) would be to say, "So much the worse for morality."  In that sense, I don't particularly care about morality, and don't think anyone else should either.  (You should lie to the murderer at the door to save lives, and I don't care about any sense of "wrong" in which doing more good is "wrong".)  But I suspect that, if the roles were reversed, Kantians would not react analogously.  (Can any Kantian readers confirm?)

If that's right, I wonder if anything interesting follows from this observation.  Might it suggest that consequentialism matters more than other theories?  Others seem to depend upon inheriting the normative halo of "morality", whereas consequentialist properties have such clear intrinsic significance that no further halo is required!  Ha, maybe.  More neutrally, it may just be that other theories are so closely tied to some essentially social mode of thought that it's harder to make sense of them mattering without also being what matters morally -- and such a connection needn't have any implications for how much they matter.

A related question (which I owe to David Enoch): how -- if at all -- would (and should) your motivations change upon learning that the meta-normative error theory is correct, and nothing objectively matters at all?  Insofar as you had proper de re desires for what you previously took to be good, it isn't clear why these would be contingent or conditional upon your normative (or meta-normative) beliefs.  Our concern for each individual's welfare, at least, presumably should not be conditional in such a way.  Perhaps deontological concerns -- e.g. to avoid violating rights even if it would overall help people more -- are more appropriately conditional on their actually being right, leaving deontological moral motivation more vulnerable to the threat of moral scepticism?

Another possible exception, even for consequentialists, might involve distant-future non-identity cases (about which it seems difficult to muster direct concern and motivation).  Consider some case where we can either help the already-existing global poor (with no expected downstream effects), or we can bring it about that a whole generation of people a million years hence will be much better off (than the different generation that would otherwise exist in their place).  For purely theoretical reasons, I currently prefer the greater distant good.  But if I came to believe in error theory, I would be much more tempted to just go with my gut, indulge my narrow sympathetic impulses, and help the already-existing people.  (Perhaps a more virtuous consequentialist would have a less theoretically-contingent commitment to the distant future!)

Any thoughts?  Which of your moral motivations do you regard as theoretically contingent in these ways?

Against Conventional Moral 'Decency'

"It’s obviously horrible to value the economy over human lives." - Regina Rini echoes a common moral conviction.  I am so deeply baffled by this.  As though "the economy" was nothing but a number on a spreadsheet, mere "corporate profits", disconnected from human lives and livelihoods. (Progressive-approved history textbooks will no doubt have to cut all mention of the Great Depression on grounds that it was no great tragedy, of interest only to fat cats and their cronies.)

The real, underlying moral question is instead how we should weigh specifically medical interests against human wellbeing more broadly.  If anything is "obviously horrible", to my way of thinking, it is to claim that nothing matters beyond sheer existence -- that happiness, mental and emotional wellbeing, quality of life, prosperity, and human connection, all count for naught in comparison to keeping our bare biological mechanisms running smoothly.  I think few people would really endorse such a claim on reflection, but it seems to be implied by the unthinking prioritization of "saving lives".

No Utility Cascades

Max Hayward has an interesting paper, 'Utility Cascades', forthcoming in Analysis.  We're told: "Utility Cascades occur when a utilitarian’s reduction of support for an intervention reduces the effectiveness of that intervention, leading the utilitarian to further reduce support [...] in a negative spiral." (p.1)  The basic puzzle Hayward sets up involves the following additional assumptions:

(1) Holding fixed their (practical) normative commitments, accurate/rational updating can (often!) tip utilitarians into a utility cascade.

(2) This "negative spiral" predictably makes things worse than if the utilitarian had stuck with their initial level of support, even given that the intervention is less effective than initially believed.

Putting these together, we obtain a surprising apparent tension between epistemic and practical normativity for utilitarians.  For, Hayward suggests, it would often promote predictably better results for utilitarians to bury their heads in the sand rather than rationally updating on new evidence of the sort that might trigger a utility cascade.

It's a fun argument.  But I don't see how (1) and (2) could both be true.  After all, if the act of reducing support for intervention X would predictably have worse results than maintaining initial levels of support, then utilitarianism straightforwardly requires the latter.

Why does Hayward believe otherwise?  He describes Bill, supposedly an act utilitarian / effective altruist agent, who "supports highly effective initiatives in proportion to their effectiveness score [expected value]." (bold added)  As a result, any drop in expected value (e.g. due to new evidence) necessarily reduces Bill's investment in an intervention.

But this is not an accurate representation of utilitarian (or effective altruist) normative commitments.  You should not allot half as much funding to a charity that's half as good as others.  You should give the suboptimal charity nothing, and instead send every dollar to the very best charity (until it is so saturated with funds that it no longer offers the best marginal return on your next dollar, which should then instead go to the next (now-best) intervention, and so on).

As a result, it's entirely possible that a slight reduction in its expected value makes no difference to how much Bill should fund X, so long as the (marginal) expected value of funding X to this degree still exceeds the expected value of shifting any of this funding to some alternative intervention Y. In such a case, assumption (1) above fails to hold. Otherwise, (2) is false: if shifting some funding to Y is really (expectably) for the best, despite reducing the effectiveness of the remaining X-funds (if any), then it's not true that it would've been (predictably) better for Bill to ignore the evidence and keep funding X at initial levels.  In neither case does epistemic rationality for the utilitarian agent make things (predictably) go worse.  Contra Hayward, in ordinary cases (i.e., barring an evil demon who punishes epistemic rationality, or the like), there are not "utilitarian reasons to adopt ostrich behaviour".

Update: Here's a simple way to bring out the incoherence in Hayward's argument.  Consider his central example of "effectanol" funding.  Draw up a table which lists the expected values of donating in the following ways: <$10k to effectanol, $0 elsewhere>, <$8k, $2k>, <$6k, $4k>, ... , <$0 to effectanol, $10k elsewhere>.  After all the evidence regarding its effectiveness has come to light, which line on the table has the highest expected value? To get the cascade going, he needs each subsequent row on the table to have higher expected value than those above it. But then we are to suppose that it would have been predictably "better" to stick with fully funding effectanol, implying that the first row has the highest expected value. There's no way of filling out the numbers so that both of those claims come out true.

Who's Responsible for Offset Harms?

Here's a fun puzzle (that I owe to Caspar Hare): Polluter is trying to work out how to dispose of her toxic waste barrel economically, when she sees her neighbor about to pour his waste barrel into the river.  Delighted, she interrupts her neighbor and pays him to find a more eco-friendly way to dispose of his waste.  Having offset this harm, Polluter now feels free to dump her own waste into the river.  The downstream farm is ruined.  Who is responsible?

Tempting answer: Polluter! She dumped waste, while her neighbor (Paid-Off) didn't.  Polluter clearly caused the harm, and is the only eligible agent to be held morally responsible.

I think this tempting answer is importantly mistaken.

A New Paradox of Deontology

[Update: see related discussion at PEA Soup, and a much improved version of the argument in my paper, 'Preference and Prevention: A New Paradox of Deontology', summarized on Good Thoughts.]

There's something odd about the view that it'd be wrong to kill one innocent even to prevent five other (comparable) killings.  Given plausible bridging principles, this implies that we should prefer Five Killings over Killing One to Prevent Five.  But that seems an odd preference: how can five killings be preferable to one?  The deontologist (like Setiya) must think that agency is playing a crucial role here.*  While we should prefer one gratuitous killing over five, there is (on this view) a special kind of killing -- killing as a means -- where the good results of the killing don't get to count. So Killing One to Prevent Five is treated as morally akin to Six Killings, rather than to One Killing.

This is odd enough, but I think it gets worse.  For compare some variations of the case.  First note that if the good results of the killing-as-a-means don't get to count, then it seems it shouldn't matter to our moral verdicts whether the intended good results actually eventuate or not.  So consider Killing One in a Failed Attempt to Prevent Five (KOFAPF).  Clearly, KOFAPF is much worse an outcome than Killing One to Prevent Five (KOPF): it has the same agential intervention, but with six killings instead of just one.  So we should strongly prefer KOPF over KOFAPF.  But then how can we coherently prefer Five Killings over KOPF?

When is Inefficacy Objectionable?

There's something I find puzzling about the dialectic on this issue.  Many philosophers suggest that there is an "inefficacy problem" or objection to consequentialism. But we need to take care to correctly diagnose what is supposed to be problematic.  If we truly are incapable of securing some good outcome, after all, it would hardly seem fair to fault a theory that (correctly) tells us that we needn't bother.  Our practical inefficacy per se cannot sensibly be held against a theory; it may just be a sad fact of life.

Really the issue here concerns a kind of mismatch between individual and collective verdicts that appears to result from collective action problems (voting, polluting, etc.) in which we combine (apparent) individual inefficacy with (apparent) collective efficacy.  But even here, care must be taken in our identification of the relevant group or 'collective'.  Suppose that everyone else is determined to bring about the collectively harmful outcome, and are certain to do so no matter what I do.  Then there's no point in delusional attempts to "cooperate" that are guaranteed to fall on deaf ears.  Permitting laxity in the face of such inefficacy is not a "problem", it's sensible -- the plainly correct verdict in this case.  The fault here clearly lies with the bad actors, not with our moral theory.


So we further need to specify that we're concerned with collective harms that could result from those who are successfully following the target moral theory.  (This clarifies why the previous scenario was not objectionable: the one follower of consequentialism did not, as a "group" of one, actually do any harm.)  More generally, the key structural feature is to generate a kind of "each/we dilemma" in which each person acts rightly in bringing about a situation that they collectively abhor.  The agents' shared moral theory would then be a failure by its own lights, in a tolerably clear and important sense: it would be (as Parfit showed common-sense morality to be) collectively self-defeating.

Curiously, the recent literature on the inefficacy objection largely focuses on arguments which, even if successful (in establishing inefficacy), would not establish collective self-defeat.  The strongest arguments for thinking that individual consequentialists shouldn't bother Φ-ing are, I think, equally reasons for thinking that consequentialists collectively shouldn't Φ.  So there is then no real mismatch between the individual and collective moral verdicts.

Consider the arguments mentioned (from Nefsky's recent survey piece) in my previous post:

Does Welfare have Diminishing Marginal Value?

Many people are drawn to the Prioritarian view that "Benefiting people matters more the worse off these people are." (Parfit 1997, 213)  Importantly, this is not just the (utilitarian-compatible) idea that many goods have diminishing marginal value, so that better-off people are likely to benefit less than worse-off people from a certain amount of material goods.  Even after accounting for all that, the idea goes, the interests of the worse-off just matter more; we should even give a lesser benefit to the worst off rather than a (genuinely) slightly greater benefit to someone who is already well-off.

This view always struck me as deeply misguided. By effectively attributing diminishing marginal value to welfare itself, you can end up implying that even considering just a single individual, it might be "better" to do what is worse for him (e.g. giving him a smaller benefit at a time when his quality of life is lower, rather than a greater benefit at a different time).

But there is a closely related view that is more theoretically cogent.  Rather than attributing diminishing marginal value to welfare itself, you attribute it to the basic goods that contribute to one's welfare (happiness, etc.).  This goes beyond the familiar idea that material resources have diminishing instrumental value, say for making you happy.  We are now introducing a kind of non-instrumental diminishing value, by saying that happiness itself makes more of a difference to your welfare the less of it you have.

The Aim(s) of Practical Deliberation

My new paper on 'Deontic Pluralism' argues for "a maximizing account of the ought of most reason, a satisficing account of obligation, and a scalar account of the weight of reasons."  One question that emerges towards the end of the paper is whether we really need all of this.  Can we identify a sense of 'ought' that has primacy in virtue of its special relevance to first-personal deliberation—i.e., as the sense of ‘ought’ that a conscientious agent has in mind when they ask themselves, “What ought I to do?”

I've previously cast doubt on the idea that the deliberative question has a suitably fixed and determinate meaning. But even just focusing on the choice between the ought of most reason and the ought of minimal decency (or blamelessness), we aren’t obviously forced in either direction here, e.g. by the constitutive norms of agential deliberation. Some agents in some contexts are particularly concerned to at least meet the standards for minimal decency, whereas others are more morally ambitious. We can certainly say that it’s better for agents to do better. But it isn’t clear that there’s much more we can say beyond this trivial evaluative observation. In particular, I see no clear basis for insisting that there is just one proper aim of deliberation.

Deontic Pluralism [draft paper]

Abstract: Consequentialist views have traditionally taken a maximizing form, requiring agents to bring about the very best outcome that they can.  But this maximizing function may be questioned.  Satisficing views instead allow agents to bring about any outcome that exceeds a satisfactory threshold or qualifies as "good enough".  Scalar consequentialism, by contrast, eschews moral requirements altogether, instead evaluating acts in purely comparative terms, i.e., as better or worse than their alternatives.  After surveying the main considerations for and against each of these three views, I argue that the core insights of each are (despite appearances) not in conflict. Consequentialists should be deontic pluralists and accept a maximizing account of the ought of most reason, a satisficing account of obligation, and a scalar account of the weight of reasons.

* * *