Many people don’t like utilitarianism. They advocate alternative (often deontological) approaches to ethics. In 2020, we saw the immense costs of some of those misguided ethical systems.
Scott Aaronson has an excellent post that begins with a discussion of why he believes our response to Covid was inexcusably slow. He discusses challenge trials of vaccines, and also a WWII-style plan to build manufacturing capacity just in case the vaccines were successful. But he also considers possible objections to his arguments, such as the fact that moving faster imposes risks:
Let me now respond to three counterarguments that would surely come up in the comments if I didn’t address them.
1. The Argument from Actual Risk. Every time this subject arises, someone patiently explains to me that, since a vaccine gets administered to billions of healthy people, the standards for its safety and efficacy need to be even higher than they are for ordinary medicines. Of course that’s true, and it strikes me as an excellent reason not to inject people with a completely untested vaccine! All I ask is that the people who are, or could be, harmed by a faulty vaccine, be weighed on the same moral scale as the people harmed by covid itself. As an example, we know that the Phase III clinical trials were repeatedly halted for days or weeks because of a single participant developing strange symptoms—often a participant who’d received the placebo rather than the actual vaccine! That person matters. Any future vaccine recipient who might develop similar symptoms matters. But the 10,000 people who die of covid every single day we delay, along with the hundreds of millions more impoverished, kept out of school, etc., matter equally. If we threw them all onto the same utilitarian scale, would we be making the same tradeoffs that we are now? I feel like the question answers itself.
And it’s not just vaccine development; we’ve also prioritized “ethics” over saving lives in the distribution of the vaccine:
Update (Jan. 1, 2021): If you want a sense of the on-the-ground realities of administering the vaccine in the US, check out this long post by Zvi Mowshowitz. Briefly, it looks like in my post, I gave those in charge way too much benefit of the doubt (!!). The Trump administration pledged to administer 20 million vaccines by the end of 2020; instead it administered fewer than 3 million. Crucially, this is not because of any problem with manufacturing or supply, but just because of pure bureaucratic blank-facedness. Incredibly, even as the pandemic rages, most of the vaccines are sitting in storage, at severe risk of spoiling … and officials’ primary concern is not to administer the precious doses, but just to make sure no one gets a dose “out of turn.” In contrast to Israel, where they’re now administering vaccines 24/7, including on Shabbat, with the goal being to get through the entire population as quickly as possible, in the US they’re moving at a snail’s pace and took off for the holidays. In Wisconsin, a pharmacist intentionally spoiled hundreds of doses; in West Virginia, they mistakenly gave antibody treatments instead of vaccines. There are no longer any terms to understand what’s happening other than those of black comedy.
Everyone is entitled to choose their own preferred ethical system as a guide to their daily life. But there is only one reliable ethical system to be used in public policy—maximizing aggregate utility. As soon as you ignore that goal, you end up killing lots of people for no good reason.
In retrospect, none of this should have been a surprise (although I admit to being caught off guard.) I had assumed that our disgraceful policy of banning kidney markets was a one-off exception. Now I see that the same instinct that leads to tens of thousands of excess deaths of people with kidneys disease also pervades our entire public health system.
Aaronson understands that this failure goes well beyond one individual or even one country; it’s a broader failure of society:
Furthermore, I could easily believe that there’s no one agent—neither Pfizer nor BioNTech nor Moderna, neither the CDC nor FDA nor other health or regulatory agencies, neither Bill Gates nor Moncef Slaoui—who could’ve unilaterally sped things up very much. If one of them tried, they would’ve simply been ostracized by the other parts of the system, and they probably all understood that. It might have taken a whole different civilization, with different attitudes about utility and risk.
At the same time, I do believe that utilitarianism is gradually gaining ground. But there’s still much more work to be done.
HT: Matt Yglesias
READER COMMENTS
Airman Spry Shark
Jan 2 2021 at 2:38pm
FWIW, I subscribe to rule utilitarianism (act utilitarianism seeming too brittle due to complexity), and find deontological formulations a convincing approach to identifying utility-maximizing rules.
Scott Sumner
Jan 2 2021 at 4:42pm
I’m also a rules utilitarian. And yes, freedom is often a good rule of thumb.
Russ Abbott
Jan 2 2021 at 2:47pm
No doubt utilitarianism provides a good starting point for public policy. But I would like to see you discuss its limitations. For example, markets in kidneys may lead to offers made to desperately poor people that those people feel obliged to take, e.g., for the sake of their families. Are you ok with that? What about a person selling their second kidney, again for the sake of their family’s survival?
These are just top-of-the-head examples. I’m sure a professional ethicist could come up with even more difficult ones.
How do you draw the line that defines where utilitarianism should and should not be applied?
Jon Murphy
Jan 2 2021 at 3:16pm
It’s not clear why you think the kidney example is 1) unethical and 2) a difficult moral dilemma. Is it really more ethical to say that the person’s family must suffer?
KevinDC
Jan 2 2021 at 3:24pm
This is a fair question. But the corollary question is also fair. Forbidding markets in kidneys will prevent poor people who might otherwise gain the ability to financially support their families from being able to do so, and will keep sick people from becoming healthy and cause many of them to needlessly die. Are you okay with that? The two sets of questions are different sides of the same coin. Both have unfortunate aspects to them. But if I have to pick which scenario is better, it is hands down the situation where the sale is allowed to take place.
Mike Munger has spoken on the sort of thing you’re talking about – you envision a poor person thinking of doing something drastic like selling a kidney to support their family, and it’s upsetting. You think “They shouldn’t have to do this.” But it’s a non sequitur to go from “They shouldn’t have to do this” to “they shouldn’t be allowed to do this.” If a person is in a bad situation with one desperate option available to them, its an odd way to express concern for them by taking that option away. And the person dying from renal disease deserves consideration too.
Scott Sumner
Jan 2 2021 at 4:45pm
You could have a minimum income level for donors. And you could forbid selling second kidneys. And yet even that system would be 100 times better than what we have now. We simply are not thinking clearly about these issues.
In general, I am comfortable with any public policy that makes the world a happier place.
Henri Hein
Jan 2 2021 at 8:05pm
What about the Inquisition? It probably made medieval Spain a happier place. Certainly it made the inquisitors a lot more happy, and the non-heretic population also gained a benefit from the process. With the relatively small amount of victims in a given year, and the seriousness of people’s faith at the time, it’s possible the sum total utility was positive.
Given you said “in general,” it may not be fair to grab an example from history that may be an aberration (though I would argue history is full of those). But the question does arise, if we are faced with such examples, how to know if a particular policy should be evaluated on general principles, or falls in one of the exceptions? Especially since the current pandemic is also an aberration.
Philo
Jan 2 2021 at 11:21pm
“What about the Inquisition?” That can’t be anything but trolling; shame on you!
Henri Hein
Jan 4 2021 at 3:43pm
Faced with the Inquisition, Utilitarianism faces two problems. The first is computational: how do you evaluate the pain and misery of burning one heretic against the aggregate small satisfaction that 3 million faithful Catholics get from it? You cannot.
The second relates to justice: if you somehow could calculate and compare the utility to the two groups, it is at least theoretically possible Utilitarianism ends up justifying the process.
If I am wrong and it is clear-cut that Utilitarianism always produces the just answer, then it should be possible to explain why that is. Perhaps the principle is that we should never burn anyone alive, anywhere, for any reason. That sounds like a good principle. But then we have stepped away from Utilitarianism and into some other territory.
Scott Sumner
Jan 3 2021 at 1:14pm
You asked:
“What about the Inquisition?”
It made Spain a far more miserable place. I don’t even view the issue as debatable. And there were lots of victims, not a “small number”. Indeed I oppose the inquisition precisely for utilitarian reasons.
Put simply, don’t be cruel.
Henri Hein
Jan 3 2021 at 2:29pm
The Inquisition tribunals processed a few dozen heretics at a time. The biggest processions had 100-200 victims, and they were rare. Out of a population of millions, I would call that a relatively small number.
Given how many spectators appeared at the processions and how far they traveled to witness one, the general population must have received positive utility from it. Recall also that many of them believed their souls would be in danger if they did not persecute the heretics.
You may be correct that a Utilitarian calculus still comes out against the Inquisition, but I don’t think there is anything obvious about it.
Jon Murphy
Jan 3 2021 at 3:32pm
Henri, why are we assuming that only the people prosecuted were made unhappy by the Inquisition?
Henri Hein
Jan 3 2021 at 6:35pm
I don’t think we have to assume that. Lots of jews and marranos (converted jews) tried to immigrate in order to escape, so they must have been disturbed. Jews and marranos still represented a small minority (maybe 10%-15% at most).
Established Catholics had much less to fear. We don’t have surveys from that time, but the Catholic Church continued to enjoy widespread support and involvement. People participated in the proceedings. They reported on their marrano friends, unprovoked.
The visitors to the public proceedings is a pretty strong revealed preference. They could number in the 100s of thousands, a large number even by modern standards. People sometimes traveled multiple days to be there. That is a considerable investment of opportunity cost at any time, let alone in a medieval economy.
When an Italian heretic, De Seso, walked to his death in 1568, King Phillip II. were watching. As De Seso passed by, he called out to ask how the king could let the friars treat a gentleman in such a manner. The king replied: “I myself would bring the [censored] to burn my own son were he as perverse as you.”
Every indication we have is that this sentiment was prevalent in the traditional Catholic populace.
Jon Murphy
Jan 4 2021 at 11:26am
Perhaps, but it doesn’t logically follow they were made happy by such sentiments. Indeed, the sentiments expressed by the King you cite suggests to me they were upset by the proceedings. Viewed as necessary, perhaps, but upsetting nonetheless.
Henri Hein
Jan 4 2021 at 3:46pm
I agree they were upset by the presence of heretics. Given heretics, they preferred to have proceedings against them.
Thomas Sewell
Jan 3 2021 at 12:10am
Yes, I’m perfectly fine with both of those being legal. We’re not that person’s master and they aren’t our slave, for us to tell them what to do with their body, even if we do it under the color of law and via voting for someone to rule over them.
I’m also not going to object to you (or anyone else) just giving them the same money if they agree not to sell their kidney because you think it’s a poor decision. So no one is stopping you, should you choose to make them that offer.
Why do you think your preferences for someone’s kidney should overrule their own decision-making?
If you (or others, singly or collectively) are so concerned about their family’s survival, then presumably there are other actions you can take to resolve that concern other than forbidding them from solving it if they wish.
KevinDC
Jan 2 2021 at 2:59pm
I’m among those who don’t hold utilitarianism in high regard, but I’m not a fan of deontology either. Hard deontologists would argue that consequences don’t matter, which I consider absurd. The idea that “consequences matter” seems intuitively obvious. Consequentialism, on the other hand, is the stronger claim that “only consequences matter,” which is necessarily less likely to be true, and also seems much less plausible. Utilitarianism is a subset of consequentialism, which says “one particular kind of consequence is the only thing that matters,” which seems even more implausible. And even within utilitarianism, there are disagreements about whether we should be trying to maximize total utility or average utility or whatnot, which further narrows the claim to “one particular kind of aggregate measurement of one particular kind of consequence is the only thing that matters.” This is a coherent idea, in the sense that it’s not internally contradictory, but it seems extremely implausible on the face of it. It would need really strong arguments to overcome the initial implausibility, but the various defenses of utilitarianism I’ve read fall wildly short of that.
(Scott Alexander, in a rather frank assessment, said of his commitment to utilitarianism that “It seems to boil down to something like this: I am only willing to accept utilitarianism when it matches my moral intuitions, or when I can hack it to conform to my moral intuitions. It usually does a good job of this, but sometimes it doesn’t, in which case I go with my moral intuitions over utilitarianism.” Me too, along with basically every other moral theory. Personally, I find Michael Huemer’s Ethical Intuitionism to pretty solid, in large part because it lacks a moral theory, but it’s been years since I’ve read it. Maybe I should give it a re-read.)
My whingeing about utilitarianism aside, I do agree with Scott that public policy should be far more in tune to consequences than it currently is. Too many people react to the idea of viewing public policies through the lens of cost-benefit analysis with abject horror. This is especially true of kidney sales.
Also, Israel’s approach to vaccination has been amazing, and it demonstrates a major advantage of the decentralized process vs the central planning approach. In Israel, they were just given the vaccine and a mandate – get the population vaccinated. Period. Work out for yourself how to do that best, make whatever decisions are appropriate on the ground from where you stand. I read a news story just today about a clinic in Israel that had some doses of vaccine that needed to be used, glanced outside, saw a pizza delivery guy, and yelled out the window to him asking if he wanted to be Covid vaccinated. And he was, right on the spot. Meanwhile, here in the US, we have central committees of people in authority debating with each other over the “right” way to administer the vaccine, to whom, in what order, taking time and care to make sure that everything is administered According To Plan, for The Plan has been determined as the One Correct Way Things Can Be Allowed For Everyone Everywhere. If anyone finds the difference in outcomes between these two approaches surprising, they really shouldn’t.
Scott Sumner
Jan 2 2021 at 4:48pm
Thanks for that anecdote about Israel.
I have yet to see a public policy that makes the world a less happy place that I favor.
KevinDC
Jan 2 2021 at 6:29pm
Here’s one plausible example that I’ll outsource to Scott Alexander, because I’m feeling lazy this Saturday evening:
Allowing this antidepressant plausibly would make the world a less happy place. Even if an angel of the lord descended from on high, and informed me with absolute assurance that the utilitarian calculus was applied correctly and allowing the antidepressant would make the world a less happy place on net, that wouldn’t change my mind that it should still be allowed anyway.
Scott Sumner
Jan 3 2021 at 1:16pm
Of course I do believe that drug bans make the world less happy, but if they made the world happier I’d be fine with the idea.
Brian
Jan 3 2021 at 10:31pm
Are you sure this story suggests utilitarianism should not be applied here?
It sounds appalling to suggest that addicts are acting rationally but perhaps they are. If you are not addicted, how can you appreciate the situation of the addicted?
robc
Jan 4 2021 at 7:49am
It is absurd because it isn’t true. Consequences do matter, they just can’t justify an immoral means. If you are choosing between two moral means, then choose the one with the better end results. But, the ends cannot justify the means. Ever.
I states my problem with utilitarianism in a recent post. If it leads to the same means as deontology, then it (or deontology) is unnecessary. If it leads to different means, then pretty much by definition, that means the utilitarian means is immoral. And while people come up with ridiculous examples (see: The Inquisition), there have to be some minor immorality slipping thru for utilitarianism to differ from deontology.
KevinDC
Jan 4 2021 at 10:13am
It’s unclear to me what you mean here. If you’re saying that the idea that consequences don’t matter is untrue, then I agree. If, by contrast, you’re claiming that it’s untrue to say any deontological theorists have claimed consequences don’t matter, then that’s just false. The old phrase “Fiat justitia ruat caelum” (do justice and let the skies fall) was a common rallying call for people who believed just that. They believed that if the only possible way you could prevent the complete destruction of civilization was by committing a slight injustice, it would be better to let the world be destroyed. To them, it was an ideal. To me, it seems like a reductio ad absurdem.
I also think the other part of your comment is assuming away the very point under dispute. To a deontologist, an action is moral or immoral to the extent that it abides by the rules (these rules might be derived from individual rights or the rules of justice or the categorical imperative or divine command theory and the Will of God or whatever). To a consequentialist, an action is moral to the extent it results in better consequences. To say that a strike against consequentialism is that it allows “immoral means,” then, is to already take for granted at the outset a deontological definition of what makes something moral or immoral in the first place. It’s pure question-begging. This is particularly clear when you say:
By whose definition though? By a deontologist’s definition. A utilitarian could just as easily say that because deontology recommends things that are immoral as morality is defined by utilitarianism, deontology is therefore “by definition” immoral. Round and round we go, each side asserting a definition, then declaring that their stipulated definition proves that the other side is wrong – by definition. I agree with Eliezer Yudkowsky that argument by definition is a uniquely bad way of analyzing pretty much anything.
I don’t find this convincing when merely asserted, nor have I encountered any persuasive arguments in favor of this. What would you recommend as the strongest defense of this idea? Book, journal article, etc. Point me to the best, most well crafted argument you can identify for this proposition and I’ll check it out.
robc
Jan 4 2021 at 11:09am
My post was a bit sloppy, so just add in somewhere “Assume the categorical imperative.” I know some moderns have tried to shoehorn Kant into consequentialism or whatever, but that is really stretching things.
Combine my 1st and 3rd sentences, that explains it. For many (most?) of our actions, the CI doesn’t apply or doesn’t entirely restrict our options, so sure, choose based on consequences. I think in Kant it is the The Principal of Right vs the The Principal of Virtue.
KevinDC
Jan 4 2021 at 2:10pm
I’m far from being a scholar on Kant, and that’s a personal deficiency I don’t think will be rectified soon, so I have no dog in that fight 😛
If I understand you correctly, this is what you contend:
Assuming I have accurately summed up your position, then I stand by my initial claim. Here’s why.
If two things both have value, it necessarily follows that one cannot be categorically more valuable than the other, in the sense that there is no ratio at which they can be traded off against each other. Means might be as valuable as diamonds and ends only as valuable as pennies, but enough pennies will be more valuable than any diamond. To say that no amount of the value of ends can outweigh the value of means is to say that the ratio of value between them is such that “E / M = 0”, where E is the consequential value of the end, and M is the moral value of the means. For E / M to be 0, either the value of E must also be 0, or the value of M must be infinity. If E is 0, then consequences must be without value, QED. And if M is infinity, that also reduces the value of E to zero. If one action has an E value of 100 and the other has an E value of 10, and both are have an infinite M value, then both options are of equal value. Saying “Option one is has a value of infinity plus 100, so it’s better than option two which is only as valuable as infinity plus ten” is incoherent. Both options are equal infinities. The only way a larger E can make a difference between these two options is for M to not be of infinite value – which entails that M is not categorically more valuable than E. So, holding both that “consequences have value”, and “consequences can never be traded off against means” entails a logical contradiction.
robc
Jan 4 2021 at 2:50pm
I would say that ends and means are measured on separate scales.
Means are a binary (its probably more complicated than this, but not much): Either they are moral or they are not. At the least there are discrete levels.
Ends are measured on a different scale and are not discrete. They are measured in utils or whatever.
So, yes, in some ways you are right. There is no amount of end value that can override an immoral means. Which may mean, in some extreme circumstance, that the continued existence of humanity can’t outweigh the choice we have to make.
But, as the means scale is discrete, there are many actions on the same level, I don’t think there is such a thing as “more moral” in that sense, so we often have choice in the matter. And in that case, the tiebreaker is consequence. It is like doing a sort by two columns in excel. The 2nd sort can’t overcome the results of the first sort, but it doesn’t mean the 2nd column is worthless.
KevinDC
Jan 4 2021 at 3:52pm
I don’t see how this escapes the criticism I made. To say that “There is no amount of end value that can override an immoral means” necessarily entails that you think E / M = 0. Specifying that E is a non-discrete value measured on a continuous scale of utils and that M is a binary scale of either infinity or negative infinity doesn’t rebut anything in the argument I made. (Negative infinity for immoral actions, incidentally, because that must be the value of immoral means, to maintain your contention that positive ends can never outweigh negative means.) No gains in E, no matter how huge, could outweigh the M when M is immoral – which requires that an immoral M is negative infinity. Similarly, no losses in E, no matter how destructive, can ever outweigh the value of M when M is moral, which necessitates that a moral M is positive infinity. To claim that finite sums can act as tiebreakers between infinities is incoherent. Analogies with other finite entities like Excel columns are not applicable when talking about infinite values.
There’s another odd implication of this view that moral and immoral are on a binary, yes/no measure. No amount of good consequences can ever outweigh the negative infinity of an immoral means, and no amount of bad consequences can ever outweigh the positive infinity of moral means. This entails, as you’ve already said, that moral actions are all equal. There is no particular option that is “more” or “less” moral than another – all moral options are equally moral, as you put it. But this would also entail that all immoral actions are equally immoral as well, which is absurd. It’s wrong to TP my neighbors house, and it’s wrong to burn their house down while they are sleeping inside, but one is obviously much worse than the other. But if immoral actions are all negative infinity, these two would have to be equally bad. I think that’s obviously false. And if immoral acts aren’t all equally bad, then the badness of immoral acts cannot be negative infinity – and if the badness of acts isn’t infinite, then it necessarily can be outweighed by other considerations that are sufficiently valuable.
AJ
Jan 2 2021 at 4:55pm
Where I struggle with utilitarianism – even where you argue it should be the only reliable ethical system for public policy – is which pleasure or happiness ought to be my guide? There are simply too many different pleasures or kinds of happiness that are to a large degree incommensurable and appeal to pleasure or happiness as an ethical norm can’t help me choose between competing enjoyable activities. If this is true for an individual, I suspect it’s true for public policy as well.
John Hall
Jan 2 2021 at 5:10pm
The issue with maximizing aggregate utility is that it ignores cross-sectional differences in utility, i.e. the public policy that raises utility by a little for a large number of people but then reduces utility by a lot for a small number of people. Conditional value-at-risk (aka expected shortfall) provides a good way to think about this downside risk. Think of it like a portfolio management optimization, except you replace variance with a number reflecting the average loss in utility in the x% worse case (you can also replace this with an average across several thresholds).
Henri Hein
Jan 2 2021 at 7:54pm
I agree with a lot of what KevinDC said above (comment @2:59PM). I have not seen good Utilitarian arguments to counter the standard objections: Utility monsters, the computational problem, Pascal’s mugging.
What I don’t understand here is how the examples above support Utilitarianism. Under a Deontological regime, Pfizer and Moderna would be free to produce and market the vaccine to the highest bidders, or whatever other scheme they preferred. Anyone could set up a challenge trial if they could find willing participants. The restrictions we have in place against these processes are generally supported by Utilitarian arguments, not Deontological ones.
KevinDC
Jan 3 2021 at 9:33am
Scott’s answer to this is that utilitarianism is a guide to public policy questions we face in the reality. Utility monsters and Pascal’s mugging are scenarios that simply don’t occur in real life and are not questions of public policy in the real world. What Scott is interested in is a theory that fits the bill of “recommends good courses of action in situations we actually face.” Demonstrating that the theory would recommend bad courses of action in imaginary scenarios we never actually face is, to Scott, irrelevant.
It’s also not clear to me that the issues currently plaguing the vaccination process in the US are due to utilitarian vs deontological moral theories. One example that comes to mind is the debate among planners described in this New York Times article. We have a myriad of people attempting to centrally plan the One Correct Way that the vaccine “should” be distributed, and their criterion certainly doesn’t appear to be a utilitarian approach, like one along the line of “save the most lives possible.” Indeed, it appears to be deontological reasoning at the core of it, but no doubt deontological principles you would reject. From the story I just linked to:
Among the arguments along these lines was that if you vaccinate 1000 elderly people, you’ll statistically vaccinate 100 black people. But if you vaccinate 1000 younger people, statistically you’ll vaccinate 150 black people. But, since the elderly are a couple orders of magnitude more likely to die or suffer major complications due to Covid, their own data predicted that prioritizing the younger crowd wouldn’t just result in more deaths overall, it would also result in more deaths among black people too. But while the absolute numbers of black people dying would be greater, the relative percentage would be smaller, and therefore it would be more “racially equitable,” whatever that means.
In the ensuing debate, the story notes that “some people believe it is wrong to give racial and socioeconomic equity more weight than who is most likely to die.” But, plausibly, the “prioritize vaccinating those who are most likely to die” is a more utilitarian approach, and “prioritize the vaccination procedure according to racial and socioeconomic equity, even though it means more people will die” is a more deontological approach.
(I really should learn to leave shorter comments.)
Henri Hein
Jan 3 2021 at 3:01pm
We do face the computational problem constantly. As AJ pointed out, it’s difficult to evaluate the cost and utility of various efforts and pleasures at the individual level. It seems pretty much impossible to do so across a population.
I disagree we don’t face the utility monster problem. A version of this was used as one of the arguments to sustain slavery in the antebellum South. If slaves give up only a small marginal utility for giving up their freedom, but the utility to the slaveholder is enormous, then the institution is warranted. Of course, the premises themselves are faulty here, but I’m uncomfortable with a framework that would sanction slavery even under those premises.
At the end of Owning Mahowney, the therapist asks him to evaluate the thrill from gambling on a 1-100 scale. Mahowney says gambling is a 100 for him, and the largest thrill outside gambling is a 20. Since a tax on the rest of us to pay for Mahowney’s gambling would be negligible, Utilitarianism says we should support his habit. This is a Utility Monster problem.
I enjoy your comments, so please don’t abridge yourself on my account. 🙂
KevinDC
Jan 4 2021 at 11:08am
I agree the computational problem is real and important. Honestly, I don’t have any clear idea of what “maximizing aggregate utility” would even mean. Matt Zwolinski summed up a number of objections to utilitarianism I share here, including how the idea of “utility” itself has no well defined meaning or agreed upon definition:
Regarding the utility monster argument, it sounds like with the slavery example, you’re saying that in practice the utilitarian calculus didn’t actually justify slavery (“the premises are faulty here”), but in principle it could have. If so, then we don’t actually disagree on this point. What Scott’s understanding of utilitarianism entails (and correct me if I’m wrong here, Scott!) is that he approves of the policies that utilitarianism actually suggests in practice, and he’s uninterested about how they could be made to recommend a bad policy in principle. After all, any moral theory can be made to recommend horrible things in principle with a carefully crafted thought experiment. I think Scott’s advocacy of utilitarianism could be summed up as “When you can show me a plausible, real world situation where a policy that does horrible things also makes the world a happier place, then I’ll abandon utilitarianism, but in the real world, policies that do horrible things make the world a less happy place, and asking me to imagine horrible policies making the world a happier place is a pointless exercise for guiding real world decisions.” (Again, correct me if I’m wrong, Scott.)
Regarding the gambler, I think this is where the “rules vs act” utilitarian distinction comes in for Scott. Maybe for an individual act, keeping the gambler subsidized would increase total utility. But establishing a rule that gambling addicts (or other forms of unsustainable self destructive behavior) should be endlessly subsidized at everyone else’s expense would, in the real world, decrease total utility, so he’d be opposed to it.
Henri Hein
Jan 4 2021 at 4:11pm
One quibble I have is that the Confederates used the argument to defend slavery in practice, not just in principle. To which you and Scott could reasonably argue that they used Utilitarianism incorrectly. But I’m still left with a problem that if Utilitarianism was used to produce bad results in the past, why should I trust it to only produce good results in the future?
In general, I am actually more interested in the hypotheticals. For topical issues, we tend to have our minds made up and use the theoretical frameworks to rationalize them. We are not vested in the issues from the thought experiments, so in some ways it is easier to think clearly about them. For a similar reason, unlike robc and Philo, I find it useful to test the theory against historical examples like slavery and the Inquisition. To modern sensibilities, these were clearly unjust. So what happens when we apply the theory against them?
KevinDC
Jan 4 2021 at 5:21pm
Hey Henri –
Yes, Scott would say that the utilitarian case for slavery was a bad case and therefore not a mark against utilitarianism. And Scott would also likely say that the problem of people using utilitarianism badly in the past should be counteracted by using utilitarianism well in the future, not by dropping utilitarianism altogether. And Jason Brennan argues (successfully I believe) that concerns that a moral theory will be misapplied or used incorrectly does not constitute an argument against the truth of that theory. Funnily enough, he uses utilitarianism as an example, although he himself is not a utilitarian. Also, in that same post, he notes:
In a way this is ironic because in other posts, Scott has argued the opposite. He views utilitarianism as being a method for making decisions, and not a criterion of right, and that he considers it to be true only to the degree that it’s useful for making decisions. That is to say, he defines utilitarianism to mean something like the opposite of how it’s used by everyone else in moral philosophy. I and others have taken turns yelling at him in the comments over his nonstandard terminology, but he seems wedded to his definition so eventually I just have to make sure to remember to keep his definition in mind when he writes about these topics.
I also agree that extreme and even unrealistic thought experiments and hypotheticals are useful for analyzing moral theories. Naturally, Scott Alexander, being both smarter than me as well as a better writer, has the best explanation about why.
Scott Sumner
Jan 3 2021 at 1:26pm
You said:
“The restrictions we have in place against these processes are generally supported by Utilitarian arguments, not Deontological ones.”
I strongly disagree. In the calculus or bioethicists, the risks to test participants count for far more than “errors of omission”, the risks to the broader public in not have a vaccine.
Henri Hein
Jan 3 2021 at 3:02pm
Fair enough. I just meant that arguments for having an FDA and a CDC in the first place tend to be Utilitarian. At least the ones I have heard.
robc
Jan 4 2021 at 7:54am
What is the opposite of deontology, where you choose the most immoral means possible? I may be over-cynical, but I would put most ethicists in that category.
Phil H
Jan 3 2021 at 3:39am
I agree with the thrust of Scott’s argument, that the only reasonable rationale for public policy is utilitarian. And yet I’m much more willing to accept conservative constraints on utilitarianism. I’m not sure if this is right, but I’ll try to articulate why I think this way.
It’s that bad mistakes have been made with markets (and other ideologies) before, and they’ve been really catastrophic. Extinction-level. And it’s not obvious to me that a mistake with ethical policy now wouldn’t lead to similar disasters.
For example: Allowing compensated organ donation. Sounds fine, and to be honest, I’d probably vote for it. But it’s still not obvious to me that 50 years down the track, it won’t lead us to a Never Let Me Go scenario. I’m not confident of that, so I feel no urgency to campaign strongly in favour of organ donation.
Skimping on the testing of vaccines: sounds great in the current crisis. But it’s not obvious to me that it won’t allow commerce to cannibalize medicine over the next 50 years, and take us back to the snake oil ages.
I’m this cautious because it feels like many of the great moral disasters of the world started this way. Exploring foreign countries was a great idea. I doubt it was obvious to Columbus that his discovery would lead to the decimation of the American peoples. But it did. And the Industrial Revolution, with its satanic mills, was a non-obvious result of the invention of the steam engine. But quality of life cratered for generations before it finally rose again.
The reason I’m not certain I’m right is that imperialism and the IR ultimately led to the modern world, which is better. So perhaps they were necessary evils? I dunno. But I’m less willing to throw away some old-fashioned ethical qualms in the name of utilitarianism than Scott is because it feels to me like those ethical qualms are part of a system that is not disastrous; and many other systems are disastrous.
Scott Sumner
Jan 3 2021 at 1:22pm
You said:
“Skimping on the testing of vaccines:”
That’s not what I and others propose. We propose challenge studies that do the same testing, but faster. We’d simply have the test participants take a tiny bit more risk. The vaccine would still be fully tested.
You said:
“I doubt it was obvious to Columbus that his discovery would lead to the decimation of the American peoples. But it did.”
He also didn’t expect it to lead to a much larger world population, more people being able to enjoy life than before—but it did. (Especially in countries like India and China, where new world foods allowed for much larger populations.)
Of course the techniques used by the Spaniards were evil, and non-utilitarian.
Phil H
Jan 4 2021 at 2:13am
OK, those make sense. Sorry, I had perhaps confused your suggestions with David Henderson’s. I agree with you that challenge trials are a reasonable kind of investigatory tool. It’s not obvious to me that they would be much better and faster, and I’d want to let the medical researchers make that decision. I see very little evidence that they have been constrained from making that decision. Medical ethicists aren’t a weird cabal with power over big corporations! So I’m not sure that there’s a problem to be fixed here.
On the evils that led to modernity – yes, I agree that that’s relevant history. I can’t take that as a good reason to embrace evils, though. Should we be celebrating the prison camps in Xinjiang because they’re part of China’s modernisation process? This isn’t a tenable way of thinking.
Scott Sumner
Jan 5 2021 at 12:36pm
You said:
“On the evils that led to modernity – yes, I agree that that’s relevant history. I can’t take that as a good reason to embrace evils, though. Should we be celebrating the prison camps in Xinjiang because they’re part of China’s modernisation process? This isn’t a tenable way of thinking.”
Since this doesn’t relate to anything I said, in any way, shape, or fashion, you are better off not making this sort of comment.
Neil S
Jan 3 2021 at 12:29pm
I believe that the challenge regarding development and distribution of a Covid vaccine have little to do with moral reasoning and a great deal to do with bureaucracy. The US is particularly challenging due to both the division of power between states and the federal government and the diversity and size of the population.
Scott Sumner
Jan 3 2021 at 1:23pm
Bureaucrats are engaging in non-utilitarian reasoning.
Neil S
Jan 3 2021 at 6:16pm
This reminds me a bit of arguments about communism…in an ideal world, with perfect information and people who make decisions without regard to their own self interest, communism works. Likewise, bureaucrats who are willing to make decisions without regard to their own self interest and who have excellent information should apply utilitarian reasoning.
Given that in the real world, we are dealing with bureaucrats who are primarily interested in protecting their own self interests and who have limited ability to weigh the consequences of their actions in a utilitarian fashion, this seems like a rather pointless discussion.
Scott Sumner
Jan 5 2021 at 12:33pm
Exactly. That’s why I favor deregulation.
Ken P
Jan 4 2021 at 1:26am
Challenge studies aren’t the speedy solution many assume. The challenge dosing studies and administration approach create delays to efficacy initiation. But a bigger issue is the complexity it brings to decision making. The approval for the Pfizer vaccine was not unanimous for even such a fairly simplistic study with strong results. If the review includes a challenge model with potentially conflicting results during development and scientific arguments surrounding the relevance of the model to real world exposure, then there will be many discussions that wade out into the weeds and require weeks to months to come to agreement.
The current data has barely moved my Bayesian prior from my perspective after the strong neutralizing antibody results. To me, that was sufficient to allow emergency use in an en Ironman where 1000 people per day are dying. But am I more confident in protection now? No. I have the same questions I had previously. Does the vaccine only prevent cell entry by virus or will it leverage the immune response if the infection actually takes hold? Does it create a T cell response? If so, what clonotypes are expanded and are those clonotypes correlated with positive outcome in real world naturally infected patients?
Ideally, I would like a spec sheet like I would look at if I was buying a new stereo receiver or a new smart phone.
Ken P
Jan 4 2021 at 7:25pm
should be “environment”.
Also, I should point out that I do agree with Scott that no company was going to release a vaccine that was not fully tested. I’m not sure how many people would have to die per day for any company, the FDA or group of politicians to make that decision. I still think there is a point (maybe 10,000 per day?) where the death rate would be high enough for institutions to think it might be worth it to shave off several months of deaths.
Scott Sumner
Jan 5 2021 at 12:32pm
Why not simultaneously do both? A challenge study and a regular study. After the challenge study start vaccinating people that that studies suggests could be helped, such as younger workers in essential industries.
This crisis should have triggered an “all hands on deck” approach—do everything.
Tiago
Jan 4 2021 at 10:03am
“Everyone is entitled to choose their own preferred ethical system as a guide to their daily life. But there is only one reliable ethical system to be used in public policy—maximizing aggregate utility. As soon as you ignore that goal, you end up killing lots of people for no good reason.”
Excellent. Joshua Greene calls this “metamorality” – the ethics about how to choose between ethical frameworks.
Dylan
Jan 4 2021 at 6:12pm
I feel outclassed in this discussion by a large degree by people who are smarter, better writers, and have thought about these issues much more than I have. I want to say a particular thanks to KevinDC and Henri for an illuminating discussion. Personally, I largely agree with what KevinDC wrote in his first post. However, I think Scott’s view is nevertheless useful. This is how I think of it. No ethical theory is going to be uniformly applied, or really come even close to it. This probably goes double for politicians. But, perhaps naively, I do think politicians are often trying to do the right thing (at least in cases where the incentives aren’t pointed in the exact wrong direction). In those cases, a politician that had a slightly more utilitarian outlook is more likely to get things right, than someone with a dentological instinct.
All of the edge cases and theoreticals are interesting and can be illuminating, but the reality is, our hypothetical politician is likely to get those wrong no matter what moral theory they are operating under. In the more normal decisions that they have to make, if they are trying to look at both costs and benefits, observed and hidden, we’d likely get better outcomes.
Thomas Strenge
Jan 6 2021 at 9:59am
It was utilitarianism that convinced the Nazis of the necessity to kill all Jews. After all, it was for the good of the German people.
Comments are closed.