Are Condoms Murder?

Utilitarianizing abortion

So I guess we’re talking about abortion again, huh? Fine by me. I don’t usually try to approach ethical questions with a top-down, “my theory says so,” approach. First because it’s not gonna be very convincing to people who don’t agree with my normative ethics, even when it could be; and second because it’s good to not get lost in the normative-ethical sauce, and lose touch with what you actually think about things outside your theoretical framework. This time, however, I’ll make an exception.

Now, I have some good news and some bad news regarding what utilitarianism says (specifically total utilitarianism, for obvious reasons). Good news first? Well, the good news is that utilitarianism gives an incredibly elegant answer on the abortion question, and avoids the hell that is personhood-discourse. Bad news? What it says seems a little crazy, at least on first impressions.

I’ll first explain why the answer it gives is so elegant, and slightly disturbing, and then try to see if we can’t make it a little more palatable.

Personhood

Non-utilitarian discourse around abortion faces a big problem. We want to find out whether abortion is murder, and since murder ostensibly requires that there’s a person there to murder, we need to find out when persons begin, and under which conditions. And be careful! It’s surprisingly hard to give conditions that don’t give the wrong answers in other cases. This leads to a barrage of increasingly intricate examples and counter-examples, to pinpoint the exact necessary and sufficient conditions for personhood in the morally relevant sense.¹

Is conscious experience required? Then what about someone asleep or in a coma? Does it require some sort of more abstract capacity for experience (however you want to spell that out)? Then what about a person whose brain would regenerate 1, 10, or 1000 days after being destroyed (either by itself or through surgery)? Does it require the above plus also having had experience in the past? Should we then settle reincarnation before we can settle whether life begins at conception?

All of this is one big mess!² Luckily utilitarianism is ready to swoop in and give you some damn answers. “Forget about perssssonal identity” it hisses in a sultry voice, teeming with temptation. “Jusssst look at expected value differencesssss” it tells you. And so we shall.

This gives a very easy answer to all of the above questions—and the correct ones at that, it seems. Is it bad to kill someone sleeping or in a temporary coma? Yes, because they’ll miss out on the goodness of their future life.³ What about a person whose brain will regenerate in a long time? Again, they’ll miss out on the life they could have had if you kill them, so it’s wrong. Do we need to settle reincarnation before we can settle abortion? No, because whether or not you had a past life doesn’t change how good your future life might be.*

*Small caveat: Reincarnation does affect the issue insofar as reincarnation affects how bad any death is—if you’ll be reincarnated, dying won’t lead to a future of 0 utils (for better or for worse). But then this is obviously the correct judgement in this scenario anyways. If dying isn’t the end, that should clearly factor into our decisions. What shouldn’t factor in is whether you have already existed for some time. Anyways, back to the studio.

Thank you, intermission-Silas. What I was about to say was that I find these judgements to all be what I would intuitively judge anyways. On top of that we avoid the endless epicycles and counter-counter-counterexamples of the above dialectic, instead having an easy general answer. Unless…

Oops, proved too much

Trouble starts once we look back on the question we were actually here to answer to begin with: What is the moral status of abortion? Let me just crunch the numbers for a moment annnd… Whoops. Looks like not having a child is about as bad as having an abortion. Wait a second, let me just run a few more numbeeers… Double whoops. Also looks like killing a newborn is about as bad as having an abortion, which is about as bad as not having any child to begin with.

Now, “about as bad” isn’t a transitive relation, so we’re not in completely deep shit. Still, something fishy is clearly going on here. Generally we don’t think that people who use contraception are moral monsters in even the same ballpark as someone who smashes a newborn with a sledgehammer to “try again,” all Peter Singer-style.

Utilitarianism sort of gets flanked from both ends on this question. Looking at it from before procreation, it seems clearly permissible to use condoms or “morning after” pills. However, looking at it from after birth, it seems like it’s obviously not permissible to kill a baby—in fact it seems about as bad as (if not worse than) killing a toddler, or a full-grown adult. So what can we say about this?

I.

First step is getting rid of words like “permissible,” “wrong,” etc. If utilitarianism is right, these are not fundamental moral categories, but sometimes-useful paraphrases over better-than and worse-than relations.

Now, we can certainly admit that procreating would be better than not, all else being equal. Similarly (in fact for the exact same reason) not getting an abortion would be better than getting it. Still, for both of these I think we can give an account of why the structure here can allow us to say that they’re “permissible” (using our folk language).

This is because we can construct an ersatz distinction between doing-vs.-allowing within this utilitarian framework. Consider two courses of action:

1. Giving Bob $1000.

2. Giving Bob $1000 and then kicking him in the balls.

3. Giving Bob $700.

Suppose Bob values not being kicked in the groin at $300. Then (2) and (3) are strictly speaking equally right on utilitarianism. Still, there is something special about the structure of (3) that makes it more objectionable in a different way: You’re going out of your way to do something bad, which you could just not have done. A way of cashing this out is that not kicking people in the balls dominates your choices—whatever you do, you might as well just not kick someone in the balls, whereas there are reasons for and against giving Bob $300 extra dollars (you’d enjoy them, but he would too).

II.

From this you probably suspect that I would think utilitarians should oppose abortion, be pro-life, or whatever you’d like to call it. There are a couple of things to be clear about here though:

1. The abortion debate is very tightly bound up with legal questions, and your position on the moral question doesn’t automatically translate into a certain position on the legal question.

2. Despite an abortion ostensibly being a “doing,” you have to remember that the account of doing-vs.-allowing is ersatz, and not actually tracking the intuitive deontological categories. What matters is not the doing-vs.-allowing part, but rather how much the act involves “going out of your way,” so to say. And while having an abortion is going out of your way to some extent, being pregnant certainly also involves going out of your way—in fact, it involves a much larger commitment than having an abortion does.

A better analogy than ball-punishing is setting up a monthly bank-transfer to Bob’s bank-account. Having an abortion would then be like cancelling it before it really got started, and not having a kid would be like never setting it up to begin with. Again, we can suppose it’d be really freaking good to give Bob this money—perhaps allowing him to afford malaria nets for him and his family. We can even say that it’d be “wrong” not to give him the money if we like. Nevertheless, it seems like cancelling the transfer is not needlessly cruel in the same way as giving him a firm foot to the groin would be.

Not procreating would be like not setting up the bank transfer to begin with, having an abortion would be like cancelling the transfer, and killing someone would be like siphoning money out of a third party’s bank account. Strictly speaking you’re acting as rightly if you give Bob money and steal money from someone else, as if you just did nothing. But there’s still a clear difference: In the former case, you could just not go out of your way to steal the extra money, which makes it especially objectionable, though not especially “wrong.” Mutatis mutandis for procreation and killing.

III.

However, now I’m sort of left with a feeling that we’ve gone back to not having solved the thing we set out to. If anything’s a big commitment, it’s having a child; and so it’s hard to see how pregnancy being demanding could make abortion count as “allowing,” without also making killing your kid later on count as the same.

I don’t really have any sexy theoretical solution to this. I mean, it should actually just be pretty obvious that on utilitarianism nothing magical happens at the moment where the baby rears its ugly head from the mother; neither does anything magical happen somewhere between the father █████ his fucking █████ all over the mother’s ██████ and then ████████████████, leading to the egg being fertilized, and 9 months later.⁴ Whether something good happens, or something happens that will lead to something good is neither here nor there—what matters is just the good.

We can still say some standard utilitarian handwavy practical things here, to make this pill slightly easier to swallow. Specifically, something magical actually sort of does happen when a baby is born, and in the time that follows—namely that they become “rooted” in the world, so to say. What I mean is that they start forming connections with others and with the world around them: People start caring about them, people start to learn about their personality and having an interest in seeing them grow up, and most importantly they themselves start having desires, hopes, and dreams.

I don’t think death is inherently bad. There is something bad about it, though, beyond just the potential life that is missed out on. Namely, it is bad when someone’s desires are thwarted; when their projects, hopes, and dreams are destroyed; and of course when a relationship that was supposed to go on is ended. A preference satisfaction account of welfare captures this quite naturally, though it is not the only way of doing so.

If a person is never created, or if a fetus is aborted before it takes root in the world, then we never get to this badness. To be clear, it is sort of unavoidable once a life has really started—death will generally be bad in this way whether you’re 10 or you’re 80, and you’re simply postponing it. But the point is that if you never have the kid to begin with, or if you abort it before it is born, you avoid having to incur this badness whatsoever, and thus there is something particularly bad about killing someone who is already in the world.

When I reflect on what makes me feel more disturbed about an actual life being ended than a potential one, then—barring obvious emotional reactions—this actually seems to be a large part of it. There’s something particularly bad about ending a life once it’s already started. Utilitarianism of course judges that it’s much better to make a full happy person than not; but if you choose to end a person, do it before they have something of substance to lose.

This goes hand in hand with considerations of non-fungibility which

Richard Y Chappell

has done a great job at making clear. The value of a person is not fungible in the same way as the value of a dollar. If you lose a dollar and gain a dollar, what happened is neutral. If you kill a person and gain a person, then sure, adding up the values, we can suppose they sum to 0. But nothing neutral happened here—rather something tragic happened, although it was combined with something very good.

And unlike actual lives, potential lives are plausibly fungible. There is no definite person yet who you’re refraining from creating—instead you’re refraining from creating some value, and we can’t say who would’ve enjoyed it. This doesn’t change what the right and wrong actions are from a utilitarian perspective, but it does change our attitude towards certain outcomes.

Keeping Score (Or: Conclusion)

I can’t lie, though, I honestly still find what utilitarianism says here quite counterintuitive. It’s just obvious that it’s much more wrong to have a kid and kill a kid than to never have a kid to begin with, and whatever bells and whistles you explain that away with, it’s hard to give up that judgement. At the same time, when I reflect on the issue more abstractly, it’s also very hard for me to see why exactly I should think that there’s any hard difference here. A person coming into existence is a very continuous process, and it would be strange to think that there’s some clear moral line that is crossed somewhere along the way.

(Oh yeah, and for many people, getting an abortion now, or not choosing to have a kid now won’t change how many kids they have, just when they have them. It would be better if they had a kid now and then also later, but in choosing between having a kid now vs. having one later, it’s not better to do it now—it may actually be worse.)

---

Epilogue: Returning to Personhood

In the face of these uncomfortable conclusions, a utilitarian might be tempted to give up on the surface-level elegance of their theory, and jump into the fray of personhood-debates once again. “Sorry guys” the utilitarian says “I was a little too quick to point fingers, and maybe there’s a point to this personhood stuff after all.” For example, you might try to say that things can only be good if they affect actual people, or that you should at least only count the welfare of actual people. IIRC, Peter Singer says something along these lines.

Maybe some moral theories can get away with this. E.g. you can say that you can only violate the rights of someone who actually exists. For a utilitarian, this move is extremely ill-advised.

I have given this kind of example before: Suppose you have a button in front of you. Pressing it will create a person in 5 minutes, who will go to hell and suffer a million years of listening to anti-vegan talking points (or fire and brimstone, if you think that’s worse). In return you get a lollipop. Should you press it? On this view, then the answer is a resounding: Yes! I mean, you get a lollipop, and nothing bad happens, as people who don’t exist yet don’t matter.

This is bad for two (or more) reasons: 1) That’s obviously the wrong verdict!! 2) In 5 minutes you’ll regret your decision very much, and be willing to sacrifice a lot more than a lollipop to undo it. Additionally you know that this is what you’ll think in 5 minutes. Nevertheless, this theory tells you that that doesn’t matter, and that you should still press it.

For someone like Peter Singer, who has argued strongly for ethical veganism, things get even worse, as there’s a very sharp tension here. If you buy animal products, you’re not actually harming any currently existing animals. Rather you’re creating demand for more animals with bad lives to be created in the future. Hence it seems like there’s nothing wrong with buying animal products anymore!

Whatever a utilitarian can say to avoid these verdicts would seem to apply equally to abortion. To avoid these conclusions, we have to care about the value of future non-actual lives, yet in doing so the conclusion from above falls out: It’s about as bad to kill a person as to refrain from creating one.⁵

---

You Might Also Like:

1. The progression of examples here is almost plagiarized from

Moralla W. Within

’s excellent post on abortion and personhood, which was the impetus for this post.

2. As with all things personal identity, the correct answer is the soul. But that only gives an in principle answer, not one that’s particularly helpful.

3. If they’re in a coma they’ll never wake up from, then this won’t work. But it also seems like it’s not very wrong to kill them at that point anyways.

4. Well, if it’s always determinate whether some being is conscious, and egg and sperm are not separately conscious, then there will be some moment where the fetus becomes conscious. But this won’t matter morally for the utilitarian—there’s no difference between harming someone now and harming someone who’ll exist in 5 minutes (more on this in the epilogue).

5. You might try to avoid this by arguing that the suffering of future non-actual people can be bad, but the pleasure of future non-actual people can’t be good (i.e. the antinatalist asymmetry). I think this is very implausible. See here for my post on the subject, or

Bentham's Bulldog

’s excellent post or paper.