Vacchablogga

September 7, 2019

The argument from predictive success

Introduction

On episode 17 of the 80,000 Hours Podcast, MacAskill presents three arguments for utilitarianism. One is the argument from predictive success:

When you look at scientific theories, how do you decide whether they’re good or not? Well, to a significant part by the predictions they’ve made. We can do that to some extent, but with a much smaller sample size, with moral theories as well. For example, we can look at the predictions, the bold claims that were going against common sense at the time, that Bentham and Mill made. Compare [them] to the predictions, bold moral claims, that Kant made.

When you look at Bentham and Mill, they were extremely progressive. They campaigned and argued for women’s right to vote and the importance of women getting a good education. They were very positive on sexual liberal attitudes. In fact, some of Bentham’s writings on the topic were so controversial that they weren’t even published 200 years later…

Contrast that with Kant. Here are some of the views that Kant believed. One was that suicide was wrong. One was that masturbation was even more wrong than suicide. Another was that organ donation is impermissible, and even that cutting your hair off to give it to someone else is not without some degree of moral error.1

Of course, Bentham, Mill, and Kant weren’t actually making predictions when they made moral claims. MacAskill’s idea is just that the “bold” or controversial moral claims made by the early proponents of a moral theory count for or against it, depending on whether those claims are true. He thinks that this is analogous to how the predictions of a scientific theory count for or against it, depending on whether those predictions are realized.

Markovits, though not a utilitarian herself, presents the same argument in a Wi Phi lecture:

[Bentham] was an early defender of economic liberalization, freedom of expression, the separation of church and state, women’s rights, animal rights, the right to divorce, the abolition of slavery, the abolition of capital punishment, the abolition of corporal punishment, prison reform, and even the decriminalization of homosexual acts. Remember, this was 1789! Bentham recognized the moral importance of these rights, many of which are now uncontroversial. In this, he was well ahead of his time, and in large part, I would think, because of his embrace of utilitarianism. That, for me, counts heavily in favor of it as a moral theory.

I’ll first say a bit more to explain the core idea of the argument. Then I’ll state the argument more precisely, and raise some challenges to it.

The core idea of the argument

Most of the views that MacAskill and Markovits attribute to Bentham and Mill are no longer controversial. Mainstream proponents of every major moral theory, including Kantian ethics, accept them. Nonetheless, on utilitarian grounds, Bentham and Mill accepted these views when they were widely rejected. This, I take it, is what MacAskill and Markovits think supports utilitarianism. It’s not just that Bentham and Mill were right. It’s that they were right when most people were wrong.

Why is this supposed to support utilitarianism?

Well, suppose that you’re developing a moral theory in 2019, when women’s suffrage is widely supported. If you notice that your theory entails that women should not have the right to vote, then you will simply change your theory—possibly in an ad hoc way—to avoid this consequence. So the fact that your final theory is right about women’s suffrage doesn’t necessarily reflect some insight that your theory has about morality. It might just reflect the fact that you copied from a better moral theory than your own.

In contrast, suppose that you’re developing a moral theory in 1789, when women’s suffrage is widely opposed. If your theory entails that women should have the right to vote, that’s probably not because you made an ad hoc change to it to get this consequence. If anything, you would have to resist the urge to change your theory to avoid this consequence. So unless you just got lucky, the fact that your theory is right about women’s suffrage probably reflects some insight that your theory has about morality.

Assessing the argument (non-contrastive version)

That’s the core idea of the argument, but I still haven’t stated the argument precisely. Here is one way to make it more precise:

  1. Many of the bold moral claims made by the early utilitarians are true.
  2. The best explanation of (1) is that utilitarianism itself is true.
  3. So, probably, utilitarianism is true.

(1) should be uncontroversial. Again, the claims in question are things like “women should have the right to vote” and “girls and women should have the same access to education as boys and men do”.

The action is at (2). This is where the core idea comes in. Why is the best explanation of (1) supposed to be that utilitarianism is true? Well, since the claims in question were widely rejected, Bentham and Mill weren’t copying from some generally accepted moral theory. So one alternative explanation of (1) is ruled out. Another alternative explanation of (1) is that Bentham and Mill just got lucky. But this seems unlikely, especially because so many of their bold moral claims are true.

Still, there are other explanations of (1). The most obvious alternative explanations are those that say that utilitarianism is false, but that it has at least some insights about morality. (You will be drawn to these explanations if you already think that utilitarianism is false.) There are many possibilities here, including:

  • Consequentialism is true, but hedonism is false. Nonetheless, happiness is one good thing, and maximizing happiness often ends up maximizing the good.
  • Consequentialism is false, but only because there are side-constrains that limit what you can permissibly do in pursuit of the good. But side-constraints are irrelevant to the cases in question. For example, women’s suffrage doesn’t violate anyone’s rights.
  • Utilitarianism is badly misguided. Nonetheless, it is right to reject arbitrary distinctions between groups of people, including men and women. And this was enough to get Bentham and Mill to reject traditional sexist beliefs.

In order to defend (2), a proponent of the argument must rule out these other explanations of (1). And it is not obvious to me how to do this without making an independent argument for utilitarianism.

How big of a problem is this for the argument? The first two alternative explanations of (1) might not be a huge problem. Although they are inconsistent with utilitarianism, they grant that it is at least pretty close to the truth. So a proponent of the argument could accept that they are plausible explanations of (1), and then weaken the argument’s conclusion to say that utilitarianism is at least pretty close to the truth. This is still an important conclusion, since many people think that utilitarianism is not even close to the truth.

It’s the third alternative explanation of (1), and explanations like it, that are a bigger problem for the argument. This explanation denies that utilitarianism is even close to the truth. So you can’t grant that this is a plausible explanation of (1) while still concluding that utilitarianism is at least pretty close to the truth. The explanation does concede that utilitarianism is not wrong about everything, but even the staunchest opponents of utilitarianism were already willing to concede this.

To be clear, the “no arbitrary distinctions” explanation is just an example of the sort of explanation that is a problem for the argument. By itself, it can’t account for all of the successes of Bentham and Mill. For example, being anti-capital punishment doesn’t seem to have anything to do with rejecting arbitrary distinctions. So an anti-utilitarian would have to either come up with a better explanation of (1), or add to the “no arbitrary distinctions” explanation an additional explanation that accounts for the other successes of Bentham and Mill.2 But it’s not obvious to me that this is impossible.

Assessing the argument (contrastive version)

That’s one formulation of the argument. There is a different formulation of the argument, which is suggested by the contrast that MacAskill draws between the claims made by Bentham and Mill and the claims made by Kant. On this formulation, the argument aims to show not that utilitarianism is true, but rather that it is closer to the truth than Kantian ethics is:

  1. The bold moral claims made by the early utilitarians are, for the most part, closer to the truth than the bold moral claims made by Kant.
  2. The best explanation of (1) is that utilitarianism itself is closer to the truth than Kantian ethics is.
  3. So, probably, utilitarianism is closer to the truth than Kantian ethics is.

Here I think (2) is plausible, provided that (1) is true. The problem is with (1). In the above quote, MacAskill defends it by comparing the greatest hits of Bentham and Mill (for example, support for women’s suffrage) with the greatest misses of Kant (for example, opposition to organ donation). That’s not a fair comparison. To accurately assess (1), we would have to compare a representative sample of Bentham’s and Mill’s bold claims to a representative sample of Kant’s bold claims.3

This is not to say that (1) is false. Maybe even if we were to do a fair comparison, we would conclude that the bold moral claims made by Bentham and Mill are, for the most part, closer to the truth than the bold moral claims made by Kant. But at the very least, MacAskill doesn’t adequately defend it in the above quote. (To be fair to MacAskill, he was presenting the argument informally in a podcast, where he had lots to talk about apart from this argument.)


  1. This is a slightly cleaned-up quote from the episode transcript.
  2. If an opponent of the argument takes the second option (that is, if they give different explanations of the different successes of Bentham and Mill), then the anti-utilitarian explanation of (1) arguably starts to look less plausible than the simpler and more unified explanation that utilitarianism is true.
  3. An obvious challenge here is figuring out what a “representative sample” of someone’s bold moral claims includes. Proponents of Kantian ethics might want to emphasize Kant’s more general claims about the importance of human dignity and autonomy. Opponents of Kantian ethics might, like MacAskill, want to emphasize Kant’s claims about suicide and organ donation.