Book review: The Life You Can Save, by Peter Singer.
This book presents some unimpressive moral claims, and some more pragmatic social advocacy that is rather impressive.
The Problem
It is all too common to talk as if all human lives had equal value, yet act as if the value of distant strangers’ lives was a few hundred dollars.
Singer is effective at arguing against standard rationalizations for this discrepancy.
He provides an adequate summary of reasons to think most of us can easily save many lives.
He’s a bit too optimistic when he suggests we know how to save lives in the third world at a few hundred dollars per life. But he points us to GiveWell, which has since produced a more realistic and rigorously supported picture, and that picture is close enough to Singer’s description for most purposes.
The biggest weakness in his argument that aid is effective comes when he tackles the possibility that aid just fuels population growth in regions that are trapped in Malthusian conditions. He provides reasons to think the world as a whole isn’t close to Malthusian conditions. But what if some regions are in such a trap? His main answer seems to be: if you think that’s the case, donate to contraception charities instead.
But that opens a whole can of worms – it suggests a large chance that I’ll waste money or cause harm by guessing wrong about regional conditions. And if I do correctly guess that poor countries are in Malthusian traps, Singer doesn’t convince me there are good aid options (I’ve seen hints that most mothers are able to limit their family size without contraceptives). (I expect that overpopulation isn’t a problem in most places, but I’m dissatisfied with the quality of the evidence that I’ve seen). (GiveWell has some evidence that aid doesn’t much affect population growth – that suggests aid isn’t causing Malthusian conditions. But that leaves me concerned that, due to existing Malthusian conditions, aid just postpones problems.)
Moral Claims
Singer tries to argue that we’re morally obligated to be more altruistic than our behavior suggests we are.
For some expansive meanings of the term moral (e.g. strict utilitarianism), his conclusions follow, but I don’t see any remotely convincing answer to the question “why should I follow that moral system?” [1].
And Singer’s arguments fall flat for notions of morality that reflect rules that most people are comfortable obeying.
These real-world moral rules are clearly somewhat focused on rules for which it’s easy to get a clear answer about whether I’m obeying them. They’re also somewhat focused on rules which most of society is willing to follow. See DeScioli and Kurzban for why selfish people want to follow such rules, and for evidence that human morality follows those patterns.
Singer’s moral arguments imply that a large fraction of people in wealthy countries ought to give more than half of their income toward helping the poor. He doesn’t pretend that he can convince many people to adopt that level of altruism. Those moral arguments also imply a good deal of uncertainty about whether I’m following them: if it looks like I could work harder to earn more money to send to the poor, how can you tell whether it would impair my mental health?
Singer’s key points are the famous drowning child example, and the multi-child, multi-rescuer version. The simple version posits an uncommon situation where a child’s life clearly depends on an identifiable rescuer.
The multi-person version slides us down what Singer implies is a slippery slope toward a more common situation where there are many drowning children and many potential rescuers. If I accept the argument for the single child, single rescuer example, then Singer claims I must also accept the multi-child, multi-rescuer example. But the ride down that slippery slope felt bumpier to me than Singer expected.
Folk-morality tends to focus on the boundary between what you are obligated to do and what you are not obligated to do. I say that this boundary is where it becomes hard for an observer to decide whether a specific child’s life depended on a specific rescuer (i.e. most likely when there’s more than one potential rescuer).
I expect that’s not a very popular answer to the question of what obligations one has to needy people, because it emphasizes the selfish reasons for folk-morality, by prioritizing the desire for predictable moral judgments over the desire to save lives of strangers [2].
My impression is that only a small fraction of people who hear Singer’s moral standards make serious attempts to obey them. And those that do, tend to burn out, because the conscious parts of their minds adopted those standards without buy-in from all of the subconscious parts, causing internal conflicts which undercut their motivations to accomplish most tasks.
I expect I’d feel lousy if I tried to adopt Singer’s morals, because moral rules are typically used to decide whether a person is good enough to belong in their community. So my intuitions would frequently tell me that, since I’m likely to fail at being a perfect saint, I’m at constant risk of not belonging in my community.
Idealism
“Doesn’t it [buying good stereo] make a mockery of any claim to believe in the equal value of human life?”, Singer asks. Yes, but so do the actions of people who the book (rightly) portrays as heroes, who devote half their income to effectively helping strangers, but still treat their children as more valuable than strangers.
Singer wants to pressure us into resolving discrepancies between our professed ideals and our actions by deciding that our ideals are right.
We know that our deliberative system is right
. Our deliberative systems know that our deliberative systems are right, but our emotional systems also know that our emotional systems are right. Our deliberative systems may have fewer biases, but if that explained the main difference between our altruistic rhetoric and our donations, I’d expect to see a larger number of people saying oops, and becoming effective altruists.
Some of more altruistic versions of Singer’s advice seem to be illegal in much of Europe (children have some rights to their parent’s wealth, although it’s probably more complicated than that post indicates). I find that hard to reconcile with the claim that we know Singer’s advice is right.
A more credible explanation is that the rhetoric is mostly signaling. In which case, it’s tricky to figure out the effects of pressuring us to take our ideals seriously.
Singer’s approach makes it harder to profess ideals without acting on them, which seems quite good. But it also creates resentment from people who want to use charity for signaling purposes – how do I evaluate the costs of that?
C.f. some critics of effective altruism consider it obvious that we’re not obligated to care much about distant strangers – enough so that they don’t feel a need to defend that belief, and assume it’s enough to point out that the standard approach to charity is under attack.
See Nate Soares for a somewhat different response to this part of Singer’s beliefs.
Practical Strategies
But the moral claims are a smaller part of the book than I expected from a philosopher.
The best parts of the book try to create a culture in which people who earn enough to be wealthy feel social pressure to (or gain social status from) donating some semi-standard amount to global poverty charities.
This approach avoids the problems associated with trying to expand moral rules to their breaking point, and better reflects what motivates philanthropists.
He describes a fairly good set of rules for donating at levels that would save many lives if a moderate fraction of people adopted them, and they mostly would feel comfortable about following those rules.
They would work roughly like a tithe, so the starting point is simple, and sort of a Schelling Point. But there’s a delicate balancing act between that simplicity and the expectation that the wealthy should give a larger proportion of their income/wealth. Singer maybe goes a little too far in the direction of IRS-like complexity, to the point where I have trouble remembering his exact recommendations.
These parts of the book may have influenced my donations, but any influence was indirect enough that it’s hard to say. I was mostly influenced by GiveWell’s claims that they had identified effective charities. I’m not aware of any connection between this book and me noticing GiveWell, but I noticed GiveWell less than a year after the book was published, so I wouldn’t be surprised if the book caused some of the attention that led me to notice GiveWell.
Singer notes that there seems to be a positive correlation between giving and happiness, and concludes that becoming happier is one more reason to give. My intuition says that the causality works mostly the other way – I don’t see much reason to think giving at roughly the level that Singer recommends has any effect on my happiness, but I have a moderately clear impression that becoming happier makes me more generous. I suspect the giving-causing-happiness effect is limited to situations where the donor directly interacts with the recipient. (Note: I haven’t studied this carefully, and I suspect the evidence is messy).
Comparative Altruism
Singer devotes a bit of effort to comparing wealthy nations.
Why?
Is he trying to persuade/shame the US into imitating Europe? If so, that has confusing implications for how we should expect that section to influence Europe. Whatever his reason, the comparison seems weak in contrast to comparisons to what people say they should give, or how much they think the U.S. government has given.
Singer is a bit misleading about how much wealthy nations give to poor parts of the world, as he omits remittances ($574 billion in 2016). Those are at least as important as the charity that Singer counts, although that doesn’t much affect his overall arguments.
Singer argues that wealthy nations have harmed poor nations. I’m a bit confused about why he mentions the topic. Does he want to use guilt as a general-purpose tool to motivate charity? Does he think that wealthy nations owe enough specific compensation to be worth prioritizing that? He doesn’t convince me that recent harms are large enough to justify anything resembling the kind of aid he wants. Older harm (mainly from the slave trade) seems larger, but I don’t feel much responsibility for injustices that old.
Conclusion
Singer tries to argue in ways that appeal to people with a variety of moral systems. But his moral claims are convincing only under some relatively unpopular moral systems. To the limited extent there’s a consensus about moral systems, it either rejects or is carefully evasive about Singer’s moral claims.
If I overlook the moralizing, and treat the book as a social advocacy argument, then it’s quite a nice book.
I don’t need to be very egalitarian to be upset that the lives of strangers on distant continents are valued around one thousandth that of strangers who are lucky enough to live in a country with wealthy philanthropists. A very ordinary commitment to egalitarianism should motivate us to reduce that difference.
Singer has helped to shift status gains away from low-value philanthropy, and toward higher impact charities that value lives on distant continents or further away.
[1] – I do see decent arguments for social/cultural agreements to change our moral systems to more closely approximate utilitarianism, but I expect that to always leave us valuing our own welfare more than we value the welfare of distant strangers.
Maybe someday everyone will agree to rewire their minds to adopt the one true utilitarian utility function (creating a Borg-like collective?), but I don’t feel obligated to support that.
See Alex Mennen’s Against the Linear Utility Hypothesis and the Leverage Penalty for hints as to why total utilitarianism is likely to conflict with observed human preferences in situations such as Pascal’s Mugging.
See also Artificial Utility Monsters as Effective Altruism. I endorse the existence of open source utility monsters, but don’t endorse moral systems that let them override my rights.
[2]
To be clear, I am implying that people who follow folk-morality tend to act as if, in situations where it is unclear who is responsible for doing something, they are not personally obligated to do the thing. I interpret this as implying a stronger desire for predictable moral judgements than the desire to save lives of strangers. The person’s true motivations are optimizing how their action will be judged by others who follow their morality, not the actual results of the actions.
If this is still unclear, then I recommend reading the DeScioli and Kurzban paper. I find the ideas in that paper important for thinking clearly about moral rules.
