Five Books Influential On How I Think About Global Poverty
Creating Capabilities, Seeing Like A State, The White Man's Burden, Poor Economics, and Why Nations Fail
I keep meaning to write about my broad approach to global poverty and near-term causes more generally, but it turns out this is kind of a broad topic and hard to write about. So as an IOU I'd like to provide a list of the books that most influenced how I think about effective altruism, specifically the global poverty cause area. Several of these books are not popularly read in effective altruist circles, so I hope my unusual recommendations will give you a couple of interesting books for your list.
These books all take different approaches to the same broad topic, and many of them disagree with each other. I have my own thoughts on how I combine them, but unfortunately that’s the really long blog post I’m not writing. I do think they’re all insightful and worth reading, and reading them together provides interesting nuance.
Creating Capabilities: Many effective altruists are utilitarians: that is, they believe that the ethical action is to maximize "utility," defined as happiness or pleasure or preference satisfaction or the absence of suffering. However, it is very difficult to measure utility. Our best studies produce counterintuitive results, such as that income only increases life satisfaction to the extent that you are richer than people around you. There are enormous theoretical issues. We cannot know that my 2/10 is the same as your 2/10, and there are solid reasons to think it's not. For example, a person who experiences extraordinarily painful cluster headaches might rank "doesn't have a headache" as a 3/10, no matter how miserable the rest of their experiences. I, conversely, have never had a cluster headache and don’t know to consider this possibility and so rate myself a 1/10.
The capabilitarian approach is a pluralistic approach to what matters. The capabilitarian approach is more fundamentally politically liberal than utilitarianism: it holds that each person should be allowed to choose their own life path as seems best to them, even if it is a life path which makes them unhappy or overall leads to more unsatisfied preferences or more suffering. In this way, it is humble. We don’t assume that we know better than other individuals what they should do.
However, people’s decisions about their life paths must be a meaningfully free choice. Mere negative freedom, or the absence of coercion, is not enough. It’s your own decision whether to fast, but it’s not a free choice unless there is affordable food for you to buy; it ‘s your decision whether to leave the house, but it’s not a free choice if your husband or father will beat you for leaving; it’s your decision whether to vote, but it’s not a free choice if you live in an autocracy without free and fair elections; it’s your decision which job to choose, but it’s not a free choice if you are illiterate because of a poor educational system and thus are shut out from many jobs.
Still, it seems that maximally free choice is not necessary. I am not able to choose to buy a megayacht, but few people would argue that my ability to choose my own life path as seems best to me is meaningfully limited by this impairment. Instead, we should protect a certain list of central capabilities particularly important for being able to choose your own life, which we choose through societal discussion and debate.
The list of central capabilities may seem rather arbitrary. Who are we to say that bodily autonomy or the right to vote or the ability to enjoy recreation matters? Nussbaum leaves the exact criteria of societal discussion frustratingly unclear. However, these problems are problems that also exist in utilitarianism.
It’s extraordinarily hard to conclusively prove that increased consumption or not being infected with worms or having the right to vote actually makes people happier. Indeed, the arguments are often made by an appeal to common sense: obviously, I would be happier if I were not infected with worms and had more than a dollar a day to live on. In utilitarian arguments, this arbitrariness is often hidden in the back, as if an embarrassing family heirloom one can’t get rid of but wouldn’t like to display for company. In capabilitarianism, conversely, the arbitrariness is presented right up front.
A pluralistic approach leads to many difficult tradeoffs. How many years of education is equivalent to not getting malaria? How much starvation is equivalent to experiencing female genital mutilation? How much contraceptive access is equivalent to having the right to free speech and freedom to practice your religion? Again, these are concerns that utilitarians also face. But I think having a specific list of what we do and don't care about makes these tradeoffs easier to conceptualize and perhaps approximate a reasonable answer to.
Seeing Like A State: The most important concept for effective altruists from James C. Scott's excellent Seeing Like A State is that of legibility.
To understand legibility, consider the difference, in the average household, between legal property rights and de facto property rights. Legally, each object has one owner, who is permitted to use it, sell it, loan it to others, or destroy it; property may also be held jointly by married couples. But de facto the situation is much more complicated.
An unmarried couple may consider the silverware, the tea kettle, and the fancy cheese to belong to them both equally. Technically, in a community-property state, your underwear belongs to your husband as much as it belongs to you. Even if legally you and your wife both own her deceased father's treasured record collection, you absolutely don’t have the right to destroy it and she may forbid you from listening to them if she’s concerned you will scratch them. A teenager may not legally own her books, but de facto she does own them. And so on and so forth.
Why are legal and de facto property rights different? The state has to decide things like "who gets the silverware in the divorce?" and "in what situations is spending money from someone else's bank account theft?", so it has to have some system for understanding who owns what. But at the same time every household has its own system of de facto property rights. For one household, the silverware is communal; for another, it's mine that I'm graciously letting you borrow. For one household, teenagers own nearly all their possessions; for another, everything they have is at the pleasure of their parents; for a third, only items the teenager buys with their own money are theirs. The state cannot go investigate each of a hundred million households' particular stances on silverware ownership. So it comes up with a general rule like "the silverware belongs to the person who bought it, unless you bought it while married, in which case it belongs 50% to each half of the married couple," which does not capture all the complexities of each individual household's system.
Legibility inherently removes information. A diagnosis of depression or migraine removes information about the specifics of an individual's disease. The categories "married," "single," "divorced," and "widowed" remove information about the exact details of your romantic life and what commitment means to the couple. Shelving books in genres like "science fiction" or "romance" simplifies the complexity of any particular story. Of course, legibility is particularly likely to be a Procrustean bed to those of us who are weird-- whether because of the science dragons in our books or because of our polygamous marriages-- but at least a little bit of information is lost for everyone.
It’s easy to assume that we shouldn’t attempt to be legible and should instead understand the particulars of each individual case. But large-scale knowledge of or change in the world requires legibility. You can experiment with drugs on a particular patient, but if you want to do a randomized controlled trial of an antidepressant you’ll have to study depression. You can encourage a particular person to stay in their unique primary relationship, but if you would like people in general to stay together-- or for that matter break up-- you’ll have to work with the imperfect categories of "married" and "divorced."
Effective altruism is an institution trying to make change in the world; for this reason, it can only study legible concepts. However, we must remain aware of the information lost through legibility. It’s important at all times to keep in mind what precisely we're studying, and the ways that it may diverge from what we care about.
It may be wise to stick to concepts where little information is lost, such as "malaria," and steer away from concepts like "subjective well-being" or "life satisfaction" that leave out a lot of things we would like to know. We should aim for robustly good interventions, such as treating diseases, which are beneficial even if we are mistaken about important aspects of how an individual society works. Approaches such as GiveDirectly's, which put power in the hands of the global poor, may compensate for our ignorance.
Legibility is a product of institutions trying to measure, understand, and affect individuals; in the present day, it’s mostly a product of the state trying to do this. Countries without a strong state, such as many developing countries, are illegible. If the government is corrupt, useless, or nonexistent, you have no reason to develop property rights that make sense to the government. Instead, all property rights can work in the way that property rights over the household silverware work for Americans: complex, contextualized and uniquely adapted for local conditions.
Property rights are just an example; many features of many developing countries work similarly. Therefore, in studying these societies, we’re particularly likely to lose information through legibility, and more caution and humility is required.
Because the American government is powerful, Americans are used to property rights working in a legible fashion, so property rights actually are legible in nearly all circumstances. It’s easy to assume that this legibility is just how property works, instead of being a culturally contingent feature of our local context. When thinking about global poverty, Americans should keep this fact in mind.
Given how little we know, should we be helping the global poor at all, instead of helping our friends and family where we know more and don’t have to rely on legible information which leaves out important considerations? I think yes. The global rich are tremendously more wealthy than the global poor, and so have more power to act in the world; we can help them a lot with relatively small amounts of money. The global rich also have some knowledge that the global poor do not; the global poor rarely read epidemiology and economics papers and so do not have an informed opinion on whether schistosomiasis treatment improves school attendance. Still, we must be aware of our ignorance.
The White Man's Burden: I originally read this book to become more informed about the aid skeptic opinion but discovered, to my shock, that I actually mostly agreed with it. Therefore, bafflingly, I am an aid skeptic who donates ten percent of my income to global poverty charities.
Today, most foreign aid is directed at the lofty objective of developing poor countries and ending global poverty. To do this, we set very ambitious goals: to achieve universal primary-school enrollment by 2015 (this did not happen); to achieve universal access to water and sanitation by 2015 (this did not happen either); to halt the spread of HIV/AIDS by 2015 (in an enormous surprise, this also did not happen). The universal failure of aid agencies to meet their very ambitious goals does not lead to rethinking and self-examination. Instead, we simply set new very ambitious goals.
This pathology is exactly what we would expect, because the very goal of the planners is impossible. Aid can’t end poverty. It would be nice if aid could end poverty. I myself certainly wish aid could end poverty. But what we have learned from decades of attempts is that it can't. In an ideal society, we’d notice this and switch to a different goal we can actually achieve. But, unfortunately, people would like to be able to say nice words about ending poverty forever which make them feel good about themselves.
There is no accountability on any level-- from individual aid workers to globe-spanning agencies-- because you cannot have accountability if you would like to delude yourself into thinking you are able to reach an impossible goal.
There’s also no accountability because big utopian plans are, by their very nature, anti-empirical. You can't exactly run a randomized controlled trial of achieving universal primary-school enrollment, much less of ending poverty forever.
A person sitting in a nice air-conditioned office sets the goal of ending global poverty, then decides that a necessary step towards ending poverty is enrolling everyone in primary school, and then decides that in order to do that we need to give all girls menstrual products, and then decides that to do that we should fund government programs to give girls menstrual products. At no point in this process do any of these decisions touch reality. It’s reminiscent of Soviet Five-Year Plans—the quixotic attempt by the USSR to centrally plan the entire Soviet economy—and has a similar rate of success.
The current approach to aid is also extraordinarily patronizing. When someone says "it's up to us to end poverty," the thing they mean is not "it is up to people in general to end poverty"; the thing they mean is "it is up to us wealthy white people to end poverty." The poor are passive recipients of our largesse.
However, the global poor are the best people to end poverty. Since they are the ones who suffer, they are motivated to make sure their efforts are directed towards ending poverty instead of attending large summits and making the global rich feel good about ourselves. And since they are more familiar with the situation on the ground, their plans are better-informed. Countries that successfully developed did not develop through deliberate planning on the part of wealthier countries; development was home-grown and bottom-up.
Of course, these are far from the only problems with utopian aid: for example, such aid tends to be stolen by corrupt governments. But I hope this is a sufficient sketch of the problems with the system.
Instead of vast utopian plans, foreign aid should be directed towards the things it is actually good at: vaccines and antibiotics, roads and fertilizer, textbooks and nurses, piecemeal civil-service reforms and macroeconomic advice for those leaders who are actually interested in hearing it. Each program is small and directed at a single goal, rather than The End Of All Poverty Forever. All these programs should have concrete goals, monitoring to ensure the goals are met, and accountability for failure to achieve their goals. These programs will not end poverty, but they will keep poor people from being sick and hungry and illiterate, and that also matters. An achievable low goal is far better than wasting money on a goal aid will never achieve.
Why Nations Fail: Why Nations Fail is about how some countries reach prosperity while others fail-- that is, when the global poor are trying to solve the problem of global poverty, what exactly are they doing? The difference, Why Nations Fail argues, is related to its economic and political institutions: the fundamental way that a country is organized. Some institutions are extractive, while others are inclusive; extractive institutions tend to lead to poverty, while inclusive institutions tend to lead to wealth. (This is the single review that contains the most of my own thoughts, and should be particularly not assumed to reflect Acemoglu's and Robinson's viewpoints on all issues.)
Extractive institutions enrich a small elite at the expense of everyday people. Of course, this behavior on the part of institutions is objectionable in itself. But it's also bad for society as a whole.
If you are a buggy whip manufacturer, you’re hardly going to be delighted by the development of cars. You’re likely to campaign to make cars illegal or tax them so heavily that no one can afford them, and if you have control of the government you may succeed.
Similarly, you don't want anyone to set up any competing buggy-whip factories, so you’ll try to get the government to create a monopoly so only your business can produce buggy whips. Then you can provide low-quality whips and terrible customer service at high prices. What are your clients going to do, leave?
You might also decide that this whole business where committing crimes is illegal is good for the lesser folk, who might commit crimes against you, but you personally ought to be allowed to commit as many crimes as you want. It's not like you will be hurt by all the crimes you’re doing constantly!
All of these dynamics-- and many others-- tend to leave a country poor. Throughout nearly all of history, all institutions were extractive, and as a result, throughout nearly all of history, all countries were very poor.
Conversely, inclusive institutions-- where many people have a say in institutions, which are more-or-less governed for the benefit of everyone-- tend to enrich a country. If everyone has say in the government, they will tend to say that the same laws ought to apply to everyone (the rule of law), which means that everyone is secure in their people and property and the elite cannot carve out special exceptions to enrich themselves at the expense of everyone else.
It's easy to assume that "inclusive institutions"is a synonym of "democracy," but in fact democracy is necessary but not sufficient. Protection of property rights, for example, is an important part of inclusive institutions. The economy is also an institution, which many people should have a say in. If property rights are respected, then each person can choose what to do with their own property, and thus gets a say in the economy. They’ll reward businesses that work well and punish businesses that work poorly; individuals can start enterprises secure in the knowledge that if they're successful they will get a reward. (Similarly, some level of redistribution is required for inclusive institutions-- if there is a high level of income inequality, the poor won't have much say in how the economy works.)
Protection of individual rights is also necessary for inclusive institutions. If individual rights are not protected, it is much easier for the elite to run rampant over the powerless. For example, if people don't have civil liberties related to the criminal justice system, it’s easier to frame the poor for crimes. The media and the free marketplace of ideas offer a check on elite power, so freedom of the press and freedom of speech are necessary to keep elites from dominating society.
(This provides a grounding for some of the essential capabilities discussed by Nussbaum: society must provide individuals with the rights necessary for inclusive institutions.)
Therefore, to end global poverty, societies should develop more inclusive institutions. Unfortunately, switching from extractive institutions to inclusive institutions is very difficult. In general, inclusive institutions tend to stay inclusive, and extractive institutions tend to stay extractive.
It’s easy to assume that we, the global rich, should force good institutions upon the global poor: this is very common in the ambitious utopian plans the White Man’s Burden discusses. But inclusive institutions cannot be imposed from above: if people from outside the country force extractive institutions to e.g. hold elections, the elites will just steal the elections. Instead, the populace must demand inclusive institutions-- often at great personal cost. Exceptional individuals such as Seretse Khama may be able to guide their countries towards inclusivity.
Poor Economics: Poor Economics is a fascinating book and an important review of the randomista research on development. The engaging and informative descriptions of various important development economics studies would be more than enough to cause me to recommend it.
But more important is the understanding that the global poor are, fundamentally, no different than the global rich. Many bad decisions made by the global poor are made out of simple ignorance. They don't know about the benefits of vaccination; they think learning to read doesn't matter if you don't get a high school degree; they are confused about which sex acts are at highest risk of HIV. If you thought that vaccination wouldn't protect your children from measles, you wouldn't vaccinate either (...as we have unfortunately learned in wealthy countries).
The global rich also have many decisions made for us. The water that comes out of my tap is purified; I’d have to go out of my way to drink non-purified water. My food is fortified with many of the nutrients I need. If I had to purify my own water, I would often forget and get sick; if I were responsible for managing my own iodine consumption, it would go on the "I fully intend to deal with this someday" list next to cooking dinner from scratch. When the global poor make decisions that seem dumb or thoughtless, they're very often decisions that we would make in their shoes. This perspective builds empathy. The global poor are no different from you or me.
Poor Economics also provides concrete examples of the sort of change-on-the-margins that the White Man's Burden discusses. It’s prohibitively difficult to economically develop a country. But we can provide malaria nets, inform voters about how well their politicians are doing, and monitor whether teachers actually show up to work-- and these small changes provide concrete improvements to the lives of the global poor. And as people begin to expect that teachers will show up to work and that they will be informed of politicians' actions, these changes can be self-perpetuating—creating a long-term and concrete improvement.