Banerjee and Duflo’s Poor Economics

First things first: Two thumbs up for for Esther Duflo and Abhijit Banerjee’s Poor Economics. Read it. It’s a quick 260 pages that quite inexplicably manage to contain (1) the basics of the theory of development economics, (2) interesting recent research that complicates the basic theories, and (3) compelling and memorable stories taken from their actual interviews with actual very-poor people around the world.

The way the book proceeds, more or less, is to look at a problem of poverty — disease, malnutrition, under-education, lack of saving, and underinvestment —  and ask three kinds of theoretical questions about it: (1) is there a ‘poverty trap’?  (2) is there a demand problem or a supply problem? (3) why are the poor making the choices they are making with respect to this problem? Answering those three questions is essential to determining the right policy response to the problem. If we start with this theoretical approach, we can understand the rest a lot better.

(1) Poverty traps: Under ‘normal’ circumstances, economic theory suggests that the poorer you are, the easier it should be for you to get relatively richer. If wages are very low in your hometown, capital should flow in, driven by owners and investors who want to take advantage of cheap labor. And the poorer you are the more ‘low-hanging fruit’ is available to you in terms of using extra wages to invest in extra education that can raise your productivity, etc. This is the basis for the belief that ‘economic convergence’ should happen pretty quickly and smoothly in an open global economy. But obviously it hasn’t always happened smoothly everywhere, most notably for the ‘bottom billion.’ Why not? There could be a ‘poverty trap.’ That is, perhaps a lower-middle income country should be able to ‘catch up’ to a rich country very quickly, but a very poor country below a very low level of income could get stuck. Think about how this could work on a micro-level: If your income is so low that you can afford little food and are very small and weak, then you may not be capable even of a basic factory job. But getting the factory job may be the only way to raise your income so you can afford more food. So you’re stuck. Determining where and when ‘poverty traps’ exist is really important because it has implications for proper development policy: If there is a poverty trap, a large initial, charitable investment to get the poor beyond the point the trap point is essential; but if there isn’t actually a poverty trap then the right response is probably just to put in place the right institutions and then get out of the way of markets. (The belief that the bottom billion are in a ‘poverty trap’ is generally associated with Jeff Sachs; the alternative view generally associated with William Easterly.)

(2) Demand or Supply: The next theoretical problem to consider when approaching a poverty problem is “is there a supply problem, or a demand problem?” For example, we know that there’s a high correlation between a country’s income and its citizens’ educational attainment. It’s easy, therefore, to assume that low educational attainment causes poverty, and so we can fix the problem by building more schools and paying more teachers — i.e., the assumption is that there is a ‘supply’ problem, namely an undersupply of educational institutions. But this might be wrong: It could be that there are too few opportunities for educated workers in the country to make investing in education an attractive option for parents and students. If your country doesn’t have the ability to employ lawyers and computer scientists, then it might actually make sense to skip school and just get to work from an early age — i.e., there could be a lack of ‘demand’ for education in the country’s labor markets. Obviously, the question of whether it is better policy to (i.) airlift teachers and schools into a poor country, or (ii.) focus on improving other institutions so that, one day, the country could employ computer scientists, hinges on whether its low levels of education stem from undersupply of education or low demand for educated workers.

(3) The Rational Poor: The above feeds into the most novel aspect of Banerjee and Duflo’s work — their efforts to speak to actual poor people and understand their reasoning for the decisions that economists, several levels of abstraction away, would normally consider plainly irrational. For example: It’s clearly irrational to get pregnant and drop out of school, yes? And so, if someone does get pregnant as a teenager, we must attribute it to ignorance or lack of access to proper contraception, both of which could be fixed by more and better aid workers, yes? Well, yes and no. Certainly there are areas in which contraception and knowledge are undersupplied. But Banerjee and Duflo, in typically provocative fashion, isolate a few Kenyan towns in which this just can’t account for the teenage pregnancy rates. They conclude that what is actually happening is that the teenagers there are actively choosing to get pregnant because, as they would themselves claim in interviews, the prospect of “getting a man to take care of them” is more attractive than the prospect of “continuing to burden their families” for the sake of a very uncertain education payoff. Does this mean that it is in vain to supply more contraception or information? Certainly not. It just means that undersupply of aid is not the only problem, and we need to take the poor’s choices and decisions more seriously and think about how those are affected by the broader institutional structure and opportunities that await them. It means we need to stop simply conceiving of the poor simply as hapless uninformed victims.

Beyond those overarching theoretical questions and approaches, the book is largely just an assembly of a lot of the coolest latest research — including some amazingly clever experiments — in development economics. And so your correspondent presents various appreciations of the insights that stood out, in no particular order:

1. Banerjee and Duflo open the book with the book with the provocative argument that the global poor generally aren’t actually desperate for more food. This sounds horribly cruel. But Banerjee and Duflo propose a test for this hypothesis: If the global poor were very hungry, they should generally use windfalls in their income to consume more calories. But this just isn’t the case. In fact, most such windfalls are used to consume tastier goodies, which, counterintuitively, in many of the studies they highlight, leads to reduced caloric consumption. The takeaway of this not that the poor or okay or dumb, or anything like that. The takeaway is that we need to get beyond the mindset of just expressing sympathy through familiar phrases, and actually try to inhabit the desires and motivations of the poor. And many of these poor would actually prefer a $100 T.V. set that could help distract from the boredom of life in an economically depressed region, where work opportunities are few and far between, to $100 of extra food. This all ties back into theme (2) above.

2. A lot of the ‘wrong’ choices the poor make have to do with the fact that we humans are social animals. The poor don’t actually spend their days thinking about how to maximize their productivity and effect economic convergence. They care about status within their community, saving face, doing what their neighbors do, observing conventions.

3. (Extremely broad point): Truth is an extremely, extremely valuable economic asset. Basic, simple immunizations for basic diseases, available at charitable health clinics, could be an enormous boon to poor people. But very many of the global poor refuse to take advantage of this immunizations, because (i) they do not have the rudimentary scientific understanding to get how they work, and (ii) do not — often due to a history of political action in bad faith to, e.g., sterilize the poor — trust the public authorities to give them accurate information about the value of immunizations. We in the developed West sort of take for granted how much trust we have placed in public knowledge that we simply accept.

4. A major problem in educating the poor comes from a basic human cognitive flaw: availability bias. That is, the very poor are surrounded by very poor people like themselves, and they seem images of very rich people in magazines and on posters and T.V. As a result, they are inclined to look at their children as a kind of lottery. Each child is likely to end up very poor, but just might be extravagantly successful. This inclines them to “pick a winner” among their children, and devote all their resources to him. This belief is, of course, wrong: There are very high returns to marginal investments in education even at a low level, and so poor families as a whole would be better off if they invested in their children more evenly. Schools for the poor have this same false belief, and do the same wrong thing: focusing on only the very best students in the class. Ironically, a false belief like “education is a lottery” can itself create a poverty trap.

5. Another problem for global education today and a generically Interesting Thing about the Modern World is that, for the first time in history, we have a lot of professions other than teaching that are demanding the skills of smart, bookish, analytic types who are not well-connected (especially women). There is a very real possibility that the global pool of teachers may be getting worse on average, just in terms of raw intelligence, as a result.

6. Basic feminism is a really big deal. Your correspondent is always self-conscious about saying this, because he fears coming off as one of those guys who, ironically, tries to advertise his feminism in order to seduce women. But seriously: It’s just a plain statistical truth. Women are at least 50% of the basic human capital and capability of any society, and a society that dehumanizes women or delegitimizes their talents is not only being intrinsically horrible and cruel, it is also depriving itself of 50% of the human resources that can contribute to its flourishing. A particularly memorable example from the book showed how, in a natural random experiment, some poor Indian houses had both husband and wife’s name on the title, while others only had the husband’s name. The former households did better on all kinds of measures of flourishing for generations down the line, just because that one little paper gave the wife dignity that improved the household’s life in myriad ways.

7. The book has (inevitably) a long section on micro-finance. The debate over microfinance is so tired by now that there’s nothing to add here, other than to say that Baneerjee and Duflo’s take is really excellent and measured and balanced.

8. This book, like so many others, highlights some the huge potential of better communications technologies. Just one example: One of the main reasons that there is very little ‘microsavings,’ (i.e., the other side of banking for the poor) is the fixed transaction costs involved in managing very small savings accounts. So the very poor often can’t find anywhere to save, or must often pay to save, which, naturally, makes them less inclined to save at all. But with better communications technologies, we could drastically reduce these costs (i.e. if you can deposit on your mobile phone, the bank doesn’t need to put a branch and a teller near your rural village). Some smart policy changes, such as letting local shopkeepers take deposits on behalf of banks, if the banks trust them and delegate that responsibility, could give a save place for the poor to save.

9. A lot of corruption, the authors argue, can be fought relatively easily, just by making more information about, say, how much money has been given to a local government to build a road, available. A lot of the corruption we have in the world exists now because ireally nobody seems to care — like, care at all. This ties into the big takeaway of the final section of the book, with their political economy prescriptions: This section differentiates INSTITUTIONS and institutions. INSTITUTIONS are big abstract things like “the rule of law,” which, they admit, it’s really hard for external forces to quickly change. But we can change ‘institutions’ on a more micro, piece-meal level, such as forcing local governments to actually use cash for what it has been tendered by publishing that tender. And this is the right place to start.

10. In the final chapter, they offer 5 modest takeaways “in place of a sweeping conclusion”: (i.) The poor “lack critical pieces of information and believe things that are not true.” Accordingly, basic dispersion of basic medical knowledge is definitely some good low-hanging fruit. (ii.) The poor “bear responsibility for too many aspects of their lives.” I.e., we in the West don’t need to remind ourselves to splash chlorine in our water every morning, and the poor shouldn’t either — policies should make the chlorination automatic. (iii.) Sometimes there are “good reasons certain markets do not exist for the poor” — they just can’t be made profitable at such a low scale, including, arguably, health insurance. And that is where, very simply, we need continued charity and public intervention. (iv.) There are many really good, ground-level, very doable changes we can make by changing the micro-institutions within the broader institutional structure of underdeveloped societies — and that’s the place to start with fixing the political economy problems. (v.) Expectations become self-fulfilling prophecies.


One thought on “Banerjee and Duflo’s Poor Economics

  1. Having finished The End of Poverty between this post and your previous one, I was going to ask for something on developmental econ. You were on it already. I’m going to read this book. Any other recs?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s