Today UKaid was distributing sacks of maize to needy people at the Mpyupyu trading center on the northern edge of TA Mwambo. As I understand it, this is a 6-month program to alleviate hunger and maize shortages. Relative to inaction, this is a really great idea – by all reports the harvest was awful here last year and maize prices are at near-record highs for this time of year. The rains were poor last year and there’s no reason they’ll necessarily be much better this time. Even if they are, the next harvest is a long ways away.

My question is whether this is really the best option. Most of the research I’ve seen on food crises and famines points to giving cash as the superior option, for two reasons. First, with cash people can buy what they really need instead of taking whatever you hand them. Staple foods seem like an obvious choice, but who are we to decide that for people? Even if we had awesome, timely data on people’s needs, it’s hard to beat people’s own choices.

The other, more compelling reason is that nothing I’ve seen indicates there is any shortage of maize, either in Malawi as a whole or in this area. I can buy more nsima than I know what to do with at prices that aren’t much different from last year. Rather, what’s going on right now is limited purchasing power on the part of many farmers in Mwambo. The story I’ve heard is that they commonly sell a lot of their crop right after the harvest in order to meet pressing needs, and then need to buy maize later in order to eat. At that later date, the supply is lower and prices go up, making it even harder to afford food out of their limited remaining cash holdings. This pattern would be strongly exacerbated by a drought, when owners of small plots might start out with too little to make it through the year. On the flip side, that maize was being sold to someone, and in many cases constitutes a big chunk of people’s liquid wealth.

Imagine you’re one of the larger farmers in this region, and you have your own silo. You hold on to your maize, and maybe even buy more as an investment since prices tend to rise over the interim period between harvests. Even though the rains were bad, you’re doing okay, and might turn a profit this year once you pay back the loans you took out to buy extra fertilizer from ADMARC. Suddenly, however, UKaid shows up and starts just giving maize away. Now the market value of your maize is close to zero. You can still eat it – you’re not starving, not right away. But you can’t pay off your debts or invest in the next harvest. The next year, you’re going to be one of the people lining up for free food. This innocuous act of charity just made you poorer and caused a bunch of problems down the road. Not problems as big as the one they were trying to fight – the prospect of mass starvation – but problems nonetheless.*

Cash handouts do have their downsides. I mentioned that possibility to one of my enumerators and he said that you can’t do that because “there are too many drunkards.” People squandering the money on stuff like booze is a real danger. There’s a growing literature on the importance of who gets the money for household outcomes. The whole idea of targeting women for microloans hasn’t generated the amazing benefits we thought it would, but giving them the cash transfers may be a way to limit how much of the money gets blown on alcohol and transactional sex. We have the research and the knowledge; I think we can do better than handing out staple foods for free.

This argument is essentially just a paraphrase of Sen, for any development economics nerds playing along at home

My wordpress account was just hijacked by by some kind of spambot. I have deleted that post but if you received it as an email please do not click the links or images in it. Kudos to my mom for pointing this out as soon as she saw it.

I’m not 100% sure what I did that allowed this to happen; I had something similar happen to my seldom-used twitter account a while back. As a precaution I’ve changed my wordpress password, but given that this happened today I suspect I may have picked up some malware during an ill-advised but unavoidable interaction with an infected local USB thumb drive at my photocopier this morning. He had lost my originals and time was of the essence in getting replacements printed; there was no time to find a different flash drive. I thought I had made it out scot-free but apparently I might have taken one on the chin.

The spam message itself was something about getting paid to take surveys. This is somewhat ironic as I’ve complained before on this very blog that American researchers almost never pay their survey respondents, and that they really should. Suffice it to say that you won’t make any money from taking surveys, and that that link is highly likely to do nasty things to your computer.

I apologize for letting this happen, and hope no one was duped into clicking anything in the post.

“Ati” is a Chichewa word that literally means “which”. Technically the prefix has to change according to the noun class in question. I am absolutely horrible at noun classes used by Bantu languages, which are akin to the gender of nouns in Romance languages except that there are way more, and they somehow seem to make even less sense.

But it’s also something I hear people say out in Mwambo Traditional Authority all the time in contexts where it makes no sense to be asking “which” about anything. Specifically, about half the time if I say something they repeat it back to me, prefaced by “ati”. If I tell kids “osakoma galimoto” (don’t hit the car*) then at least one almost invariably says “Ati osakoma galimoto!” (a direct translation would be
“which don’t hit the car!”) in an amused and fairly incredulous tone of voice. I get it from adults as well.

Stephen Paas’s excellent Chichewa-English dictionary is of surprisingly little help here. It lists the definitions of “ati” as “ati? 1.which?; anyamata ati? = which boys?; 2.isn’t that right?;”. “Isn’t that right don’t hit the car!” doesn’t make much sense either. So I started running this enigma by my native-Chewa-speaker friends. Their explanation, as I understand it, is that the people around here speak the Chinyanja dialect rather than Chichewa, and this is a local slang term meaning roughly “he/she says”.

So my sense of what’s going on here is that if I try to tell folks that, for example, I’m busy working in their native tongue, they are mostly blown away that this crazy white guy knows how to say stuff that they understand. “He says he’s working!” they yell to their friends, and the subtext is that there’s an mzungu that can actually talk, and that that’s the funniest thing that’s happened all week.

* I don’t like being a jerk to the kids, who honestly just want something fun to do, but I have to be fairly diligent in keeping them from screwing around near the vehicle. They haven’t managed to do much to actually mess it up yet, but since they love sneaking up behind it to mess with the back bumper where I can’t see them, they can get badly hurt. A couple of days ago a group of kids we had chased off came back as we were pulling away, and one got his hand caught on the bus – nothing serious, but a closer call than I’m comfortable with.

I had a conversation about designing experiments a while ago that I thought was worth putting up here, in case some other stressed-out development economist Googles this same issue. A friend of mine, who I’ll leave anonymous, was worried that people would find something to critique about their experimental design. Rather than using a random number generator, their design assigned respondents to different experimental arms using an alternating pattern. Is this okay? The worry is that someone might notice it during a seminar and use it to make them look bad.

The answer is that it’s absolutely fine unless someone has rigged the order of your respondents in order to try to thwart you. Formally, let’s suppose there are two treatments, so T is equal to 1 or 0, and the outcome variable is income, I. You want to run regressions of the form I = a + bT + e, where e is the so-called omitted variable that captures all other reasons I varies besides T. You will consistently estimate b unless T is correlated with e. You’re alternating between T = 1 and T = 0, so you’ll only get into trouble if for some reason your respondents are sorted in such a way that they themselves alternate between those that are ceteris paribus rich and poor. Note that clusters of rich and poor people are fine – it’s only an issue if a large portion of the stack of surveys or whatever are exactly flipping back and forth between rich and poor. Real life doesn’t work that way – if you collected data in an actual village, it would have streaks of rich and poor households but not precise alternation.

We can go a step further: a lot of experiments don’t control T directly, but rather affect some other variable V that in turn drives T (which we tend to call an instrument or instrumental variable). The key assumption for using V to study the effect of T on I is that V is “excludable” – that it affects I only through its impact on T. Equivalently, V must be uncorrelated with e. If V itself is binary, and assigned in an alternating pattern, it’s easy to see that we’re fine – there’s no reason to think that the unobserved factors that influence I would be alternating in a fixed pattern.

But why take the word of some guy with a blog? An alternate source is the methodology section of arguably my favorite paper, Kremer and Miguel’s “Worms”, which helped kick off the current RCT-centric paradigm in development economics. Ironically, although it launched a thousand randomization ships, the worms study does not itself employ randomization:

The schools were stratified by administrative subunit (zone) and by their involvement in other nongovernmental assistance programs, and were then listed alphabetically and every third
school was assigned to a given project group.

There is no serious question as to whether this approach is valid; any issues would imply that there is some other reason why every third school, alphabetically, experienced a decline in intestinal helminths and a rise in school attendance. The goal of experiments is to achieve excludability (or exogeneity in ITT specifications), not randomization per se. We often use random assignment to achieve that goal – but that’s not the only way to do it, and other approaches can work just fine.

Chris Blattman links to a piece by Joe Nocera that criticizes the tyranny of U.S. News college rankings. He promotes an alternative ranking by the Washington Monthly that emphasizes over-performance in graduating students and encouraging graduates to give back to society. Fighting the U.S. News rankings is a noble endeavor – they’re generally oversimplified, game-able and wrong – so I’m inclined to like anything that attacks them. But then Nocera wraps his argument up in the standard conventional wisdom about higher education getting too expensive:

Those who don’t land a prestigious admission feel like failures. Those who do but lack the means often wind up taking on onerous debt — a burden that can last a lifetime.

I’ve pointed out before that this is nonsense: the sticker price of college has risen rapidly but actual prices paid are falling over the last few years, due to an increase in financial aid. Moreover, the gap between actual and sticker prices is biggest at “prestigious admission” schools. When I was admitted to the Leland Stanford Junior University (organized 1891!) my actual cost of attendance was substantially lower than the out-of-state tuition I would have paid at its rival across the bay. It was Berkeley, and not Stanford, that I could not afford to attend. That discrepancy has only grown since 2002 – most ultra-wealthy schools at the top of the college rankings now don’t require students from the poorest families to pay anything at all. Stanford is a follower, not a leader, on this massive level of financial aid; Princeton has been doing it for nearly a decade, and even prior to that financial aid was very generous at all these institutions. So I don’t believe this quote one bit, unless it was based on fairly ancient history:

The author, who was not part of the cheating scandal, had succeeded in getting into a “Desirable University,” as she put it, but her parents had been unable to afford the tuition. She wound up, deeply embittered, at a state school.

Now, not being able to pay the price of your child’s university attendance and not wanting to do so are different animals. It’s entirely possible that the latter was the case. Just because a (fairly wealthy) family can afford the high sticker price of a rich private school doesn’t mean that they are willing to do so. But emphasizing the “elite schools are too expensive” story is harmful to low-income students, who may be discouraged from applying to schools that are cheaper than the alternative. Washington Monthly rates UCSD highly, partly on the back of cost-effectiveness; might poor students from other states assume that it is cheaper (for them) than Stanford? That would be a shame, and an expensive one. For sufficiently poor applicants Stanford is free, and it’s pretty hard to beat free in a cost-effectiveness comparison.

As a side note, my alma mater came in third in the Washington Monthly rankings. Given the emphasis on post-graduation public service, this isn’t a huge surprise, but we may have been underrated: one component of that ranking is ROTC participation, and (originally due to “Don’t Ask Don’t Tell”) Stanford doesn’t allow ROTC activities on campus. It also doesn’t include, for example, becoming a development economist or other life-long careers in public service. What about teaching, for example?

In the latter part of the Mutharika administration, Malawi was plagued by increasingly frequent fuel shortages. Joyce Banda’s decision to devalue the currency (and allow fuel prices to rise) solved that problem. Until this week. This was the view from my project minibus tonight:

I was waiting in Zomba in a queue that went along M3 from the Engen station all the way to the robot (which, and I am not making this up, is the local term for the stoplight) and down to the taxi stand. Not the longest fuel queue I’ve ever seen, but it had the feel of summer 2011. So did the actual pumping of fuel: they refused to sell me more than 5000 Kwacha worth (about twenty bucks), and tons of people were filling jerrycans. Based on some people I’ve talked to, I have reason to believe this renewed fuel crisis will be resolved shortly, but for now it’s pretty terrible. It can’t be good for Banda, either – a lot of her popularity derived from her resolution of the fuel and foreign exchange crises that beset the country prior to her inauguration. Even before this, I was increasingly hearing nostalgia for Bingu wa Mutharika, which seemed inconceivable just a few months ago. If fuel remains in short supply that sentiment will only grow.

The current best practice in development is to conduct rigorous evaluations to determine how well different interventions work, and to pick the ones that are the most cost-effective. It is entirely right and good that we should do this; without question, it beats the old method, still common in some circles, of just assuming that an approach is a good one and running with it. I identify strongly with this movement, which can loosely be described as the RCT approach (or the “randomista” camp) and in no way want to see it fail.

If anything, it still has a ways to go in winning over the broader development community. To pick one example, many people are convinced that a lack of access to sanitary products – things like menstrual pads/cups or tampons – are a major limiting factor in girls’ school attendance and success in poor countries, and so providing these things will help girls stay in school and excel academically. My advisor thought that might be the case as well, and so in this paper with Emily Oster she tried it out in Nepal, to see whether that approach works. The answer is a fairly resounding “no” – the girls liked the menstrual cups and used them, and benefited in other ways, but school attendance was unaffected. That has not stopped tons of NGOs from still focusing on this intervention (and they still insist it will help with school attendance). The impact evaluation movement still has much work to do, especially in publicizing its results – it’s pretty frustrating to see people cling to approaches that don’t work.

But I look at most of the projects people evaluate, including the ones I work on myself, and my perception is that they won’t really do what we want them to. What we want, really, is “development” – an improvement of people’s lives. Sen notwithstanding, there is now a strong consensus that that means reducing poverty. It is very hard, if not impossible, for people living on less than dollar per day to improve their lives in the ways that really matter, like health, education, and improving their living conditions. So people need more money, that is, larger incomes.

Most of the poor people in question are subsistence farmers in rural parts of the tropics, especially Africa. That means that anything that really promotes economic development – with all its attendant benefits in health, education and happiness – needs to either 1) increase the income people get from farming or 2) involve them substituting from farming into something else. Those are the only options, plus some convex combination of the two. If neither occurs, incomes cannot rise.

Pure logic has taken us this far, but with a little practical knowledge we can go further. First, the only real way to get more income from farming in a place like Malawi is to improve the productivity of a given patch of land. Most of the arable land is already in use. So we’re talking about increasing per-acre productivity here, and nothing else. Second, in rural Africa farming is pretty much the only game in town. Around this area some people fish, but in most places even that is rare, and in any case probably can’t be increased sustainably; people already catch a ton of fish in the local lake. Not farming, then, means going someplace else.

The two options for improving livelihoods for most of the world’s poor are therefore either improved agricultural productivity or people moving away from the countryside, i.e. to cities. If that sounds familiar, it probably should: every industrialized nation that I can think of has follow this exact path. Many people migrate to cities, while a few take advantage of massively improved productivity on farms to grow tons of food with few workers. Everyone gets much richer. The transition from rural to urban can be ugly and painful, but we forget how ugly and painful the status quo is: subsistence farming is hard, dangerous, and not really getting much better.

It’s hard to see how most of the projects people work on or evaluate could possibly help people either produce more from the soil or get out of the countryside. Raising incomes isn’t the be-all of development; it’s worth helping people even if we can’t permanently solve their problem. But we should look harder at ways to improve farm productivity, and, much more important, help people move to places where there are other jobs.

In Southern Malawi, mangoes and drought march arm in arm. I’ve been fairly obsessed with mango trees ever since starting data collection for my dissertation: my minibus becomes an oven without shade, making it uncomfortable to work in there and possibly overheating and destroying my laptop. Moreover, as a card-carrying white dude, there are limits to how much sun I can take in a place with some of the highest UV levels in the world. Mango trees provide the only decent shade in TA Mwambo outside of man-made structures, and by decent I mean crazy awesome. In a place where almost all the plants are dead, you can still find massive mango trees with full, green leaves. This is because of their huge taproots that dig through the dry, sandy soil for up to twenty feet in search of groundwater: many of these trees are as deep as they are tall.

A few days ago, some of my staff started saying that a drought was coming because there are too many mangoes on the trees.* I wasn’t sure what to make of this – an old wives’ tale, like the idea that pets can predict earthquakes? As it turns out, I had misunderstood – they meant a drought in terms of a falling water table, driven by limited rains in the mountains that feed the rivers here, and by poor rainfall last year. Apparently this causes mango trees to bear more fruit. Now, I have no idea why this would be the case, or even whether it’s true – Google turned up nothing definitive, and I’m no botanist. But it’s definitely true that the trees are covered in tiny, green fruit that bang on the roof of my bus when I’m trying to escape from the sun. I’m inclined to believe them, just because Malawi’s rainfall is so variable that it would pay dividends to understand water here.

The other side of poor rainfall and lots of mangoes is that I’ve started to see, in the words of one of my employees, “kids carrying a shirt full of mangoes, holding up the bottom of their jerseys”. Think about the shirt-carry that I’m sure we all did lots as kids. The kids are taking these green, unripe mangoes home, where they will be boiled and eaten for lunch. A sign of hunger – we are far from the last harvest, which wasn’t a good one. I don’t really think of boiled unripe mangoes as being a food, but you learn something new every day. Now I see kids with mangoes everywhere, even eating them raw, and am reminded of the desperation of the fruit pickers toward the end of The Grapes of Wrath.

*I’ve noticed during this trip that even Malawians who grew up and live in cities still have a broad range of expertise in many agriculture topics. I chalk this up to very limited specialization – people are good at everything but rarely great at any one thing, whereas in America we tend to be awful at nearly everything except our own professions.

Being out in the field has limited my access to newspapers, which is particularly unfortunate in Malawi, where a lot of information still flows through the print media rather than online. I’ve argued previously that Malawi’s newspapers are quite good in many ways: they are concise and seem to me to put more emphasis on  But all media has trouble interfacing with foreign cultures – indeed, America’s usually doesn’t even try, focusing instead on the exciting, violent story of the week – and Malawi’s is no exception. American culture can be particularly baffling, especially our weird sports. I spotted the following gem in the “kids activity section” of last weekend’s edition of The Nation:

 

Which ball is used in American soccer? A soccer ball, of course. The listed answer, number 3, is used in American (a.k.a. gridiron) football, which no one calls soccer. The confusion here is that the author is aware we have our own crazy sport we call “football”, and so, he presumes, we must also call it soccer since that’s another word for football.

The following article is even better, though – tons of Malawians wear second-hand American clothes, and I often wonder if they know what the logos mean. The answer is generally no, but the author of this op-ed takes hypothesizing about the origins of a shirt to an extreme. He spotted a man wearing a “VOTE FOR PEDRO” t-shirt, which became popular in reference to a shirt worn by the title character of Napoleon Dynamite to support his friend’s student government bid.

Semphere supposes that Pedro is a Latin American politician, and uses the anecdote to launch into a discussion of the effects of media on youth, and from there into being a good influence on your child. I’m still holding out the possibility that this is a tongue-in-cheek joke – he does discuss movies for a while – but honestly it’s pretty hilarious either way. On the one hand, Americans are similarly clueless about other cultures (cf. the plethora of half-understood Chinese character tattoos), but on the other, doesn’t The Nation have internet access? How about a Google search?

Anyone with even a passing interest in development has a litany of reasons that diamonds, and especially diamond engagement rings, are downright awful. The most famous of these is the possibility that you may be buying a “blood diamond”, a term which was neutered into the euphemistic “conflict diamond” by the gem industry’s successful PR push on the topic totally selfless and completely ingenuous Kimberley Process scheme.

A related but more subtle point is that the Kimberley Process – which certifies certain gems as “conflict-free” – cannot, in principle, wash the notional blood off of your engagement ring. The problem is that blood diamonds and regular diamonds are near-perfect substitutes for most purposes. Deciding to purchase a Kimberley-certified diamond will tend to raise the price of blood diamonds as well, and hence to incentivize their production.

Another problem with diamonds, as with any easily-capturable natural resource, is that the diamond trade can promote the “resource curse”, wherein a reliance on particular kinds of export can induce corruption and undermine political institutions. And of course there’s the problem of engagement-rings-as-arms-race: to some extent a diamond ring is just a positional good, whose value is determined by its size relative to those of your peers (for women) or peers’ fiancees (for men). In that case we can burn up tons of resources for no gain at all, since everyone else is trying to win the same competition.

But the simplest reason to disdain the diamond engagement ring is that, frankly, it’s kind of silly. We (and by “we” I mean “other Americans besides myself”) pay exorbitant amounts of money for shiny, fairly fragile rocks with no resale value. I had had a passing familiarity with the history of this odd cultural institution, but didn’t know the full story until I came across this thirty-year-old piece from The Atlantic that describes how De Beers invented the concept of the diamond engagement ring, and then used a massive marketing campaign to indoctrinate generations of Americans in the belief that the giving of diamonds is the only legitimate way of proposing marriage. Here’s one of the more telling portions:

In 1951, N. W. Ayer found some resistance to its million-dollar publicity blitz. It noted in its annual strategy review:

The millions of brides and brides-to-be are subjected to at least two important pressures that work against the diamond engagement ring. Among the more prosperous, there is the sophisticated urge to be different as a means of being smart…. the lower-income groups would like to show more for the money than they can find in the diamond they can afford…

To remedy these problems, the advertising agency argued, “It is essential that these pressures be met by the constant publicity to show that only the diamond is everywhere accepted and recognized as the symbol of betrothal.”

The whole thing is interesting – De Beers also worked to shape demand and supply in other nefarious ways. But it is the invention of a preference for diamond engagement rings that I find the most disturbing. Economists tend to take preferences as given – de gustibus non est disputandum – and assume that any distribution of goods that optimally satisfies existing preferences is efficient. The example of diamond engagement rings exposes a deep flaw in that unspoken central principle of economics: what if our preferences are bad? What if powerful companies use their vast resources to invent a preference for somethig that objectively looks like a stupid waste of money and time? What if we blow billions of dollars annually on symbolic crystals to win arms races within our social circles, inadvertently leading to brutal killings and the undermining of foreign governments? Maybe it’s time to rethink whether it’s okay to judge preferences on their merits.