Goodbye to a gratuitous exercise in understanding global poverty

It's been emotional. Credit: Tumblr

I've exited my MOOC (massive open online course) adventure via the exit survey and can now claim to be an MIT alumnus according to my pdf certificate (classy). Time for some reflections on 'The Challenges of Global Poverty'. I'll split this into two: first, comments on the MOOC itself for those who might be interested in doing one/it, then some nuggets on content. Feel free to skip one or the other section, or indeed both.

The MOOC itself was a beauty. 14 weeks, each consisting of a couple of video lectures, split into bitesize chunks with (mostly multiple choice) questions after each to check you were following. It was perfect for that time after lunch when you don't want to do anything productive but feel you kind of should. Then each week there was a (again mostly multiple choice) homework which sometimes required dipping into the 'further reading' section, which was basically just Poor Economics so not too taxing. Overall perhaps 5 hours a week. Finally on top of that, a final exam (you've guessed, multiple choice) and a 'project' or 600-word essay on a topic of your choice, which was self-marked (I nobly docked myself one mark out of twelve). They got the level of questioning pretty much bang on in that it wasn't so hard that you had to continually go back to the videos, but generally you couldn't answer the questions without having watched the videos at least. I found it verged on the easy side but that's probably also a function of the fact that I'm a native English speaker, doing a related degree, who had already read the book which was the foundation of the course. I could have chosen a harder MOOC.

In terms of the MOOC experience more generally, it was perfect for what I wanted: a way to channel my slightly workoholic desires towards a topic of interest which would, unlike the course I'm taking simultaneously on international development at university here in Madrid, leave me with more knowledge than when I started. The baby homeworks and multiple-choice exercises keep you hooked with enough momentum to reach the end of the course and there's even a little section with percentages and bar charts to track your 'progress' and boost your ego as you nail yet another self-graded question. I really didn't engage much with the 'discussion forum' or 'community' sections, although they were pushed a fair amount and this podcast suggests that's where MOOCs are going, as they attempt to become more like 'real' unviersity, social interacion and all. I remain unconvinced.

Abhijit Banerjee speaks really slowly so I had to double the speed (literally) on the video to avoid boredom. Content was good though. Credit: The Guardian

As for the 'Challenges of Global Poverty' themselves, they are, predictably, many: from education to entrepreneurship, food to family planning. But in essence, this was a course about poverty from the point of view of social policy, where the basic message is that you need to pay attention to the details. The world's poor often don't obey the theories of policymakers, unfortunately, due to their being normal human beings. In the course and in their book, then, Banerjee and Duflo (B&D) try to get to the micro level, using experimental studies to identify why people act in certain ways at what that means for policy.

As a final farewell, I'll note their counter-argument to those like Bill Easterly who say that randomistas like B&D, in focusing on micro-level policy questions miss the bigger picture (seamlessly taking up a theme from an earlier post - see argument #6). If, as many profess, Institutions like Democracy and Property Rights (capitalisation explained later) are what ultimately makes USA better off than Chad, then this tinkering around the edges is surely just re-arranging deckchairs? Well no. Because even if you can't precipitate a social revolution and re-engineer a country's social fabric (engineering fabric?), institutional change (small-i) is possible even when Institutions (big-I) are bad. B&D cite a study that found improvements in outcomes when limited levels of democracy are introduced in village councils in China (but wait, China's communist?), and likewise mention effective health and education policies in 'bad institutional environments' like Indonesia in the 1970s. If you're lucky, policies can even sometimes change politics. So although the 'big questions' addressed by the development high priests like Sachs, Easterly, Acemoglu and Robinson are often fun to think about, it's at the ground level where the important stuff often happens. B&D argue that the question should be “where can we make a small push and see if we can make a change?” After completing their MOOC, I'm inclined to agree.

Are the extreme poor hungry?

An obvious question maybe. The WFP reports that there are 842 million undernourished people in the world today. Policymakers, too, seem to think so: at Davos last month global leaders committed to the zero hunger challenge. MDG 1 refers to 'eradicating extreme poverty and hunger'. Ayuda en Acción also have a cool campaign on hunger right now which reminds me of Where the Wild Things Are. So the extreme poor are hungry, right?

But are the extreme poor hungry?

This question is the the central problem posed by Banerjee & Duflo (B&D) in Chapter 2 of Poor Economics, and of the second week of the 'The Challenges of Global Poverty' MOOC of theirs I'm currently taking. So, inspired by Development Intern's weekly report on Sachs' rival course on Sustainable Development, I'm going to do something similar, hopefully mashing up course material with ideas from their book and other sources sufficiently to sidestep anticipated legal issues (it's a free course available to anyone so not sure 'revealing course content' has any meaning but whatever I'm a coward).

Getting back to the question, B&D suggest that we could be looking at a food-based 'poverty trap'. I.e. poor people don't have enough food to be productive, meaning they earn less and have less money to buy food, and so the cycle continues. Meanwhile someone with more money can buy more food, be more productive, and thus escape from poverty. In the class B&D use concepts like the S-shape graph and capacity curves and elasticity so you feel like a real economist: it's fun.

But if we actually look at the actual behaviour of the poor in developing countries, we get a story different to the WFP and poverty trap narrative. B&D find that in Udaipur, the poor could spend up to 30% more on food than they actually do by cutting expense on festivals, alcohol and tobacco. Another study cited by B&D found that a 1% increase in income only translates into a 0.67% increase in food expenditure. Significance: there's room for the poor to be buying more food. A further study in China even found that when the price of rice declined, people ate less rice (that is to say, fuck you conventional economics). In general, Deaton and Dreze find that poor Indians have been eating less and less over the last few decades while the percentage of people complaining that they do not have enough food dropped from 17% in 1983 to 2% in 2004. Not what you would expect from hungry people stuck in a food-based poverty trap. So maybe poor people in general have enough food. As B&D put it, “most adults, even the very poor, are outside of the nutrition poverty trap zone”. We can offer some explanations - technological change means that people aren't doing as much physical labour and so require less food; people are healthier so lose fewer calories to hungry parasites or to fighting illness.

But the Freakonomics-style journey doesn't end there, because it's not just that the WFP is totally wrong and everyone is fine food-wise. Roughly half the children under five in India are stunted, a horrific stat which suggests something is going seriously wrong with their nutrition. So, skipping a few steps, we arrive at our 'solution': it's quality not quantity. When 40% of pregnant women are anaemic according to the WHO, it doesn't matter how much rice you give them, they need iron. And this is particularly bad in India because many people are vegetarian and the staple is rice (which actually inhibits iron absorption) creating a perfect, iron-free, storm. The results of addressing nutrient deficiencies are huge - B&D cite a study which found that deworming children for two years (equivalent of giving them the nutrients and calories taken by the worms - and yes it does work like that) led to a yearly increase in income of 20% and a lifetime gain of $3269, compared to a cost of $1.36 per deworming treatment. Bam.

Sweet potatoes: yummy and nutritious

So we need to get kids dewormed (already happening in Kenya), nutrients to pregnant women and to young children, because that's where the real 'hidden hunger' is, say B&D. Another nail in the coffin of USAID's food aid policy based on getting US-produced staples to supposedly needy countries (not that it needed another one). And bad news too for the Campaign for Boring Development with its focus on increasing incomes: money isn't the issue here, so having more of it won't make a big difference. Fortified fish-sauce is already cheap and within reach of the poor in Indonesia (relative to its benefits and average income). Instead we need to directly target mothers-to-be during pregnancy with fortified foods, perhaps sprinkle children's food at school with micronutrient-rich salts, and more sustainably, develop nutritious foods that people actually want to eat (see DFID bloggers' on exactly this topic.
Orange fleshed sweet potatoes and iron-rich beans are where we're heading apparently).

The most interesting part of this I've left until last. If the benefits are so great and the costs so small, why aren't poor people doing this anyway? And this is where B&D are at their best, taking us away from high-level, often introspective debates driven by Western academics (whatever its merits, I still find the fact that Nina Munks just wrote a book on Jeff Sachs kind of surreal) and towards the realities of the poor. As they drily put it, “the natural place to start to unravel the mystery is to assume that the poor must know what they are doing”. Unpacking all the reasons why the poor might not take to foreign experts' advice would, and probably will, take another blog post (think mistrust of the state and NGOs, difficulty of verifying supposed benefits, human psychology). But one aspect of the explanation becomes obvious when you think of the extreme poor as normal people. Because people aren't machines made to maximise their productivity or long-term income - I'm sure I'd be better off productivity-wise giving up coffee, sugar and alcohol but I'm not going to. Now imagine you work a lot harder than I do, and on top of that don't have the internet, books, podcasts and a load of leisure time to spend with friends, you might be justifiably tempted to spend your money buying tastier but less nutritious food, or even saving up for a TV, instead of investing in things that might, in the long-term, be objectively 'beneficial'. B&D quote George Orwell's The Road to Wigan Pier:

“When you are unemployed, you don’t want to eat dull wholesome food. You want to eat something a little tasty”.

So a couple of conclusions from this week's class: 1) quality not quantity. And 2) the extreme poor in developing countries know what they're doing more than we give them credit.