Dean Karlan and Jacob Appel’s book, More Than Good Intentions, is perhaps the most readable discourse available on development economics’ current obsession, randomized controlled trials (RCTs). For anyone new to the field or the concept – and for those who want to read an enjoyable book and learn something, too – it’s the perfect entry point, blending humor, an engaging writing style, and interesting RCT results into a cohesive whole. For those who are already well-versed in the field – or those who have read Esther Duflo and Abhijit Banerjee’s book Poor Economics – it’ll be less useful from a learning perspective, but still an insightful and entertaining recap.
Karlan is a Professor of Economics at Yale and founder of Innovations for Poverty Action (IPA), a research organization that tests interventions and scales the ones that work most effectively, and stickK.com, a commitment website. Appel is his writing partner, found by Karlan because of the engaging, smart writing on his blog (incidentally, a career path that I would very much enjoy, for all of you authors looking for a co-author!)
A randomized controlled trial is a means of inferring causality of an intervention. To oversimplify a bit, a population is split into an intervention group – those that receive a drug, a new process, or a treatment, maybe – and a control group – those that don’t. A coin is flipped for each person in the population; heads go to the intervention group, tails to the control group – hence “Randomized.” With a large enough population, each group will be, on average, extremely similar at the outset, so any differences perceived after the intervention would be the result of the intervention. This approach has been used for ages, and is considered the “gold standard” of research methodology – but has only been recently transposed to development economics.
What drives Karlan and Appel towards this approach seems to be a distrust of abstraction and a preference for small, provable, incremental change:
“Up in the realm of high-minded concepts and metaphor – choices, opportunities, dignity, fishing – the air is thin and there are no actual poor people to be found. This isn’t where development needs to be. It needs to be on the ground. If we want to solve poverty, we need to know what it is in real – not abstract – terms. We need to know how it smells, tastes, and feels to the touch”
After reading a number of rather all-or-nothing development authors (William Easterly and Jeffrey Sachs come to mind), it’s refreshing for authors to acknowledge that small wins are still wins – big wins for those whose lives are directly improved.
Karlan and Appel are tinkerers at heart, emphasizing skepticism of “best practices” and a preference for continuous iteration. Some conditional cash transfer programs seem to work really well, but could they work better? How can we create a microsavings program that maximizes impact? Using a combination of RCTs, common sense, and behavioral economics, they hope to make already-good programs great, and already-great programs excellent.
To find answers to some of these questions, they take the reader around the world to visit many of the interventions they have worked on through IPA or those that were done by friends and colleagues (the names of Esther Duflo, Sendhil Mullainathan, Michael Kremer, and other stars of the field pop up continuously in the book). From El Salvador to Accra, Bogota to Bangladesh, Karlan and Appel highlight interventions that have been shown to work at reducing poverty and making the lives of individuals in developing countries better off – and then they ask how each could be improved.
They spend more time on microfinance than Banerjee and Duflo did in Poor Economics. Even if you have only a passing interest in development, you likely know the story of microfinance’s inception: in the 70s, Muhammad Yunus, of Bangladesh, decides to offer small loans to groups of women. With that capital, they have the funds needed to start a small business – maybe they can buy a sewing machine, or bead and string to make jewelry. They repay the loan, take another, and prosper, pulling themselves out of poverty.
It’s a great story – but unfortunately, only a part of a more complex reality. As Karlan and Appel discuss at length, microfinance simply doesn’t pull most from poverty; most people don’t want to be the entrepreneurs highlighted by microfinance institutions the world over. What it does do is make life a bit easier for those who need some cash to pay for transportation, medical bills, food, or past loans. It’s popular with Westerners, aid organizations, and the media, but is less popular with its intended beneficiaries: the poor.
This is an area of the book where the authors’ guiding philosophy shines through. Rather than assume that the relative unpopularity of microfinance amongst its beneficiaries is an unfortunate inevitability, they ask how it can be improved, in the process breaking down the reigning orthodoxy.
One of the main tenets of microfinance used to be that, in order to tap the “social collateral” of a group and ensure repayment, the group had to be liable for each individual loan made. For the individual that always paid back her loan, this was annoying, as she’d sometimes be on the hook for another person. In other words, the best borrowers were dis-incented to join the group. Thanks to research by Karlan, we now know that group liability wasn’t necessary for repayment rates to stay high, and that individually-liable loans within a group were just as likely to be repaid.
Or at least we think we know that. One of the issues with RCTs, as Karlan and Appel acknowledge, is that they can’t tell us if results will have “external validity” – whether they will be applicable anywhere but the area the research was conducted. The microfinance research I just referenced was done in one specific area of the Philippines; do the results apply to another area of the Philippines, or in Kenya? We don’t know. Their solution is to conduct trials of similar interventions in different areas of the globe and then infer external validity from the results. If an intervention passes the common sense test and appears to have similar results in dissimilar environments, we can conclude with some certainty that it is externally valid and scale it further.
In addition to microfinance, the book discusses studies done in the realms of agriculture, sex, health, and conditional cash transfer programs. There’s a lot to learn in each section, and the authors’ engaging writing style makes it easy to read; I blew through the book in a few sessions and – even though in barroom debates I occasionally cite the research they discuss – I still learned quite a bit.
More Than Good Intentions is a readable, entertaining entry point to development economics, blending engaging writing with heartfelt human interest stories, cutting-edge development research, and humor. Whether you’re new to the field or relatively well-versed, you’ll find it an interesting read that you’ll learn something from.