Syndicate content

Evaluating The Best Ways to Give to the Poor: Guest post by Bruce Wydick

What are the best things ordinary people living in rich countries can do to help poor people living in developing countries? This is the question the editors of Christianity Today assigned to me for a special issue this month on world poverty. It is a question many people like my parents worry about, people who would like to give money to causes that help poor people overseas, but to put it simply, have little clue about what actually works. (Except for the fact that their son strives to keep them painfully abreast of such matters.) 

A plethora of options exist today for ordinary people with a little bit of extra means to help the poor: donating a yak to a Nepalese family in the name of the “rich uncle who has everything,” making a microfinance loan to a Nigerian cobbler through Kiva, sponsoring a child in virtually any developing country in the world. Courtesy of the web, today it is not just big governments and multinational organizations that do all of the heavy giving. Ordinary people give to the overseas poor now more than ever.

So among the popular means through which ordinary people try to help the poor overseas, which are the most effective? I decided to take a poll of our peers and have each rate, on a scale of 0-10, some popular interventions.   Specifically, if someone wanted to donate $X to alleviating poverty in developing countries, which of these approaches would be the most (10) and least effective (0) means of increasing the welfare of the poor in developing countries?  

Survey participants included Chris Barrett (Cornell), Alexander Tarozzi (Duke), Dean Karlan (Yale), Ken Leonard and Pamela Jakiela (Maryland), Betty Sadoulet, Alain de Janvry, and David Levine (Berkeley), Pascaline Dupas (Stanford), Jon Robinson (U.C. Santa Cruz), Sarah Baird (George Washington University), Paul Glewwe (University of Minnesota), Judith Dean (Brandeis and World Relief), Chris Ahlin (Michigan State), Julie Schafner (Tufts), and David McKenzie (the World Bank ). 

The results of the poll appeared in the cover story article for the magazine this month.  A couple of the lesser-used interventions were omitted for space considerations in the main article, but I include them here with an abbreviated synopsis of the commentary. 

Here are the rankings:

1. Providing Clean Water to Rural Villages.  (Average Rating: 8.31) A colleague of mine at USF had his large undergraduate global economics class take the poll as well. Even the undergrads came up with the same answer for the #1 ranking, so it seems you don’t have to be a development economist to get this one.  A million children die from drinking unclean water each year. One famous WHO study found clean water in a rural village reduces infant mortality by 35% to 50% at a cost of roughly $10 per person per year. Bang for the buck, clean water is hard to beat.

2. De-worming Treatments for Children.  (Average Rating: 7.81) This one was probably ranked high for three reasons. First, because de-worming is so effective at a medical level, it’s cheap, and also because this early intervention was the subject of one of the most famous early randomized trials. Miguel and Kremer’s celebrated study found that regular de-worming treatment in worm-infested areas of the developing world could reduce school absenteeism by 25% at a cost of $0.50 per year. While impacts of this magnitude are not likely to have external validity in areas with lower worm infestation, there are plenty of places left where deworming impacts would likely be huge.

3. Donation of Insecticide-treated Bed Nets. (Average Rating: 7.31) The Sachs / Millennium Challenge Corporation efforts along with Pascaline Dupas’ excellent work in this area have created quite a buzz around insecticide-treated bed nets in the last few years. Malaria is a leading killer of children in developing countries, accounting for nearly one in five deaths of children under five in sub-Saharan Africa. The new nets last for several years without needing to be re-treated, and reduce instances of malaria by 50% and malaria mortality by 20%. Bed nets are also incredibly cheap—only $5-10 dollars. Another big bang for the buck.

4. Child Sponsorship.  (Average Rating: 6.86) Paul Glewwe, Laine Rutledge, and I did a BASIS-sponsored study on the long-term impacts of Compassion International’s child sponsorship program.   We used a regression discontinuity/IV framework based on age eligibility at program rollout that examined adult life outcomes for children who were sponsored in six countries through Compassion’s program during the 1980s and 1990s. The adults who as children providentially lay on the right side of the age-eligibility cutoff were more likely to complete secondary school, have a white-collar job, marry and have children later, and have a house with electricity. Evans and Kremer’s (2008) shorter-term study in Kenya also finds big impacts.   Expensive, but effective. 

5. Efficient Wood-Burning Stoves. (Average Rating: 6.00)   Indoor air pollution is worse than many people previously thought: about 1.6 million people are believed to die prematurely each year from it.   Poor stoves also lead to more rapid deforestation, where 5.8 million hectares of tropical rainforest are being lost annually. Improved wood-burning stoves deal with both of these issues in one go, with big impacts on both. Some improved stoves are super-cheap. For example for $15 you can buy one of the new high-tech rocket stoves.

6. Microfinance.  (Average Rating: 4.19)  Development economists now are nearly all aware of one major fact about microfinance: it’s astounding growth over the years is significantly larger than its impact.  Most of the recent studies of microfinance using RCTs, including the relatively recent 2011 papers by Duflo et al. study in Morocco, and the new Attanasio et al. study in Mongolia  find modest impacts: increases in entrepreneurialism and business investment, but little or no increase in income. Microfinance clearly exhibits some modest positive impacts, but it isn’t the silver bullet in the heart of poverty some had hoped it might be. At least by some measure, development economics and microfinance have lost that lovin’ feeling.

7. Cleft Palate and other Reparative Surgeries.  (Average Rating: 3.87) There is little doubt that disabilities such as cleft palates and cataracts create huge obstacles to leading a normal life for kids in developing countries. We have all seen the shocking pictures of disfigured children in magazines that ask for a donation for organizations such as Smile Train, which does corrective surgeries for a $250 donation. Little research has been done by economists to evaluate the impacts of these surgeries in developing countries, but medical research shows excellent long-term results. Intervention is expensive, but respondents had little doubt about impacts.

8. Farm Animal Donation.  (Average Rating: 3.81) The marketing of organizations like the Heifer Project are second to none. Who can resist the picture of the happy Bolivian girl hugging the alpaca? Chris Barrett and I have been working on an evaluation project with Heifer based on non-experimental data. Our results are just preliminary but indicate positive impacts on dairy consumption from donated dairy cows, but little measurable impact from the smaller animals. Organizations like Heifer face difficult issues related to targeting. They would like to target the poorest of the poor with farm animals, but in many cases these households lack the resources to properly care for one of the major farm animals. We’ve noticed this in the field.

9. Libraries and Literacy.  (Average Rating: 3.44) Literacy and a love of reading and ideas are key to economic development and escaping poverty. Many development practitioners feel that small libraries in rural villages strongly promote literacy, learning, and development.  There is some evidence that libraries may be a cost effective alternative or supplement to investments in schools, where the cost of a child reading an extra book a year is estimated to be between $0.74-$1.30. Smaller donations of a few dollars buy books, larger ones of hundreds of dollars establish the village libraries themselves. I thought this one might be rated higher.

10. Fair-Trade Coffee. (Average Rating: 1.94) How can fair-trade coffee, as it is commonly practiced, continue to thrive when the most rigorous evidence shows that it doesn’t work?  (and even clearly explains why). In a paper certain to amplify the caffeine jitters in the fair-trade coffee industry, De Janvry, McIntosh, and Sadoulet (2011) find zero average impact on coffee grower incomes over 13 years of participation in a fair-trade coffee network. Low impact is due to extremely poor institutional design: growers have to pay for fair-trade “certification,” recouping these costs only in years when the coffee price falls below the $1.41 price floor. (The New York Coffee price is about $2.55 right now.) Moreover, here is a program that promotes the cultivation of coffee, when it is obvious that coffee growers would be collectively better off if everyone grew less. Clearly a case where development economists are aware of the deficiencies of a program, but word hasn’t “filtered down to the average Joe.”

11. Laptops for Kids. (Average Rating: 1.80) The One Laptop per Child program gives away laptops to kids in developing countries who have at least a thousand more pressing needs. The bang for the buck on this one is at best a whimper, at worst inaudible. A recent study presented at the NEUDC by Cristia et al. at the Inter-American Development Bank evaluated the impact of the One Laptop per Child program in Peru, finding zero impact on learning or on expectations about future education.

12. Menstrual Pads for Young Girls. (Average Rating 1.60) Lack of access to sanitary pads can result in lost days of schooling and work, and are thought to reduce education in girls and have numerous other negative consequences. SHE (Sustainable Health Enterprises) asks donors to provide $28 every 28 days to promote small businesses that make locally produced menstrual pads available to girls and women in developing countries.   Many in the pool had read the current research: In a recent study, Oster and Thornton find no detectable impact on girl’s school attendance. 

What does this all add up to? I see three lessons here for development researchers doing impact work. First, there is considerably more awareness and consensus within our own circles about the effectiveness of different types of development programs than there is within the general population. We need to do a better job of getting the word out about what works and what doesn’t. Sacrificing a research project here and there to spend more time doing public writing and speaking would probably constitute a public-welfare-improving reorganization of time for many of us.

Second, there appears to be almost an inverse relationship between marketing hype and rigorously measured impacts. Over the decades there have been movements to bring greater levels of transparency into the consumer product industries, the medical industry, and now the financial industry. The aid industry needs to be next. As a profession we should be insisting that these consumer-oriented aid organizations gain third-party certification of impacts for the interventions they market to the public. 

Third, we need to diversify our collective research portfolio. While many of us continue to work on issues like microfinance, water, and education, there are a number of interventions that have had very little research carried out on them, some included here. The work done by McIntosh et al. on fair-trade coffee and by Cristia et al. on One Laptop Per Child should serve as examples of rigorous empirical work that out to help staunch the flow of funds to ineffective programs before more money is wasted. The opportunity costs are too high.

Comments

Excellent post and set of survey participants. Like most development economists my friends/family ask me where to make their donations. In terms of the most impact ones, I'm not sure which charities do good work on 1-3 and 5? Anyone have any insights? I wonder two things. Are we as economists too dependent on a few studies? I think 6,8 and 12 seem understudied to me. I have my own priors, but even if I think deworming or clean water is best bang for the buck I still spread my money around since I'm uncertain. What about poverty traps. 1-3 target the poorest. Targeting those slightly up the development ladder like heifer's recipients might help a long term escape of poverty. Finally, apologies for the self promotion but two of my recent papers with co-authors do show some positive impacts of fair trade coffee World Development (http://www.sciencedirect.com/science/article/pii/S0305750X10001427) and forthcoming in J. Development Studies (http://ideas.repec.org/p/tow/wpaper/2010-14.html), but not out of line with the above cited study.

Hi Seth, Bruce's piece in Christianity Today links to a number of charities under each point above. I agree that one issue in comparing all these efforts is that they support development and poverty reduction in different ways. The child sponsorship paper he links to is one of the few studies that look at long-term outcomes here (although there are now also some long-term outcomes of the deworming work too).

Submitted by Brigid on
When it comes to where to give, though, there is a difference between causes/interventions and the actual charity. An intervention may be the right one, but no on-the-ground charity may be doing it correctly (example coming to mind are water charities that built wells and left without providing infrastructure for continued maintenance) As a layperson, I take my info from GiveWell, which seems most accessible (easy on the academic-speak) and transparent. On water-related interventions, they haven't yet recommended a charity: http://givewell.org/international/health/water The most interesting paragraph here was the last one noting the value of debunking interventions that have no impact, so money is no longer diverted poorly. In strictly economic terms, there is an opportunity cost, yes, but I have observed that usually there's quite a few egos involved, too. Defensive arguments can get heated, diverting conversation from what works, and of course media loves a scandal.

Interesting list - a couple of quick comments from someone working/researching in the water sector. Firstly, most people in the sector would say that sanitation and handwashing are probably even more important than the actual water itself. Secondly, the water and sanitation sector is a great example where the potential benefits are huge, but the sector needs to use its existing resources better as well as getting more money. Achieving a rural water service that lasts forever (instead of just some infrastructure that breaks down and is abandoned within a few years) is a much bigger challenge. See "Admitting failure: the “naked truth” for water and sanitation?" http://wp.me/p1jV1a-46 for how the sector is gradually acknowledging and addressing these problems. Disclosure: I'm doing my PhD in collaboration with WaterAid on these issues

Submitted by Meleesa Naughton on
Excellent points - and I am very happy you brought up both the issue of sustainability and that of sanitation and handwashing, which have more benefits than single access to water in terms of impact on neglected tropical diseases too (Esrey et al 1991). On a related topic, I am surprised that no one has commented on local menstrual pads' place at the bottom of the list. Is school attendance in itself an appropriate indicator for the success of such programmes? Giving young girls pads but nowhere safe/ dignified to get changed at school is likely to be a factor leading them to stay home while on their period. For this reason, I believe menstrual health should be fully incorporated in water, sanitation and hygiene programmes and evaluated as such. Another consequence of poor menstrual hygiene - using inappropriate material- is that it leads to infections. The consequences of using sanitary pads on menstrual and genital health should also be measured.

Submitted by Reykia Fick on
Thanks for the interesting article. It’s certainly important for us to consider to best make an impact. However, your argument against Fair Trade coffee is based on a study which examines a single coffee cooperative. Meanwhile, a 10-year review of academic literature on Fairtrade (33 case studies) found “higher returns and more stable incomes as clear benefits enjoyed by Fairtrade producers,” and also added that “non-income impacts of Fairtrade are at least as important as income benefits for smallholder farmers… (e.g. capacity building, stability of income, market information and access, self-confidence, access to credit, national representation etc)”. More information here if you're interested: http://www.fairtrade.net/impact_studies.html

Thanks for the great comments. Responding to Seth and Reykia: The McIntosh et al. paper on fair-trade coffee is particularly important and convincing for a couple of reasons. The first is that it specifically focuses on price returns to coffee growers, which is, and should be, the principal focus of any fair-trade effort. The second is that it looks at these returns over a long period of time (over a decade). They demonstrate how a grower can benefit from being in fair trade temporarily (when coffee prices plunge), but that this benefit is transitory, just like we teach in Econ 101. The surplus the receive goes away as more growers become certified to grow fair trade when it pays to become certified. What they observed from this single cooperative (that more people enter into fair trade when it is profitable to do so) is exactly what we would expect in other situations as well. That they find zero long-run benefit to fair trade makes perfect sense after you think about it a while: why should there every be excess returns in a market with free entry? Fair trade coffee efforts, as generally practiced today, is grounded in a flawed economic mechanism that essentially guarantees that it won't work in the long run. The fact that the case students mentioned show that the side benefits are greater than the income benefits illustrates the point: there probably aren't significant income gains, and given that this is true, the non-income gains are likely to be higher. There are so many better ways to help coffee growers than what the FLO and Transfair are doing presently: scholarship fund for coffee grower kids, some of the top-5-rated interventions mentioned here, other things to wean them away from coffee growing so that prices increase to those who continue in coffee will benefit from the reduced supply. It would be so easy to develop an alternative fair-trade mechanism that far more efficiently transfers income from drinkers to growers. The question is when will the fair trade folks wake up and smell the coffee. With respect to Brigid and Stephen's comments on water. Yes, absolutely. I have heard anecdotally the same--there are innumerable wells with broken pumps drilled by well-meaning NGOs that sit dormant because of lack of maintenance and a plan to keep them running in the long term. No doubt. But water is so primal among human needs, and the welfare payoff is so high per dollar, we might want to focus on doing water better than doing something else.

Submitted by MattJ on
This is a very interesting and educational article for development in poorer countries! It shows great statistical data to support the rankings of the ideas. I can think of no better example than this article to display William Easterly’s, Planners and Searchers shown in The White Man’s Burden. This shows even clearer the effect of how Planners (an overwhelming majority) believe they know what’s best to give for aid, whereas, the Searchers seek deeper explanations and look for what is needed by the people to better them. A perfect example is the coffee trade. Coffee is easily cultivated; however, not everyone needs to be growing coffee – basic economics teaches us an oversupply creates a reduction in price, resulting in lower profits. The coffee has had zero effect on development. When a solution such as providing clean water has such a positive impact for development, the world needs to be educated – not only to be a solution, but also not to oversupply and create abundance. The Planners have great intentions, but they need to be educated by some of the policies of the Searchers, such as following through and making sure their aid was given to the proper cause, in order to construct a better developed country. This article does a great job of proving that.

Bruce, Thanks for your comments and well written article. I'm assigning it in my class later this semester. I need to read the De Janvry, McIntosh, and Sadoulet piece more closely. They have an excellent design/sample for impact evaluation. My co-authors on fair trade projects have another paper that fair trade "reveals that yields rather than price premiums are most important for increasing net cash returns for coffee growing households" http://www.sciencedirect.com/science/article/pii/S0305750X11002890 What I meant was the story we are finding is consistent with the de Janvry, McIntosh, & Sadoulet story (ie prices don't matter), but additionally fair trade might also be agricultural extension. I think 3ie is doing a systematic review of agricultural extension impacts and ag extension might also be added to the list of ways to help.

Submitted by William A Stovall on
I think Bruce makes an excellent point saying that we need to diversify our collective research portfolio. So much focus is on the first things that come to mind- money, water, and educational opportunities- which is good, but we need to diversify our focus to other things, some listed above. It’s great that more “ordinary” people are wanting to help than ever, especially in the economy that we’re currently in. Often times, planners (Easterly's book) are charged with lack of knowledge of the bottom. By actually figuring out what these countries need, which some of the people noted above (laptops, cleft palate surgeries, etc), we can make more of a difference and not spend 2.3 trillion dollars with little to show.