How can we best help extremely poor people, today?

sp001px

Helping the world’s poorest people escape poverty is, in principle, a simple matter: Give them cash!

The trouble is, there are too many of them: About 700 million people -- more than twice as many people as live in the US -- are thought to live on less than $1.90 a day, according to best estimates from the World Bank. Probably the only practical way to end poverty on that scale is with robust economic growth. That will take time.

In the meantime, though, unless we choose to ignore their suffering,  it would be helpful to figure out how governments, foundations, donors and profits should use their wealth to lift the incomes of the poor. Should the wealthy give cash grants, gifts of livestock, low-interest loans, job training or something else to the extreme poor?

It’s an important question. While longer term investments in public goods -- roads, ports, education, health care systems, good governance -- can help spur economic growth, so can helping poor people be less poor.

As Jeremy Shapiro, a founder of nonprofit GiveDirectly, has said: “Poverty alleviation through redistribution--if it acts as a stimulus or engine for human capital--can further economic development.”

Two recent studies -- both about nonprofits that I’ve praised on this blog -- have got me thinking about short-term poverty alleviation. One is a randomized evaluation from Innovations for Poverty Action, a research and policy nonprofit that often works with development economists, that looked at a “graduation” program run by Village Enterprise in Uganda. The study found that the graduation program--a multi-faceted attempt to improve livelihoods--lifted the consumption, assets, nutrition and self-reported well-being of the poor, and that it did so better than cash alone. The Village Enterprise program included a cash grant of about $100, classes in business practices, two years of mentoring and a savings group that enabled small-scale entrepreneurs to pool their resources. Graduation programs have been well-studied, notably in a landmark 2015 study of six programs that produced encouraging results. Even Nick Kristof wrote about it! Despite the high-touch nature of the Village Enterprise program, IPA’s evaluators said that “cost-effectiveness appears high.” Village Enterprise, understandably, spread the word:

Graduation approach proved more effective than cash. Third party evaluation results now available!

— Village Enterprise (@village_ent) February 22, 2018

A Devex story on the study ran under the headline: “Ultra-poor graduation model results show more than just cash is needed.”

It’s not that simple. As the story goes on to explain, graduation programs are expensive. Their results vary, depending on the design of the program and the context. (One program studied in Honduras did almost no good.) Unless graduation programs can be scaled down and delivered more cheaply, they will be able to reach only a limited number of those 700 million people in need.

The other study on my mind evaluated the effects of direct cash transfers administered by GiveDirectly in the Rarieda district of Kenya in 2011. It followed recipients for three years. Most of the households got a one-time $300 grant. According to Michael Cooke, research director at GiveDirectly:

Overall the findings are encouraging. The treatment effects on all the main outcomes (assets, earnings, expenditure, food security, and psychological wellbeing) were sustained after 3 years. Gains on an education index that were not significant at 9 months also becomes significant at 3 years, driven by increased spending on school fees, uniforms, books and supplies.

This was only the latest study of cash transfers, which in various ways have been tried by governments in Africa, Latin America and Asia. The results have generally shown that if you give poor people money, they will use it wisely to improve their lives. Today, GiveDirectly typically gives $1,000 lump sum grants; just the other day, the group said it had enrolled its 100,000th household, which is great, but, again, a long way from 700 million.

So what’s a confused donor (like me) to do? There’s no simple answer.

Presenting IPA's study of Village Enterprise at the World Bank last month, Richard Sedlmayer, a doctoral candidate at Oxford who lead the research, said: “The question of how graduation relates to cash transfers is a contentious one….It’s a very hot topic right now.”

“I would not say that this study should be taken as the definitive word on how to design anti-poverty programs,” he went on. “But it has given me a lot of confidence that there is a case to be made for mentorship and training.”

Nathanael Goldberg, who directs IPA’s social protection program, agreed. But he noted that graduation programs are costly. The 2015 study says that per-household costs range from $358 in India to $2135 in Ghana and $2697 in Peru. Future research, he said, may provide guidance on how to reduce those costs: “We have to figure not only which of the components are critical, but how much of each.”

Jeremy Shapiro, who co-authored the evaluation of GiveDirectly’s work in Kenya, says that while randomized controlled trials (RCTs) can be designed to make side-by-side comparisons of different poverty interventions, they have more often been used to evaluate individual programs and improve their design. Fortunately, he said, more studies are beginning to compare different anti-poverty interventions.

Shapiro, who is now president of the Busara Center for Behavioral Economics, pointed me to another study, this one a 2016 report from IPA and CGAP, a network of anti-poverty groups, that compared cash transfers, graduation programs and livelihood training. It found:

Among all 48 programs reviewed, lump-sum cash transfers were found to have the highest benefit-cost ratio, though there are very few lump-sum cash transfer programs that serve the extreme poor or measure long-term impacts. Livelihood programs that targeted the extreme poor had much lower benefit-cost ratios. Graduation programs are more cost-effective than the livelihood programs that targeted the extreme poor and measured long-term impacts (i.e., at least one year after end of interventions).

With respect to the comparison of cash transfers and graduation programs, Shapiro told me: “You can say, generally, that both do good things for poor people. It’s not abundantly clear that one does better than the other.”

Meta-charities don’t offer decisive guidance on this question either. GiveWell has designated GiveDirectly as a top charity; it briefly endorsed Village Enterprise years ago but no longer recommends nonprofits delivering graduation programs. Other NGOs that deliver versions of graduation programs include BRAC, Trickle Up and the BOMA Project.

ImpactMatters, which conducts “impact audits” of nonprofits to rigorously estimate their philanthropic impact, praises Village Enterprise for “putting its flagship microenterprise program under the microscope of an RCT and contributing to the evidence base for poverty alleviation programs.” It said the program delivered “modest, statistically significant increases in consumption, assets and income for participants.”

Just how modest? The cash grant, training, mentoring program and savings group, taken together, raised the incomes of the extreme poor by about 4 percent, according to Elijah Goldberg, the co-founder and chief operating officer of ImpactMatters.

IPA’s Nathanael Goldberg (no relation) said the same thing: “It’s very clearly a modest impact. But it’s a modest impact for the people who need it most.”

This shouldn’t be surprising or disappointing. Here in the US, neither the government nor most nonprofits have a stellar track record when it comes to lifting the incomes of those in need, whether they be laid off coal workers in West Virginia or high school dropouts in inner city LA. Job training programs, whether domestic or global, don’t seem to work well.

Again, what's needed is economic growth or encouraging migration.

Two final thoughts: First, it’s great that organizations like IPA are trying to figure out what works when it comes to helping the poor. More research is on the way. In Ghana, IPA is working with Heifer International on a study of its programs, which typically involve gifts of livestock, to evaluate different variations--without coaching visits, with cognitive behavioral therapy and a cash transfer along. USAID has also commissioned an RCT from IPA to compare a nutrition and WASH (water and sanitation) program to cash transfers in Rwanda. I'll be reporting on that, from Rwanda, later this spring.

Second, donors who care about the world’s poorest people can’t go wrong with GiveDirectly, Village Enterprise or other NGOs, like Trickle Up or BOMA, that have been independently evaluated. Unlike better known global anti-poverty groups that try to do many things--CARE, Save the Children, or, goodness knows, Oxfam--these groups are focused, transparent, effective and, importantly, ready to learn.

--Background for potential donors: My blogpost about Village Enterprise and the first development impact bond for poverty alleviation in Africa is here. Lots of previous posts about GiveDirectly, including here and here.