« Back to news

Sharpening a powerful anti-poverty tool

Dean Yang

Dean Yang

Tuesday, April 26, 2011

Development economist Dean Yang wields "gold standard" research design to boost the impact of the wages migrants send back home.

Official development assistance, the amount contributed worldwide to promote the welfare and development of emerging economies, fell to $120 billion in 2009—down from $128 billion before the economic recession began. In the same year, remittances—gifts working migrants sent home to their loved ones—amounted to nearly three times that amount: $307 billion. Clearly, remittances are a powerful tool for combating extreme global poverty.

A powerful tool for combating extreme global poverty

For the families that receive them—roughly one-third of the developing world—remittances are a lifeline. They help ease poverty, increase food security, purchase critical medicines, finance educational investments, launch entrepreneurial ventures, and build family savings.

While these basic outcomes are well documented, little is known about policies that can boost the size and impact of remittances. Dean yang, associate professor of public policy and economics at the Ford School, is designing powerful studies to find out.

"Politically, increasing our national contribution to development aid presents a huge challenge in today's economic climate—however much it's needed," says Yang. "But if we can find programs and policies that encourage the number, value, and impact of remittances, we could have a substantial impact on poverty in developing countries."

Randomized controlled trials (RTCs) and why they matter

To test the impact of remittances and other microfinance tools, Yang employs what some are calling today's "gold standard" in economic research—randomized controlled trials (RCT) that are very similar in design to the clinical trials used in medicine. "One of the most important innovations in economics in the last fifteen years or so has been the rise of a new methodology for economics—the use of randomized controlled trials in real-world settings," says Yang.

Let's say researchers want to better understand how a particular development tool—microloans, for example— works in the world. In an RCT, they would begin by identifying a hundred neighborhoods that look very similar from a demographic standpoint. They would survey those neighborhoods to capture baseline data about entrepreneurism, consumption, and savings. Based on the results of those surveys, they would randomize their selection of 50 neighborhoods to receive the microloan offer, and work with local banks to market and administer the loans in those neighborhoods. Six months or a year later, they would return to conduct a follow-up survey of entrepreneurism, consumption, and savings. Sometimes, they would follow up with another survey to capture longer-term outcomes.

Sounds like a lot of work? It is.

"RCTs aren't easy to do," admits Yang, explaining that they require substantially more donor funding and up front work than the traditional econometric evaluations conducted by most development economists. The advantage, though, is threefold. First, in many cases RCTs make it possible to clearly and precisely identify the impact of a microfinance innovation that could be offered more broadly. Second, without isolating controls and testing variations, we can't fully understand the impact of development interventions—and those impacts aren't always what we'd expect them to be. Finally, policymakers and practitioners get the methodology, says Yang, which is often not the case with more sophisticated econometric evaluation techniques. As such, RCTs are making big waves in real-world policy spheres.

In the microfinance field, for example, RCTs have provided a much more nuanced understanding of when micro lending can be beneficial, and when it might not be. In households with an entrepreneurial bent, RCTs have shown that microloans increase business profits. But in other households, they increase consumption, and have little impact on the future success of the family—leaving loan recipients in debt that's difficult to recover from. For policymakers interested in micro lending, this means it's important to screen for entrepreneurial qualities, and to offer financial literacy and entrepreneurial education programs in conjunction with all loans.

Understanding how savings can best benefit Africa's rural farmers

While microloans have received considerable attention from the media and policymakers, opportunities to save are at least as important, if not more so, because they provide the poor with resources for investment, without the burden of loans. To understand the impact of savings, Yang recently received funding from the World Bank Research Committee and the Bill and Melinda Gates Foundation to offer rural farmers in Malawi a way to set aside a portion of their harvest earnings. Test groups were randomized, with half given an opportunity to open a standard savings account (one that could be accessed at any time) and half given an opportunity to open a commitment account (one that would restrict access until a date determined by the farmer—typically just before the next planting season).

Yang's finding: While Malawian farmers enthusiastically took advantage of both account types—not surprising, since many in developing economies are willing to pay high fees for an opportunity to save—those offered commitment accounts both saved more and spent more on agricultural inputs like seeds and fertilizer in the subsequent planting season. This is impressive evidence that commitment savings accounts produce strong benefits for Africa's rural farmers. Yang's next question, of course, is to quantify how those agricultural inputs affect the subsequent crop yield. As of press time he's still crunching those numbers.

Boosting the development power of remittances

Dean Yang

Dean Yang talking with farmers in Nkhotakota, a town on the western shore of Lake Malawi


Yang thinks his most distinct work, however, is the research he's done—and is continuing to do—on remittances, which he's been studying since 2004. In June 2009, he completed a two-year study that explored how remittance fees and joint bank accounts influenced the size, frequency, and impact of remittances among Salvadoran migrants in Washington, DC.

Going into the project, Yang had two theories. First, that the frequent policy recommendation of reducing remittance transaction fees (in El Salvador, these can range from 1.25 percent to 16.5 percent for a $200 remittance) would encourage migrants to increase their remittances. Second, that giving migrants more control over their contributions could potentially lead to expenditures with longer-term development impacts—like investments in education, housing, businesses, and savings.

Transaction costs, Yang discovered, do indeed have a dramatic impact on remittances. He tested price points that ranged from $4 to $9 per transaction, and found that every $1 reduction in transaction fees led to 11 percent more remittances per month, and $25 more in contributions per month. The takeaway? As the policy community expected, increasing competition between the companies that provide remittance services, and offering an easy way for migrants to comparison shop, has the potential to boost remittance contributions significantly.

Control, too, makes a big difference. In a pre-study focus group, DC-area Salvadorans shared stories about problems they or their acquaintances had experienced with remittances. One would only send money to the family members he trusted to spend it wisely, fearing other relatives would spend it on liquor. Another had an acquaintance who had sent tens of thousands of dollars to his mother to purchase a home in El Salvador, only to find out she had lent the money to another relative who threatened to kill him if he ever returned.

Beyond the matter of trust, Yang learned that migrants and remittance recipients disagree over how the money itself should be spent. The most dramatic differences are in the areas of consumption and savings. Remittance recipients felt 65 percent of the funds should be spent on food and other daily consumption needs, with only 3 percent reserved for savings. Migrants, on the other hand, felt only 42 percent of their contributions should be spent on consumables, and that seven times as much of the money (21 percent) should be reserved in a savings account for future needs.

To test how increased migrant control would impact savings and expenditures, Yang worked with a Salvadoran Bank, Banco Agricola, to offer DC-area migrants three savings tools. While the control group received no savings account, channeling remittances with the standard fee- per-transaction method, the others were assigned to one of three account types: 1) an individual remittance recipient account that wouldn't be accessible to the migrant donor; 2) a joint account that would allow both the donor and recipient to withdraw funds and check balances; and 3) both a joint account and a migrant-only account that would allow the migrant to both monitor the remittance account and build his or her own savings.

Yang's discovery? Six months later, amounts saved in accounts only accessible to remittance recipients were negligible, but accounts accessible to both migrants and recipients held $211 on average—roughly a 50 percent increase over baseline savings in remittance-recipient households. The takeaway here? That channeling remittances into savings accounts doesn't necessarily promote savings accumulation, but channeling them into joint savings accounts works wonders.

Influencing policymakers and practitioners

The multilateral Inter-American Development Bank (IDB) is excited by these findings and is hoping they will influence policymakers throughout Latin America and the Caribbean.

Ernesto Stein, lead research economist in the IDB's Research Department, believes that any studies that contribute to a better understanding of migrant remittances are important—particularly to Central American countries where it's not unusual for remittances to provide between 10 and 20 percent of Gross National Income. However, Stein says Yang's research—because it's so methodologically rigorous—is especially attractive.

"Policymakers may be inspired to subsidize the cost of remittances, to foster increased competition among transmitting banks, or to place restrictions on allowable remittance fees. They might also put in place vehicles so that the sender can have some measure of control over what the money is used for," says Stein. "These are exciting findings that will likely inspire new ways of thinking about remittances."

The IDB is working on a book to publicize these and other remittance research findings, and will be promoting them more broadly both within the Bank and to an audience of policymakers in Central America.


Below is a formatted version of this article from State & Hill, the magazine of the Ford School. View the entire Spring 2011 State & Hill here.

Open publication