By Stephanie Beasley
In its nearly two decades of operation, GiveDirectly, the direct cash payment charity started by MIT and Harvard graduate students, has delivered more than $800 million to more than 1.6 million people, largely in sub-Saharan Africa.
The effort to allow financially struggling people to use aid money as they see fit — rather than as directed by faraway grant makers — has seen success and failure, both in its work abroad and in its first seven years in the United States.
Operating in the United States has been less logistically complex than providing cash grants in some the world’s poorest and most unstable places. Yet what works in Kenya does not necessarily work in Kansas — and GiveDirectly has become enmeshed in debates about universal basic income, the no-strings-attached cash payments intended to even the U.S. economic playing field. Its home turf story is one of learning through failure, collaboration, and identifying the communities where it can have the biggest impact.
When GiveDirectly launched in 2009 to transfer cash through mobile phones to Kenyans and Ugandans living on less than $2 a day, it quickly became a darling of tech philanthropy. The fledgling nonprofit embodied the effective altruist movement — to identify and “rationally” (or technocratically) solve the world’s most pressing problems. GiveDirectly launched with a $2.4 million startup grant from Google and then received $25 million from Good Ventures, a private foundation operated by Dustin Moskovitz, Facebook’s co-founder, and his wife, Cari Tuna. Other major donors have included Twitter co-founder Jack Dorsey, eBay founder Pierre Omidyar, and Amazon co-founder MacKenzie Scott, who recently gave the organization a rare fifth grant of an undisclosed amount. She has given the nonprofit more than $125 million since 2020.
Give Directly was not originally intended to work in rich countries like the United States. However in 2017, at the request of philanthropists Laura and John Arnold, the Houston-based couple interested in tackling persistent problems, the nonprofit agreed to adapt the lessons it had learned to respond to domestic disasters. That year, GiveDirectly collaborated with the Texas-based Arnold Ventures (then known as the Laura and John Arnold Foundation) to deliver nearly $10 million in cash assistance to families affected by Hurricane Harvey.
Over the past seven years, the organization has provided more than $270 million in cash payments to over 220,000 Americans. It has found that cash transfers are a highly effective way to help people several weeks after a disaster. And it has continued to experiment with different groups of aid recipients, such as pregnant mothers and homeless families.
“We’re a learning organization. We are committed to research, we’re committed to candor,” said Sarah Moran, a vice president at GiveDirectly. “We don’t want to be doing things that don’t work.”
The U.S. Pivot
GiveDirectly has learned in a typical way: through trial and error.
During the pandemic, the nonprofit found its $1,000 payments to nearly 200,000 households barely made a dent in the lives of U.S. recipients. There was no measurable difference between the households that received the cash grants and those that did not, GiveDirectly declared in a blog post based on academic studies from the University of Michigan. One reason, according to the study, was that most cash grant recipients also received government assistance, which made it harder to discern the impact of the GiveDirectly money alone. The apparent lack of positive impact surprised Luke Shaefer, the principal investigator behind the assessment of Covid cash programs.
“It was a really humbling experience for me because I thought it was going to be really straightforward,” he said.
GiveDirectly has vowed to learn from its mistake. So far, that has meant targeting populations such as pregnant mothers, infants, and families experiencing homelessness for whom regular, modest cash impacts have proven to be life altering. Academic studies and government data have shown that low-income mothers who receive cash transfers during pregnancy were less likely to have low-birth-weight children, among other problems.
GiveDirectly’s new approach is getting broad support, even from conservative politicians such as Michigan Republican Sen. John Damoose. Conservatives have historically opposed unconditional cash grants and guaranteed basic income programs over concerns that they might discourage people from working.
But cash aid to people dealing with environmental disasters and pregnant mothers has been less politicized than programs targeting broader swaths of low-income people.
Last year, GiveDirectly continued its emergency aid program to environmental disaster victims, delivering $1,000 emergency grants to thousands of families impacted by hurricanes in Florida and North Carolina. In recent weeks, GiveDirectly raised money for survivors of the Los Angeles wildfires, bringing in more than $2.1 million. It also is sending cash directly to L.A. families participating in SNAP, formerly the food stamp program, through an existing benefits app.
GiveDirectly has also taken a tech approach to disaster aid that has included artificial intelligence. The nonprofit says AI has enabled it to identify the low-income and disaster-plagued areas that need the most aid six times faster, with families receiving payments within hours rather than weeks.
Cash Transfers for Michigan Moms
GiveDirectly likes academic scholarship. When it launched its next U.S. initiative, the nonprofit relied on research about the lifelong benefits of cash transfers for pregnant women and children from Shaefer, a public policy and social justice professor at the University of Michigan, and Mona Hanna, associate dean for public health at Michigan State University who was the pediatrician who exposed the Flint water crisis in 2015.
In January 2024, GiveDirectly partnered with Shaefer and Hanna to administer Rx Kids, a public-private maternal and infant cash transfer program based in Flint in collaboration with the City of Flint, the State of Michigan, and the Michigan Department of Health and Human Services.
Over the past year, Rx Kids has provided more than 1,300 pregnant moms an initial $1,500 no-strings cash grant and $500 a month during their baby’s first year of life. Flint has a population of around 80,000 people and a median income of about $36,000. It is one of the poorest cities of the nation with an overall childhood poverty rate of approximately 50 percent. The most recent survey of the Rx Kids program shows the bulk of participating mothers make less than $10,000 a year.
Rx Kids funders include the Charles Stewart Mott Foundation, the Michigan Department of Health and Human Services, and the Stryker Johnston Foundation. GiveDirectly, in its effort to raise money for Rx Kids Flint, has received $37 million in grants and philanthropic contributions for the project, which accounts for 85 percent of the total budget. GiveDirectly administers the program — including enrollment, sending payments, and following up with recipients — and supports related research. The remaining 15 percent of the total $44 million budget goes directly to Michigan State and the University of Michigan for research, advocacy, communications, and marketing.
What’s different about Rx Kids is that it isn’t a pilot, Hanna said. It’s not working with a sample size of 100 or 200 people; it’s providing cash to a whole community.
“There are many reasons why there is excitement about this now by lots of different folks,” Hanna said. “It’s low cost. It is so administratively efficient. There’s no income requirements. There’s no policing of folks’ spending. It’s unconditional.”
Shaefer believes the outcomes of Rx Kids will be easier to track than the Covid relief program. Last month, Rx Kids expanded to Kalamazoo, Mich., a city of roughly 73,000 with a 27 percent poverty rate, and seeks to help the mothers of more than 800 babies there this year. Next on the list for expansion is Michigan’s Eastern Upper Peninsula, a primarily rural, white population with a 15 percent poverty rate and median household income of about $56,000.
Not a day goes by that the Rx Kids team doesn’t hear from other communities in red and blue states that want to create a similar program, Hanna said. In Michigan, RX Kids supporters include Senator Damoose, a self-proclaimed “pro-life person.”
“We’ve been accused for years about being pro-birth, not pro-life. And I think that’s not without merit,” he told NPR. “We need to put our money where our mouth is and support these children and support mothers.”
Basic income programs have rarely received that kind of bipartisan support. The right-leaning Heritage Foundation, for example, criticized a GiveDirectly-backed basic income pilot that provided $500 monthly cash payments to thousands of low-income families in Chicago and the surrounding suburbs in Cook County from 2022 to 2024, saying the payments could discourage people from working. Conservatives also have acted to block basic income programs in states like South Dakota and Texas amid concerns that the programs would make recipients permanently rely on assistance. Additionally, some philanthropists have been reluctant to support direct cash aid, primarily due to concerns about how recipients would spend the money.
“People really believe that guaranteed income depresses work. It’s become controversial, especially in red states, because people are convinced that people get guaranteed income and they’re lying on their couch painting their nails and eating bonbons,” said Mary Bogle, a researcher at the Urban Institute who has studied the effects of cash-based social policies on low-income parents.
The Alaska Permanent Fund, the longest-running universal basic income program in the country, has largely not suffered from such negative perceptions, Bogle said. Notably, that program provides unconditional cash payments to all Alaska residents rather than just to those considered low-income, which alleviates the stigmatization, she said.
More Basic Income Pilots
Bogle doubts that limiting guaranteed basic income programs to certain groups, as GiveDirectly is doing, will further the case for broad adoption of universal basic income programs in the United States.
“I think any time you target to low-income people, people in this country are going to find a way to stigmatize it,” she said.
Nonetheless, GiveDirectly continues to move forward with its selective approach for cash support in the United States and to collect new data to support its work. Last year, Illinois selected GiveDirectly to administer the third round of a pilot to help families experiencing homelessness. More than half of 628 families will receive a $9,500 lump sum, while the rest will be in a control group receiving a $500 payment.
This is another population that “if we can do it right, the impact that we stand to have on someone’s life and on their families’ lives is immense,” said Laura Keen, GiveDirectly’s U.S. program director.
Families participating in the pilot are living in shelters, which can strain their physical and mental well-being and well as their education outcomes, increasing the odds that they will be unhoused in the future, Keen said.
“If we’re able to give the amount of money that actually exits people [from the shelter] and gets them into stable housing, the return on investment is really high,” she said.
GiveDirectly expects to have results from that pilot later this year. Ultimately, the organization is trying to push forward the tech-driven strategies that succeeded in Africa and South Asia and prove they work among the American poor.
