Give Poor People Cash

There’s a simple way to reform welfare: Send money to those who need it, without conditions.

Before he was a leading contender for the Republican presidential nomination, Ben Carson appeared on The View to talk welfare reform—an issue that continues to be debated in GOP circles on and off the campaign trail. He argued that the social safety net could breed dependency among America’s poor. “You rob someone of their incentive to go out there and improve themselves,” he said, and that’s “not doing them any favors.”

What would be more empowering, Carson suggested, “is to use our intellect and our resources to give those people a way up and out.” That’s surely correct. And the good news is that growing evidence around the world suggests there’s a simple design for a safety-net system that may not create dependency—and may help lift people up and out of poverty: Give poor people cash without conditions attached, and it turns out they use it to buy goods and services that improve their lives and increase their future earnings potential.

It’s a system that policymakers in many countries are loathe to try. They worry, in part, that recipients will waste the money—spending it on, say, flat-screen televisions, cigarettes, and alcohol rather than nutritious food or school supplies. For example, the United States has a very (very) small cash-transfer program called Temporary Assistance for Needy Families, which provides a maximum of $497 per month to a family of four. Even though this cash assistance amounts to only about 8 percent of average household income in the United States, lawmakers frequently feel it necessary to limit how beneficiaries spend the money. Take the Kansas legislature, which in April passed a law specifying that the assistance could not be used to get a tattoo, go to a movie, get your nails done, buy lingerie, or purchase cruise tickets.

It isn’t just the United States. When former Brazilian President Luiz Inacio Lula da Silva introduced a cash-transfer system called Bolsa Familia in 2003—a combination of unconditional cash transfers to Brazil’s poorest families and payments conditional on vaccinating kids and keeping them in school—opposition politicians and newspapers first suggested that the program would be mistargeted and open to fraud. But when evidence suggested the money was well-targeted, mostly reaching poor recipients, opposition moved on to focus on dependency, waste, and disincentives to work. One newspaper editorial bemoaned that “billions are distributed as alms without improving the social status of beneficiaries.”

That concern with handing out money has led to complex systems of in-kind welfare support—or, as Carson put it, “Let me give you housing subsidies, let me give you free health care because you can’t do that.” In the United States, programs range from SNAP (electronic food stamps) and free school meals to Medicaid and rental assistance. In India, the poorest can buy subsidized grains or kerosene. Especially in the developing world, these systems are often inefficient and expensive to run. The Indian government, for instance, has estimated that two-fifths of the kerosene involved in its subsidy scheme goes missing before it is distributed and only half of what is left flows to the poorest families.

But more to the point, the programs are almost certainly less effective at reducing poverty than simply giving poor people cash.

When governments give people in-kind support like food, it frequently costs more to deliver that support than it would to distribute cash—and for the same or even a lesser impact. Jesse Cunha of the Naval Postgraduate School conducted a randomized trial of cash versus in-kind transfers in rural Mexico. In addition to finding that cash recipients didn’t spend more on tobacco or alcohol, Cunha learned that those who received cash experienced the same improvements in nutrition and child-health measures as those who received food. But the food program cost at least 20 percent more to administer, and the cash program led to significantly higher non-food consumption by recipients. In other words: At less cost to the government, cash programs led to the same health outcomes as food-based programs, but also provided additional resources for recipients to spend on schooling, medicine, and transport.

This is not a one-off finding. In many cases, cash programs are simply much more effective than in-kind transfers at turning dollars spent into positive nutritional outcomes. A 2013 survey by Sarah Bailey for the Canadian Foodgrains Bank—involving Zimbabwe, Ecuador, Malawi, and Yemen, among other countries—found that cash transfers usually led to far greater increases in a “food consumption score” of dietary diversity and food frequency than did similarly priced food delivery. In Malawi, the food consumption score increased by 50 percent for cash recipients compared to 20 percent for food recipients. This despite the fact that households in the countries surveyed only report spending between 45 and 90 percent of the cash they receive on food, with the rest going to expenses like debt repayment, household items, and school fees.

Cash also has a larger multiplier effect. Bring food from elsewhere to an area, and the impact of that food stops with those who eat it. Give people cash and they spend it on goods provided by local farmers and traders, who are often poor themselves and benefit as well. A 2010 study in Zimbabwe by Cormac Staunton of Concern Worldwide and Micheal Collins of Trinity College Dublin compared food transfers to cash transfers, and estimated that each dollar provided by cash transfers circulated 2.59 times around the local economy before being spent on goods and services from elsewhere. That compared to the 1.00 multiplier of food that was simply consumed.

Perhaps most important, cash transfers often lead to productive investments. Consider the charity GiveDirectly, which transfers cash from rich people in the West directly to poor people in Africa using mobile-phone payments. A randomized evaluation in 2011, co-designed by a GiveDirectly co-founder, evaluated the organization’s activities in Kenya and focused on one-off, unconditional payments to families that ranged from $404 to $1,520. Four hundred dollars was more than twice the average local monthly household expenditure. In relative terms, it would be the equivalent of handing $12,000 to a household in the United States. As long as 14 months after the transfer, survey evidence suggested that households were still spending more on food, health, and education than non-recipients. One reason why is that they had invested in physical goods, particularly in metal roofs to replace thatched shelter and in livestock to provide milk and meat. That translated into rising incomes from farming and enterprises in the short term, and—thanks to higher spending on nutrition, health care, and education—the hope for greater earnings potential in the long term as well. As in the Mexican cash-transfer program, spending on alcohol and tobacco did not rise after the transfer. Perhaps that’s because recipients felt less need for a pick-me-up: They reported feeling happier after the transfer, and tests of cortisol in saliva revealed lower biomarkers for stress.*

In India, a pilot program between 2011 and 2012 transferred cash—roughly $4 to $6 for adults, and half that amount per child—once a month to every household in select villages in the state of Madhya Pradesh. According to evaluations in 2014 by India’s Self Employed Women’s Association, households in recipient villages proved more likely than those in non-recipient villages to have modern toilets and to use public taps or hand pumps for water rather than wells. They also used cooking fuels that produced less indoor air pollution, which is linked with poor respiratory health. Along with money spent on food, all this helps explain why children in transfer villages were healthier. Program villages saw twice the rate of progress in reducing the number of underweight girls as control villages. The proportion of 14- to 18-year-old girls in school was 65 percent in villages that received transfers, compared to 36 percent in villages not benefiting from the program. As in Kenya, the cash transfers were associated with people working longer hours and making more money thanks to investments in assets including livestock.

The Indian and Kenyan experiments suggest that cash transfers aren’t just a safety net—they increase physical and human capital in poor households. That conclusion is bolstered by work from Columbia University’s Chris Blattman and colleagues in Uganda, where a $150 cash grant to poor women in the northern part of the country doubled their earnings within a year, while one-off $382 transfers to 16- to 35-year-olds were associated with 40-percent higher earnings four years later.

If resources for indirect subsidies from housing through food were redirected toward cash payments to the poorest, more (and more sustainable) poverty reduction could be achieved at less cost.

To be clear, the most generous plausible cash-transfer program would do little to make a poor Kenyan or Indian as rich as the average American. The hundred-fold rise in average incomes that requires would take massive political and institutional reform alongside many billions invested in health, education, and infrastructure. But cash transfers can encourage the world’s poorest to play a larger role in the economic transformation of their countries, helping them evolve from recipients to participants.

The United States, for its part, tried an unconditional cash-transfer program 40 years ago and found it worked, too. The “negative income tax” provided cash to low-income recipients across five states in four different experiments between 1968 and 1980. As in the developing world, the payments were associated with reduced child malnutrition, improved school attendance, and growth in household assets. The transfers also had significant effects on children’s test scores. Unlike outcomes in Kenya and India, the results in the U.S. indicated a small decline in household working hours among beneficiaries. But this occurred primarily among second- and third-earners in a family rather than the primary (usually male) worker, and was concentrated among women who responded to the transfers by taking more time to return to the workforce after having a child.

These programs, alas, were never scaled up, forgotten in the welfare-reform battles of the 1980s and 90s. But is it time to revive the idea—not only in America, but around the world? As a tool to encourage consumption and investment, it should appeal to bleeding hearts and the up-by-the-bootstraps crowd alike.

---

* This article originally stated that an evaluation of GiveDirectly measured the stress levels of beneficiaries through blood tests. They were saliva tests. We regret the error.