It’s a conundrum faced around the world: Why do some government-sponsored cash assistance programs contribute to unhealthy behaviors? A series of studies by Stanford Medicine epidemiologist Sanjay Basu, MD, PhD, has revealed a simple answer: Payments made to recipients are often ill-timed. By demonstrating the big impact of small changes to payment systems, Basu is helping governments at home and abroad to improve the health of their low-income citizens — changing the broader discourse around nutrition assistance programs in the process.
Basu began to explore the issue after a 2013 international meeting on chronic disease prevention, where he met representatives from several middle-income countries with an alarming rise in high blood pressure and diabetes. Among those countries was Peru, which became one of Basu’s major project sites. He and former postdoctoral scholar Justin White, PhD, started to examine the effects of cash payment frequency on the purchase of “temptation goods” — including alcohol, tobacco and sweets — that are typically associated with chronic diseases like diabetes, hypertension and high blood pressure. Basu and White, now an assistant professor at UC-San Francisco, paid particularly close attention to these purchases by low-income Peruvians who live in rural areas and eke out an existence by farming the steep hillsides.
“We’re very interested in exploring this issue in countries like Peru and India, because they’re especially well-positioned to take full advantage of prevention programs,” says Basu, an assistant professor of medicine at the Stanford Prevention Research Center. “They’re middle-income nations in a state of so-called epidemiologic transition from infectious to chronic diseases. Yet many chronic conditions haven’t fully ramped up in these areas, so prevention measures can be really effective.”
Basu and his team analyzed a recent change in the payment schedule of Peru’s cash assistance program: The government moved from twice-monthly to once-monthly payments, without changing the overall payment amount. Under the new schedule, the team observed a striking increase in the purchase of alcohol and sweets in particular — nearly 70 percent in some cases. Their results were published in January in the Journal of Health Economics.
“We suspect two behavioral issues may be to blame for this increase,” Basu says. “One, there’s the ‘payday’ effect, where people tend to make larger discretionary purchases around receipt of a regular income stream. And two, going without temptation goods for a longer period of time may lead people to binge once they finally do receive a payment. As they say, never shop on an empty stomach.”
Based on Basu’s results, the Peruvian government is revising its program’s payment schedule to every two weeks — and may even increase the timing to weekly, should the initial results prove positive. Basu is also in conversation with Brazil about adapting the program for the country’s low-income communities.
“This approach is very translatable,” notes Basu, who has now incorporated lessons learned from the Peruvian study into a new program for San Francisco’s low-income neighborhoods — hoping to learn whether more frequent voucher distribution helps to mitigate the purchase of temptation goods in these communities. Starting in the city’s Tenderloin, South of Market, Bayview and Outer Mission districts, participants will receive $20 in food vouchers per month, with recipients randomly split into two groups: those who receive a $5 voucher at the beginning of each week, and those who collect a $20 voucher once a month. All participants will provide Basu with detailed dietary information.
A pilot study of 250 adults from the Tenderloin and South of Market districts, which Basu conducted with UCSF associate professor Hilary Seligman, MD, has already shown positive effects. “Some of the folks in our pilot had chronic conditions that are really hard to control,” Basu says. “We see that these conditions profoundly worsen when people face ‘food insecurity’ — another term for hunger — from inappropriate resources to purchase foods.”
In addition to increased payment frequency, Basu has identified other factors that likely influence outcomes in these types of programs. “The success of patients who receive vouchers in a program like this is often related to other factors in their lives, such as when they have to pay the rent, or receive their work paychecks,” he says. “In other similar programs we’ve studied, we’ve discovered that in many locations like the Bay Area, rent prices vary widely, but voucher programs and other nutrition assistance programs are not adjusted for the cost of living, which profoundly affects what families are able to afford.”
Basu points out that adjusting for these factors could improve nutrition in ways that reduce longer-term costs of caring for diseases like diabetes, which has made federal agencies, like the Centers for Medicare and Medicaid Services, more open to subsidizing these programs. He has also begun working with the U.S. Department of Agriculture to develop another pilot for its Supplemental Nutrition Assistance Program, testing out alternative ways to distribute SNAP’s food stamps.
“This approach has the potential to change the national discourse around food stamp programs,” Basu says. “The National Academy of Medicine has identified a large research gap in translating the results of nutrition studies and poverty studies to workable programs that benefit communities.”
Basu plans to continue to dig into the various political, social, fiscal and other considerations that contribute to optimal health in low-income communities. “For example, the food stamp program is considering an increase in the amount of assistance it provides to low-income residents of high-food-cost areas, so those residents can use the extra assistance for transportation to stores with less expensive food,” he says.
“There’s certainly no one-size-fits all strategy when you’re dealing with the public health aspects of chronic disease,” says Basu. “We’re taking a precision health approach based not on genetic codes, but on ZIP codes — where very different people from very different parts of the world receive ‘treatment,’ in the form of tailored assistance programs, that is specific to their circumstances. If we get to the point where we can customize to the neighborhood or even to the person, that’s very policy amenable.”