How the Robin Hood Foundation Uses Data Science to Fight Poverty

January 18, 2017 Scarlett Swerdlow

Two of the most effective federal anti-poverty programs in the United States are the Earned Income Tax Credit (EITC) and the Supplemental Nutrition Assistance Program (SNAP). The EITC, a tax credit for low- to moderate-income working individuals, lifted 6.2 million people out of poverty in 2013, according to the Center on Budget and Policy Priorities. SNAP, formerly known as food stamps, lifted 4.7 million people out of poverty in 2014.

Despite the effectiveness of these programs, roughly one in five people who are eligible do not participate. Nationally, the IRS estimates 20 percent of eligible households did not claim the EITC in 2013, and the USDA estimates 17 percent of eligible households did not enroll in SNAP in 2014.

The Robin Hood Foundation, the largest poverty-fighting organization in New York City, recently approached us with the goal of closing the participation gap in these programs in New York City. They were beginning to plan an ambitious multi-year campaign to enroll more people in these programs, but they needed to know who was in the gap, where they lived, and how to reach them.

Information already exists on who participates in SNAP and the EITC: the agencies that administer these programs know how many people are enrolled, who they are, and where they live. What is unknown is what the larger eligible population looks like—and, importantly, the differences between those who participate and those who do not.

Combining microdata from the American Community Survey with data from our partnerships, and using advanced statistical modeling techniques, we were able to shed significant new light on the eligible nonparticipating population in New York City:

  • Citywide, 76 percent of individuals who were eligible for SNAP in 2014 enrolled in the benefit and 89 percent of households that were eligible for the EITC in 2013 claimed the credit.
  • Some of the communities that have the most-eligible nonparticipants as a share of the population are diverse areas undergoing rapid demographic change. For example, Jackson Heights in Queens is a community where we estimated roughly 13 percent of the population is eligible for but not claiming the EITC. More than half of the population is Hispanic, and about 20 percent is Asian. More than 100 languages are spoken in the community.
  • Communities in which the plurality of the population is Asian, such as Flushing, Queens, also came up as potential target areas, particularly for SNAP. Unlike predominantly African-American or Hispanic communities, where program participation tended to be high among those populations, we found that communities with significant Asian populations tended to have lower participation in SNAP.

The Robin Hood Foundation recently announced its campaign to close the SNAP and EITC participation gaps in New York City. Armed with new insights into the eligible nonparticipating population, Robin Hood is better poised to assist the hundreds of thousands of New Yorkers who are eligible for, but not enrolled in these key programs.

The post How the Robin Hood Foundation Uses Data Science to Fight Poverty appeared first on Civis Analytics.

Previous Article
Super-Fast, Hyper-Accurate Market Research Made Easy
Super-Fast, Hyper-Accurate Market Research Made Easy

Six months ago I sat down with Dan Wagner, our CEO, and had a heart-to-heart on the state of survey science...

Next Article
How We Fine Tuned Lyric Opera’s Marketing Efforts
How We Fine Tuned Lyric Opera’s Marketing Efforts

Like many nonprofits in the arts, Lyric Opera of Chicago faces the familiar challenges of aging audiences, ...