Students who took a college-entrance exam more than two weeks after their families received food stamps scored more poorly than those who took the test soon after receiving the aid, a new research paper finds.
The results have implications for public policy, the researchers argue. Other social science research has shown that households tend to increase their food spending and consumption immediately after receiving food stamps, but then taper off before their next disbursement. In effect, the research indicates, students with the misfortune of a bad testing date are probably hungrier and not as able to put their best foot forward on the high-stakes SAT.
While the declines in test scores are modest, they can have very real consequences for low-income students, who are less likely to attend a four-year college right after high school, the research concludes.
The study is among the first to connect the inner workings of vital nutrition policies to specific education outcomes.
“It is a huge life impact for a small, random policy that students had absolutely no control over,” said Timothy Bond, an associate professor of economics at Purdue University and one of four researchers who conducted the study. “We have a quirk with a huge and random impact on people’s long term lives. It’s just sort of stunning to think about.”
It also potentially raises new questions about the relative importance of college entrance exams, which have come under increasing criticism in some quarters.
States, not school districts, control disbursement rules for food stamps, formally known as the Supplemental Nutrition Assistance Program or SNAP. But schools could potentially minimize the effects of SNAP benefit timing by offering the SAT during the week—when low-income students typically receive school meals—rather than on Saturdays.
In all, the findings underscore the link between child nutrition and test scores, even as the COVID-19 pandemic lays bare the extent of food insecurity among Americans.
Fewer students are currently signing up for school meals, probably for logistical reasons amid the pandemic, and more families say they sometimes go without food, troubling signs that have prompted a federal response.
‘Winners and losers’ when test dates and food stamp disbursals don’t align
The research was released as a working paper from the National Bureau of Economic Research. Bond’s Purdue colleague Jillian Carr, as well as Analisa Packham of Vanderbilt University and Jonathan Smith of Georgia, also contributed to it. It has not yet been published in a peer-reviewed journal.
States use different ways to determine which day of the month a family receives its food stamp benefits; some use case ID numbers, while others use the first letter of the family surname. The researchers limited their study to the seven states that use the latter policy.
They collected data on SAT scores, college attendance, and college selectivity data for high school cohorts between 2009 and 2014 from a variety of sources, including the College Board, which administers the SAT.
No database directly connects SNAP and education data, so the researchers used household income information reported by the students, tallies of students who received fee waivers on the exam, and U.S. Census data to determine which students were most likely to be low income and to qualify for SNAP benefits.
Then the researchers used disbursement schedules to see how students who received their SNAP benefits just before an exam date fared on the test, compared to those who received them just after, controlling for race, gender, and other factors that could skew the results.
They found that students in households receiving less than $60,000 in annual income who sat the exam at the end of their benefit period scored about six points lower, an effect size of about .06 of a standard deviation.
That is generally considered a modest effect size, and social scientists have debated how relevant such effects are for public policy. But the researchers make a case for why: The declines were concentrated among lower-achieving students, not higher achieving ones.
“We’re talking about kids who are already at the margin,” Bond said. “These are kids where 10 points could put them above an automatic admission threshold at their state flagship colleges.”
It’s “a small thing,” he said, but the randomness of benefit cycles yields “a very arbitrary selection of winners and losers.”
Small effect has large consequences
Indeed, the researchers found evidence that low-income students who took the exam near the end of the benefit cycle also were more likely to attend a two-year college, less likely to enroll in a four-year one, and enrolled in less selective institutions. They estimated that about 1,150 fewer students initially attended a four-year college in that time period, out of about 170,000 students in the overall sample, because of the test timing.
There are various ways to minimize the harm to students who don’t end up taking the SAT on a date that aligns well to the SNAP benefit cycle. A more complicated one would require states to smooth out benefits, to avoid peaks and valleys throughout the month.
But school districts could also play a role by agreeing to hold the SAT during the school week, rather than on weekends, for the simple reason that low-income students would likely be able to access school meals on those days.
About half of the class of 2020 took the exam on a school day, according to the College Board, and the nonprofit has encouraged schools to offer the SAT during the week; about 10 states also provide the exam at no cost during the school day. 46 percent of school day test-takers also attend high-poverty schools, the organization said.
Testing centers could also consider offering breakfast when the exam is given on a Saturday, Bond suggested, though that would incur some new costs.
The study also has implications for testing research in general.
Other studies have shown that a constellation of factors, like heating and air conditioning, exposure to violent crime, and sleeping patterns, can all affect how students do on measures of achievement. College-entrance exams in particular have been criticized because of the link between family income and scores; the University of California last year made the decision to phase out the exams by 2025.
The pandemic has further upended the industry, with scores of colleges saying they will forgo the scores for college admission in the 2021-22 school year.
But many of the other factors that play into college admissions—selective academic tracks, extracurriculars, essay preparation, and so forth—are also highly dependent on family resources.