SNAP Math Needs an Overhaul
Outdated formula makes it hard to put food on the table
The Supplemental Nutrition Assistance Program (SNAP) is critically important, yet severely inadequate, as much of it stems from the way the United States defines and measures poverty. Yes, SNAP is the largest and most effective federal nutrition assistance program, providing a crucial lifeline for low-income individuals and families and mitigating the worst of food insecurity for the 41 million people who participate. But it is simply not enough. SNAP benefits tend to run out about 3 weeks into the month, leading to a surge of demand at food pantries as families scramble to find enough food to get through the last week.
The formula that determines the amount that people get in SNAP benefits needs to be overhauled. There is no reason for Americans to go hungry when we have the means to ensure everyone has enough to eat. Here are three key problems with the way the United States defines poverty and calculates SNAP benefits, and my recommendations for fixing them.
The definition of poverty is absurdly outdated.
SNAP benefits are tied to the federal definition of poverty, which is based on a formula developed in 1963 by economist Mollie Orshansky. She estimated the cost of a bare-bones but nutritionally adequate diet–now called the Thrifty Food Plan (TFP)–and assumed that households spend one-third of their income on groceries. From there, she defined poverty as three times this minimal grocery budget.
Poverty, and therefore SNAP benefits, are still calculated in essentially the same way. But the American economy and household finances look completely different now. The assumption that households spend one-third of their income on food is based on data from the 1950s, and has never been updated, despite the sharp rise in the cost of housing, health care, and child care in the intervening 70 years. According to a 2013 NIH report, a realistic definition of poverty would use a much higher multiplier: “based on changes in consumer expenditures, the multiplier should have been four in 1960-61, six by 1988, and around seven today.” In other words, today’s poverty guidelines should be more than twice as high as they are.
Solution: Revising the poverty measure is a complex endeavor, but we should start talking about raising the multiplier to be more realistic, at a minimum.
The poverty limits are the same everywhere in the continental United States.
There is no geographic variation in the poverty limits, aside from higher limits for Alaska and Hawaii. This assumes that households have the same cost of living in Boston, Massachusetts and Boston, Indiana. For reference, the cost of living in Boston, Massachusetts is 50% higher than the national average; in Boston, Indiana it is 17% lower. Saying that poverty is the same in both Bostons is plainly false, but this assumption undergirds the federal definition of poverty, which drives the allocation of hundreds of billions of dollars of federal spending, including SNAP benefits.
Like the poverty definition, SNAP benefits are based on the TFP. The consequence of this is that households in Boston, Massachusetts and Boston, Indiana get the same SNAP food budget, though the Indiana town has food costs 13% lower than the national average and the northeastern hub has food costs 14% higher.
Solution: Benefits should not (and cannot without an act of Congress) be reduced in lower-cost locales, but they should be increased in places with higher food prices. Various tools exist to compare the cost of living in different places, but there is not an official data source or consensus on how to measure this. Developing consensus on measuring local food price variation is step one. In the meantime, the current SNAP benefit algorithm could be tweaked to account for some geographic differences in price.
SNAP benefit calculation discourages work.
The TFP determines the maximum SNAP benefits that households could get. But SNAP households that have other income do not get the maximum SNAP benefit, because their benefit is reduced by 30% of their income, reflecting the assumption that households spend 30% of their own income on food. This is called the benefit reduction rate (BRR).
The BRR discourages work, because every additional dollar earned triggers a 30 cent decrease in SNAP benefits. The earnings deduction offsets this partially. The amount of a family’s income that counts in the SNAP benefits calculation is reduced by a 20% earnings deduction—for every dollar earned, only 80 cents is counted when calculating SNAP benefits, thereby increasing benefits. But the 30% BRR cuts benefits by even more.
What would SNAP benefits be if they reflected a more realistic assumption about food spending? Using the TFP, let’s imagine a family–the Smiths–has $1,000 in monthly income and needs to spend $976 on food in a month. The way SNAP is calculated, the Smiths’ benefit would be $976 minus the BRR (30% of income, or $300): $976 - $300 = $676. (SNAP benefits incorporate numerous other deductions, like the earnings deduction, that I am not going into here, so the benefit may be higher depending on household circumstances.)
If the Smiths actually spend 14% of their income on food, instead of 33%, then their SNAP benefit would be $976 minus 14% of their income: $976 - $140 = $836. This would give the Smiths $160 more every month, getting them much closer to meeting their real food needs.
The SNAP benefits calculation also assumes that one adult has the time and energy to shop efficiently and prepare healthy meals. The TFP and the BRR both come from a time when one adult (usually a woman) was staying home to raise kids, shop for food, and cook for the family. This is rarely the case today, both because more women have chosen to enter the workforce and because living expenses are too high for many two-parent families to rely on just one income.
The benefits calculation–like the poverty definition on which it is built–is out of date and out of touch. SNAP needs a raise to meet people’s actual food needs, and the calculation needs to be modernized. The calculation should use a much lower benefit reduction rate, reflecting the current reality of household food spending, and participants living in higher-cost parts of the country should get higher benefits.
Solution: reduce the BRR to 14% or lower, reflecting current research on actual household food spending.
These are among the recommendations made by the Institute of Medicine and National Research Council in their 2013 report, Supplemental Nutrition Assistance Program: Examining the Evidence to Define Benefit Adequacy, which concluded that the U. S. Department of Agriculture (USDA) “should evaluate whether there is a need to adjust downward the current benefit reduction rate… to reflect the current purchasing behaviors of U.S. households.” The report also argued that USDA “should evaluate whether there is a need to adjust the design of the net income calculation to better reflect the ability of SNAP participants to purchase food within the boundaries of their incomes.”1
In other words, make the math of SNAP work.
National Academies of Sciences, Engineering, and Medicine. 2013. Supplemental Nutrition Assistance Program: Examining the Evidence to Define Benefit Adequacy. Washington, DC: The National Academies Press. https://doi.org/10.17226/13485.