Skeptics of work requirements in the Supplemental Nutrition Assistance Program (SNAP) often argue that most families receiving SNAP benefits are already working. A recent NBCNews article, for example, claimed that four out of five SNAP households have at least one working person in the household, and that 10 percent had three or more workers in 2021, citing the US Census Bureau as its source. Other organizations, such as the Center on Budget and Policy Priorities, have produced similar statistics, also sourced to the Census Bureau.
However, a closer look at the data suggests these statistics are misleading. When understood properly, they support the importance of encouraging employment in SNAP.
The main problem with these claims is that they rely on household survey data collected by the Census Bureau, which, according to an expert review, “suffer[s] from an alarming and growing extent of survey error and few studies attempt to correct for this error.” Many recipients simply fail to report receiving government benefits when asked on surveys, with research suggesting that between 25–50 percent of SNAP recipients (depending on the survey) do not report participating, while some non-recipients report receiving it. This misreporting should give researchers caution when presenting the results of any analysis that relies on Census data alone.
To make matters worse, researchers often misrepresent employment among SNAP households even when using Census data by including employment outside of the time the household receives SNAP. For example, the NBCNews claim likely reflects employment at any time in the past 12 months, not necessarily when the household was receiving SNAP. (Although even this is unclear because the authors did not respond to multiple requests for methodological clarification.) In fact, the way the Census Bureau survey asks the survey question, it is possible that the “working person” was not even present in the household when they received SNAP.
Collectively, these problems vastly overstate employment among households while they are receiving SNAP. Below, we show the number of workers per SNAP household according to three different data sources—SNAP quality control (QC) data, the Current Population Survey (Census), and the American Community Survey (Census). SNAP QC data contains a representative sample of real SNAP cases, and shows that less than one-third of SNAP households include a worker (Figure 1). According to Census data, between 55–60 percent of households receiving SNAP at any point in the year contained a working person at the time of the survey. When we expanded the definition of employment to working at any time in the past 12 months, the rate of households with a “working” person comes closer to 80 percent.
Why do these differences matter? It should be unsurprising that many SNAP households contain adults with some work history. But the relevant information is whether household members work while they receive SNAP, and to what extent SNAP contributes to low employment. And given the particularly low work rates documented by SNAP caseworkers in the QC data, policymakers are rightfully concerned that so few are employed while receiving benefits, especially when the economy is strong and employers are struggling to find workers.
There is also an inconsistency in presenting these data to make a case against SNAP work requirements. As already noted, survey error is a problem. But beyond survey error, the use of a broad definition of employment to include work outside the time of SNAP receipt changes the analysis. This also raises the question for opponents of work requirements: If employment is common among SNAP households, why would households struggle to meet work requirements? Burdensome paperwork is the oft-cited culprit, but most working adults should be capable of reporting work hours to a government agency.
Although low-income families face barriers to employment, including disability and caretaking, we recently found that even after accounting for these constraints, a large share of SNAP recipients were not working, not caretaking, and not disabled, raising questions about why employment was not more common.
Because employment is one of the most important ingredients for escaping poverty, policies should encourage work among those SNAP recipients who are able. Policymakers may have different approaches for accomplishing this goal, but we must first agree on the facts—especially when the livelihoods of American families are at stake.