The Supplemental Nutrition Assistance Program (SNAP) serves as a safety net for more than 41 million low-income families, but only about 80 percent of eligible individuals participate.[1] Among SNAP-eligible agricultural workers, take-up is likely even lower.[2] In a recent study, we explored the seasonality of agricultural employment and the extent to which its associated administrative burdens impact households’ SNAP eligibility and participation. To measure households’ attachment to SNAP, we used ‘churn’—exit and subsequent re-entry into SNAP—among Fresno County, CA households. We found a significant negative correlation between SNAP churn and agricultural employment seasonality.

Understanding the relationship between SNAP churn and the seasonality of agricultural employment would likely improve access to the safety net for a vulnerable population, as well as reduce SNAP implementation costs for both households and the state.

DOWNLOAD THIS BRIEF

Key Facts

• Eligible farmworker households are at least one and a half times as likely to exit and re-enter the SNAP program (‘churn’) as non-farmworker households.
   
• Farmworkers are especially relatively likely to churn shortly after peak agricultural employment season, indicating that seasonality in agricultural employment makes SNAP participation more challenging and complex for farmworkers than non-farmworkers.
   
• Policymakers should consider changes to SNAP administration that would likely increase participation among agricultural workers and decrease program costs.

Background

SNAP has documented benefits for participants, including reducing food insecurity.[3] Given the policy objective of SNAP—to ensure adequate nutrition for all individual—and high rates of food insecurity among farmworkers,[4] it is important that all eligible farmworkers participate in SNAP. However, rates of SNAP participation among farmworkers are low—about 13 percent.[5]

Low SNAP participation among agricultural workers can be explained in part by the share of non-citizen and undocumented workers. However, citizens account for 36 percent of farmworkers, suggesting that a mismatch between program requirements and employment structure may be contributing to low rates of SNAP attachment among these workers.

Research has shown that missing paperwork and appointments are substantial drivers of SNAP exits and churn.[6] For example, once enrolled, households are required to re-certify regularly to demonstrate that they still meet the eligibility requirements, submitting proof of monthly income below 180 percent of the federal poverty guideline. This costly and time-consuming re-certification process may be difficult to manage during peak season, when agricultural workers have a particularly high workload and may be working further from home. Higher earnings during busy periods may also make these workers temporarily ineligible for SNAP, only for them to become eligible again when peak season has passed. These factors likely contribute to churn.

Yearly certification costs from churn have been estimated at between 2.6 to 9.3 percent of total annual program costs. Given that the national annual budget for SNAP exceeds $119 billion, this suggests that reductions in churning costs could yield decreases in program costs of between $3.1 and $11.1 billion.

In our study, we took Fresno County as a case study. Fresno makes a good case study of SNAP churn among farmworkers because it has both a large farmworker population and a large SNAP participant population. Fresno County ranks 7th within California for SNAP issuance and participation, representing 0.6 percent of national SNAP participation. Fresno also employs the most agricultural workers of any county in California—the largest agricultural state in the country—representing about 9.5 percent of California’s farm employment.[7] Agricultural employment in Fresno is highly seasonal and pays wages that fall below SNAP gross income limits.

Examining the relationship between seasonality and SNAP churn 

We used data shared by the Fresno County Department of Social Services (DSS), spanning from January 2016 to December 2020. The data included monthly observations of SNAP participation for each individual or family unit that receives SNAP, as well as descriptive information about each family unit. To identify instances of churn, we observed gaps in participation or the assignment of a new application ID for the same case ID. We defined a case as churning if we observed a one-, two-, or three-month gap in participation. We characterized a household as a farmworker household if they are a migrant farmworker or list farming as an occupation. We also used the Quarterly Census of Employment and Wages to observe monthly employment in non-citrus fruit and tree nuts in Fresno County. Non-citrus fruit and nuts—largely peaches—are a key agricultural industry in Fresno County. This sector relies on manual labor and pays wages at or below the SNAP eligibility threshold.[8]

We analyzed the churn rate among farmworker and non-farmworker households, producing odds ratios to quantify the comparative difference. We also highlighted seasonal variation in churn relative to agricultural employment patterns and demonstrated consistency between the timing of churn and the employment patterns of likely SNAP-eligible agricultural workers. We considered the interaction of farmworker churn with employment in crop production and plotted total monthly employment in each quartile against monthly instances of farmworker churn.

SNAP churn more likely among farmworkers than non-farmworkers 

Odds ratios show that farmworker households were at least one and a half times as likely to churn as non-farmworker households (see Figure 1). From October to December—the months in which farmworkers that exit SNAP during the harvest season would be re-entering and thus measured as churning—farmworkers were more than twice as likely to churn as non-farmworkers. The increase in the odds ratio from September to October (after the end of peak harvest season in Fresno County) was very sharp, jumping from 1.8 to 2.5. The odds ratios peaked in November, when farmworkers were 2.6 times more likely to churn relative to non-farming households, falling again after the three-month post-harvest window, to an odds ratio of 2 in January and back to 1.7 in February.

Figure 1. SNAP participation and churn among farmworker and non-farmworker households

We also saw spikes in farmworkers’ churn starting in the month that agricultural employment declined after seasonal peaks. This suggests that farmworkers were re-entering SNAP when crop-production employment was low after the busy work season. Non-farmworkers did not demonstrate the same seasonal patterns as the farmworkers in our data, indicating that the variation is specific to agricultural employment rather than underlying seasonal trends. 

Refine SNAP administration to reduce churn among farmworkers and reduce program costs

Using administrative data from Fresno County’s SNAP program, we examined the relationship between highly seasonal agricultural employment and churn in SNAP. We found a strong negative correlation between agricultural employment and the amount of churn in SNAP. As agricultural employment decreased, the amount of churn in the county increased significantly. Farmworkers’ probability of churning was higher than non-farmworkers’ at all times but was especially higher in the three months immediately after harvest season, when farmworkers were two and half times more likely to churn relative to non-farmworkers. Our findings suggest that the cyclic nature of agricultural employment increases SNAP churn and contributes to low rates of participation among eligible agricultural workers.

Churn is costly to both households and the state. Many farmworkers who churn are likely continuously SNAP eligible. For others, who become temporarily ineligible for SNAP during harvest season due to higher earnings, policymakers should consider annual income, as used in safety net programs like the EITC, rather than monthly income. Outreach to farmworkers at peak season, alternative processes for expediting farmworker applications, or minimizing documentation required for recertification could all mitigate churn caused by administrative burden and/or insufficient knowledge of program rules. Implementing these changes would likely help to ensure food security and financial stability for agricultural workers while also reducing SNAP’s program costs.

Casey J. McNichols is a graduate student in the Department of Applied Economics at the University of Minnesota.

Charlotte E. Ambrozek is an assistant professor in the Department of Applied Economics at the University of Minnesota.

Timothy K. M. Beatty is a professor and chair in the Department of Agricultural and Resource Economics at UC Davis.

References

1. USDA SNAP. SNAP Data Tables, 2023.
https://www.fns.usda.gov/pd/supplemental-nutrition-assistance-program-snap.

2. A. Gold, W. Fung, S. Gabbard, and D. Carroll. Findings from the National Agricultural Workers Survey (NAWS) 2019-2020: A Demographic and Employment Profile of United States Farmworkers. Research Report 16, U.S. Department of Labor, Employment and Training Administration, Office of Policy Development and Research, Jan. 2022.

3. Gundersen, Kreider, and Pepper. 2017. “Partial Identification Methods for Evaluating Food Assistance Programs: A Case Study of the Causal Impact of SNAP on Food Insecurity.” American Journal of Agricultural Economics 99 (4): 875—893.

4. S. Brown and C. Getz. Farmworker Food Insecurity and the Production of Hunger in California. In A. H. Alkon and J. Agyeman, editors, Cultivating Food Justice, pages 121-146. The MIT Press, Oct. 2011. ISBN 978-0-262-30021-6. doi: 10.7551/mitpress/8922.003.0010.
https://direct.mit.edu/books/book/4423/chapter/189391/Farmworker-Food-Insecurity-and-the-Production-of.

5. A. Gold, W. Fung, S. Gabbard, and D. Carroll. Findings from the National Agricultural Workers Survey (NAWS) 2019–2020: A Demographic and Employment Profile of United States Farmworkers. Research Report 16, U.S. Department of Labor, Employment and Training Administration, Office of Policy Development and Research, Jan. 2022.

6. T. Homonoff and J. Somerville. Program Recertification Costs: Evidence from SNAP. American Economic Journal: Economic Policy, 13(4):271-298, Nov. 2021. ISSN 1945-7731, 1945-774X. doi: 10.1257/pol. 20190272.
https://pubs.aeaweb.org/doi/10.1257/pol.20190272.

7. State of California Employment Development Department. Employment by Industry Tables, 2023.
https://labormarketinfo.edd.ca.gov/data/employment-by-industry.html.

8. Fresno County Department of Agriculture. 2022 Fresno County Annual Crop and Livestock Report. Technical report, The County of Fresno Department of Agriculture, Fresno County, CA, 2023.
https://www.fresnocountyca.gov/Departments/Agricultural-Commissioner/Annual-Crop-Livestock-Report.