Predicting National Rates of Food Insecurity in the Absence of Official Data Collection

According to the United States Department of Agriculture (USDA), more than 47 million people in the United States (14.3%) lived in a food-insecure household in 2023, meaning that they “limited or uncertain access to adequate food.” Their experiences of food insecurity put them at higher risk of several health problems and economic hardships in both the short and long term. Looking historically, food insecurity has also become more common over the past decades, rising by 17% between 2001 and 2023. Food insecurity is thus a vital measure of national economic well-being that is imperative to monitor given its rising prevalence. In September 2025, however, the USDA announced that it would no longer collect the data needed to measure food insecurity, and that the next food security report, scheduled for release on October 22, 2025 and based on 2024 data, would be the last. As of the timing of writing, however, the ongoing federal government shutdown has delayed this final report release. 

The cancellation of future food security data collection and postponement of the 2024 report leave a critical gap in our national data on economic well-being. While the 2024 data could eventually be released, the delay is a preview of the situation we will confront next year once this data is no longer available. This report addresses this critical data gap by presenting a model predicting national food insecurity rates in the absence of the USDA food insecurity data. The model draws on secondary data measuring key correlates of food insecurity – specifically poverty, unemployment, and food-specific inflation – and predicts food insecurity rates at the population level, among children, and among adults. The model’s accuracy is tested across several years, and its predictions align remarkably closely with actual food insecurity rates. The model is also used to predict food insecurity rates for 2024. Predictions for 2024 indicate that food insecurity rates have not fallen, but continued to climb since pandemic-era lows. While continuing to measure food insecurity using the method employed by USDA since 1995 is the only way to guarantee consistent data on this critical indicator, the model presented here may prove useful in estimating food insecurity in future years when this USDA data is unavailable.

Key Findings: 

• The cancellation of the USDA’s food security survey from 2025 onwards, and the 2024 data delay, leaves a critical gap in national indicators of economic well-being. 
   
• A model drawing on secondary data on poverty, unemployment, and food-specific inflation to predict food insecurity rates may prove useful in the absence of this data.
   
• An example model presented in this report yields predictions that align remarkably closely with actual food insecurity rates in years for which such data exists.
   
• Predictions for 2024 suggest food insecurity did not decline between 2023 and 2024, but continued to climb since pandemic-era lows.

Suggested Citation:

Collyer, Sophie. 2025. Predicting national rates of food insecurity in the absence of  official data collection. Poverty and Social Policy Brief, vol. 9, no. 18. New York: Center on Poverty and Social Policy, Columbia University. 

Published on October 22, 2025

This brief largely draws on a model more comprehensively described in a recent working paper.