Skip to main content
Skip to main content

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Due to the Radical Left Democrat shutdown, this government website will not be updated during the funding lapse.

President Trump has made it clear he wants to keep the government open and support those who feed, fuel, and clothe the American people.

Food Expenditure Series - Documentation

This page provides the following information:

Scope and Coverage

The Food Expenditure Series (FES) provides annual and monthly national expenditures and annual State expenditures estimates, organized by type of product, outlet, final purchaser/user, and valuations.

National Estimates

Coverage: Expenditures are classified into four product types:

  • Food at home (FAH): Packaged food for off-premise consumption.
  • Food away from home (FAFH): Prepared meals and snacks for immediate consumption.
  • Alcohol at home (AAH): Packaged alcoholic beverages for off-premise consumption.
  • Alcohol away from home (AAFH): Alcoholic beverages for immediate consumption.

Each product type is further allocated by outlet type:

  • FAH: Grocery stores, convenience stores, other food stores (specialty stores), warehouse clubs and supercenters, other stores and foodservice, not elsewhere classified (NEC), direct sellers (sales by farmers, manufacturers, and wholesalers directly to consumers for final purchase or use), and home production and donations.
  • FAFH: Full-service restaurants, limited-service restaurants, drinking places, hotels and motels, retail stores, recreational places, schools and colleges, other FAFH sales, NEC (food sold on trains, at hospital and nursing home cafeterias, at veterans canteens, and at office buildings); and food furnished and donated.
  • AAH: Liquor stores, food stores, and other NEC locations.
  • AAFH: Eating and drinking places, hotels and motels, and other NEC locations.

FES distinguishes between final purchasers (who pays for food) and final users (who consumes the food). This distinction is important mainly for food assistance programs such as the Supplemental Nutrition Assistance Program (SNAP), the USDA Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), the National School Lunch Program (NSLP), and the Summer Food Service Program (SFSP). The Government is considered the final purchaser of foods purchased with food assistance funding, and households/individuals are considered the final users of these foods.

  • FAH has three final purchasers—individuals/households, home production, and Government.
  • FAFH has three final purchasers—individuals/households, Government, and businesses.
  • FAH has two final users—individuals/households and Government.
  • FAFH has three final users—individuals/households, Government, and businesses.

Reporting units: Presented with and without sales taxes and tips, and in both nominal (not adjusted for inflation) and constant dollars (with 1988 as the base year). Food expenditures by households are expressed per household and as a share of disposable personal income. Food expenditures by all purchasers are expressed on a per capita basis.

Time series: Annual estimates available from 1997 forward.

Monthly National Estimates

Coverage: Expenditures are classified into two product types:

  • FAH: Packaged food for off-premise consumption.
  • FAFH: Prepared meals and snacks for immediate consumption.

Each product type is further allocated by outlet type:

  • FAH: Grocery stores, warehouse clubs and supercenters, and all other establishments.
  • FAFH: Full-service restaurants, limited-service restaurants, and all other establishments.

Exclusions: Estimates are for sales only and exclude food furnished, donated, home produced, and served at educational institutions.

Reporting units: Presented with sales taxes and tips, and in both nominal and constant dollars.

Benchmarking: Monthly totals in nominal dollars are adjusted so they sum to the annual estimates, except for the most current year.

Time series: Monthly estimates available from 1997 forward.

State-Level Estimates

Coverage: Expenditures are classified into two product types:

  • FAH: Packaged food for off-premise consumption.
  • FAFH: Prepared meals and snacks for immediate consumption.

Exclusions: Estimates are for sales and food served at educational institutions only and exclude food furnished, donated, or home produced.

Reporting units: Presented with and without sales taxes and tips, total and per capita expenditures, and in both nominal and constant dollars (using regional consumer price indexes (CPIs) with 1988 as the base year).

Benchmarking: State totals in nominal dollars without taxes and tips are adjusted so they sum to the annual estimates.

Time series: Annual estimates available from 1997 forward.

Methods and Data Sources

National Estimates

Primary data sources:

  • Economic Census (Census Bureau) – provides detailed sales data by industry, including the percentage of sales for food and alcohol purchases.
  • Annual and monthly Census Bureau surveys – including the Annual Retail Trade Survey, Annual Wholesale Trade Survey, Annual Survey of Manufactures, Service Annual Survey, Monthly Retail Trade Survey, Monthly Wholesale Trade Survey, Monthly Manufacturing Trade Survey, and Quarterly Survey of Services.
  • Other Federal sources – Bureau of Economic Analysis (BEA) Personal Consumption Expenditures, USDA’s Food and Nutrition Service (FNS) program data, USDA’s National Agricultural Statistics Service (NASS) Census of Agriculture, National Center for Education Statistics, Bureau of Prisons, Department of Veterans Affairs, and Amtrak annual reports.
  • Proprietary sources – Circana data used for outlets not fully covered in Census Bureau’s surveys (e.g., commissaries and exchanges).

Adjustments to sales:

To produce accurate estimates of food expenditures, FES applies several adjustments to raw sales data:

  • Nonfood Sales Adjustment: The Economic Census’s North American Product Classification System (NAPCS) is used to separate food and beverage sales from total sales within each industry (North American Industry Classification System (NAICS) code). These shares are also applied to State-level sale estimates.
  • Double Counting Adjustment: To avoid overstating totals, sales from food retailers and warehouse clubs to other businesses (e.g., restaurants, other stores) are excluded. Only sales to final purchasers (households, Governments, and institutions) are counted. This adjustment is applied consistently in national and State-level estimates.
  • Direct Seller Adjustment: While most sales by wholesalers and manufacturers are purchased by other businesses, a small share is sold directly to households or governments. Using the Economic Census’s “Class of Customer” data, FES estimates these direct sales. The same adjustment applies to State-level estimates.
  • Multiple Outlet Adjustment: Some foodservice contractors and concession operators (e.g., at schools, hospitals, recreation facilities, transit terminals) operate across multiple outlet types. Economic Census’s “Miscellaneous Subjects” data are used to reallocate these sales to the appropriate outlet type (e.g., reallocating concessions from outlet type “restaurants” to “recreation facilities”).

Other adjustments:

  • Retail and foodservice sales are augmented by sales taxes and tips to reflect total expenditures. This adjustment applies to State-level estimates.
  • Government food programs (e.g., SNAP, WIC, school meals, commodity donations) are included, distinguishing between final purchasers (government) and final users (households).
  • Home production is estimated using ERS’s Farm Income and Wealth Statistics and household consumption patterns from the National Health and Nutrition Examination Survey (NHANES).
  • Food furnished to employees and the military is estimated with personal consumption expenditures from BEA, while the cost of food furnished to institutionalized populations (hospitals, nursing homes, psychiatric facilities, and prisons) is based on information from the U.S. Bureau of Prisons and various trade associations.
  • Donations (through the Commodity Supplemental Food Program, the Emergency Food Assistance Program, and Food Distribution Program on Indian Reservations) and some Federal food and nutrition assistance programs (e.g., Child and Adult Food Programs, Summer Food Service Program) is based on administrative data from FNS.
  • Other FAFH outlets (commissaries, exchanges, Veterans Canteen Service, higher education, farms with direct sales to consumers, and train food services) are estimated with data from Circana, the U.S. Department of Veterans Affairs, the National Center for Education Statistics, NASS’s Census of Agriculture, and Amtrak annual reports.

State-level Estimates

The State-level Food Expenditure Series adapts the national methodology using employment-based measures.

Primary data source:

  • Bureau of Labor Statistics’ Quarterly Census of Employment and Wages (QCEW), which provides employment counts by NAICS industry for all states.

Estimation approach:

  • Sales are estimated using sales-to-employee ratios derived from six rounds of the Economic Census for each NAICS code.
  • Beginning with the 2024 revision, a regression-based smoothing method interpolates ratios between Economic Census years (replacing linear interpolation).

Revisions and Updates

To improve the accuracy and timeliness of FES, major revisions were implemented in 2018. These revisions include:

  • Inclusion of new source data and updated methods.
  • Introduction of advance, revised, and final estimates.
  • Reorganization of published tables.
  • Benchmarking of the monthly sales series to the annual series.
  • Establishment of a consistent timetable for data release.

Advance, Revised, and Final Estimates

FES produces three types of annual estimates:

  • Advance estimates: Published 1 year after the reference period, based on incomplete source data. Missing information is extrapolated from revised estimates using monthly/quarterly surveys or price indexes (CPI or Producer Price Index (PPI)).
  • Revised estimates: Published 2 years after the reference period, based on more complete data that includes the Annual Survey of Retail Trade, Service Annual Survey, Annual Wholesale Trade Survey, and Annual Manufacturing Survey.
  • Final estimates: Published 3 to 8 years after the reference period, incorporating Economic Census data. Final estimates remain subject to revision if underlying data change.

Adjustments for Industry and Product Changes

Key updates from the 2017 Economic Census, reflected in the 2022 FES update, include:

  • Dissolution of the “Mass merchandisers” NAICS code (452112), with sales reallocated to warehouse clubs and supercenters and other stores and foodservice.
  • Transition from Product Line codes to North American Product Classification System (NAPCS) codes for calculating the share of sales devoted to food and beverages, with minor historical adjustments back to 2012. 
    The Census Bureau published a concordance document to link the product codes previously used to their NAPCS equivalent. The Census Bureau’s What’s New for 2017 Economic Census web page provides additional information.

Key updates from the 2022 Economic Census, reflected in the September 2025 FES update, include:

  • Dissolution of the “Mail in Order and Home Delivery” NAICS codes (4541 and 45439), with sales reallocated to convenience stores, other food stores, warehouse clubs and supercenters, and other stores and foodservice.

Other updates:

  • In 2024, the Bureau of Economic Analysis revised its estimates for income, resulting in the share of disposable personal income spent on food estimates in the FES changing each year since 1997.
  • In 2025, the Census Bureau revised the 2022 Annual Retail Trade Survey, resulting in updated estimates since 1997.

Strengths and Limitations

Accuracy and Limitations

  • Sampling error: Survey-based inputs (e.g., annual and monthly surveys from the Census Bureau) are subject to sampling error. The Economic Census is a complete enumeration of employer establishments and thus has no sampling error, though it excludes nonemployer businesses.
  • Nonsampling error: Estimates may be affected by coverage error (businesses misclassified by NAICS code), measurement error (misreporting of sales), nonresponse error (imputed data for missing survey responses), and processing error (coding or aggregation issues).
  • Forecast error: Advance and revised estimates rely on incomplete data, extrapolations, and price indexes (CPI/PPI). These estimates may differ from final values.
  • Extrapolations and assumptions: Between Economic Census years, some estimates used in the adjustments are interpolated linearly, which may not fully capture structural changes in the economy. State-level estimates use sales-to-employee ratios that are interpolated between Economic Census years using a regression-based smoothing method, assuming stability in productivity across years and States.
  • Coverage limitations: National estimates include food donations, home production, and food furnished to employees and institutionalized persons; State-level and monthly estimates exclude these categories.
  • Timeliness verses accuracy: Advance estimates provide more timely information but are less reliable; final estimates are more accurate but lag by several years.

Strengths

  • Comprehensive: Covers all major components of food acquisition, including food purchased at retail and foodservice outlets, food produced at home, Government food programs, donations, and food furnished to employees and institutionalized populations.
  • Flexibility: Estimates available annually at the national and State levels, with monthly national updates for FAH and FAFH.
  • Multiple perspectives: Data presented by type of product, outlet, and purchaser/user, and available with and without taxes/tips and in nominal and constant dollars.
  • Timeliness: Monthly estimates released with a two-month lag and annual estimates with a one-year lag.

Reconciliation

As part of the comprehensive revision of FES in 2018, the following comparisons were made to reconcile the series.

  • Estimates were reconciled with previously published FES data and with Personal Consumption Expenditures (PCE) and Consumer Expenditure (CE) surveys to assess accuracy.
  • Advance estimates, such as those for 2016, were compared to revised estimates to evaluate errors from using more timely but lower-quality data.

Resources

Relationship to Other ERS Data Products

FES complements other ERS data products, including:

Recommended Citation

U.S. Department of Agriculture, Economic Research Service. Food Expenditure Series.