Food Availability Documentation
The food availability data estimate the amount of food available for human consumption in the United States and Armed Forces overseas by measuring the supply of several hundred foods moving from production to marketing channels. For each commodity, ERS calculates the residual of the total annual supply available by subtracting measurable uses, such as farm inputs (feed and seed), exports, ending stocks, and industrial uses.
The food availability data series is a popular proxy for food trends and is the only source of time series data on U.S. food availability in the country. This data series also provides per capita availability data for hundreds of commodities. This page covers general information, methodological concepts, and a detailed discussion of the supply and use estimates.
- History Behind the Data
- Construction of the Data
- Data Sources
- Level of Measurement
- Strengths and Limitations of the Data
- Sources of Error
- Usefulness of the Data
- Estimating Supply and Use of Major Foods
History Behind the Data
Although USDA has collected and published information on food production since the 1860s, information on food consumption did not appear until much later. Interest in food consumption was stimulated by surpluses in agriculture following World War I. The need for accurate data became apparent in analyzing and administering production planning programs under the Agricultural Adjustment Act of 1933. One objective of these programs was to maintain adequate supplies of food for domestic consumers. Droughts in 1934 and 1936 and consequent fears of food shortages further aroused interest in analyzing the national food supply.
USDA issued its first estimates of per capita food availability in 1941 for use in appraising food requirements and resources in the war emergency. Since then, estimates of the availability of major foods have been published annually with only a few exceptions. Historical series on food availability were developed to analyze long-term trends, shifts in demand, and nutrients provided by foods. Data were estimated back to 1909 for many foods.
Construction of the Data
Food availability data measure the use of basic commodities, such as wheat, beef, and shell eggs for food products at the farm level or an early stage of processing. They do not measure food use of highly processed foods—such as bakery products, frozen dinners, and soups—in their finished form. Ingredients of highly processed foods, however, are included as components of less processed foods, such as sugar, flour, fresh vegetables, and fresh meat.
The food availability series is based on records of annual commodity flows from production to end uses. This involves the development of supply and disappearance balance sheets for each major commodity from which human foods are produced. In general, the total annual available supply of each commodity consists of the sum of production, imports, and beginning stocks. These three components are either directly measured or estimated by government agencies using sampling and statistical methods. For most commodity categories, measurable nonfood uses are farm inputs (feed and seed), exports, ending stocks, and industrial uses.
The amount of food available for human consumption is calculated as the difference between available commodity supplies and nonfood use:
Available commodity supply (production + imports + beginning stocks) - Measurable nonfood use (farm inputs + exports + ending stocks, etc.) = Total annual food supply of a commodity.
In a few cases, supplies for human food use are measured directly and one of the other use components becomes the residual. This is the case for wheat, in which flour production is measurable and available from manufacturers' reports on flour milling and therefore, livestock feed use becomes the residual.
Per capita food availability is calculated by dividing the annual total food supply during a specific time period by the U.S. total resident population plus Armed Forces overseas in a given year:
Total annual food supply of a commodity / U.S. population for that year = per capita availability.
Yearly population estimates are from the U.S. Department of Commerce, Bureau of the Census (Census Bureau). For commodities not shipped overseas in substantial amounts, such as fluid milk and cream, ERS uses the resident population as the base. No adjustments are made for changes in the demographic makeup of the population.
Because food availability data are presented for calendar years, the supply and use tables are adjusted from crop year on January 1 stocks or monthly marketings by producers, to calendar year, using data provided by USDA's National Agricultural Statistics Service (NASS). Crops not marketed by the end of the calendar year are assumed to be marketed during the following year. That is, estimates of ending stocks are used as beginning stocks in the next period. For perishable products such as various types of produce, ending and beginning stocks are not components of the balance sheets.
ERS develops commodity supply and use balance sheets for raw and semi-processed agricultural commodities—wheat, corn, red meat, and fluid milk, for example—that are made into food products. These balance sheets use data from a variety of government and private sources. See Estimating Supply and Disappearance of Major Foods for more information on data sources for each commodity group.
Level of Measurement
Data on various components of the supply and use balance sheets come from different sources, and measurements are not always at the same point in the production and marketing system. Before a balance sheet can be constructed, all components must be converted to a common unit, or primary weight, to measure production. The structure of the marketing system and the availability of data dictate the point in the marketing system where production is measured. For some commodities, such as fruit and vegetables, the primary measurement level is at the farm gate. For others, such as margarine, it is at the processing or manufacturing plant.
Once the primary weight of production is selected, quantities of other components from the balance sheet are converted to the same level of marketing, using appropriate conversion factors. For example, production data for meats are based on slaughter plant data; therefore, carcass weight is the primary weight for meats. Meat imports, usually measured in retail or processed weights, are converted to carcass-weight equivalents before aggregation in the balance sheet.
For many food groups, ERS converts food availability figures from a primary weight to a secondary weight or retail-weight equivalent, using conversion factors that account for further processing, trimming, shrinkage, or loss in the marketing and distribution system. For example, ERS provides estimates of the per capita availability of red meat, poultry, and fish on a boneless, trimmed-weight basis in addition to the estimates on a carcass-weight basis. The boneless weight excludes all bones and some separable fat from red meat, poultry, and fish. However, no adjustment is made in the food availability data series for cooking loss, plate waste, or spoilage.
Strengths and Limitations of the Data
Data are collected by USDA directly from producers and distributors using techniques that vary by commodity. The data are not collected from individual consumers, and thus provide an independent basis for examining dietary trends without the problems implicit in consumer survey data. If waste and other losses in the system are relatively constant over time, the data can provide an independent measure of changes in consumption patterns. Thus, trends in per capita availability data can also be used to test hypotheses that government and general sources of diet and health information are affecting consumers' food choices.
The series measures food supplies available for consumption in all outlets—at home and away from home. It measures food use of basic commodities without identifying all end use products, thereby eliminating common problems with food intake survey data, such as decomposing compound foods like lasagna or beef stew back into commodity ingredients. However, final product forms and consumption locations are usually unknown, and little data exist on supplies of more highly processed foods. In short, relatively good data exist for many food ingredients (for example, flour, sugar, or eggs) but not for common food products (for example, bread, cookies, or beef hot dogs).
In addition, food availability data provide good estimates of the annual per capita availability of kidney beans, for example, but provide no information on the different ways that the beans were processed for consumption (canned or dried), where the beans were marketed (supermarket, hospital, school, restaurant, or food manufacturer), how they were consumed (in burritos, chili, or salad), or how they were prepared (made from scratch or reheated from canned). The data do not show where the food is consumed because the data are derived from food production statistics rather than from direct observations of consumption. The poundage represents food available for consumption from all sources, including grocery stores, restaurants, and government food and nutrition assistance programs.
To correctly interpret the data, it is also important to understand that the data are aggregates for the United States and that no data are available for use as a proxy for consumption within States or regions or by socioeconomic or demographic categories (such as those of the consumer who ultimately ate the food).
Sources of Error
Several potential sources of error may affect interpretation and use of the balance sheets and food availability data. Because food use is generally estimated as the residual of the balance sheets, food use data are subject to the various types of errors present in each of the balance sheet components. Primary sources of error are incomplete reporting, inaccurate conversion factors, and inappropriate estimation techniques. In compiling the data, ERS makes substantial efforts to maintain consistency in methods used to measure availability trends and to avoid introducing new sources of error.
The scarcity of information on the components of supply and use introduces one source of error. For example, data on stocks are not available for some commodities. The only available data for estimating stocks for some commodities come from farmer marketings of crops, and it is assumed that stocks are equal to the proportion of the crop not marketed by the end of the calendar year. Moreover, stocks do not include inventories of retailers and wholesalers because those data are not available.
Perhaps more importantly, the data may overstate the amount of food actually ingested by humans by capturing substantial quantities of nonedible food portions and food lost through waste and spoilage in the home and marketing system. The series also includes unknown quantities of foods that are used as ingredients in processed foods that are exported, such as soft drinks, baked goods, and cereal products. For example, the food supply series for caloric sweeteners includes some high fructose corn syrup (HFCS) used by U.S. beverage manufacturers to make soft drinks for export. As a result of the inclusion of these foods and over-counting, the average calories provided by the food supply are generally well above those needed to meet the energy needs of the U.S. population. Therefore, ERS also provides loss-adjusted food availability data to more closely approximate actual intake (see Loss-Adjusted Food Availability Documentation).
Usefulness of the Data
Per capita food availability data are used extensively for analytical and comparative purposes. Economists use the series to estimate effects of changes in price, income, and information on food consumption. Market researchers use the data to study changes in consumption and market shares for food commodities.
The per capita food availability data are most commonly used as a proxy for actual food intake or consumption. In particular, they are used to:
- Measure the average level of food consumption in the country
- Show year-to-year changes in the consumption of major foods
- Estimate long-term consumption trends
- Assess changes in estimated food consumption relative to major nutrition or policy initiatives
Food availability data are also used to construct two other data series:
- Loss-Adjusted Food Availability: Because the per capita food availability data do not account for all spoilage and waste that accumulates in the marketing system and is discarded in the home, the data typically overstate actual consumption. Therefore, ERS calculates a set of estimates that account for food loss prior to ingestion. Per capita food availability poundages are adjusted for food loss, including spoilage, inedible components (such as bones in meat and pits in fruit), plate waste, and use as pet food. These estimates also include the loss-adjusted daily number of calories consumed (per capita) and the daily food pattern equivalents (previously called servings and MyPyramid equivalents). This second series is considered preliminary because the underlying food loss assumptions and estimates require further improvement.
- Nutrient Availability: This data series is no longer being updated, but data are available in archived tables. USDA's Center for Nutrition Policy and Promotion (CNPP) computed a related statistical series on total nutrients available for human consumption per capita per day. This series combined the detailed ERS estimates of per capita food availability and CNPP information on the nutrient content of foods, including inedible waste, such as bones and excess fat in meat. The resultant time series provided data on the effects of changing food use and composition of nutrients available for consumption. This data series can be found on both the ERS and the CNPP websites.
Estimating Supply and Use of Major Foods
This section describes methods and data sources used for developing the supply and use balance sheets and per capita food availability tables for each commodity and/or commodity group. The composition of each commodity and/or commodity group, the conversion from primary to retail weight, and special problems related to coverage are discussed. The following commodity groups are covered:
- Fishery Products
- Boneless Red Meat, Poultry, and Fish
- Dairy Products
- Added Fats and Oils
- Vegetables and Pulses
- Potatoes and Sweet Potatoes
- Fruit and Tree Nuts
- Added Sugar and Sweeteners
- Coffee, Tea, and Cocoa
- Miscellaneous Beverages
- Spices and Herbs
ERS compiles and publishes supply and disappearance tables annually and quarterly for most red meats: beef, veal, pork, and lamb and mutton. Meat availability estimates include fresh and processed meats used and sold through grocery stores and restaurants.
Meat production data are usually derived from three sources: slaughter under Federal inspection, other commercial slaughter, and slaughter on farms. Data on the number and weight of animals slaughtered under Federal inspection are obtained through meat inspection programs administered by USDA's Food Safety and Inspection Service (FSIS) but are reported by USDA's NASS. NASS also collects slaughter statistics on meat production in plants not federally inspected and on the number and weight of animals slaughtered on farms. Beginning and ending stock data are from NASS. Import and export data are from the U.S. Department of Commerce.
Production data are based on carcass weight in pounds of product at the slaughter plant. Commercial stocks and most imports and exports are on a product-weight basis but are converted to carcass weight for use in the supply and disappearance balance sheets. ERS also converts meat data to retail-weight and boneless, trimmed-weight food availability equivalents. The retail weight measure represents supplies on a retail-weight basis converted from the carcass-weight equivalent basis. The boneless-weight measure excludes all bones but includes separable fat sold on retail cuts of meat. Conversion factors are used to account for processing, trimming, shrinkage, or loss in the distribution system when converting between carcass-, retail-, and boneless-weight measures. In most cases, food availability estimates at the carcass level include pet food because data are not available to separate it from food for human consumption.
The conversion factors for the different types of meat are periodically changed to reflect changing production and marketing practices. For example, the amount of trimmed fat in beef has increased over time. In addition, yield grades have gradually improved, requiring less fat to be trimmed. Yield grades predict the yield of trimmed cuts from a carcass—lower numbers mean higher yields. As yield grades improved, the average carcass weights increased, and both factors imply a strong tendency for cattle carrying less fat. The conversion factor used to estimate retail-equivalent weight from the carcass weight of pork has gradually increased over time, reflecting a reduction in the fat content of hogs.
ERS calculates per capita red meat availability for a specific year by dividing annual total disappearance of a particular type of meat by the Census Bureau's estimate of the U.S. total resident population plus Armed Forces overseas. ERS provides the per capita availability data on a carcass-weight, retail-weight, and boneless, trimmed-weight basis.
Per capita food availability estimates for poultry products (broilers, mature chicken, and turkeys) are published in several sources. Broilers are young chickens of either sex produced for meat. The terms "broilers," "fryers," and "young chickens" are interchangeable. Estimates of per capita availability are published monthly in USDA's World Agricultural Supply and Demand Estimates (WASDE) report. This report contains the latest monthly revisions to the quarterly supply and demand estimates, which form the basis for estimating per capita availability. Historical per capita availability data are reported on a monthly, quarterly, and yearly basis in ERS's Livestock & Meat Domestic Data. Per capita poultry availability estimates are actual estimates of domestic disappearance (implied consumption or availability) using secondary data sources rather than primary observations of individual consumption.
ERS uses a similar procedure to construct supply and disappearance tables for broilers, mature chicken, and turkeys. The first step is to estimate domestic production for the three poultry meats. The domestic production estimates come from NASS's monthly Poultry Slaughter report. This report contains estimates of the domestic production of the three poultry meats on a ready-to-cook (RTC) basis. The estimates for domestic production are multiplied by a coefficient to estimate the amount of production condemned after processing. This estimate is subtracted from overall production to derive net production on an RTC basis.
The second step is to estimate poultry meat products in cold storage at the beginning of the period (monthly, quarterly, yearly). Estimates of cold storage holdings come from the NASS Cold Storage report.
The third step is to estimate poultry meat imports. These estimates are derived from U.S. Census Bureau data. The data, originally in a large number of categories, are aggregated into estimates of imported broilers, mature chicken, and turkeys.
The estimates of net production, beginning stocks, and imports are added together to arrive at the total supply of poultry products available for consumption. Estimates of poultry products exported and ending stocks in cold storage are then subtracted from the total supply figure to estimate implied domestic availability. This figure is then divided by an estimate of the total resident population of the United States plus Armed Forces overseas to derive per capita availability on a carcass-weight basis. This estimate of availability of broilers and mature chickens is a proxy for consumption of whole birds. Since a large percentage of availability is of chicken parts, these estimates are multiplied by a coefficient to arrive at a per capita availability estimate on a retail weight basis. Turkey availability has no conversion factor between RTC and retail weight.
See Amber Waves article "U.S. Per Capita Availability of Chicken Surpasses That of Beef," September 2012.
ERS compiles supply and disappearance tables for eggs using mostly NASS data. To exclude eggs for hatching from the calculations, ERS estimates numbers of hatching eggs from numbers of chicks hatched and a hatch percentage calculated from weekly eggs placed in incubators and chicks hatched. Data on stocks, exports, and imports of dried, liquid, and frozen eggs are reported by product weight, with weights converted to shell-egg equivalents for use in the supply and disappearance balance sheet. While the balance sheet is based on dozens of shell-egg equivalents, data are also available in cases (30 dozen per case) and pounds (1.57 pounds per dozen).
Egg availability includes fresh and processed uses by manufacturers and institutional outlets, such as hospitals, hotels, and restaurants. Egg availability also includes use as a culture medium for lab use because data are not available to distinguish this use from the total estimate for human consumption.
The National Marine Fisheries Service (NMFS) of the U.S. Department of Commerce compiles data on supply and disappearance of fishery products. The total U.S. supply of imports and landings is converted to edible weight, and decreases in supply—such as exports—are subtracted. The remaining total is divided by the U.S. resident population plus Armed Forces overseas to estimate per capita availability. Data are derived primarily from secondary sources and are subject to incomplete reporting; changes in source data or invalid model assumptions may also have a significant effect on the resulting calculation. NMFS publishes separate balance sheets on an edible-weight basis for fresh, frozen, canned, and cured fish, as well as for total fish and shellfish. The series Fisheries of the United States on the NMFS website contains related supply and disappearance data.
Production data for fresh or frozen fish and shellfish from NMFS surveys relate only to commercial landings of major cultured species (for example, salmon, sardines, and tuna). Commercial processors prepare regular reports on canned and cured seafood. Alaskan and Hawaiian production of fresh and frozen fishery products have been included since 1960, consistent with reports for most other commodities. Canned production includes production from Alaska in all years, Hawaii since 1952, Puerto Rico since 1953, and American Samoa since 1954. Cured fishery products have been included from Alaska since 1955 and from Hawaii since 1960. The production data have been included for cultured catfish since 1973, trout since 1991, and salmon, tilapia, striped bass, and shrimp since 1996.
The Census Bureau provides foreign trade data on fishery products. Imports of fresh, frozen, and cured fishery products are adjusted to eliminate duplication resulting from domestic production of canned and cured fish products from imported fish. Exports of fishery products include both domestic and re-exported products.
Data for stocks of fresh and frozen fish and shellfish held in commercial cold storage facilities have been used since 1917. Data for stocks of canned fish were incomplete and use was discontinued in 1999.
Boneless Red Meat, Poultry, and Fish
Since 1986, ERS has developed and published a series on availability of meat, poultry, and fish on a boneless-weight basis. These boneless-weight estimates serve as a proxy for consumption and are mainly used to make comparisons of quantities of meat types consumed. Analysts compare quantities based on boneless rather than retail weight to estimate whether, for example, more turkey is consumed than fish. Data on fish are available only on a boneless-weight basis.
Factors for calculating boneless and trimmed weight were derived from USDA data on the quantity of boneless meat obtained from a carcass. These factors are based on values from ERS's Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products (June 1992, see link below) and current ERS estimates. The conversion factors for different kinds of meat are included in the supply and disappearance spreadsheets for particular meats (see the far right-hand column of Red Meat: Supply and Disappearance). The boneless-weight measure for red meat excludes all bones but includes separable fat sold on retail cuts of meat. Boneless-weight figures for poultry are derived from RTC figures, using USDA food composition data.Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products
Milk's various components are transformed into a tremendous variety of dairy product basics, such as butter, cheese, yogurt, and condensed and dry milk. Dairy products are consumed directly but are also used as ingredients in a vast number of foods. Analyzing the supply and demand conditions for farm milk requires some way of adding dairy products together.
To avoid confusion, ERS aggregates dairy products on a common basis, by choosing a particular component (or a cluster of related components) of milk and adding products based on the level of that component in the product. Any component could work, but milkfat, skim solids or protein, and calcium are the most common bases. Milkfat has traditionally been most commonly used because it is the most valuable component and the least likely to be wasted or fed to animals.
The concept of milk equivalent was derived because a quantity of milk is easier to grasp for most people than a quantity of a component. More accurately, a milk equivalent, or the milkfat basis of a product, is the farm milk required to provide the milkfat in that product. The simplest way to obtain a factor to convert product weight into a milk equivalent is to divide the fat percentage of the product by the fat percentage of farm milk. For example, a fat content of 27.5 percent in Swiss cheese and 3.67 percent in farm milk generates a factor of 7.49. In practice, many of the conversion factors were derived by more intricate, but conceptually close, procedures.
No single aggregation of products is likely to be satisfactory, at least in the short run. Changes in milkfat-based product markets (like butter) can be quite different from changes in skim-based product markets (for example, nonfat dry milk). For this reason, total dairy product availability is best understood if simultaneously measured by a milkfat basis and a skim-solid basis.
Avoiding double counting
For dairy products, the total is generally less than the sum of the parts. Dairy products commonly are used as ingredients in the production of other dairy products. For example, ice cream might contain fresh milk and cream, condensed and dry milk, buttermilk, whey, and butter. Unless extraordinary measures are taken to adjust for duplication, adding availability of individual dairy products into total dairy availability results in double counting. An easier and more robust approach is to calculate aggregate availability similarly to individual product availability (see Dairy products: Per capita availability). Stocks, trade, and the other factors needed for the calculation are first aggregated into totals that are free of duplication (because the components can only be in one product at a time), and then total availability is calculated. USDA's NASS estimates milk production and stocks; the Census Bureau reports imports, exports, and shipments to U.S. territories.
Storable dairy products
Availability of most storable manufactured dairy products is estimated by relatively simple food disappearance calculations (see commodity supply and disappearance tables for American cheese, other cheese, total cheese, condensed and evaporated whole milk, nonfat dry milk, and butter). Disappearance estimates for these products generally involve few interpretation problems as most manufactured dairy products undergo relatively little further processing. The traditionally high cost of these products and their straightforward marketing flow leads to relatively minor wastage between manufacturing and consumer purchases. For example, a considerable amount of cheese is trimmed off rectangular blocks that are cut into specialty shapes such as "longhorns," but this trim is then used in processed cheese products.
Perishable manufactured products
Availability of perishable manufactured products, such as ice cream or cottage cheese, is set equal to production. Two problems exist with this approach with no pragmatic alternative. First, stocks and trade may be significant, particularly for ice cream. On an annual basis, the magnitude of error is probably fairly small but could be sizable for shorter periods. Second, spoilage occurs in the distribution channels. At one time, waste was considerable; however, longer shelf life, better packaging, and improved refrigeration have mitigated losses considerably.
Sales of fluid milk, cream, and specialty products
Data for sales of fluid milk, cream, and specialty products are compiled from Federal and State regulatory sources and estimates of the very minor amounts of unregulated milk (see Dairy (fluid milk and cream): Per capita availability). For beverage milks, the data represent quantities sold by fluid processors net of any returns from retailers. At one time, returns were quite significant, but improved raw milk quality, better pasteurization, and improved distribution have substantially reduced the number of returns. Beginning in 2000, availability data for fluid milk, cream, and specialty fluid items changed from a net sales basis to a production basis.
See an Amber Waves data feature "Trends in U.S. Per Capita Consumption of Dairy Products, 1970-2012."
Also see an Amber Waves article "Americans' Dairy Consumption Below Recommendations" and "Cheese Consumption Continues to Rise," and "Behind the Data: Measuring America's Cheese Consumption."
Added Fats and Oils
Due to the termination of select Current Industrial Reports (CIR) by the U.S. Census Bureau, data on added fats and oils (except butter) cannot be updated in the Food Availability Data System.
In 2000, the number of firms reporting vegetable oil production to the U.S. Census Bureau increased noticeably. As a result, estimates of the availability of salad & cooking oils and shortening spiked. ERS surmises that it is unlikely that this unusually large increase in the number of firms occurred in one year; it is more likely that the number increased incrementally over time. The spike suggests that the availability estimates in a few years before 2000 could have been somewhat underestimated. However, ERS does not have additional information from the U.S. Census Bureau to support or clarify that supposition. Therefore, the data must be interpreted with care when discussing particular estimates for salad & cooking oils and shortening around 2000 or when discussing aggregated numbers based on these estimates, such as estimates for total vegetable fats and oils, total added fats and oils, and total calories from added fats and oils and from all foods. For more information, see page 13 of the following publication:Dietary Assessment of Major Trends in U.S. Food Consumption, 1970-2005
ERS constructs supply and disappearance tables for oilseeds, such as soybeans, cottonseed, sunflower seed, canola seed, and peanuts, and for the primary oil and meal products derived from oilseeds and animal sources. Data for oil crop products are kept on an October-September crop-year basis. These data are published in the Oil Crops Yearbook. Data for stocks and crush of oilseeds and the supply and disappearance of oilseed products are derived from Current Industrial Reports of the U.S. Department of Commerce.
The U.S. Department of Commerce also provides monthly information on use of primary fats and oils in related products. ERS summarizes the data according to primary oil products. For example, soybean oil is distributed among various processed products such as margarine and shortening. ERS estimates the amount of each primary vegetable oil and animal fat used in the production of margarine and publishes the results in its Oil Crops Yearbook. Analysts use data on the distribution of primary oils in processed products in studying the demand for particular oil crops. The summary by final product is often used in estimating changes in the fatty acid content of the fats and oils products consumed in the United States.
ERS also compiles supply and disappearance data for the major manufactured fats and oils products, including margarine, edible tallow, lard, shortening, and salad and cooking oils. These data are used in the ERS Food Availability Data System to develop per capita availability estimates. Food use data include availability of fats and oils from all sources, whether purchased by consumers or used by manufacturers and restaurants to produce bakery and other food products. Food disappearance figures for lard and tallow reflect only direct use by consumers, restaurants, institutions, or manufacturers. Indirect use of lard and tallow in margarine and shortening is accounted for in the disappearance figures for margarine and shortening, which helps avoid double counting when estimating total food fats and oils disappearance. For example, the Food Availability Data System reports estimates for margarine and shortening but does not report estimates for the soybean oil, other vegetable oils, and animal fats used to make these products because that would double count the amount of added fats and oils available for consumption.
Disappearance may not be a reliable indicator of change in consumption of fats and oils. Evidence suggests that the waste (or nonfood use) portion of fats and oils disappearance has increased during the past three decades with the growth in away-from-home eating places, especially fast food places. Foodservice establishments that deep fry foods can generate significant amounts of waste grease, referred to as "restaurant grease." A study by SRI International indicated that the quantity of used frying fat disposed by restaurants and made available for use in animal feeds, pet foods, industrial operations, and for export amounted to about 6 pounds per capita, or about 10 percent of the total disappearance of food fats and oils in that year.
Data on the supply and disappearance of peanuts come from USDA's NASS reports and from trade data compiled by the Census Bureau. Annual production data are reported in the NASS publication Crop Production. The total supply for each peanut marketing year (August to July) is the sum of production, imports of shelled and in-shell peanuts, and the beginning stocks for that year as reported in the NASS publication Peanut Stocks and Processing.
Peanut availability (use) data are broken down into exports, seed and residual use, peanuts crushed for vegetable oil and protein meal, and the largest category-food use or disappearance. The food use and crush data are reported in the NASS publication Peanut Stocks and Processing. The peanut domestic food use calculation is based primarily on NASS's manufacturers survey data on peanuts used to make peanut candy, snack peanuts, peanut butter, and other products, plus the apparent disappearance of "roasting stock" peanuts (from Peanut Stocks and Processing), with some adjustments for trade. Summary data on peanut supply and disappearance on an in-shell ("farmer stock") basis are reported by ERS in the monthly Oil Crops Outlook. ERS uses a conversion factor of 0.75 to convert in-shell data to a shelled basis.
Vegetables and Pulses
ERS compiles and publishes supply and disappearance statistics for a wide variety of commercially produced fresh vegetables and melons. Although the supply and disappearance tables are essential for industry analysis, their primary purpose is to estimate food disappearance, both in total and on a per capita basis. Largely compiled on a calendar-year basis, the per capita disappearance-or food availability-statistics generated from these tables are published in ERS's bi-monthly Vegetables and Pulses Outlook. Many of the supply and disappearance tables are released annually in the Vegetables and Pulses Yearbook data product.
For decades, USDA, ERS vegetable analysts have projected various components of commodity supply and availability whenever traditional national data from USDA, National Agricultural Statistics Service (NASS) or other crop industry sources were temporarily (and sometimes permanently) unavailable. ERS typically made these projections based on available data from industry, State agricultural bulletins, and census-based interpolations. Although USDA, ERS would prefer official published annual data over alternative estimation techniques, the loss of critical data negatively impacting market analysis or important long-term time series (such as per capita availability) sometimes has required ERS to “fill the gaps” where feasible.
Changes over the years in the availability of data for some vegetable commodities from both USDA, NASS and industry sources influenced the level of granularity of ERS supply and use estimates for some vegetable commodities. An example of this is the loss of industry-supplied stocks data for canned vegetables in the early 1990s. ERS vegetable analysts continued estimates for commodity stocks using “rules of thumb” percentages but discontinued this method after it became too disconnected from the last known data points. Similarly, after USDA, NASS dropped canning and freezing breakouts for processing vegetable annual estimates for technical reasons after 2015, ERS continued to estimate these breakouts using historical relationships. After 2019, ERS discontinued these estimates and folded canning and freezing into “all processing” supply and availability for the affected crops (asparagus, snap beans, lima beans, carrots, sweet corn, green peas, and spinach). Following that merge, new tables for Vegetables (processed) were introduced. The tables contain aggregated data for all processed versions (i.e. canned, frozen, dehydrated) of vegetables.
Because of this data limitation, loss-adjusted food availability estimates for the seven affected vegetables—such as per capita daily quantities, calories, food pattern equivalents (or servings), and food loss at the retail and consumer levels in the United States—cannot be calculated beyond 2019.
More information can be found in the special article “Fill the Gaps: Supplementing Annual Domestic Specialty Crop Production Estimates” in Vegetables and Pulses Outlook: April 2022, VGS-368.
Supply and disappearance coverage for vegetables:
Fresh vegetables: artichokes, celery, iceberg lettuce, spinach, asparagus, collards,* Romaine/leaf lettuce, squash, snap beans, sweet corn, green lima beans,* tomatoes, broccoli, cucumbers, mushrooms, turnip greens,* brussels sprouts,* eggplant, mustard greens,* pumpkins,* cabbage, endive/escarole,* okra,* carrots, garlic, onions, cauliflower, kale,* bell peppers, potatoes, sweet potatoes, and radishes.*
Vegetables for freezing: asparagus, carrots, green peas, snap beans, cauliflower, green lima beans,* broccoli, sweet corn, spinach, potatoes, and miscellaneous.*
Vegetables for canning: asparagus,* carrots, mushrooms, tomatoes, snap beans, sweet corn, green peas, potatoes, beets,* pickling cucumbers, Chile peppers, cabbage for kraut, green lima beans,* spinach,* and miscellaneous.*
Other vegetables: dry edible beans,** dry peas and lentils,* potatoes for chips and shoestrings, and onions for dehydration.
- *Only per capita disappearance or availability data are published for these items; detailed supply and disappearance tables are not available.
- **Dry edible beans consist of supply and disappearance tables for 14 classes of beans aggregated to an all-bean total.
In general, commodity disappearance data are derived by adjusting total production or use for trade (exports minus imports), stocks (inventories), and other uses where applicable (seed, feed, shrink, and storage losses). Disappearance data divided by the total annual U.S. population (including Armed Forces overseas) yield an estimate of per capita availability. Per capita availability data for fresh vegetables are presented on a farm-weight and retail-weight basis.
The primary data sources used in determining vegetable and melon supply and disappearance include NASS (production, frozen stocks, pickling cucumber stocks, census acreage, and onion shrinkage), the Census Bureau (import volume, export volume, and population estimates), and industry sources (processing tomato stocks and onion stocks). The data cover U.S.-produced vegetables and melons for fresh markets, freezing, canning, and dehydrating (onions).
Supply and disappearance estimates for fresh market vegetables can be divided into three categories:
- Estimates based on NASS national production estimates;
- Estimates based primarily on State-supplied production estimates, and/or USDA, Agricultural Marketing Service (AMS) shipments (for example, radishes, eggplant, green lima beans, endive/escarole, and brussels sprouts); and
- Estimates based largely on Census Bureau acreage with interpolated intercensal years (for example, okra, collards, kale, mustard greens, and turnip greens).
Annual fresh-market supply is largely determined by NASS production estimates (except for selected crops where production is estimated by ERS using alternative sources) plus import volume reported by the Census Bureau. NASS fresh vegetable production estimates cover the majority of harvested quantities destined for sale in commercial markets; these data exclude produce from home gardens and output from States that primarily serve local markets for limited time periods.
Prior to the 1990s, imports usually entered the market during the winter and early spring when domestic supplies were low. Although the majority of volume still enters during the Winter-early Spring market window, imports of fresh vegetables (such as tomatoes and asparagus) are increasingly seen outside their traditional season. Onions are the only fresh vegetable for which stocks data exist (supplied by industry). NASS frozen stocks are included in the brussels sprouts estimate because the supply and disappearance table for this dual use (fresh and processing) vegetable covers all uses. Shrink and loss estimates are subtracted from onion and fresh cabbage supply. Although NASS provides annual estimates for onions, ERS estimates cabbage shrinkage and loss as 2 percent of production.
Per capita domestic disappearance or availability calculations for most fresh-market vegetables are straightforward. U.S. imports are added to domestic production to arrive at total supply. Except for onions and brussels sprouts (frozen stocks), stocks are not entered into the supply and disappearance equation for fresh vegetables. U.S. export volume, shrinkage, and loss for onions and cabbage are subtracted from total supply to yield net domestic use. Domestic use is then divided by the July 1 estimate of the U.S. population (including Armed Forces overseas) to arrive at the per capita proxy for consumption.
Vegetables for freezing
The annual supply of vegetables destined for frozen products (excluding potatoes) is largely determined by NASS production estimates. Until 2005, the American Frozen Food Institute (AFFI) estimated the production of frozen vegetables (frozen pack), which was used by ERS to estimate production of carrots and miscellaneous vegetables. The ERS estimate of production of carrots for frozen items is 66 percent of the NASS estimate of carrots for processing. This percentage is based on the average share of NASS carrots for processing accounted for by AFFI pack during 1996–2000.
Since 2004, production estimates for miscellaneous vegetables for freezing are based on 1980–2004 estimates of pack, imports, and ending stocks. Import volume and beginning stocks are added to production estimates to complete the annual supply estimate. Domestic use is then calculated by subtracting export volume and ending stocks from total supply. The miscellaneous category consists of items such as collards, kale, mustard greens, okra, black-eyed peas, pumpkin, rhubarb, summer squash, turnip greens, turnips, and other vegetables. Since the supply and disappearance tables of vegetables for freezing are presented on a fresh-weight basis, all frozen product weight data for imports, exports, pack, and stocks are converted to a fresh-weight basis using conversion factors published in the following report:Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products
Vegetables for canning
The annual supply of vegetables destined for canned products (excluding potatoes) is largely determined by NASS production estimates. For carrots, the NASS production estimate for processing minus the ERS estimate of frozen pack yields the production estimate. No estimate of miscellaneous vegetables for canning exists, so domestic use is estimated simply as net imports (import volume less export volume). Due to program cutbacks since 2001, NASS ceased production estimates for beets and cabbage for kraut, both of which are now estimated by ERS using available State data and Census Bureau acreage. Canning supply and disappearance for vegetables are calculated in the same manner as supply and disappearance for freezing. Imports and beginning stocks are added to production to arrive at total supply, with exports and ending stocks subtracted from supply to yield domestic disappearance.
In the 1980s, the National Food Processors Association began to phase out the reporting of canned vegetable stocks (all estimates were dropped after 1989). Because of processor consolidation, these estimates were dropped to lessen the potential for disclosure of individual firms' operations. Inventory movements provide year-to-year stability in total disappearance and per capita availability estimates. When stocks are dropped out of the supply and disappearance estimate, the year-to-year per capita use series varies substantially because disappearance estimates swing with production adjustments (which move based on stock levels and market prices).
To maintain integrity in the year-to-year disappearance series, ERS began estimating ending stocks for the major canning vegetables based largely on historical patterns between stocks and production. Despite an increased likelihood of errors in these estimations, ERS decided to continue estimating December 31 stocks as a fixed share of production. The share of production represented by ending stocks in 1999 was carried forward beginning in 2000 for sweet corn, snap beans, green peas, asparagus, and cabbage canning. Although stocks data do not appear in supply and use estimates for chile peppers, carrots, spinach, beets, and green lima beans—ERS does include NASS stocks data in estimates for pickling cucumbers. In 1992, the California League of Food Processors, in cooperation with tomato processors, began to report quarterly stocks of processing tomatoes held in California warehouses. These data have been essential for determining national supply and disappearance of processing tomatoes—a crop that accounts for about 70 percent of all vegetables for canning.
Onions for dehydration
ERS compiles calendar year estimates of the supply and disappearance of onions used for dehydration. More precise data became available in 1992 when NASS began to explicitly separate out California onion production for processing in NASS's annual onion estimates. Previously, ERS relied on industry estimates and rules of thumb to determine the share of California's summer storage crop dedicated to processing (virtually all onion dehydrating takes place in California). The weak link in these data, as with data for many canned vegetables, is the lack of finished stocks. Stock estimates contained in the supply and disappearance table represent raw onions to be processed.
The supply and disappearance table is similar to that for fresh onions in that total supply is the sum of NASS production, Census Bureau import volume, and an estimate for beginning dry bulb onion stocks. Domestic disappearance equals total supply less the sum of dehydrated onion export volume, ending dry bulb onion stocks, and shrinkage and loss of raw onions for processing (estimated as 50 percent of the shrink in California's summer onion crop).
Dry edible beans
Dry edible beans cover more U.S. area than any other single vegetable or melon crop. Dry beans is actually a catchall category containing dozens of dry bean classes, including: pinto, navy, Great Northern, light-red kidney, and black beans. Virtually all of these classes constitute separate markets that operate independently of each other. As a result, ERS now maintains separate supply and disappearance tables for each of 14 classes of NASS dry bean production estimates. Total supply is calculated as NASS production plus import volume plus estimated beginning stocks.
ERS largely estimates stocks based on the share of production that is marketed the following calendar year. Thus, ending stocks on December 31, 2007, for example, would be equal to production in 2007 less the share of the crop marketed during September-December—the first 4 months of the 2007/08 crop year. This estimation method is imprecise because dry beans can be held over more than 1 season, and in years of large crops, ad hoc adjustments may be necessary. NASS publishes monthly marketing percentages by State for all dry beans at the close of each crop year. ERS estimates the shares that apply to each bean class based on the State where the majority of the crop is produced. The exception is for beans produced primarily in California (e.g., baby limas, large limas, and black-eyed peas), where stocks are reported by industry.
Net domestic use is calculated by subtracting export volume, seed use (area planted in the following year times an estimate of seed use per acre), and ending stocks (on December 31) from total supply. As is done for most other vegetables, net domestic use is then divided by the July 1 estimate of the U.S. population (including Armed Forces overseas) to arrive at the per capita food availability estimates.
ERS compiles crop year (July 1–June 30) supply and disappearance data for fresh market and processing mushrooms. The calculation of fresh market data follows the same procedure used for most fresh market vegetables, with the exception of the population figure used to calculate per capita use or availability. ERS uses a January 1 population figure for mushrooms, as that date falls in the middle of the mushroom crop year. The procedure for processing mushrooms (used mostly for canning) is similar to that for fresh mushrooms. There are no stock data for processing mushrooms, so data on production and net trade are used to determine net domestic use. The annual NASS report, Mushrooms, provides production data for both agaricus-type mushrooms (the majority of mushroom output) and specialty mushrooms, such as Shiitake and Crimini. Import and export volumes of processed mushrooms from the Census Bureau are converted to a fresh-weight basis using a conversion rate of 1.538 for canned, 1.5 for frozen, and 10.0 for dried/dehydrated mushrooms.
Limitations of vegetable supply and disappearance data series
The first limitation is that disappearance or use cannot be termed "vegetable consumption" per se. Rather, disappearance represents the apparent net use of vegetables produced on the farm. Although the supply and disappearance series does not directly measure what people eat, it nonetheless provides a useful measure of consumption patterns and trends. In addition, the farm weight series is not adjusted for factors such as loss during transportation from the shipping point, shrinkage during retailing (such as spoilage, trimming), and products thrown out unconsumed.
Second, ERS does not capture the entire universe of vegetables produced and/or consumed by Americans. Despite the large number of vegetables and melons included in the supply and disappearance series, coverage is not complete. Many commodities are not included due to a lack of information on which to base a solid estimate—these include fresh green peas, various Asian vegetables (such as bok choy, turnips, and rutabagas), fresh herbs (such as dill and parsley), fresh beets, parsnips, leeks, scallions (green onions), rhubarb, domestically produced greenhouse vegetables, and other specialty and dehydrated vegetables. For canned and frozen vegetables, ERS does maintain a miscellaneous supply and disappearance table to capture the pack of miscellaneous frozen vegetables and to account for net imports of canned and frozen vegetables not specifically estimated.
Third, the information underlying per capita estimates is not always complete. U.S. trade statistics have not consistently included commodity-level detail over time. For example, data for fresh sweet corn exports were not reported by the Census Bureau until 1978; previously, sweet corn was included in a miscellaneous vegetable export category. Thus, the supply and disappearance table used to calculate per capita fresh sweet corn use or availability contains no data for exports prior to 1978. Domestic use may be overstated prior to 1978 by the unknown amount exported.
Fourth, ERS must sometimes estimate principal data crucial to understanding supply and disappearance due to discontinued reporting by primary agencies. In 1989, for example, the industry discontinued reporting pack and stocks of most canned vegetables. Although the tomato processing industry soon resumed reporting raw-equivalent stocks, ERS estimates changes in canned stocks for such crops as sweet corn, green beans, and green peas. If stock changes in the supply and disappearance of these canned vegetables were eliminated, year-to-year variations would likely become wider and more unrealistic.
It has been crucial to fill in missing 1980s data. Following the 1981 season, NASS ceased to report national production estimates for selected vegetables and melons due to budget cuts. National production data were not reinstated for these items until 1992, with the exception of asparagus and cucumbers for pickles, which were both reinstated in 1984.
To continue monitoring as much of the vegetable sector as possible, ERS generated estimates of national production for those commodities dropped from the NASS program in 1982. These estimates were based on data from State Departments of Agriculture (in cooperation with the NASS State office) that continued to collect production information for their State. In many cases, the States that continued to maintain their full vegetable data series in the 1980s accounted for more than half of the U.S. total in 1981. As a result, returning to NASS-supplied U.S. production estimates in 1992 was a smooth transition and required few statistical adjustments.
See Amber Waves article "Meeting Fruit and Vegetables Dietary Recommendations Will Impact Agriculture."
Also see Amber Waves article "Behind the Data: Estimating Per Capita Domestic Use of Head Lettuce."
Potatoes and Sweet Potatoes
Supply and disappearance data are available for fresh and processed potatoes and for all sweet potatoes. NASS provides survey data on production and frozen stocks, and the Census Bureau provides trade data. ERS estimates disappearance of potatoes and sweet potatoes based entirely on production and net trade because data on stocks are not available, except for frozen potatoes. Utilized production data have been available since 1959 for fresh-market potatoes and potatoes destined for key potato products. ERS has published potato supply and disappearance data since 1960—including the farm-weight equivalent of fresh, frozen, canned, chip, and dehydrated potatoes. ERS estimates domestic use of each of these products by adding corresponding imports to, and subtracting exports from, NASS estimates of utilized production. Only in the annual frozen potato supply and use table does ERS add the difference between beginning and ending stocks (derived from the NASS Cold Storage report) to utilized production.
Disappearance estimates for sweet potatoes are available on a farm-weight basis. Domestic use or availability is calculated from production after adding imports and subtracting exports, as well as subtracting estimates of seed and feed use, shrinkage, and loss. Stocks of canned sweet potatoes are directly accounted for through 1989, after which the industry discontinued reporting canned vegetable stocks because of canner consolidation (there were too few firms to allow data to be published). Seed use is estimated by ERS as acres planted (for the coming year) multiplied by an average seeding rate per acre. After use data were discontinued in 1984, feed use, shrinkage, and loss estimates have been assumed to be 5 percent of production. Per capita use of both potatoes and sweet potatoes is calculated as total annual domestic disappearance or availability divided by total U.S. population (including Armed Forces overseas) on July 1, as reported by the Census Bureau.
Fruit and Tree Nuts
ERS compiles and publishes supply and disappearance tables for a broad range of commercially produced fruit and tree nuts. Both total and per capita domestic food disappearance, or food availability, estimate the amount of food available for domestic consumption. These estimates do not represent the actual amount of a specific commodity or product consumed; however, they may be used to make indications about demand trends. Other historical data presented in each of the supply and disappearance tables such as production, stocks (where applicable and available), imports, and exports provide a basis for market analysis of a specific fruit or tree nut industry. Often supply and disappearance tables and/or food availability estimates for specific fruit or tree nuts are presented in the Fruit and Tree Nuts Outlook published five times a year. Annual publication of most of these supply and disappearance tables is through the Fruit and Tree Nuts Yearbook.
Product Aggregates and Weight Equivalents
Because most fruit commodities have versatile food uses, supply and disappearance tables for fruit are available for fresh and major processed products, including canned, dried, juice, frozen, and wine (for grapes only). All fruit included in the ERS food availability series have food availability estimates for fresh use but not all have similar processed uses. Individual commodity supply and disappearance tables for the major processed products are based on the major processed product uses of the fruit and on the availability of reliable and consistent data to support a supply and disappearance balance sheet. For instance, specific citrus commodities (oranges, grapefruit, lemons, and limes) in the ERS food availability series estimate one product—juice—on the processed side because processed citrus products consist mostly of juice. Major noncitrus fruit such as apples and grapes, on the other hand, have food availability estimates for canned, dried, juice, and frozen. Domestic supplies of most tropical fruit commodities included in the ERS food availability series are marketed primarily for fresh use and the lack of data on the processing sector limits supply and disappearance balance sheets to fresh only, except for pineapples (which include canned and juice).
Per capita availability data are presented on a farm-weight basis for fresh fruit. ERS uses various conversion factors to present availability of canned, dried, juice, frozen, wine, and "other" processed fruit on both a product-weight equivalent and a farm-weight basis. These conversion factors are listed as footnotes in the supply and disappearance tables for the various processed fruit products. All availability data on tree nuts for domestic consumption are presented on a shelled basis.
In developing the commodity supply and disappearance tables, data on domestic production for fresh and processing use of fruit and tree nuts are mostly drawn from the Citrus Fruits Summary and Noncitrus Fruits and Nuts Summary published by NASS. Fruit growing wild and in noncommercial areas are not estimated, except for wild (low-bush) blueberries grown in managed low-bush fields in Maine. Shipment data is used to represent domestic production in the supply and disappearance balance sheets for raisins and prune juice. Shipment data for raisins come from the Raisin Administrative Committee and those for prune juice from the Prune Marketing Committee. Data on the amount of packed frozen produce had historically come from the American Frozen Food Institute until they recently ceased reporting this data. Various other sources now provide this data. The Florida Citrus Processors Association provides stocks of processed citrus fruit juices.
The commodity supply and disappearance tables also use NASS data on stocks of frozen fruit. Stocks of processed noncitrus fruit juices are not available from any source. Stocks of tree nuts come from various commodity trade groups. Fruit and tree nut trade data come mostly from the Census Bureau, except for data on exports of almonds (from the Almond Marketing Board), fresh cranberries (from the Cranberry Marketing Committee), and dried prunes (from the Prune Marketing Committee).
Data Limitations and Adjustments
Reporting of domestic production ceased for mangos in 1998, limes in 2003, and pineapples in 2007. Imports comprise the vast majority of available supplies of these tropical fruit in the United States. The mango and lime industry in Florida did not recover from the losses from 1992's Hurricane Andrew. Many trees were not replaced and production continued to decline significantly in the following years, leading the Florida NASS Field Office to drop mangos and limes from the annual commodity production survey. Reporting was also recently discontinued for pineapple production in Hawaii to avoid disclosure of individual operations as the local industry now consists of one major corporation.
Industry tabulation and publication of canned fruit inventories ceased in 1988. Estimates of the availability of canned fruit for consumption tend to follow an alternating pattern, increasing one year and decreasing the next. Without reported data, ERS will have difficulty making quantity comparisons between categories of processed products. Certain other valuable fruit data have also become unavailable for use in the balance sheets. The Pineapple Growers Association of Hawaii stopped furnishing information on canned pineapples and juice in August 1982.
In 2003-04, the Prune Marketing Committee stopped reporting prune juice and concentrate shipments separately. Now prune juice shipments are reported under "byproduct for manufacturing," which includes shipments of other byproducts such as baby food. Because prune juice shipments have historically made up a major share of all byproducts for manufacturing, prune juice shipments dictated trends in manufacturing byproducts. Thus, year-to-year changes in manufacturing byproducts are being used to estimate current prune juice shipments.
The 2004 Frozen Food Pack Statistics was the final release of annual pack statistics from the American Frozen Food Institute (AFFI). ERS has replaced some of these estimates with annual NASS figures on production for freezing, such as for raspberries, apples, apricots, tart cherries, peaches, and plums. Specific conversion factors are applied to the NASS figures to convert farm-weight production to frozen-weight equivalents. In the case of strawberries, frozen pack statistics now come from the Processing Strawberry Advisory Board of California. Other AFFI pack data cannot be directly replaced by other data sources, requiring ERS to extrapolate estimates based on year-to-year rates of change in the specific commodity's total processed production reported by NASS. ERS extrapolated estimates for blackberries, blueberries, and boysenberries. For sweet cherries, the year-to-year rate of change is based on production for other processed products. With changes in data sources, adjustments to several of the historical time series have been incorporated.
For some fruit, quantities used in processing products such as jam, jelly, vinegar, wine, and juice are very small and are not listed separately in processing reports. For apples, sweet and tart cherries, and peaches, production of jam and other minor products is listed as "other" processing uses, whereas for grapes, it is listed under production used for juice. Production of these minor items is excluded from the supply and disappearance table, except for apples (which has an "other" supply and disappearance table) and grapes (which is incorporated in the grape juice supply and disappearance). With grapes being the only fruit having a significant proportion of production going into the manufacturing of wine, grapes for wine form a separate processing category apart from the major categories—canned, dried, juice, and frozen. However, in this data system, wine estimates in the beverage file are from the Wine Institute.
In 2004, NASS included fresh sliced production among the processed product categories for apples. However, a supply and disappearance balance sheet cannot be created because there is no data for imports and exports. Presently, fresh sliced apple production is combined with other processed apple production.
See Amber Waves article "Almonds Lead Increase in Tree Nut Consumption."
Data on supply and disappearance of grains are organized by primary use. ERS maintains balance sheets for the major food grains (wheat and rice) and the major feed grains (corn, barley, oats, and sorghum). ERS also maintains balance sheets for rye through USDA's World Agricultural Outlook Board. Food availability data are presented as grain equivalents. USDA's NASS, the Census Bureau, and other government agencies provide the data used to construct the food grain supply and disappearance tables.
ERS maintains supply and disappearance data for five major classes of wheat: hard red winter, soft red winter, hard red spring, white, and durum. These data are published in the Wheat Outlook and Wheat Yearbook tables and are compiled on a marketing year basis (June-May). Data on production and stocks are collected by NASS. Food use of wheat is derived from Census Bureau data on production of wheat flour; the flour data are adjusted for imports and exports of wheat products from Census Bureau data. For more information, see the topic: Wheat: Estimating Wheat Supply and Use.
Due to the termination of select Current Industrial Reports (CIR) by the Census Bureau, data on durum flour cannot be updated beyond 2010. The absence of data for durum flour is not critical in the Food Availability Data System since data is still available at a higher level of aggregation (for example, wheat flour).
Per capita estimates for rice are unavailable beyond 2010 due to a large and unexplained decline in the implied total domestic and residual use estimate. Residual use accounts for all unreported losses in the milling, transporting, and marketing of rice, and also offsets any statistical error in another supply and use account.
Data on U.S. rice production and stocks by State and class from 1934 to 2010 are reported by NASS. Trade data are reported by the Census Bureau. Estimates of domestic rice use are derived from several sources. First, seed use for next year's crop is calculated as expected planting multiplied by the seeding rate. Seed use is reported as a separate use category in the rice balance sheet. The rest of domestic use is reported as a single term—Food, Industrial, and Residual (FI&R).
The FI&R term is calculated to equate total supply with total demand for all market years in which NASS released a survey-based estimate of actual ending stocks. For historic market years, the FI&R term is calculated to equate the sum of beginning stocks, imports, and production with the sum of domestic disappearance, exports, and ending stocks. Domestic disappearance equals the FI&R term plus seed use. Seed use is calculated by multiplying the next year's planted area by the per acre average seeding rate.
Prior to the release of the ending stocks estimate by NASS in late August (completing the August 1 to July 31 market year), the FI&R term is forecasted from a statistical model. The forecasted FI&R term is based on historic FI&R estimates and expectations regarding U.S. population growth and ethnic composition, changes in per capita rice consumption, price movements, and income levels.
USDA does not report separate estimates for the three components of the FI&R term. Only the aggregate FI&R term is an official USDA estimate. USDA does, however, develop internal estimates for all three FI&R components—food use, industrial use, and the residual—to assist in forecasting the FI&R term prior to the release of the ending stocks data by NASS in late August. After the release of the ending stocks estimate, the FI&R term is revised to equate total supply and total use in the U.S. rice balance sheet.
Data from two non-USDA sources are used to support internal USDA food and industrial use estimates and to justify any revisions. First, monthly shipments of rice for use in manufacturing beer—the bulk of industrial use—are reported by the U.S. Treasury Department's Alcohol and Tobacco Tax Trade Bureau. (In this data system, per capita beer estimates are from the Beer Institute.) A substantial time lag occurs between actual shipment of the rice to U.S. brewers and the release of the data.
Second, data on U.S. milled rice shipments for domestic food uses (including direct food use, processed foods, and pet food) are available from an annual survey of U.S. rice mills and repackagers by the USA Rice Federation. The survey is reported in the Federation's annual U.S. Rice Domestic Usage Report. Data from the annual milled rice distribution survey are used to support historic USDA internal food use estimates and to justify any revisions.
Domestic food use estimates reported in the survey typically do not match USDA's internal food use estimates due primarily to lack of survey participation by some U.S. rice mills. There is a substantial time lag between the end of a market year on July 31 and the release of the milled rice survey data. The survey data are not directly used in forecasting the FI&R term.
The final component of the FI&R term is the residual, which—for market years with a NASS-reported ending stocks estimate—is calculated so that when added to internal USDA food and industrial use estimates (to yield the FI&R term), total supply will equal total use. The residual includes unreported losses in handling, processing, and transporting, as well as statistical errors in any component of supply and disappearance. ERS provides annual FI&R estimates and further information on domestic rice use in its monthly Rice Outlook report.
Use of food grains for feed and alcohol production is estimated as the residual component of the balance sheet and thus is subject to errors in other balance sheet components. ERS compiles supply and disappearance balance sheet tables quarterly for corn, sorghum, barley, and oats. In 2007-08, livestock feed and residual accounted for about 57 percent of the total domestic use of these four feed grains and for 46 percent of total use. NASS publishes estimates of feed grain production in its monthly Crop Production reports. Stock estimates are included in its quarterly stock reports.
Feed grains are processed into a number of food and nonfood products. Corn, for example, is processed into many food and nonfood products, often in the same manufacturing process. Some products, like cornstarch, are used by both food and nonfood industries in further manufacturing. ERS estimates food and industrial use from census data and other sources. The nonfood use of feed grains includes quantities for processing into beverages and industrial alcohols, industrial starches, and for seed and feed. About 83 percent of the starch production is purchased for industrial uses.
Use of oats and barley for food is derived from Census Bureau reports on production of final products, and industry estimates augment these reports. Feed grains and rice used for alcoholic beverages are estimated from U.S. Department of Treasury data.
Per capita disappearance data for grain products are reported for several levels in the manufacturing process. In the balance sheets, food use is presented on a grain-equivalent basis. These are inexact estimates of food available for consumption. Wheat flour and rice data are measured at the point of milling and include food use in all forms, whether purchased directly or consumed as bread, cereal, or other processed products.
Data on the production of some processed grain products are available from the Census of Manufacturers. To derive estimates of the food available for consumption, ERS adjusts the production figures to account for imports and exports. Products estimated in this manner include corn flour and meal, and hominy. The data are interpolated between 5-year census intervals. In the ERS per capita availability data, grain products include wheat flour, rye flour, rice, barley products, and corn products.
See Amber Waves articles "Will 2005 Be the Year of the Whole Grain?" and "Americans' Whole-Grain Consumption Below Guidelines."
Also see Amber Waves article "Consumer Preferences Change Wheat Flour Use."
Added Sugar and Sweeteners
Since 1941, ERS has estimated annual U.S. total and per capita availability of caloric sweeteners. The data series comprises dry-weight availability estimates of refined cane and beet sugar, corn sweeteners, honey, and edible syrups.
The estimates are based on deliveries of sweeteners by processors, refiners, and importers to U.S. food and beverage manufacturers, institutional users, wholesalers, and retailers. Food and beverage manufacturers use the sweeteners in processed products ranging from candy and soft drinks, catsup, yogurt, peanut butter, and boxed rice mixes. Food wholesalers and retailers distribute refined sugar, honey, maple syrup, and molasses for individual and household use.
ERS relies on estimates of refined cane and beet sugar deliveries published by USDA's Farm Service Agency (FSA) in Sweetener Market Data. These estimates include sugar refined from domestic and imported raw sugar as well as refined sugar imports. As required by law, all sugar beet processors and sugarcane refiners in the United States and Puerto Rico provide FSA with monthly reports on deliveries of refined sugar. USDA's Foreign Agricultural Service provides FSA with estimates of refined sugar imports.
ERS estimates deliveries of corn sweeteners (high-fructose corn syrup, glucose, and dextrose) for domestic food and beverage uses (excluding nonfood uses), using information from industry contacts, consulting firms, and Census Bureau import data.
ERS divides total deliveries of various sweeteners by total U.S. population to estimate per capita deliveries. Estimates of per capita delivery help determine whether Americans, on average, are consuming more or less added sugars over time.
Due to the termination of select Current Industrial Reports (CIR) by the Census Bureau, data on candy & confectionary products can no longer be updated. The absence of data for candy & confectionary products is not critical in the Food Availability Data System since these products are represented by individual sugars and sweeteners (for example, cane and beat sugar).
See Amber Waves article "High-Fructose Corn Syrup Usage May be Leveling Off."
Also see Amber Waves article "Behind the Data: Estimating Consumption of Caloric Sweeteners."
Coffee, Tea, and Cocoa
Except for small quantities of coffee grown in Hawaii, the United States does not commercially grow coffee, tea, or cocoa. Thus, imports supply virtually all U.S. needs for these tropical products. Since stocks data for coffee, tea, and cocoa are no longer available, supply and disappearance tables for these items include only net changes in stock levels rather than estimated beginning and ending stock levels as previously shown. The net change in stocks is estimated as a residual.
ERS estimates coffee supply by adding Hawaiian production and U.S. imports. Food availability, as a proxy for consumption, is estimated by adding domestic roastings and net imports of roasted and instant coffee. The balance sheet is reported on a green bean weight basis. Net imports of roasted coffee are converted at 1.19 pounds of green beans for 1 pound of roasted coffee. Instant coffee is converted at 2.5 pounds of green beans for 1 pound of instant coffee. Larger conversion factors were used in earlier years when the processing of instant coffee was less efficient. Per capita availability data are published on a green bean weight and retail weight basis. Retail weight is the roasted or instant weight as sold in retail stores.
Per capita availability data for all tea is on a leaf-equivalent basis. It takes about 2.5 pounds of tea leaves to make 1 pound of instant soluble tea. The supply of tea, which is based on U.S. imports, includes all forms of black tea, tea bags, instant tea, and tea mixes. Herbal teas are excluded. Disappearance is derived from the difference between imports and exports because there are no stock data for tea. This measure tends to fluctuate more than tea consumption would be expected to fluctuate because imports tend to be erratic. Therefore, ERS estimates tea availability by subtracting exports from imports and assuming that disappearance for each year is equivalent to a 3-year moving average of imports minus exports.
ERS estimates supply and disappearance of cocoa (bean equivalent), using import data for product forms such as beans, chocolate liquor, cocoa butter, cocoa powder, and sweetened products. It is assumed that 1 pound of unsweetened chocolate is obtained from 1.25 pounds of cocoa beans. Chocolate liquor contains about 53 percent cocoa butter (fat) and 47 percent cocoa powder (nonfat solids). Cocoa powder is converted to a bean equivalent, using a factor of 1.18, and cocoa butter, using a factor of 1.33.
Cocoa bean availability is estimated as the U.S. annual cocoa bean grind, plus net imports of semi-processed products (unsweetened chocolate, cocoa powder, cocoa butter) and consumer products. Per capita cocoa availability is published for both a whole-bean and chocolate liquor basis, which is 80 percent of the weight of the beans. Retail weight is the weight of the chocolate liquor.
See Amber Waves article "Coffee Consumption Over the Last Century."
All beverage data are presented in gallons per capita. ERS converts fluid milk and juice data from pounds to gallons, using factors from ERS's Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products (June 1992, see link below). Coffee is converted to fluid equivalents on the basis of 60 6-oz. cups per pound of regular roasted coffee and 187.5 6-oz. cups per pound of instant coffee. ERS assumes a conversion rate of 200 6-oz. cups per pound of tea leaf equivalent.Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products
Spices and Herbs
Most U.S. supplies of herbs and spices are derived from net imports (imports less exports) of over 20 spices plus a miscellaneous group, as reported by the Census Bureau. The remaining supply comes from the domestic production of mustard seed and dried chile peppers. Small amounts of domestic production of other spices are not included in the total. ERS assumes that all annual production is consumed the following year, with no allowance for changes in stocks of imported spices because there are no estimates of stocks.
Written by ERS commodity analysts (Mark Ash, Tom Capehart, Jerry Cessna, Nathan Childs, Kamron Daugherty, Russell Knight, Michael McConnell, David Olson, Broderick Parr, and Agnes Perez) and by Alan Lowther from NMFS for seafood. Parts of this documentation are adapted from the 1989 edition of Major Statistical Series of the U.S. Department of Agriculture: Consumption and Use of Agricultural Products.
For more information, see Readings, FAQs, and Glossary.
Download Excel spreadsheets of the Food Availability data and Loss-Adjusted Food Availability data.