Documentation

Notes: As of September 28, 2017, changes have been made to the data file “Per capita consumption of selected cheese varieties.” For the worksheet entitled “Per capita cheese consumption of selected cheese varieties since 1995,” the last two columns have been eliminated displaying cheese content of processed products and the quantity consumed as natural cheese. (A file including the last two columns, designated as “old,” will remain on the website until November 28, 2017.) The cheese content could not be justified because (1) some of the processed cheese products are not made from natural cheese and (2) the natural cheese content percentages for processed cheeses made from natural cheeses cannot be adequately estimated. (While U.S. Food and Drug Administration (FDA) standards of identity for processed cheeses require inclusion of natural cheese as an ingredient, not all processed cheeses as reported in the primary data, and reflected in the ERS data file, are made according to an FDA standard. FDA standards of identity affect how products are labeled and marketed; they are not safety regulations.)

U.S. dairy situation at a glance

Basic dairy-related data is published monthly, providing statistics for the most recent 15 months and annual statistics for the past 3 years.

U.S. milk production and related data

The report includes quarterly milk-production data, a proxy for dairy feed prices, and replacement cow prices.

Commercial disappearance for dairy product categories

ERS provides commercial disappearance tables for seven product categories: butter, nonfat dry milk, American cheese, other-than-American cheese, dry whey, whey protein concentrate, and lactose. Commercial disappearance tables include total commercial disappearance, domestic commercial disappearance, and the elements that go into calculating commercial disappearance numbers. Monthly data are provided from 1995 to the present.

Relationships are expressed in the following equations:

  • Total commercial supply = beginning commercial stocks + production + imports 
  • Net Government removals [1] = price support purchases through the Dairy Products Price Support Program + exports under the Dairy Export Incentive Program – unrestricted sales of Government stocks 
  • Total commercial disappearance = total commercial supply – net Government removals – commercial ending stocks 
  • Domestic commercial disappearance = total commercial disappearance – commercial exports
Data sources, commercial disappearance for product categories
Commercial disappearance element Source
Commercial stocks 1 USDA, National Agricultural Statistics Service (NASS); USDA, Farm Service Agency
Production USDA, National Agricultural Statistics Service
Imports U.S. Department of Commerce, Census Bureau
USDA net government removals  
Price support purchases and unrestricted sales
USDA, Farm Service Agency
Dairy export incentive exports
USDA, Foreign Agricultural Service
Commercial exports 2 U.S. Department of Commerce, Census Bureau; USDA, Foreign Agricultural Service
1 Government stocks are subtracted from total stocks to calculate commercial stocks. NASS reported government stocks through 2007, but has since not published government stock data. A large percentage of government stocks were held by USDA Commodity Credit Corporation (CCC). After 2007, CCC stocks were subtracted from total stocks to estimate commercial stocks.
2 Commercial exports are distinguished from government exports, which include exports that received bonuses under the Dairy Export Incentive Program and foreign donations of government products.

Import and export codes used for commercial disappearance product categories

The source for dairy trade data is the U.S. Department of Commerce, Bureau of the Census. U.S. Census Bureau data are compiled using a commodity classification system developed by the World Customs Organization: the Harmonized Commodity Description and Coding System, or simply the Harmonized System (HS). The Harmonized System is an international standard for recording world trade at 2-digit, 4-digit, and 6-digit levels of detail. The United States adopted a 10-digit code system to include greater product detail and began using it for U.S. trade on January 1, 1989. Codes used for U.S. exports are provided in the U.S. Census Bureau Schedule B. Codes used for U.S. imports are provided in the Harmonized Tariff Schedule of the United States (HTS), administered by the U.S. International Trade Commission (USITC). U.S. Census Bureau data for dairy products are reported in kilograms of product weight or liters of product volume. Since ERS commercial disappearance data are reported in pounds, data for imports and exports must be converted to product weight in pounds.

For Excel workbooks displaying import HTS numbers used for each product category, see the HTS codes used for imports, product categories. Descriptions listed are those provided by the USITC Interactive Tariff and Trade DataWeb. [2]

For an Excel workbook displaying Schedule B export codes used for each product category, see HS codes used for exports, product categories. Export classification in Schedule B is not nearly as detailed as import classification in the HTS. Note that the distinction between American and other-than-American cheese is not precise. We have categorized HS codes according to what we believe are the predominant varieties of cheeses within each HS classification. For example, we make a simplifying assumption that all processed cheese exported is of an American type.

Commercial disappearance of milk in all products

Four tables are provided for commercial disappearance of milk in all dairy products:

  • Milk-equivalent milk-fat basis
  • Milk-equivalent skim-solids basis
  • Milk fat
  • Skim solids 

Calculation of estimated commercial disappearance of milk in all products is very similar to the calculation for product categories. An exception is the deduction of farm use to estimate milk marketings.

Milk marketings = Milk production – farm use

Total commercial supply = Beginning commercial stocks + milk marketings + imports 

USDA National Agricultural Statistics Service (NASS), publishes farm use data on an annual basis. To estimate monthly farm use of milk, ERS prorates the annual number grounded on the number of days in each month. For each month in the current year, ERS assumes that farm use is equal to the estimated monthly farm use from the previous year.

Milk is made up of water, milk fat, and skim solids (protein, lactose, ash, and trace elements). Milk equivalents are measured on both a milk-fat basis and a skim-solids basis. The milk-fat and skim-solids content of milk varies from year to year, month to month, cattle breed to cattle breed, and even cow to cow. Most farm milk in the United States averages about 87.5 percent water, 8.8 percent skim solids, and 3.7 percent milk fat.

To account for the supply and use of milk in all products, it is necessary to either account for the milk solids (milk fat or skim solids) or the equivalent amount of milk (on a milk-fat or skim-solids basis) associated with commercial stocks, imports, exports, and net Government removals of the products. Conversion factors associated with milk fat and skim solids are used for this accounting.

Conversion Factors

For stocks and Government removals, milk-fat and skim-solids percentages are estimated for each product reported by NASS and USDA Farm Service Agency. For imports and exports, milk-fat and skim-solids percentages are estimated for each HS code, based upon information from several sources. To estimate the farm milk equivalent (m.e.) on milk-fat and skim-solids bases, estimated milk fat and skim solids are divided by estimates of milk fat and skim solids in U.S. farm milk, which vary from one month to the next.

National milk fat test data for each month are provided by NASS. A national solids-nonfat (SNF) test based on USDA data does not exist. Therefore, an estimate is created using data reported by USDA’s Agricultural Marketing Service (AMS) and the California Department of Food and Agriculture (CDFA). AMS reports a monthly weighted-average SNF test for Federal milk marketing orders with component-based pricing. California is not included under a Federal Milk Marketing Order but publishes its own SNF test each month. Therefore, a weighted average of the United States SNF test is calculated based on the California data and NASS milk production data:

Weighted monthly U.S. SNF test = { [ (CA monthly SNF test CDFA )  x (CA monthly milk production NASS ) ]

+ [ (Federal Milk Marketing Order monthly SNF test AMS) x ((US monthly milk production NASS )

– (CA monthly milk production NASS))]} / {US monthly milk production NASS }

For an Excel workbook with conversion factors for stocks, Government removals, imports, and exports—as well as the sources for the conversion factors—see Conversion factors and sources.

As an example, to convert 4,000 kilograms of cheddar cheese imports or exports into a skim-solids basis milk equivalent in a month when skim solids in farm milk equal 8.7 percent, the following formula would be used:

 (Product quantity [KG or Liters] x (Metric Conversion Factor) x (Skim percentage) / (SNF Test)

= Skim-solids basis milk equivalent (4,000 kg of cheese) x (2.204623 pounds/kg) x (29.9 % of skim solids in cheddar cheese)/ (8.7 % of skim solids in farm milk) = 30,307 pounds of m.e. on a skim-solids basis                                 

This number is divided by one million, as commercial disappearance numbers on milk-equivalent bases are published in million pound units.

Commercial Exports, DEIP, and Donations 

Dairy exports through the Dairy Export Incentive Program (DEIP) or U.S. Government donations must be subtracted from total exports reported by the U.S. Census Bureau in order to represent an accurate commercial export figure. ERS has obtained DEIP amounts and U.S. Government donations on a monthly basis starting in 2004. For years before 2004, ERS has only annual estimates of DEIP and Government donations (Government-related exports). Therefore, prior to 2004, ERS has made the simplifying assumption that monthly Government-related exports are the same proportion of total exports for each year.

Dairy Products, Per Capita Consumption

ERS provides annual per capita consumption estimates for major dairy products. For most products, per capita consumption is calculated by dividing domestic disappearance by the U.S. resident population plus armed forces overseas. July 1 population data reported by the Census Bureau are used.

Fluid beverage milk is the exception, with per capita consumption calculated by dividing estimated route disposition by the U.S. resident population. (The definition for route disposition is provided in the documentation below for Fluid milk domestic sales quantities by product.) The armed forces overseas population data are not included in the calculation because the route disposition data do not include overseas deliveries.

Fluid beverage milk sales quantities by product

ERS reports annual estimated U.S. sales quantities of fluid beverage milk. USDA Agricultural Marketing Service (AMS) provides ERS with route disposition data for milk sales quantities by product within Federal milk marketing order (FMMO) areas. Route disposition is defined in the U.S. Code of Regulations as “delivery to a retail or wholesale outlet (except a plant), either directly or through any distribution facility (including disposition from a plant store, vendor, or vending machine) of a fluid milk product in consumer-type packages or dispenser units” (7 CFR 1000.3). Class I milk sold includes “all skim milk and butterfat disposed of in the form of fluid milk products.”

While the AMS data accounts for most of the fluid beverage milk sold in the United States, there are substantial areas of the country that are not covered by the AMS data because they are not subject to FMMO regulations. Some areas outside of FMMO have State regulations that are similar to the FMMO system, with the entire State of California being the largest in terms of milk sales and population. Fluid milk sales data for California are publicly reported by CDFA.

 ERS receives fluid milk sales data from other State entities as follows: 

  • Pennsylvania Milk Marketing Board: fluid milk sales quantities by product for the area of Pennsylvania outside of the FMMO system.
  • New York State Department of Agriculture and Markets: fluid milk sales quantities by product for the Western New York Milk Marketing Area.
  • Virginia Department of Agriculture and Consumer Services: fluid milk sales quantities (not broken out by product) for Eastern, Western, and Southwestern Market areas of the State.
  • Montana Department of Livestock: fluid milk sales quantities for the State (not broken out by product).
  • Maine Milk Commission: fluid milk sales quantities sold in containers with volume greater than or equal to 1 quart (not broken out by product).

For areas where sales by product are known, per capita quantities by product are calculated using county population data covering each area. For areas where sales by product are not available but total fluid sales quantities are available, ERS assumes that proportions of quantities sold in those areas match proportions of the areas where sales by product data are provided.

There are some areas of the country for which no fluid milk sales data are currently available. These include parts of Ohio, Missouri, North Dakota, South Dakota, Colorado, and Idaho. No data are available for the entire States of Nevada, Utah, Wyoming, Alaska, or Hawaii. From 2000 through March 2004, the entire State of Utah and parts of Nevada, Oregon, Idaho, and Wyoming were included in a Western FMMO area. The Western FMMO was terminated as of April 1, 2004. For April through December 2004 and years after 2004, ERS estimates per capita consumption in these areas as follows:

           Former Western order area quantity =

[Total FMMO quantity + California quantity] X

[2000 to 2003 average of [(Western FMMO quantity)/

(Total FMMO quantity + California quantity)]

Fluid milk in Alaska and Hawaii is relatively expensive compared to other parts of the country because most of it is shipped long distances from other States. With relatively high prices, ERS assumes that per capita fluid milk consumption in these States matches the FMMO area with the minimum per capita sales for the year—usually either the Southeastern or Appalachian FMMO area. For counties with no fluid milk sales quantity data other than those in the former Western FMMO, Alaska, or Hawaii, per capita sales quantities are assumed to match the aggregate per capita quantities where data are available.

Selected soft dairy products, domestic use

ERS reports domestic use of frozen products, yogurt, cottage cheese, sour cream, and fluid cream products. ERS uses production data from NASS and import and export data from the U.S. Census Bureau as reported by the USDA Foreign Agricultural Service. Data for frozen products, reported in gallons from NASS, are converted to pounds, with ice cream products and other frozen products assumed to weigh 4.7 and 6 pounds per gallon, respectively. Where possible, domestic use is calculated as production plus imports minus exports. Where detail is not available for every term of the calculation, aggregated data for imports and export are apportioned in accordance with production of the products. Production numbers, not adjusted for imports and exports, are provided for cottage cheese and sour cream. Domestic use of fluid cream products are not reported for years after 2006 because primary data are not readily available.

Milk cows and production by State and region

For this annual report, ERS provides regional subtotals for milk cows, milk per cow, and milk production from State data provided by NASS. Data for the most recent 5 years are shown when the file is opened. If the data user unhides columns, data are displayed since 1970.

Milk production and factors affecting supply

For this table, most of the primary data are provided by NASS. For years after 2008, the slaughter price is calculated as an estimated live weight equivalent from a national price for domestic cutter cows as reported by AMS. Note that the table includes both a January 1 inventory value for milk cows and an average inventory value as reported by NASS. The former is provided in order to calculate the ratio of replacement heifers to milk cows.

Milk supply and utilization of all dairy products

This table includes annual supply and utilization tables for milk; butter; nonfat dry milk and skim milk powder; condensed and evaporated milk; American cheese; and Other-than-American cheese. Note that these numbers are similar to annual commercial disappearance numbers reported for dairy product categories and milk in all products. However, for these supply and utilization tables, stocks include Government stocks (for the period when they were reported), shipments to U.S. territories, and U.S. Government donations. The domestic human utilization numbers more closely reflect U.S. consumption than the commercial disappearance tables. While the commercial disappearance data can be computed monthly, utilization numbers as computed for this table can only be computed annually due to data constraints.

Number and size of milk bottling plants

This data file includes two tables with overlapping periods, 1960 to 2011 and 2008 to the most current year available. The plant data from or the data covering the earlier period was by ERS from FMMO and various State data sources. In recent years, this method has been judged less accurate than data from the Interstate Milk Shippers list of the U.S. Food and Drug Administration (FDA). The plant count from FDA as of July for each year was used to construct a series with data since 2008. For the table with data since 2008, plants producing the following products have been included in the total: 

  • Pasteurized whole milk, reduced fat, low fat, and skim.
  • Heat-treated (may include reduced fat, skim, low fat, or cream).
  • Ultra-pasteurized milk and milk products (except plants that only produce cream products).
  • Aseptic milk and milk products (including flavored, but not including plants that only produce cream products). 

To calculate the average size milk plant, the total quantity for the year as reported in the "Fluid beverage milk sales quantities by product" file is divided by the number of plants for the year.

Per capita consumption of selected cheese varieties

Due to changes in data availability, two tables are provided, one with data since 1995 and another with data from 1970 to 1994. Per capita consumption is calculated as domestic use divided by population of U.S. residents plus armed forces overseas. For all of the varieties included in the table, production data are available. For some varieties, data concerning stocks, imports, exports, shipments to U.S. territories, and donations are available to calculate domestic use as done for the "Milk supply and utilization in all products" table. However, for most varieties, since such detail is not available for every term of the calculation, aggregated data (for stocks, imports, etc.) are apportioned in accordance with production of cheese varieties.

Archived Historical Data

Previously published historical dairy data are available in a Dairy Yearbook accessible through the USDA Economics, Statistics and Market Information System (ESMIS), a collaborative project between the Albert R. Mann Library at Cornell University and several USDA agencies. The Yearbook contains data on production, supply and use of milk and manufactured dairy products, wholesale and retail price indexes, prices received by farmers, milk production costs, and regional shares of U.S. milk production, among other information. The Yearbook was last updated in September 2005. Data contained in archived Yearbooks may not be comparable to the dairy data tables due to subsequent changes in the methodology used to calculate supply and use.


[1] The Dairy Products Price Support Program and the Dairy Export Incentive Program ended on September 30, 2013.

[2] While the ITC descriptions provided by the USITC Interactive Tariff and Trade DataWeb are useful, they are not as accurate or complete as those in the actual tariff schedule. In the Excel workbooks, DataWeb descriptions are provided instead of the actual HTS descriptions in the interest of brevity.