United States - Agriculture Prices: Fresh Vegetables - Celery

Prices received represent sales from producers to first buyers. They include all grades and qualities. The average commodity price from the survey multiplied by the total quantity marketed theoretically should give the total cash receipts for the commodity.

Prices paid by farmers represent the average costs of inputs purchased by farmers and ranchers to produce agricultural commodities. Conceptually, the average price when multiplied by quantity purchased should equal total producer expenditures for the item.

Prices Received

Survey procedures: Primary sales data used to determine grain prices were obtained from probability samples of about 1900 mills and elevators. These procedures ensure that virtually all grain moving into commercial channels has a chance of being included in the survey. Generally, States surveyed account for 90 percent or more of total United States production. Livestock prices are obtained from packers, stockyards, auctions, dealers, and market check data from AMS-USDA, private marketing organizations, and state commodity groups and agencies. Inter-farm sales of grain and livestock are not included since they represent very small percentages of the total marketings. Grain marketed for seed is also excluded. Fruit and vegetable prices are obtained from sample surveys and market check data from AMS-USDA, private marketing organizations, state agencies, and universities.

Summary and estimation procedures: Survey quantities sold are expanded by strata to state levels and used to weight average strata prices to a state average. State prices are then weighted to a United States price based on expanded marketings. Recommendations are prepared by the State Field Offices and reviewed by the Agricultural Statistics Board in Washington, D.C. State recommendations are reviewed for reasonableness with survey data, other States, and recent historic estimates.

Reliability: United States price estimates based on probability surveys generally have a sampling error of less than one percent for the major commodities such as corn, wheat, soybeans, cotton, and rice. Current methods of summarization for non-probability commodities are not designed directly to calculate sampling errors. However, analytical measures approximate the United States relative sampling errors at around five percent. Any nonsampling errors are attributed to such things as the inability to obtain correct information, differences in interpreting questions or definitions, mistakes in coding or processing the data, etc. Efforts are made at each step in the survey process to minimize these nonsampling errors.

Prices Paid

Survey procedures: The prices paid data are obtained from establishments that sell goods and services to farmers and ranchers. Annually, about 8,500 firms are randomly selected from lists by type of item sold with an average response rate in the range of 75-80 percent. Firms are asked to report the price for the specified item "most commonly bought by farmers" or that was the "volume seller". Approximately 135 items are surveyed each March to represent all production input items purchased. The survey reference period for most items is the five business days centered around the 15th of the month. Separate prices paid surveys are conducted for agricultural chemicals, fuels, feed, fertilizer, machinery, and seed.

Summary and estimation procedures: The annual March Prices Paid Survey is summarized as a nonprobability survey. Average prices reported are aggregated to the region and United States level using weights available from expenditure data and other administrative sources. Price recommendations are prepared by the State Field Offices and Headquarters for review by the Agricultural Statistics Board in Washington, D.C. Also, the change in price level for individual items surveyed are combined to the regional and United States levels, and are published as prices paid indexes referenced to a specific base period. Prices paid indexes for new autos and trucks, building materials, farm supplies, motor supplies, and marketing containers are updated based on price changes measured in selected Bureau of Labor Statistics (BLS) indexes.

Reliability: Current methods of summarization for the March data are not designed directly to calculate sampling errors. However, analytical measures approximate the United States relative sampling errors for major items around 10 percent. Any nonsampling errors are attributed to such things as the inability to obtain correct information, differences in interpreting questions or definitions, mistakes in coding or processing the data, etc. Efforts are made at each step in the survey process to minimize these nonsampling errors.

Rebasings

Program change: Effective January 1, 1995, the National Agricultural Statistics Service updated the weights and changed the construction of its prices paid and received by farmers indexes. A Federal Register Notice dated October 28, 1994 discussed these program modifications.

A 1990-1992 reference and base period replaced the 1977 reference and 1971-1973 base weight period. Five-year moving average weights replaced the current fixed 1971-1973 base period weights. The 1910-1914=100 price indexes, required by statute for computing parity prices, were linked forward based on the changes in the new 1990-1992=100 indexes. The new indexes were constructed by multiplying the ratios of the current prices to the base period prices by the moving average weights (adjusted for seasonal marketing patterns for the prices received indexes).

Moody's Analytics supplements

We extend the history of select series.

Prices Received

Revisions: For most items, the current month's preliminary price represents a 3-5 day period around the 15th of the month. Previous month's prices represent actual dollars received for quantities sold during the entire month. Revisions are published in monthly issues of Agricultural Prices.

Prices Paid

Revisions: Any revisions are published in the monthly and in annual issues of Agricultural Prices. The basis for revision must be supported by additional data that directly affect the level of the estimate. More revisions are likely for March when separate prices paid surveys are conducted, in lieu of BLS indexes, by the USDA to measure price change.