| |
WHY ARE THERE DIFFERENCES BETWEEN DATA PRODUCED BY OTHER SOURCES AND THE NASS CENSUS OF AGRICULTURE FOR ITEMS THAT SHOULD BE THE SAME?
Differences between NASS and other sources occur for many reasons, but are primarily due to differences in the methodologies used to collect and compile the data. NASS has a data gathering infrastructure focused on U.S. agriculture and uses documented statistical methods to produce the data in the Census. Factors such as nonresponse among different groups of farmers, differences in the definitions of farming operations used by NASS and others, differences in the specific information collected or the way in which it is processed will yield different data for items which are seemingly the same. There will always be differences in results when procedures differ.
WHY ARE THERE DIFFERENCES BETWEEN THE CENSUS OF AGRICULTURE AND OTHER NASS REPORTS FOR THE SAME COMMODITIES? WHAT ARE THE "OFFICIAL" NUMBERS? WHAT NUMBERS SHOULD DATA USERS BE USING?
There will also be differences between official numbers produced by NASS if they are produced using different definitions, procedures and methodologies. Items may appear to be identical but are not. For example, the census of agriculture asks operations to report bee colonies for all purposes on December 31, regardless of size or use. In contrast, annual NASS bee and honey estimates include only honey producing colonies, and only operations that have 5 or more colonies. Thus, total colonies would be different for these 2 similar items, both published by NASS. Another example would be statistics for hogs and pigs. The census asks for inventory on December 31, while the NASS Hog and Pig report asks for inventory as of December 1.
All statistics published by NASS are "official" USDA numbers. However, each one will have its particular strengths and weaknesses. Data users should consider these when deciding which numbers are best for their particular uses. For example, NASS production and inventory numbers published in weekly, monthly, or quarterly reports are based on a combination of data collected from a small sample of operations, administrative data, and other reliable sources. These statistics will be timely but may not include the entire population of agricultural operations or statistics for small areas such as counties. The census of agriculture, in contrast, attempts to collect data directly from every U.S. agricultural operation and provides more data and a finer level of detail. But because of the complexity and scope of the census, it is only published once every 5 years and does not include data from the time period immediately preceding publication.
top
A FEW COUNTIES SHOW MORE ACRES IN AGRICULTURAL LAND THAN TOTAL ACRES IN THE COUNTY. HOW CAN THAT BE?
Data inconsistencies like this appear from time to time. Data are assigned to the county based on the county with the largest value of agricultural products produced. Therefore, if an operation reports in multiple counties on one report form, all data on the report form are published in the principal county. The summation of acreage from other counties added to the principal county acreage may cause the total agricultural land to be greater than the total acres in the county. Reweighting records for nonresponse or coverage adjusted values for land in farms may also contribute to the agricultural land in the county.
WHY ARE CASH RECEIPT ESTIMATES FOR HORSES SO DIFFERENT FROM VALUE OF SALES NUMBERS COMING FROM THE CENSUS?
Cash receipts for horses may include items that the census categorized as "farm related" income. Such activities as boarding, riding fees, and stud fees are really "farm related" income if there is no separate business. The annual cash receipts simply have these combined.
top
HOW DO WE EXPLAIN DIFFERENCES IN THE CENSUS OF AGRICULTURE AND THE STATE STATISTICAL OFFICES CENSUS OF FRUIT TREES AND GRAPES DONE EVERY 5 TO 7 YEARS?
Definitions are different when comparing the census of agriculture to the periodic Fruit Tree and Grape Inventory surveys. The census collects fruit data if an operation has at least 20 fruit trees or vines. The NASS list frame should only be carrying fruit control data for commercial fruit operations. A commercial apple operation, for example, is defined as having 100 or more bearing age trees. You could expect a lot more fruit operations shown for the census of agriculture than the Fruit Tree Inventory Survey due to these differences in definition.
HOW DO YOU EXPLAIN THE DIFFERENCE IN LAND STATISTICS BETWEEN THE NASS CENSUS & THE NRCS, NRI DATA?
The Natural Resource Inventory (NRI) is a statistical survey of natural resource conditions, land use and trends on nonfederal land. Private ownership and economic activities are not priorities of the NRI. The Census of Agriculture is an economic census of businesses engaged in agricultural production. The Census of Agriculture accounting of land is not intended to provide total land use statistics. The Census of Agriculture is an accounting of agricultural production information based on having annual sales of at least $1,000 or normally would sell during the year. The NRI data is physically and biologically based to reflect technical definitions and is statistically designed to accurately represent national and state natural resource use trends. The Census of Agriculture is an accounting of the complete universe of farm businesses contributing to agricultural production based on meeting the farm definition.
top
HOW DO WE INTERPRET DIFFERENCES - GOVERNMENT PAYMENTS VS FSA DATA?
From 1987 through 2002, the Census of Agriculture collected 56 to 71 percent of FSA's total distribution. For 2002, the Census of Agriculture collected 59.7 percent of the total 2002 FSA distribution. With the exception of the 1997 Census of Agriculture; 1987, 1992, and 2002 Census of Agriculture collected between 55 and 60 percent of the total FSA distribution. Thus, the 2002 Census of Agriculture is comparable to recent historical census figures. Please remember that the farm operators do not receive all of the FSA distribution, landlords receive some of the distribution also.
top
< 1 2 3 4 >
|