GENERAL EXPLANATION DATA COLLECTION Method of Enumeration All agriculture censuses beginning with the 1969 census primarily have used mailout/mailback data collection. Direct enumeration methods, however, continue to be used for the agriculture censuses in Guam and the U.S. Virgin Islands, and parts of Puerto Rico. The report for each area includes a discussion of the method of enumeration. Mail List The mail list for the 1992 census was comprised of individuals, businesses, and organizations that could be readily identified as being associated with agriculture. The list was assembled from the records of the 1987 census, administrative records of the Internal Revenue Service (IRS), and the statistical records of the U.S. Department of Agriculture (USDA). In addition, lists of large or specialized operations, such as nurseries and green- houses, poultry farms, fish farms, livestock farms, and cattle feedlot operations were obtained from State and Federal agencies, trade associations, and similar organizations. Lists of companies having one or more establishments or locations producing agricultural products were obtained from the 1987 census and updated using the information from the Standard Statistical Establishment List maintained by the Census Bureau. Exhaustive record linkage, unduplication, and mathematical modeling yielded a final mail list of 3.55 million names and addresses which had a substantial probability of being farm operations. Report Forms In 1992, three different report forms were used--a sample form, a nonsample form, and a screener form--to minimize the reporting burden, particularly for small farms and places less likely to be farms. The screener form was the same as the nonsample form except section 1 of the screener form listed crop and livestock items, asked if the respondent had any of the items at any time in 1992, and omitted a question on grazing permits. The sample form contained all the questions asked on the nonsample form plus additional questions in sections 21 through 26 on commercial fertilizer, chemicals, production expenses, machinery and equipment, value of land and buildings, and income from farm- related sources. The information collected in the sample sections gave the Bureau of the Census a good basis for making estimates of these data for other farms included in the census. There were 11 regional versions of the nonsample and screener forms and 13 regional versions of the sample form with the listings of crops varying by region. Appendix D in the printed publication Volume 1, State and County Data, contains copies of the sample form and the information sheet. The sample form was mailed to 1,030,000 addresses on the mail list, including all those expected to be large (based on expected sales or acreage); all farms operated by multiestablishment companies or nonprofit organizations, all those in Alaska, Hawaii, and Rhode Island; and a sample of other addressees. The screener form was mailed to 412,000 addressees. These were less likely to be farms or expected to be small farms. The nonsample form was mailed to the remaining 2,112,000 addressees. Further discussion of the criteria used to determine which form was mailed to an addressee is provided in the Census Design section of appendix C in the printed publication Volume 1, State and County Data. INITIAL MAILING The report forms were mailed in mid-December 1992 to approximately 3,554,000 individuals, businesses, and organizations on the mail list. The mail packages included a report form with a description of the purposes and uses of the census on the last page, a cover letter, an information sheet containing instructions for completing the form, and a postage paid return envelope. Additional special instructions were included with report forms sent to grazing associations, feedlot operations, institutional organizations, Indian reservations, and firms with multiple farm or ranch operations. Special instructions also were sent to producers of poultry under contract, bees and honey, fish and other aquaculture products, laboratory animals, and nursery and greenhouse crops. To provide additional help to farmers in completing their reports, copies of the 1992 Census of Agriculture Report Form Guide booklet were sent to vocational agriculture instructors, USDA county offices of the Agriculture Stabilization and Conservation Service, the Cooperative Extension Service, land- grant colleges, State departments of agriculture, and farm organizations. The Guide contained descriptions and definitions of various items in more detail than was contained in the instructions included with each report form. Representatives of the above agencies graciously consented to assist farmers in completing their report forms. Follow-up Procedures A thank you/reminder card was mailed to those on the mail list in early January 1993. Four followup letters, three of which were accompanied by a report form, were sent to nonrespondents at 5-week intervals starting in mid-February and continuing until early June 1993. STATISTICAL METHODOLOGY The 1992 Census of Agriculture used two types of statistical estimation procedures. These estimation procedures accounted both for nonresponse to the data collection and for the sample data collection. These procedures are used because some farm operators never respond to the census despite numerous attempts to contact them, and not all farm operators are requested to provide the sample data items. The statistics in this report are estimates derived from these procedures. There are two types of errors possible in these estimates, sampling, and nonsampling. Sampling errors in the census data result from the nonresponse sample and the census sample data collection. Nonsampling errors arise from incompleteness of the census mail list, duplication in the mail list, incorrect data reporting, errors in editing of reported data, and errors in imputation for missing data. Extensive efforts are made to minimize both types of errors through efficient sample designs, quality control, verification, and clerical review of specific operations. For a complete and detailed explanation of the statistical methodology and the reliability of census estimates, see printed publication Volume 1, State and County Data, Appendix C.