APPENDIX A. 

 General Explanation 



Page 



TAKING THE CENSUS A-1 



DATA PROCESSING A-2 



MAJOR DATA CHANGES A-2 



DEFINITIONS AND EXPLANATIONS A-2 



FARMS CLASSIFIED BY SPECIFIED CHARACTERISTICS ... A 7 



STATISTICAL ADJUSTMENTS A 9 



NONSAMPLING ERRORS A-10 



CENSUS COVERAGE A 10 



RELIABILITY OF CENSUS ESTIMATES A-10 



TAKING THE CENSUS 

 Method of Enumeration 



All censuses beginning with the 1969 census have been con- 

 ducted primarily by nnail. The 1978 census was the only census 

 to include a mailout/nnailback enumeration supplemented by 

 the direct interview of all households in a sample of area seg- 

 ments. This combination of the mailout/mailback enumeration 

 plus the area sample was used in 1978 to improve complete- 

 ness of coverage for U.S., regional, and State level agriculture 

 census statistics. Due to budget reductions, the area sample was 

 eliminated in 1982. 



In censuses prior to the 1969 census, enumerators were as- 

 signed to specific areas and called on all farm operators within 

 each area. Beginning with the 1950 census, copies of the report 

 form were mailed prior to the enumeration date to each box- 

 holder served by post offices in predominantly rural areas and 

 farmers were asked to complete the forms and have them ready 

 for the enumerator to pick up. 



The mailout/mailback enumeration procedure was not used 

 in taking the agriculture census in Puerto Rico, Guam, and the 

 Virgin Islands. A discussion of the direct enumeration methods 

 used there appears in the reports for these outlying areas. A 

 description of the special direct enumeration of citrus care- 

 takers is included in the Definitions and Explanations section. 



Mail List 



The mail list for the 1982 census was comprised of all in- 

 dividuals, businesses, and organizations that could be readily 

 identified as being associated with agriculture. The list was as- 

 sembled from the records of the 1978 census and administra- 

 tive records of various government agencies, primarily the 

 Internal Revenue Service and the U.S. Department of Agri- 

 culture (USDA). 



Lists of large or specialized operations, such as nurseries, 

 specialty crop farms, broiler growers, fish farms, livestock 

 farms, and cattle feedlot operations, were obtained from State 

 and Federal agencies, trade associations, and similar organi- 

 zations. Lists of multiestablishment companies having one or 

 more establishments (or locations) producing agricultural pro- 

 ducts were obtained from the 1978 census and updated using 



information from the Standard Statistical Establishment List 

 maintained by the Census Bureau. 



A preliminary census mail list was assembled using names 

 and addresses from the 1978 census and administrative source 

 lists available in September 1981. Those records which were 

 less likely to be farms were included in the 1982 Farm and 

 Ranch Identification Survey. Approximately 3 million farm and 

 ranch forms were mailed in March 1982 to names which ap- 

 peared on only one source list or selected combinations of lists 

 which had yielded a low percentage of farm operators in the 

 1978 census. As a result of this survey, nonfarm names and 

 addresses were deleted from the census mail list, new tenant 

 and successor operations were added, and the names, addresses, 

 and size information were updated for the identified active farm 

 operators. 



The final census mail list was developed using results of the 

 1982 Farm and Ranch Identification Survey, names and ad- 

 dresses from the unduplicated preliminary list that were re- 

 tained without precensus verification, and new or updated 

 source lists acquired after the preliminary unduplication. The 

 preliminary and final census mail lists were both constructed 

 by merging and unduplicating the names and addresses from the 

 various source lists on the basis of Employer Identification num- 

 bers. Social Security numbers, and names and addresses. To faci- 

 litate -processing, each name on the administrative source lists 

 was assigned a geographic code indicating the State and county 

 location of the operation and a size code indicating an estimated 

 value of sales. Most duplicates were identified and resolved 

 prior to mailing. Other duplicate names were either reported 

 by respondents or located during office processing. 



Report Forms 



In 1982, 12 regional report form versions were used. These 

 forms were tailored primarily in sections 2 through 8 to list 

 crops commonly produced in one or more States, and in sec- 

 tion 15 to list livestock specialties produced. These modi- 

 fications were made to enhance reporting of crop and live- 

 stock data and to reduce respondent burden. 



Two report form versions were used to minimize the re- 

 porting burden, particularly for small farms. Approximately 

 75 percent of all farms received the 4-page nonsample form 

 covering major items such as land use, crops, livestock and 

 poultry, market value of agricultural products sold, and 

 operator characteristics. The 5-page sample form was mailed 

 to all large and specialized farms (based on expected sales, 

 acres, or standard industrial classification), all farms in 

 Alaska and Hawaii, and approximately 17 percent of all other 

 farms. The.sample form contained all the items asked on the 

 nonsample form plus the sample items (sections 22 through 28). 



The sample form and the information sheet appear in appen- 

 dix C. The Statistical Adjustments section of this appendix 

 includes a discussion of the criteria used to determine whether 

 the sample or nonsample form was to be mailed to addressees. 



1982 CENSUS OF AGRICULTURE 



APPENDIX A A-1 



