Characterizing Discrepancies in Reported Acreage Between the Census of Agriculture and June Agricultural Survey

• Activity Number: 426
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Michael Bellow* and Heather Ridolfo
• Companies: USDA/NASS and NASS
• Keywords: Census of Agriculture ; June Agricultural Survey ; logistic regression ; matching

Abstract

The USDA's National Agricultural Statistics Service (NASS) launched a research effort to identify the causes of significant discrepancies in reporting of land related variables (in particular total farm acres operated) between the 2012 Census of Agriculture (COA) and June Agricultural Survey (JAS). NASS conducts the JAS (a probability-based sample survey of U.S. farm operators) annually and the COA (a complete enumeration of U.S. farms and ranches) every five years. JAS records were matched to corresponding ones from the COA (both unedited and edited), with those having absolute acreage differences exceeding a preselected threshold categorized as discrepancies and subjected to further investigation. The degree of influence of explanatory variables such as type of farm, number of operators and average drought level during the JAS data collection period on percentage and size of discrepancies associated with total acres operated was evaluated using descriptive statistics and logistic regression.