Partitioning the Adjustment for Nonresponse, Undercoverage and Misclassification for the 2017 Census of Agriculture

• Activity Number: 419 - Contributed Poster Presentations: Government Statistics Section
• Type: Contributed
• Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Government Statistics Section
• Author(s): Linda J Young and Bayazid Sarkar* and Habtamu Benecha and Sarah Goodale and Gavin Corral
• Companies: USDA National Agricultural Statistics Service and National Agricultural Statistics Service (NASS) and NASS/USDA and National Agricultural Statistics Service (NASS) and National Agricultural Statistics Service (NASS)
• Keywords: undercoverage; nonresponse; misclassification; Census of Agriculture

Abstract

The USDA National Agricultural Statistics Service (NASS) conducts a census of agriculture every five years. Prior to 2012, NASS reported the percent of the official estimates due to adjustments for undercoverage and nonresponse. In 2012, NASS adopted new methodology based on capture-recapture for the census. In addition to undercoverage and nonresponse, official estimates were adjusted for misclassification, and the adjustments were based on conditional probabilities instead of the marginal probabilities used earlier. In 2017, a sample of nonrespondents was identified for intensive follow-up, further complicating properly partitioning the adjustment into its sources. This poster will outline the methods used to identify the percentage of the official estimates due to undercoverage, nonresponse, and misclassification for the 2017 Census of Agriculture.