Nonresponse adjustments in NASS agricultural surveys
The National Agricultural statistics Service (NASS) currently uses two distinct adjustment methods to account for nonresponse on the list side of its Agricultural Surveys. A complex imputation program is employed for crop (acreage, production, and stock) items, while an adjusted form of reweighting is used for livestock items. This paper examines the subtle differences in the two approaches. A unified imputation/reweighting methodology for crop acreages and livestock is proposed and alternative specifications of the methodology are critiqued. Extensions to production, stocks, and the area side are also discussed.