Hierarchical Bayesian Model with Inequality Constraints for US County Estimates

Abstract In the production of US agricultural official statistics, certain inequality and benchmarking constraints must be satisfied. For example, available administrative data provide an accurate lower bound for county-level estimates planted acres, produced by U.S. Department Agriculture’s (USDA) National Agricultural statistics Services (NASS). addition, within a state need to add state-level estimates. A sub-area hierarchical Bayesian model with produce that satisfy these important relationships is discussed, along associated measures uncertainty. This combines County Production Survey (CAPS) data. Inequality complexity fitting present computational challenge full approach. To evaluate inclusion constraints, models without were compared using 2014 corn acres three states. The performance illustrates improvement in accuracy precision while preserving required relationships.