Estimating Uncertainty Bounds for Complex Demographic Models: A Case Study of Estimates of the Unauthorized Population in the United States

• Activity Number: 529
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Social Statistics Section
• Author(s): Dean Judson*+ and Derekh Cornwell
• Companies: Office of Immigration Statistics and Office of Immigration Statistics
• Keywords: uncertainty ; demographic models ; information integration ; Taylor linearization ; Posterior predictive distribution

Abstract

The purpose of this paper is to develop a set of approximate margins of error for the Office of Immigration Statistics' (OIS) unauthorized estimates. These estimates are a demographic model which is a complex function of American Community Survey (ACS) data, estimates of emigration and mortality, and other administrative data. We will start with the simplest, tractable margin of error and introduce complications in stages. The margin of error calculation will, at first, ignore nonsampling error and mortality/emigration model error, and focus only on the sampling uncertainty of this complex demographic model. The final sampling calculation will incorporate all reasonable sources of uncertainty to construct the final margin of error (MOE) estimate. Finally, there will be some discussion of methods for estimating other sources of uncertainty, and the generalizability of the technique.