Total error in PES estimates of population

Area of Science:

• Demography
• Statistical Methodology
• Survey Methodology

Background:

• Accurate population estimates are crucial for policy and resource allocation.
• Dual systems estimates (DSEs) combine census data with post-enumeration surveys (PES) for improved accuracy.
• Evaluating the accuracy of DSEs and census estimates, including undercount, is essential.

Purpose of the Study:

• To develop and apply a methodology for estimating the accuracy of DSEs, census population estimates, and census undercount estimates.
• To assess the accuracy of estimates from the 1988 dress rehearsal census in St. Louis and east-central Missouri.
• To discuss the applicability of the methodology to the 1990 U.S. census and PES.

Main Methods:

• The methodology involves decomposing total error into components like sampling error, matching error, and nonsampling errors.
• Application of the error decomposition methodology to the 1988 dress rehearsal census data.
• Utilizing capture-recapture models and addressing limitations in their assumptions.

Main Results:

• Tentative estimates of errors for the 1988 census, DSEs, and undercount estimates were produced.
• Limitations in error component accuracy, particularly in capture-recapture models, were identified.
• The study anticipates improved estimates for the 1990 census.

Conclusions:

• The developed methodology provides a framework for assessing the accuracy of population estimation techniques.
• The findings highlight the importance of understanding and quantifying various error sources in demographic estimation.
• The approach is adaptable for future large-scale census and survey operations.