The Census Bureau has a long history of measuring net coverage error to assess the overall quality of its censuses. The Census Coverage Measurement (CCM) program and Demographic Analysis (DA) both provide estimates of net coverage error. For nearly all uses of census data those net coverage error measures are sufficient. Information about gross errors such as omissions, duplications, and fabrications as well as information on whole-person imputations provide an important supplement to net errors. Measuring the components that make up net coverage allows us to better understand the processes that lead to coverage shortcomings. Census planners can use these components to identify needs for methodological changes in future censuses.

Analysis has shown that the net undercount rate for young children (aged 0 to 4) is higher than for other ages. We can use information about gross errors and whole-person imputations to help understand the possible causes. Specifically, we are interested in assessing if the high net undercount rate for young children might be because of a higher omission rate for young children (gross undercoverage) or whether it might be the result of a lower rate of erroneous enumerations (gross overcoverage) and/or a lower rate of imputed young children. The analysis in this report seeks to harness the unique strengths of both DA and the CCM to address this issue.