Content area
Abstract
Major uncertainties about the quality of elderly population and death enumerations in the United States result from coverage and content errors in the censuses and death registration. This study seeks to evaluate the consistency of reported data between the two sources for the white and the African-American populations. The focus is on the older population (aged 60 and above), where mortality trends have the greatest impact on social programs (Preston, 1993) and where data quality is most problematic. Using intercensal cohort analysis, age-specific inconsistencies between the sources are identified for two periods: 1970-1980 and 1980-1990.
In order to evaluate what combinations of coverage completeness and age misreporting patterns would produce the empirical results, a series of simulations were executed. The U.S. data inconsistencies are examined in light of both the simulation results and evidence in the literature regarding the nature of errors in the data sources.
Data for whites in the 1980-90 intercensal period were found to be remarkably consistent. Data quality up to age 95 approaches that of Sweden and the Netherlands, countries which maintain highly efficient population registers. Less consistency was observed for whites during the 1970-80 decade. The principal explanation for the pattern of inconsistencies is the combination of net undercount in the 1970 census, overcount in the 1980 census and accurate death statistics.
African-American data are far less consistent. Above age 70, the enumerated population falls increasingly below the expected population in both 1980 and 1990. It appears that the major reason for this pattern is that ages are overstated in censuses relative to death registration. Such a pattern implies that recorded death rates at older ages for African-Americans are likely to be seriously underestimated. A mortality crossover between black and white death rates may occur at advanced ages, that basing such a conclusion on census and vital registration data is hazardous.
While the catalyst for this research was to ascertain the effect of age misreporting on mortality estimates, the evaluation clearly reveals serious consequences for any research based on official population and vital registration data at advanced ages.
