Validity of centenarian data in the 1960 Census

Abstract

This paper examines the question of how many genuine centenarians there actually are in the United States as compared with those reported in the census. It is concluded that the numbers of centenarians shown in various United States censuses are definite overstatements of the number of true centenarians. It seems likely that instead of the 10,326 centenarians reported in the 1960 census there were at most only about 3,700. Overstatement of ages seems to be particularly the case among those who claim to be aged 110 or over, and it is believed that there probably are no persons who are actually this old. The analysis has been made by projecting, through the me of population life table survival factors, the populations reported cit carious advanced age groups in one census to the next census and then comparing the results with the corresponding number reported in the latter census for the same age cohort. In general, the enumerated populations at ages below 95 are reasonably close to the projected populations, especially for white persons. On the other hand, at ages 95 and over-especially for centenarians-the enumerated populations significantly exceed the projected ones. As a subsidiary part of the analysis, the paper points out the significant differences at the older ages between the "full count" age distribution in the 1960 census and the corresponding "inflated 25 per cent sample" one. This is a subject that bears further investigation and explanation. The paper also discusses centenarians on the social security benefit rolls and concludes that the present data cannot be considered of substantial accuracy with regard to genuine centenarians, particularly the oldest ones. In a number of years, however, this program will provide excellent data, became the individuals involved will have been on the benefit rolls for many years and will have had their ages proved with reasonable accuracy. © 1966 Population Association of America.