A Comparison of Official Poverty Estimates in the Redesigned Current Population Survey Annual Social and Economic Supplement (Black and White)

This paper presents a descriptive analysis of the poverty estimates from the 2014 CurrentPopulation Survey Annual Social and Economic Supplement (CPS ASEC) redesigned andtraditional survey questionnaires. The 2014 CPS ASEC utilized a probability split panel designto test a new redesigned set of income questions. The income questions were redesigned withthe goals of improving income reporting, increasing response rates, reducing reporting errors by taking better advantage of an automated questionnaire environment, and updating questions on retirement income and the income generated from retirement accounts and all other assets. Our main finding is that, among the demographic subgroups examined, most differences between the poverty estimates for the samples assigned to the traditional and redesigned survey instruments were not statistically significant but child (people under age 18) and elderly (people age 65 and older) poverty were higher in the sample assigned to the redesigned questionnaire despite the higher aggregate, mean, and median income collected in the sample with the redesigned questions compared to the sample with the traditional questions.