The purpose of this paper is to report on operationalizing the National Academy of Sciences (NAS) Panel’s basic proposed procedure in a test environment. We identify our poverty measures as “experimental,” in contrast to the current poverty measure and the one published by the Panel. In this study, we do not attempt to justify the Panel’s recommended procedure nor do we attempt to evaluate its advantages and disadvantages; instead we take the recommendations as they are presented. This work represents a joint project between the Bureau of the Census (Census) and the Bureau of Labor Statistics (BLS), the two U.S. government statistical agencies which are most likely to be involved in producing the new poverty measure if the recommendations are followed. Estimation of the thresholds work is being conducted primarily by the BLS team and estimation of resources by the Census team.

For the thresholds, we first implement the same basic recommendations that the Panel assumed to be sure we are following the same procedure. However we go beyond this and make changes in the definitions of the poverty thresholds as suggested by the Panel. We produce experimental sets of thresholds based on these changes. For our experimental work, we first develop a set of thresholds using an expenditures outlays approach to reflect transportation expenditures, accounted for in the “a little bit more.” The Panel did not state explicitly in the report that such as approach be followed but did suggest in a memorandum that this represents a more reasonable depiction of transportation costs. Second we develop another set of thresholds, building upon the first set, but using rental equivalence to assign a flow of services to homeowners to approximate their shelter costs. The Panel suggested that the flow of services for owner occupied housing be reflected in the shelter costs. They stated that “Such a definition would treat homeowners with low or no mortgage payments in a comparable manner with other homeowners and renters.” In addition, we use different multipliers to test the sensitivity of the thresholds. As recommended by the Panel, the thresholds are based on data from the Consumer Expenditure (CE) Survey.