NY24-04: Poverty Among Older Adults According to the “Principal Poverty Measure”

Abstract/Specific Aims:

On March 29, 2023, a NAS panel issued a report recommending major changes to Census Bureau poverty measures (NAS 2023). The panel’s recommended measure, the Principal  Poverty Measure (PPM), would replace the Supplemental Poverty Measure (SPM) and the official poverty measure (OPM) and become the headline federal poverty statistic. It recommended changes to the treatment of housing, health insurance, and healthcare—the largest components of the poverty  threshold—particularly for the older population. The proposed research would be the first to implement a pilot PPM and use it to estimate poverty rates for the U.S., and to estimate impacts of health insurance benefits, Social Security, housing assistance, and other social insurance and benefits on poverty across the life course. This study will estimate poverty disparities across demographic groups defined by age, racial and Hispanic identification, urbanicity, gender, and living arrangements (for the household population) and others. It will estimate benefit and program impacts on poverty rates and poverty gaps, as well as differential impacts across groups. It will show how the choice of poverty measure (PPM, anchored PPM, SPM) shapes the understanding of U.S. poverty and how it differs across the age distribution: the demographic composition of the poor, the level of poverty, poverty disparities, and program impacts. It will provide measures of direct impacts of health insurance benefits (e.g., Medicare and Medicaid) unavailable with the SPM. The PPM will use improved methods to capture  geographic differences in housing costs. The PPM holds promise to provide a more accurate description of U.S. poverty and a more accurate accounting of the impacts of social and health insurance benefits and other transfers.

Researchers have long been aware of the limitations of the U.S. Census’ Official Poverty Measure. A 1995 NAS report recommended fundamental improvements to the OPM, which became the foundation for the SPM (Short 2013; Creamer et al. 2022), the preferred measure among researchers. While the SPM, unlike the OPM, incorporates income from many non-cash benefits, SPM resources exclude health benefits, the largest in-kind transfers. And the SPM threshold does not include a need for health insurance or care (NAS 1995; Korenman and Remler 2016). Instead, the SPM subtracts from resources all out-of-pocket expenditures on medical care or insurance (MOOP), treating these expenditures as non-discretionary, like taxes. The proposed PPM revises the SPM treatment of medical care based on the Health Inclusive Poverty Measure (HIPM; Korenman and Remler 1996). The threshold insurance needed would be the price of a basic insurance plan (the unsubsidized benchmark silver plan in the ACA Marketplace or, for Medicare recipients, the lowest-cost Medicare Advantage plan that includes a prescription drug plan). The PPM housing need is based on HUD’s Fair Market Rents rather than national housing expenditures. PPM housing resources would be adjusted for homeownership to reflect both that home equity reduces the need to pay rent, an addition to resources, and that home ownership involves “user costs” (mortgage interest, property taxes, and repairs), a subtraction from resources. As housing accounts for one-third of expenditures (BLS Consumer Expenditure Survey, 2012-2021), the PPM treatment of housing could result in a different picture of the population in poverty. This research would be the first to carry out a pilot  implementation of these PPM changes.

This project will produce academic journal articles focused on major differences between PPM and  SPM poverty, and how rates, disparities, and estimated program impacts (and disparities in impacts) differ between the two measures for those age 65 and older and those under age 65. In addition, articles focused on conceptual and technical implementation of the PPM and detailed reports (RDRC working papers) with descriptions of poverty populations will be produced. To construct the pilot PPM, we will merge information on health insurance plans (ACA Marketplace Plans; Medicare Advantage  Plans) and Fair Market Rents (FMRs), with Current Population Survey ASEC microdata file for 2023. We will use the Consumer Expenditure Survey (CE) to impute user costs for homeowners. We will compare SPM results to results from the PPM and an anchored PPM that equates the PPM overall poverty rate to the SPM rate. This shows which differences between the PPM and SPM result simply from  differences in overall rates as opposed to different conceptualizations of poverty.