Association for Public Policy Analysis & Management

Official poverty statistics are used in the United States to evaluate economic well-being at the national level, and to distribute federal anti-poverty funds across states and urban areas. However, these poverty estimates are based on pre-tax cash income compared to poverty thresholds that do not take into account geographic differences in price levels. To address these concerns and provide a more comprehensive estimate, the Census Bureau has, since 2009, issued a supplemental poverty measure (SPM).*

The issue of how to account for differences in prices has long been controversial. While most would agree that the cost of living in New York City is higher than in rural Alabama, there is much less consensus on the issue of how these differences in the cost of living should be reflected in the SPM poverty thresholds. In 2011, the University of Kentucky Center for Poverty Research sponsored a workshop on geographic adjustment of the SPM thresholds.** During this workshop many economists expressed concern that higher prices in some areas might themselves be a reflection of other area amenities. Correcting for prices might therefore lead to an underestimate of the welfare levels in high cost areas if cost differences are driven by amenity differences which improve the quality of life.

Currently, the SPM thresholds produced by the Bureau of Labor Statistics are geographically-adjusted by the Census Bureau using an index based on 5-year American Community Survey (ACS) estimates of median gross rents for two-bedroom units. Separate medians are estimated for each of the 260 metropolitan statistical areas large enough to be identified on the public-use version of the Current Population Survey Annual Social and Economic Supplement (CPS ASEC). For each state a median is estimated for all nonmetropolitan areas (47) and for a combination of all smaller metropolitan areas within a state (42), resulting in 349 adjustment factors. This index is applied to only the shelter portion of the threshold and does not take into consideration differences in amenities across jurisdictions.

Although many would argue that amenities should be incorporated in the construction of this index, there is no commonly accepted methodology for doing so. The general consensus at the workshop was that the current SPM geographic adjustment (based on rent alone) may be over-adjusting thresholds because amenities would tend to be directly related to rent. In terms of poverty rates, this would mean that SPM rates for metropolitan statistical areas are too high while rates outside metropolitan statistical areas are too low. This paper makes a first effort to account for amenities by reducing the “weight” of the median rent index before estimating SPM rates. The paper then examines the poverty rates for major demographic groups with this alternative geographic adjustment. The paper will also consider other mechanisms for incorporating amenities into the adjustment mechanism.

* The SPM thresholds and poverty estimates are not intended to assess eligibility for government assistance nor for use in distributing anti-poverty funds.

** Papers commissioned for this workshop can be found at http://www.ukcpr.org/sites/www.ukcpr.org/files/Supplemental_poverty_measures.pdf