Alternative Poverty Measures in the IPUMS

In accordance with the Office of Management and Budget's Directive 14, poverty values reported in the original CPS PUMS and reproduced in the IPUMS variables POVERTY and CONTPOV are based on the official definition of poverty originally developed by the Social Security Administration in 1964, later modified by federal interagency committees in 1969 and 1980. At the core of this definition is a needs threshold based on the 1961 economy food plan, the least costly of four nutritionally adequate food plans designed by the U.S. Department of Agriculture. A 1955 USDA study determined that families of three or more persons spend about one-third of their income for food, so the poverty level for these families is set at three times the cost of the 1961 economy food plan. For smaller families and unrelated individuals, the multiplier is higher, since these people generally spend a smaller proportion of their income for food.

In 1995, a National Academy of Sciences (NAS) panel evaluated the official poverty measure and developed recommendations for improving the measure. Based on these recommendations, the Census Bureau produced and released to the public a set of alternative poverty measures through research data files. These measures have been matched to the original CPS PUMS and made available through IPUMS-CPS as continuous measures.

The alternative poverty measures differ from the official measure in three key ways: they use alternative definitions of thresholds, they use alternative definitions of resources or income, and they use different methods for updating the thresholds. While each of the alternative thresholds and income definitions are designed to incorporate tax transfers and noncash benefits, each measure differs in how it accounts for medical costs incurred by families, geographic differences in price of living, and changing prices and expenditures from year to year.

To better account for medical out of pocket (MOOP) costs, the NAS-based measures use three different approaches: medical costs are subtracted from income (MSI), potential medical expenses are incorporated into the threshold (MIT), and medical costs are simultaneously incorporated into income and the threshold through the combined approached (CMB). The MSI method subtracts actual medical expenditures from income based on likely medical expenditures for each family imputed using the Medical Expenditure Survey. The MIT method similarly incorporates expected medical expenses for each family but instead increases the threshold based on average expected expenditures rather than actual expenses. Finally, the combined method accounts for both potential and actual medical expenses. For a detailed discussion of the MOOP imputation process adopted by the Census Bureaus, see page C-16 of Appendix C in "Experimental Poverty Measures: 1999".

To better account for geographic differences in costs of living, another recommendation of the NAS panel, the Census Bureau adjusted some thresholds based on indexes developed by the Department of Housing and Urban Development. The indexes are designed to capture differences in rental housing costs based on metropolitan status for each state. For a detailed discussion and further referenced documentation of the index, see page C-3 of Appendix C in "Experimental Poverty Measures: 1999".

Finally, to Census Bureau used two methods for updating the thresholds for inflation. Some thresholds are updated using the Current Price Index (CPI-U). Other thresholds are updating using median expenditures for the most recent Consumer Expenditure Survey.

Along these three dimensions, the Census Bureau generated twelve measures shown below.

	Updated based on Consumer Price Index (P)		Updated based on Consumer Expenditure Survey (E)
	Geographic Adjustment (G)	No Geographic Adjustment (NG)	Geographic Adjustment (G)	No Geographic Adjustment (NG)
Medical Subtracted from Income (MSI)	MSIPG	MSIPNG	MSIEG	MSIENG
Medical Included in Threshold (MIT)	MITPG	MITPNG	MITEG	MITENG
Combined Method (CMB)	CMBPG	CMBPNG	CMBEG	CMBENG
No Adjustment for Medical Expenses		Official Measure (POVERTY / CONTPOV)

The IPUMS calculates CONTPOV by dividing FTOTVAL (total family income) by CUTOFF (the family's poverty threshold).  Both FTOTVAL and CUTOFF are provided just as they were in the original CPS data, except for assigning the household head's CUTOFF to the members of all related subfamilies in the 1968-1975 data (in which CUTOFF was not provided for such people).  Additionally, the IPUMS provides the experimental measures just as they were provided by the Census Bureau, except for truncating the range of codes at a low of 1 and a high of 501, in keeping with CONTPOV (and the IPUMS-USA variable POVERTY). 

This page is intended to provide only a brief introduction to the alternative poverty measures.  The Census Bureau has developed extensive documentation on the conceptualization and construction of these alternative measures; analysts who wish to use them are strongly urged to read this documentation.

References

United States Bureau of the Census, Alternative Poverty Estimates in the United States (Current Population Report P60-227 prepared by Joe Dalaker), 2005.

United States Bureau of the Census, Experimental Poverty Measures: 1999 (Current Population Report P60-216 prepared by Kathleen Short), 2001.

United States Bureau of the Census, Experimental Poverty Measures: 1990 to 1997 (Current Population Report P60-205 prepared by Kathleen Short et al.), 1999.

Citro and Michael, eds., Measuring Poverty: A New Approach. (National Academy Press) 1995.