How is poverty measured in the United States?
Current Measure
During the mid-1960s, Mollie Orshansky, an economist and statistician at the Social Security Administration (SSA), began publishing articles with poverty statistics for the United States, using a poverty measure that she had developed. Like any poverty measure, Orshansky's had two components—a set of poverty lines or income thresholds, and a definition of family income to be compared with those thresholds.
Orshansky developed her poverty thresholds by taking the cost of a minimum adequate diet for families of different sizes and multiplying the cost by three to allow for other expenses. (The minimum diet she used was the Economy Food Plan, the cheapest of four food plans issued by the U.S. Department of Agriculture. The factor of three was derived from a 1955 Agriculture Department survey.) Poor families were those whose yearly income was below the threshold for a family of a given size. She intended that the method be used for research, not to determine eligibility for antipoverty programs.
For the base year 1963, Orshansky's weighted average poverty threshold for a family of four was $3,128. She used the Census Bureau's definition of income—before-tax money income.
In 1965, the U.S. Office of Economic Opportunity adopted the SSA thresholds as a working definition of poverty for statistical purposes and for program planning. In 1969, the U.S. Bureau of the Budget (now the U.S. Office of Management and Budget) issued a directive that made the thresholds the federal government's official statistical definition of poverty.
In 1967, the Census Bureau began to publish annual poverty statistics calculating the number and percentage of persons in poverty (the poverty population and the poverty rate) by comparing the Orshansky thresholds to families' before-tax money income, using data from the Current Population Survey that is taken every year in March. For these tabulations, the thresholds are updated annually for price changes and so are not changed in real (constant-dollar) terms; in other words, the 2009 weighted average poverty threshold of $21,954 for a family of four represents the same purchasing power as the corresponding 1963 threshold of $3,128.
In 2005, the Census Bureau fully implemented a new survey, the largest household survey in the United States, called the American Community Survey (ACS). The ACS collects detailed demographic, socioeconomic, and housing information, like the long-form questionnaire of the Decennial Census, from about 3 million addresses per year.
The 2005 ACS estimates cover all geographical areas with a population of 65,000 or more; 3-year average estimates begin in 2008 for areas and subpopulations as small as 20,000; and beginning in 2010, the ACS will use 5-year averages to provide estimates for every state, county, city, town, and place, as well as for census tracts and block groups. The first 5-year ACS data are scheduled for release in December 2010.
In August 2008, the Census Bureau published a report of the findings, “Income, Earnings, and Poverty Data from the 2008 American Community Survey.” In September 2010, the Census Bureau issued a series, ACS Briefs, to highlight 2009 survey results; Poverty: 2008 and 2009 presents poverty estimates for states.
While the official poverty rate for the entire United States is based on data from the Current Population Survey Annual Social and Economic Supplement (CPS ASEC), the ACS is a reliable source of annual survey estimates of poverty for states and for substate areas with populations of 65,000 or more. To learn more, read a fact sheet the Census Bureau prepared to describe the differences between the income and poverty estimates from the ACS and the CPS ASEC.
Attempts have been made over the years to improve the measure. For example, in September 2008, bills were introduced in the Senate and the House of Representatives that call for an improved poverty measure. Senator Christopher Dodd and Representative James McDermott introduced the Measuring American Poverty Act of 2008 in both the Senate and the House. As of April 2009, no action had been taken on either bill.
A new poverty measure would likely be based in large part on the recommendations of a 1992 National Academy of Sciences (NAS) panel, on the findings of a 2004 Committee on National Statistics workshop to review NAS recommendations, and on subsequent research on alternative poverty measures. For more on federal plans to develop such a measure, visit the Federal Supplemental Poverty Measure page on IRP’s Web site.
NAS Recommendations
In 1992, the National Academy of Sciences study panel, the Panel on Poverty and Family Assistance, was established at the request of Congress to conduct a comprehensive examination of poverty measurement in the United States. Among the panel members were two IRP affiliates and UW–Madison faculty, Robert Hauser and Franklin Wilson; a former IRP director, Sheldon Danziger; and an off-campus IRP affiliate, David Betson, University of Notre Dame. In 1995, the panel issued a final report and recommendations for change in the monograph titled Measuring Poverty: A New Approach.
The study panel found the following:
Our major conclusion is that the current measure needs to be revised: it no longer provides an accurate picture of the differences in the extent of economic poverty among population groups or geographic areas of the country, nor an accurate picture of trends over time. The current measure has remained virtually unchanged over the past 30 years. Yet during that time, there have been marked changes in the nation’s economy and society and in public policies that have affected families’ economic well-being, which are not reflected in the measure.
The panel recommended the following:
The official U.S. poverty thresholds should comprise a budget for the three basic categories of food, clothing, shelter (including utilities), and a small additional amount to allow for other needs (e.g., household supplies). Actual expenditure data should be used to develop a threshold for a reference family of four—two adults and two children. Each year, that threshold should be updated to reflect changes in spending on food, clothing, and shelter over the previous 3 years and then adjusted for different family types and geographic areas of the country. The resources of a family or individual that are compared with the appropriate threshold to determine poverty status should be consistently defined to include money and near-money disposable income: that is, resources should include most in-kind benefits and exclude taxes and certain other nondiscretionary expenses (e.g., work expenses).
The procedure for updating the poverty thresholds over time is an integral part of the proposed measure…. We propose a regular updating procedure to maintain the time series of poverty statistics. We also recommend a conservative updating procedure that adjusts the thresholds for changes in consumption that are relevant to a poverty budget, rather than for changes in total consumption.[1]
The U.S. Office of Management and Budget, a unit within the Executive Office of the President, issued directives on the use of the poverty thresholds in 1969 and 1981 and presumably has the authority to impose more fundamental changes in the way poverty is measured. Yet the NAS committee’s recommendations have failed to lead to any change in the measure.
CNSTAT Review
In 2004, the Committee on National Statistics hosted a workshop to review federal research on alternative methods for measuring poverty using an income poverty measure. The workshop was requested by the U.S. Office of Management and Budget to assess progress in moving towards a new measure of income poverty as recommended by the 1992 NAS Panel on Poverty and Family Assistance.
The workshop provided a forum for comment on methods developed for key components of the CNSTAT Panel’s proposals and the degree of support for such methods. The roster was as follows (2008 affiliations are listed): Timothy Smeeding (co-chair), IRP Director and Arts and Sciences Distinguished Professor of Public Affairs and Economics, University of Wisconsin-Madison, and Director Emeritus of the Luxembourg Income Study; Barbara Wolfe, Professor of Economics, Population Health Sciences, and Public Affairs, and former Director of both IRP and the La Follette School of Public Affairs, UW–Madison; Rebecca Blank (co-chair), Under Secretary for Economic Affairs, U.S. Department of Commerce; David Betson, Associate Professor of Economics and the former Director of the Hesburgh Program in Public Service at the University of Notre Dame; and Graham Kalton, Chairman of the Board of Westat, Incorporated, an employee-owned research corporation.
Further Reading
Research section of this Web site:
Focus 19:2 newsletter issue devoted to Revising the Poverty Measure (PDF, 64 pp).
Description of IRP’s 1999 Conference to Improve the Poverty Measure.
John Iceland, 2005, “The CNSTAT Workshop on Experimental Poverty Measures, June 2004,” Focus 23(4): 26–30.
Robert Haveman, “What Does It Mean to Be Poor in a Rich Society?” Robert J. Lampman Memorial Lecture, University of Wisconsin–Madison, June 18, 2008 (see PowerPoint slides).
Robert Haveman, 2008, “What Does It Mean to Be Poor in a Rich Society?” (Conference paper and chapter in Changing Poverty, Changing Policies volume.)
[1] C. Citro and R. Michael, Eds., Measuring Poverty: A New Approach (Washington, DC: National Academy Press, 1995), p. 1.