November 15, 1999
Proposed Changes in The Official Measure of Poverty
by Kathryn Porter
The current poverty measure, developed in the late 1960s and based on data from the 1950s, is considered by most experts to be seriously out of date. At the direction of Congress, an expert panel was convened by the National Research Council of the National Academy of Sciences to study the poverty measure. In 1995, the expert panel made several recommendations for changes in the way the government measures poverty. Based on these recommendations, the Census Bureau has developed several new experimental measures of poverty.
The recommendations of the NAS expert panel primarily affect the way income and needs are measured. The panel's recommendations would not necessarily change the number of people counted as poor or affect who is eligible for government programs. The purpose of the Census Bureau experimental poverty measures is to show how the expert panel's recommendations could be put into practice and to assess the impact of these alternatives. No changes in the current poverty measure will be made before these alternatives have been throughly studied and discussed.
The Current Poverty Measure
The current measure of poverty was developed by Mollie Orshansky, an analyst in the Social Security Administration, and first published in 1963.(1)
Orshansky reasoned that the level at which a family's income was no longer sufficient to purchase a nutritionally adequate diet marked the point at which the family should be considered poor. She used the cost of the Economy Food Plan the least expensive of four food plans developed by the U.S. Department of Agriculture to determine the least amount of money needed to purchase an adequate diet. To determine the proportion of income spent on food, she used the best data then available, which showed that in 1955 the average family spent about one-third of its budget on food. Therefore, she multiplied the cost of the Economy Food Plan by three to arrive at a poverty level.
Because the cost of the Economy Food Plan varied by family size, the poverty level also varied according to the size of the family. For families of one or two persons, Orshansky multiplied the cost of the Economy Food Plan by a factor of greater than three to account for the proportionally higher non-food costs of smaller families.(2)
The poverty level set by Orshansky in the 1960s is still used today. Each year the Census Bureau updates the dollar figures used in the poverty level for inflation so that the purchasing power of a poverty-level income remains the same.
Since the 1950s, however, family expenditure patterns have changed. Costs such as housing have risen faster than food costs. As a result, food costs now make up a much smaller share of family budgets than they did in the 1950s, while expenditures for items such as housing make up a larger share.
In addition, in the 1950s, few mothers of young children worked outside the home and few families paid for child care. Today, child care costs consume a substantial share of the budgets of many low-income families. And because more families now have two working parents, other work expenses also take a larger share of family budgets.
Recommended Changes to the Poverty Measure
In the early 1990s, the Joint Economic Committee of Congress initiated a review of the poverty measure. As a result of this review, funds were appropriated for a study of the concepts and methods underlying the poverty measure. This study, which began in 1992 and continued into early 1995, was carried out by a panel of academic poverty experts, working under the aegis of the National Research Council of the National Academy of Sciences.
The recommendations of the NAS expert panel, published in 1995, included a number of changes in the methods used to determine the poverty level and to decide what income should be compared to the poverty level to determine whether a family is poor.(3) The expert panel recommended that:
- actual costs for food, clothing, and shelter, rather than a budget for food, be used to determine the poverty level;
- a better method be used to adjust the poverty level for family size;
- the poverty level be adjusted to reflect geographic differences in housing costs;
- government food and housing benefits not in the form of cash, and tax-related benefits such as the Earned Income Tax Credit, be included as family income to be compared to the poverty level;
- mandatory expenses such as taxes, work expenses, child care costs, child support payments, and out-of-pocket medical care costs be deducted from family income before comparing income to the poverty level; and
- a different Census survey, the Survey of Income and Program Participation, be used to derive annual figures on the number of poor.
The expert panel did not recommend a specific poverty level, however, as this was considered a matter of judgement that would likely be decided in the political arena. In their analyses, panel members evaluated the effects of their recommendations using a range of poverty levels.
The Census Bureau Experimental Measures
Following the report of the NAS expert panel, analysts at the Census Bureau and the Bureau of Labor Statistics began to look at ways to implement the panel's recommendations. The result of this assessment was a report on experimental poverty measures published by the Census Bureau in the summer of 1999.(4)
The Census Bureau report analyzes the effects of specific changes recommended by the expert panel. The report also combines these changes in various ways to examine the overall effect on the characteristics of the people counted as poor. For the combined analyses, the Census Bureau set the poverty levels so that the total number of people counted as poor remains the same as under the current measure.
Nowhere in these analyses does the Census Bureau make any recommendation concerning the dollar amount of a new poverty level. In fact, the report explicitly states that its purpose is to show the impact of various alternatives, not to make recommendations.
The current measure of poverty is based on a policy directive from the U.S. Office of Management and Budget. Changing the current poverty measure would require a new OMB directive (or specific direction from Congress).
Since the poverty measure is used by many government agencies and private organizations for a variety of purposes, it is unlikely that any change in the current poverty measure will be made without a great deal of public discussion. The publication of the new experimental poverty measures by the Census Bureau is intended to begin that discussion. It is important that the implications of any new measure be throughly assessed and understood before any decision on a new poverty measure is made.
1. Mollie Orshansky, "Children of the Poor," Social Security Bulletin, Vol.26, July 1963, p. 3-13.
2. In addition to variations by family size, Orshansky originally set different poverty levels for farm and non-farm families, for families headed by single women and families headed by married couples, for families with different numbers of children, and, in families of one or two people, for the elderly and the non-elderly. These differentials were based on differences in the cost of the Economy Food Plan for women and men, and for children, the elderly, and non-elderly adults. Also, farm families had lower food costs because they could grow some of their own food. In the early 1980s, the differentials in the poverty level for farm and non-farm families and for families headed by single women and families headed by married couples were eliminated. The differentials for the elderly and for numbers of children remain in the current poverty level.
3. Constance F. Citro and Robert T. Michael, ed., Measuring Poverty: A New Approach, National Academy Press, 1995.
4. Kathleen Short, Thesia Garner, David Johnson, and Patricia Doyle, Experimental Poverty Measures: 1990 to 1997, U.S. Census Bureau, Current Population Reports, Consumer Income, P60-205, June 1999. In September, the analyses in this report were updated using 1998 data. See Kathleen Short, John Iceland, and Thesia I. Garner, Experimental Poverty Measures: 1998, U.S. Census Bureau, September 1999.