Experimental poverty measures: accounting for medical expenditures

Kathleen Short is a senior research economist, in the Housing and Household Economic Statistics Division, U.S. Census Bureau, [email protected]. Thesia I. Garner is a senior research economist in the Division of Price and Index Number Research, Bureau of Labor Statistics, [email protected]. The official measure of poverty in the United States has been in place since the 1960s and has served to inform many policy debates. However, over the years, debate has ensued concerning the level and extent of poverty estimates, as well as the methodology that should be used to measure poverty. One issue that has arisen is whether medical care is or should be accounted for in poverty measurement. Based on research, and recommendations by an expert panel, experimental measures of poverty have been developed that account for medical care costs as well as other dimensions. Accounting for health care costs considerably increases the number of people who appear to be struggling to get by. Particularly, it increases the number of elderly who are considered poor, while only slightly affecting other groups, such as poor children and Blacks. This article describes and compares the size and composition of the poverty population under the official poverty measure and two experimental measures of poverty. The major focus is a discussion of methods and data used to estimate medical out-of-pocket expenses. All statistics shown in this article—poverty rates, poverty gaps, and income-to-poverty thresholds ratios— are affected by the method chosen to account for medical expenses in the measure. Results indicate that, while many groups are somewhat more likely to be classified as poor under the experimental measures, the depth of their poverty is less than is generally found under the official measure. In general, results show that alternate methods of measuring medical expenses affect our perception of the relative incidence of poverty, the depth of poverty experienced by these groups, and the number of people who are classified in extreme poverty (those with family income below one-half of the poverty threshold). Experimental poverty measures are presented here that update those presented in the 1999 Current Population Report by the Census Bureau.1 Two experimental measures that use Consumer Expenditure (CE ) data to estimate poverty thresholds and medical out-of-pocket expenses are presented. These measures and resulting poverty rates are contrasted with the official poverty measure for 2000.