8
Data Issues, Other Topics, and Future Research

DATA ISSUES

Official poverty statistics, which date back to 1959, are calculated with data from the Current Population Survey (CPS), a monthly survey of about 60,000 households conducted by the U.S. Census Bureau for the Bureau of Labor Statistics. In addition to the basic monthly CPS survey, supplementary questions are asked in February through April in the Annual Social and Economic Supplement (ASEC), which serves as a source of detailed information on income and poverty used in official poverty reports.1 The ASEC suffers from two major poverty measurement-related shortcomings: (1) it does not collect all the information needed to compute the 1995 National Research Council (NRC) report’s recommended measure; and (2) income is underreported by respondents. Because of these shortcomings, the 1995 NRC report recommended that the Survey of Income and Program Participation (SIPP) should eventually be used for official poverty statistics, because it asks more detailed income-related questions and obtains income data of higher quality than the CPS.

The SIPP is a panel survey of the U.S. civilian noninstitutionalized population, begun in 1983, which contacts households every 4 months for

1  

The ASEC sample is also larger than the regular monthly sample—roughly 99,000 households are interviewed in the ASEC.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 30
Experimental Poverty Measures: Summary of a Workshop 8 Data Issues, Other Topics, and Future Research DATA ISSUES Official poverty statistics, which date back to 1959, are calculated with data from the Current Population Survey (CPS), a monthly survey of about 60,000 households conducted by the U.S. Census Bureau for the Bureau of Labor Statistics. In addition to the basic monthly CPS survey, supplementary questions are asked in February through April in the Annual Social and Economic Supplement (ASEC), which serves as a source of detailed information on income and poverty used in official poverty reports.1 The ASEC suffers from two major poverty measurement-related shortcomings: (1) it does not collect all the information needed to compute the 1995 National Research Council (NRC) report’s recommended measure; and (2) income is underreported by respondents. Because of these shortcomings, the 1995 NRC report recommended that the Survey of Income and Program Participation (SIPP) should eventually be used for official poverty statistics, because it asks more detailed income-related questions and obtains income data of higher quality than the CPS. The SIPP is a panel survey of the U.S. civilian noninstitutionalized population, begun in 1983, which contacts households every 4 months for 1   The ASEC sample is also larger than the regular monthly sample—roughly 99,000 households are interviewed in the ASEC.

OCR for page 30
Experimental Poverty Measures: Summary of a Workshop about 3 years (depending on the panel). Panel sizes have varied from 14,000 to 36,700. About 31,700 households were interviewed at the start of the 2001 panel (for details, see Weinberg, 2004). The SIPP collects data on a number of items that the ASEC does not, such as child care and other work-related expenses. The panel design allows one to calculate poverty over different durations (such as a month, a year, or multiple years) and to track how families are doing over the life of a 3- or 4-year panel. The SIPP also has shortcomings. An important one is that many people drop out of the survey over the course of the panel, which likely introduces some bias into the poverty estimates over time. Studies have shown that low-income households are more likely to drop out of the survey than others. This bias could be overcome by reintroducing “overlapping” panels (a strategy that was dropped after the 1993 panel), in which a new 3- or 4-year panel is implemented every year. This approach would produce annual poverty estimates that come from three or four panels that are simultaneously in the field. Some of the other shortcomings in the SIPP have been addressed in recent panels. Wage and salary information tends to be underreported in the SIPP, though an improved questionnaire implemented in the 2004 panel may reduce the magnitude of this problem. Prior to 2004, the SIPP also did not have state-representative samples in all states. While the 2004 panel does have an improved design to address this issue, the small samples in a few states will produce poverty estimates that are not very reliable for those states. Reintroducing overlapping panels may help address this problem too. Finally, while the SIPP collects information on taxes, the data are of poor quality. There are efforts now under way to model what families pay in taxes (and refunds they receive from the Earned Income Tax Credit) in the SIPP; these models are somewhat similar to CPS tax models. John Czajka (Mathematica Policy Research) said that using data from the SIPP rather than the CPS has several advantages. He cautioned, however, that the SIPP still needs to be improved in a few ways. In addition to underreporting of earnings, he mentioned that data have to be released from the Census Bureau in a more timely manner. He noted that while an overlapping panels design is important for addressing the bias arising from people dropping out of the sample, it may involve making tradeoffs if, for example, each of the panels contain smaller sample sizes (which reduce the reliability of estimates from any given panel). Overlapping panels may need to be smaller because of the expense it takes to concurrently field multiple surveys.

OCR for page 30
Experimental Poverty Measures: Summary of a Workshop Several workshop participants echoed these concerns with using SIPP data, with the main issues being the timeliness of the release of data and the loss of sample over the course of the panel. Wendell Primus (Joint Economic Committee) voiced strong support for using the CPS data, noting that one advantage is the consistent CPS sample design, which provides more comparable poverty data over time than the SIPP. Rebecca Blank (University of Michigan) agreed that that the CPS should continue to be used as the core data source for poverty statistics, with the hope that SIPP data might improve and might become more usable over time. A final issue discussed in the session on data revolved around whether a new poverty measure could be implemented using American Community Survey (ACS) data. This survey, which is designed to replace the decennial census long form, could be an important source of poverty information at the state and local levels. The main problem with implementing the NRC report-recommended measure with ACS data is that the survey does not collect information on noncash benefits or health insurance status (needed to estimate medical out-of-pocket expenses). A couple of workshop participants noted that the ACS had the potential to produce valuable annual small-area poverty estimates, but only if more questions on the above items were added to the survey. OTHER TOPICS Workshop participants briefly discussed the following four topics: whether families’ wealth should be accounted for in a poverty measure, the appropriate unit of analysis to use in a measure, whether and how to account for household production, and whether to have one alternative measure of poverty or several (as is currently done in Census Bureau reports). Wendell Primus felt that it would be too difficult to incorporate people’s wealth in a poverty measure. The quality of wealth data in household surveys is generally quite poor. Timothy Smeeding (Syracuse University) added that accounting for wealth would also necessitate including family debt in the measure. He agreed that the quality of data on these items was poor. Daniel Weinberg (Census Bureau) noted, however, that the poverty measure discussed in the workshop takes wealth into account at least to some extent by making distinctions between homeowners and renters when accounting for the value of housing. David Ribar (George Washington University) noted that it is conceptually important to take wealth into account, as wealth helps smooth people’s income and expenditures.

OCR for page 30
Experimental Poverty Measures: Summary of a Workshop Some participants said that research was needed on the possibility of incorporating wealth into a poverty measure, but that it should not necessarily be a top priority in the research agenda. On the topic of unit of analysis, the issue is the most appropriate unit for which to measure poverty—the family, the household, or some other entity. While the Census Bureau definition of “families” is persons related to one another by birth, marriage, or adoption, “households” consist of all people—related or unrelated (such as housemates)—living in the same housing unit. The key question is whether people should be classified as poor on the basis of their family’s income, which is then compared with a corresponding poverty threshold based on their family’s size and composition, or whether it is more appropriate to pool incomes of all household members and use a poverty threshold based on the household’s size and composition. Rebecca Blank (University of Michigan) suggested that if better data collection efforts that clarified household relationships in complex households are a priority area for future research, the results would be helpful on this issue. In considering how to account for household production when measuring family’s resources, Nancy Folbre (University of Massachusetts) argued that the work of parents who stay at home should be valued and challenged the idea that there were economies of scale for working mothers. For example, the additional cost of child care for a second child in a day care facility is the same as the cost of the first. She noted that spending patterns of families in which a parent stays at home also differ from those where both parents work (or a single parent works); the latter types of families, for example, spend more on food (often purchased outside the home). Family time should be viewed as a basic need, and some families are “time poor.” She noted that data collected in time diaries in some surveys now provide useful information that could help impute the value of nonmarket work. Rebecca Blank agreed that this is an important topic for future research, though knowledge of how to incorporate such information in a poverty measure is still some time away. The final topic of discussion centered on whether Census Bureau poverty reports should contain multiple poverty measures. Several workshop participants argued that reports should highlight no more than two or three measures, and that a single new measure was preferable. Timothy Smeeding added that it would useful for public-use datasets to have information available that would allow analysts to calculate different variations of any measure. Constance Citro (National Research Council) mentioned that it

OCR for page 30
Experimental Poverty Measures: Summary of a Workshop should be made clear to the public that any poverty measure needs to be periodically evaluated, and perhaps improved, to incorporate new information or improved methods. Timothy Smeeding and Rebecca Blank also advocated continuing the current poverty measure time series to have some level of continuity in poverty statistics. FUTURE RESEARCH In addition to the topics for future research already mentioned, some participants indicated that further research could be helpful on some of the elements discussed above. Thus, some of the participants advocated developing improved methods for incorporating geographic adjustments to the thresholds, and others supported more research on whether equivalence scales should incorporate more than three parameters. If SIPP data rather than CPS data are to be used as the main source for poverty statistics, participants said that research is needed on the attrition problems in the SIPP. Some participants repeated their interest in future research on the use of an alternative unit of analysis other than the official family, and the feasibility and practicality of accounting for wealth and household production in a new poverty measure.