Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
APPENDIX B 413 1990âit is substantial for some population groups. In 1980, an estimated 9-10 percent of black children under age 5 were missed, as were about 15 percent of middle-aged black men (Fay, Passel, and Robinson, 1988:Tables 3.2, 3.3; Robinson, 1990). (The decision was recently made to use census-based population estimates that are adjusted for the census undercount as weighting controls for the CPS and SIPP.) Second, the ratio adjustments do not correct for characteristics other than age, sex, and ethnic origin on which the undercovered population might be expected to differ from the covered population. Fay (1989) analyzed within- household undercoverage in the CPS relative to the decennial census, using a 1980 CPS-census match. His results are suggestive of ways in which weighting adjustments do not adequately compensate for household survey undercoverage. For example, he finds that about one-fourth of adult black men who are counted in the census but not in the CPS are household heads, whose households should be categorized as married-couple households in the CPS but instead are categorized as households headed by unmarried women. The correlates of undercoverage (besides age, race, and sex) are not definitely established. However, analysis of the 1980 census postenumeration survey and of other survey, administrative records, and ethnographic data suggests that undercount rates are higher for the following groups: household members other than the head, spouse, and children of the head; unmarried people; people living alone or in very large households; and people residing in central cities of large metropolitan areas (see Citro and Cohen, 1985; Fein, 1989). In addition, there is evidence that the rate of undercount increases as household income decreases. Overall, these tentative findings suggest that minorities, unattached people, and low-income people are at much greater risk of not being covered in household surveys than other people and, hence, that undercoverage affects SIPP and March CPS-based estimates of poverty. Both the overall poverty rate and, perhaps more important, the distribution of poverty across groups may be affected. The Census Bureau has recently begun a research program to investigate the undercoverage problem in greater depth and take steps to reduce it (Shapiro and Bettin, 1992). Household and Person Nonresponse Relative to many other surveys, the CPS obtains high response rates. Yet, 4-5 percent fail to respond to the CPS, and another 9 percent of people in otherwise interviewed households fail to respond (Citro, 1991:26). In addition, a considerable number of people, although responding to the basic CPS labor force questionnaire, do not respond to the March income supplement. Nonresponse to the supplement is treated together with other cases of failing to answer one or more specific questions (see below). To adjust for whole
APPENDIX B 414 household nonresponse to the basic CPS, the Census Bureau increases the weights of responding households; to adjust for person nonresponse, it imputes a complete data record for another person with similar demographic characteristics. These procedures assume that respondents represent the characteristics of nonrespondents; this assumption has not been adequately tested. Like all household surveys, SIPP experiences household nonresponse, and like all longitudinal surveys, it suffers cumulative sample loss or attrition at each successive interview wave (some households that fail to respond at an interview wave are subsequently brought back into the survey). In addition, it experiences "type Z" nonresponseâthe failure to obtain information, either in person or by proxy, for individual members of otherwise cooperating households. Attrition in SIPP to date has been highest at the first and second interviewsâ 5-8 percent of eligible households at Wave 1 and 4-6 percent of eligible households at Wave 2. Thereafter, the additional loss is only 2-3 percent in each of Waves 3-5 and less than 1 percent in each subsequent wave. By Wave 6 (after 2 years of interviewing), cumulative sample loss is 18-20 percent of eligible households; by Wave 8, it is 21-22 percent (Bowie, 1991). The Panel to Evaluate SIPP estimated that total sample attrition at the end of 12 waves (4 years) might be 25 percent (Citro and Kalton, 1993:102). The attrition experience in SIPP is quite comparable to that in the ISDP (Nelson, Bowie, and Walker, 1987) and the PSID (with the exception that, as noted above, the PSID experienced a larger sample loss at the first two waves). Attrition reduces the number of cases that are available for analysisâ including the number available for longitudinal analysis over all or part of the time span of a panel and the number available for cross-sectional analysis from interview wavesâand thereby increases the sampling error of the estimates. More important, the people who drop out may differ from those who remain in the survey. To the extent that adjustments to the weights for survey respondents do not compensate for these differences, estimates from the survey may be biased. The available evidence does suggest that people who drop out of SIPP differ from those who stay in the survey. Studies of nonresponse from the 1984 SIPP panel show that household noninterview rates after the first wave tended to be higher for renters, for households located in large metropolitan areas, and for households headed by young adults. Individuals who did not complete all of the interview waves, compared with those who did, tended to include more residents of large metropolitan areas, renters, members of racial minorities, children and other relatives of the reference person, people aged 15-24, never- married people, and people with no savings accounts or other assets (Jabine, King, and Petroni, 1990:35-37, Table 5.4). A recent analysis of attrition from the 1990 SIPP panel obtained similar results (Lamas, Tin, and Eargle, 1994). This study found that attrition was more likely to occur among