Understanding The Current Population Survey's Insurance Estimates And The Medicaid ‘Undercount’

Abstract

The widely cited Census Bureau estimates of the number of uninsured people, based on the Current Population Survey, probably overstate the number of uninsured people. This is because of a Medicaid "undercount": Fewer people report to survey takers that they're covered by Medicaid than program administrative data show are enrolled. Our study finds that the undercount can be explained by the inability of people to recall their insurance status accurately from the previous year. We suggest that other data sources, such as Census's American Community Survey, should be studied to determine whether they would provide better estimates of the uninsured.

CHIP Reporting in the CPS

“…The fourth column gives the uninsured; i.e., no health insurance at all for all of last year. Finally, while some have argued that the CPS should be interpreted as a point in time survey, this analysis follows the analysis of Klerman et al (2009), which suggests that CPS respondents are better viewed as responding about coverage in the previous year (and that, as much as the CPS, "looks like a point in time survey" due to other response errors). Given that the question refers to a 12-month period, coverage from multiple sources would not be unlikely.…”

“…Specifically, we link individual-level CPS data to corresponding individual-level administrative data from the Centers for Medicare & Medicaid Services' (CMS) Medicaid Statistical Information System (MSIS). Earlier studies have shown that CPS Medicaid reporting has a substantial number of errors associated with it (Davern, Baugh, Call, Cox, & Klerman, 2009;Klerman, Davern, Call, Lynch, & Ringel, 2009). In this paper, we examine CHIP reporting in the CPS and we show that CPS CHIP reporting is far less accurate than Medicaid reporting, calling into question the value of the CHIP responses in the CPS for researching the CHIP program.…”

“…This would be consistent with the difficulty of the CPS's implied recall task or stigma in reporting Medicaid. Given this net finding, it is surprising that the CPS also has a large number of "false positives;" i.e., people who report Medicaid, but are not recorded as having Medicaid in the MSIS (Davern, Klerman, Ziegenfuss, 2007;Klerman et al, 2009). The recall task would not suggest this response error.…”

The Effect of the 2014 Medicaid Expansion on Insurance Coverage for Newly Eligible Childless Adults

“…Panel B of Table A.1 shows the sample size available by year and group of states. The sample size increases substantially over the first three years of the sample in part because adult poverty rates climbed sharply following the -2009Great Recession (DeNavas-Walt and Proctor, 2014. The NHIS's increased sample size after 2011 is also apparent.…”

“…The NHIS has several featuresincluding a point-in-time coverage question and the use of state-specific plan names-that should mitigate limitations identified in other surveys (Cantor et al, 2007;Klerman et al, 2009). Previous audit studies suggested that survey reporting error may have a limited impact on estimates of the uninsurance rate in our setting (Call, Davidson, et al, 2008).…”