Skip to main content
. 2008 Jun;43(3):901–914. doi: 10.1111/j.1475-6773.2007.00808.x

Table 3.

Experimental Estimates of Medicaid Persons Being Coded as Uninsured in Surveys and the Percentage Point Upward Bias in Uninsured Rate Compared with One Matching Study*

Studies and Target Population (1) Size of the Medicaid Population Studied (2) Percent of Medicaid Population Answering They Are Uninsured (3) Upward Bias in Number of Uninsured (Columns [1 × 2]) (4) Size of the Total Relevant Population (Insured and Uninsured) (5) Size of the Total Population (Insured and Uninsured) (6) Percentage Point Upward Bias in the Uninsured Rate in Relevant Population (Columns [3/4]) (7) Percentage Point Upward Bias in the Uninsured Rate (Columns [3/5])
Experimental studies
 Adults on Medicaid in CA 2004 3,309,192 10.4 344,487 25,831,000 35,394,000 1.3 1.0
 Nonelderly (<65) persons on Medicaid in FL 2004 1,868,073 4.9 92,096 14,056,000 16,921,000 0.7 0.5
 Persons on Medicaid in PA 2004 1,559,687 3.4 53,497 12,155,000 12,155,000 0.4 0.4
 Children on Medicaid in MN 1999§ 215,469 4.5 9,696 1,415,000 4,833,000 0.7 0.2
 Persons on Medicaid in MN 1999 334,378 4.1 13,709 4,833,000 4,833,000 0.3 0.3
 Adults on Medicaid in Blue Cross in MN 2003 51,473 0.6 309 3,809,000 5,054,000 0.0 0.0
 Persons on Medicaid in MD 2004** 713,692 4.4 31,402 5,493,000 5,493,000 0.6 0.6
Matching study††
 Adults (age 15–64) on Medicaid in CA (pooled 1990–2000 data)‡‡ 21.7 1.0
*

All estimates are of the noninstitutionalized population.

Based on enrollment data.

Population totals are taken from the Current Population Survey for the corresponding year, state, and age group (Table HI05). Accessed July 25, 2007.

Note: All experimental studies compared “point in time” uninsurance self-report to “point in time” Medicaid enrollment, with the exception of MD, which used an “uninsured all year” self-report to a Medicaid at some point during the year comparison.

§

Blumberg and Cynamon (1999). Note: Results from Study 1 only (MN) are included here.

††

No population totals are provided because the study design pools multiple years of data.