B

Measuring Insurance Coverage and Insurance Rates

The most commonly used survey is the CPS, conducted by the U.S. Census Bureau. The CPS provides relatively timely reports, traditionally reporting every September about insurance coverage during the previous calendar year. The Committee's report relies primarily on estimates produced by the CPS because of its annual updates and its usefulness in analyzing trends over time. The other surveys described in Table B.1 have been used to address questions that may be difficult or impossible to answer using CPS data alone.

The CPS is the principal source of information on labor force characteristics of the civilian, noninstitutionalized population in the United States. It is perhaps most widely known as the source of data for the monthly unemployment rate. Each month, about 50,000 households, representing approximately 130,000 individuals, are interviewed. In 2001, an additional 30,000 households are to be inter-viewed to provide more precise state-level estimates of the number uninsured. Despite the short (six- to seven-month) turnaround from data collection to publication, the uninsured rates published each September do not reflect the most recent economic trends, because the CPS reporting period for individuals' insurance status begins 21 months before coverage rates are published.

A supplement every March collects income information for households, families, and individuals. In 1980, the March supplement to the CPS began to ask questions about health insurance. Until the March 2000 survey, the CPS did not ask individuals directly if they were uninsured. Instead, it asked about a series of coverage types such as private insurance and Medicare. Those individuals who did not acknowledge any form of coverage were counted as uninsured.

In the March 2000 CPS an experimental question was added to the interview to verify whether a person was uninsured throughout the previous year. While the results from the experimental questions were not known when this report went to press, it is likely that the Census will revise its uninsured estimate downward, making the CPS estimates more consistent with other surveys such as the Community Tracking Survey and the National Survey of American Families that do use a verification question.

Issues that arise in the CPS include the following:

• Recall period. The CPS asks if someone was covered by a particular form of health insurance at any time over the previous year. It appears that respondents answer the question in different ways. Some may report coverage status at the time of the interview, rather than over the previous calendar year. Some may not accurately recall when coverage began or ended and thus fail to report coverage during the period in question (Lewis et al., 1998.)
• Household and family definition. The CPS defines households as all persons under a roof and family members as all related persons in a household. A family unit for health insurance purposes tends to consist of fewer persons than all who live in a household. For example, a single parent and children may live with the child's grandparents, but the grandchildren would be unlikely to be eligible for health insurance coverage through their grandparents' employers.
• Underreporting participation in public programs. Although the CPS uses state program names when asking about Medicaid, and thus refers to Medi-Cal in California and MediPass in Florida, CPS data show fewer people participating in Medicaid than do administrative records for the program.