ASPE Research NotesINFORMATION FOR DECISION MAKERS |
FOCUS ON: Health Insurance | Issued November 1996 |
Health Insurance in 1994 from the Current Population Survey: Measurement DifficultiesPDF Version: http://aspe.hhs.gov/daltcp/reports/1996/rn15.pdf (5 PDF pages) |
The March 1995 CPS (Table 1) shows for 1994 dramatic changes in health insurance coverage compared to the coverage estimates the March 1994 CPS showed for 1993. Some of these coverage changes may reflect conditions in the economy during the two years. Most seem to be caused by changes in the questions asked on the CPS. The number of insured persons increased by two million. The number with employer-sponsored insurance increased by 11 million persons, seven million of whom were dependents. The number with other private coverage fell by nine million persons. The numbers with no insurance and with Medicaid coverage were almost exactly the same in each of the two years.
In the March 1994 CPS the health insurance questions were completely redone with new skip patterns and many new questions, including responses for each child in the family. Approximately a quarter of the sample was asked the new questions. In the March 1995 CPS all the respondents were asked the new question sequence. In both cases, the responses were coded into the old format. Therefore, the responses in both surveys must be classed as preliminary. The 1995 responses have been released, but are not yet fully analyzed. We do know, however, that the number of persons with ESI as dependents will be much larger than the number shown in Table 1 while the number with other private will be about the same number smaller. Census counted persons under age 15 who had coverage from outside the household as having ESI, while they counted older persons in the same circumstance as having other private coverage. We count all such persons as having ESI since prevailing evidence indicates that more than 90% of the insurance holders who covered persons outside their household had ESI. Census is likely to make this change in subsequent years. In addition, the new questions allow persons to be shown as having both ESI and other private coverage. The old questions did not.
The old set of questions asked about public insurance plans (Medicare, Medicaid, VA, CHAMPUS) first. The private insurance questions were fourth on the list. In the new questions, the private insurance questions were asked first.
The increase in employer-sponsored coverage of dependents and the decrease in other private coverage seem to have been largely the result of question changes. The old questions asked if the respondent was covered by any private plan. If the answer was "yes," the respondent was asked if it was in his own name. If not, the interviewer skipped to the next person. If it was in his own name, he was asked if the coverage was received through an employer and who else was covered by his plan. Census then went back through the family and coded anyone else In the household who was covered by the plan as having employer-sponsored insurance. If the insurance was obtained from someone outside the household (most often a separated spouse or a parent), there was no way of determining if the insurance was from an employer. The respondent and any children in the family were by default coded as having "other private" insurance merely because they had not been asked whether or not their insurance was from an employer.
The new set begins the insurance questions by asking about coverage through an employer or union and, in a separate question, about coverage by a plan purchased directly. Further, there is much more certainty about which specific children are dependents on a policy covering the family. The previous questions ascertained that a spouse and a child or children were covered on a given policyholder's plan. Census then attributed the coverage to own children of the insured person even though many grandchildren and other relatives are covered as children under insurance plans. The new questions ask the specific line numbers of persons dependent on the plan of someone with ESI or other private coverage. Therefore, we now know with more certainty which family members (especially children) are covered as dependents by ESI and other private insurance.
This set of questions and perhaps the reordering of the questions on private and public insurance produced a major increase in the number of persons covered by employer group insurance and a somewhat smaller decrease in the number of persons covered by "other private" insurance. It seems to this writer that changing the questions probably made more difference than changing the order. The increase in coverage of dependents lends credence to this view. However, we should also not lose track of the possibility that some part of the change shown on the files may be real. The year 1994 was prosperous with low unemployment. Some of the new jobs filled during the year no doubt provided health insurance to the newly employed.
The number with Medicaid coverage was affected by the reordering of the questions and by a major change in the treatment of children under age 15. Previously, respondents were asked a global question whether all, some, or none of the children in the household were covered by Medicare and Medicaid. The responses were interpreted as Medicaid coverage since very few children have Medicare. In the new set of questions, the respondent is asked separately about the Medicaid coverage of each child in the household. In the last several years, the number of Medicaid recipients shown on the CPS has been growing by about 7-8% per year. The March 1995 number is approximately the same as the March 1994 number. Actual Medicaid records (Table 2) show, however, that while the increase in eligibles was smaller than in recent years, more than a million more people were eligible for Medicaid in FY1994 than in FY1993. The CPS result seems to be mainly the result of the reordering and changing the questions.
The new questions also seem to encourage couples with ESI to respond that they have two separate policies rather than a single policy in which one is the policyholder and the other that person's dependent. In the March 1993 CPS, there were 38 million married couples with both members covered by employer-group insurance (Table 3). Of these, 8.8 million (23%) had two separate policies. The remainder had policies in the name of one spouse while the other spouse was a dependent. In the March 1994 CPS in which only a quarter had the new question, of 37 million such intact couples, 11.2 million (30%) were coded as having two policies. In the new March 1995 CPS, of 39 million such intact couples, 11.5 million (29%) were coded as having two policies. It seems clear that the descriptions of the plans given by respondents have changed, but probably not the true plan descriptions.
CONTACT PERSON: M. Eugene Moyer, Office of Health Policy.
Circulated periodically by the Office of the Assistant Secretary for Planning and Evaluation. This paper reflects only the views of its author(s) and does not necessarily represent the position of the Department of Health and Human Services. For further information on Health Insurance, contact the author at 202-690-6871.
TABLE 1: Health Insurance Coverage in
1994 and 1993 (millions of persons--all ages) |
|||
---|---|---|---|
March 1994 | March 1995 | Increase | |
Employer-Sponsored Insurance | 148.3 | 159.6 | 11.3 |
In own name | 79.8 | 84.0 | 4.2 |
Dependent | 68.5 | 75.7 | 7.2 |
Other Private | 34.0 | 24.7 | -9.3 |
All Private | 182.4 | 184.3 | 1.9 |
Medicare | 33.1 | 33.9 | 0.8 |
Medicaid | 31.7 | 31.6 | -0.1 |
Champus, VA | 9.6 | 11.2 | 1.6 |
Insured | 220.1 | 222.4 | 2.3 |
Uninsured | 39.7 | 39.7 | 0.0 |
TOTAL | 259.8 | 262.1 | 2.3 |
SOURCE: ASPE Tabulations. NOTE: Persons with more than one kind of insurance are included more than once. Values are preliminary since responses to all new questions have not yet been released or analyzed. |
TABLE 2: Medicaid Ever Eligible From the 2082 Forms in Millions, Fiscal Year | |
---|---|
1991 | 32.2 |
1992 | 35.6 |
1993 | 38.8 |
1994 | 40.2 |
SOURCE: Actuarial Research Corporation. |
TABLE 3: Married Couples in Which Both Have Employer Group Health Insurance Coverage: Percent with 2 Plans | |
---|---|
March 1993 | 23% |
March 1994 | 30% |
March 1995 | 29% |
SOURCE: ASPE Tabulations. March 1994 and 1995 values are preliminary since responses to all new questions have not been released. |
To obtain a printed copy of this report, send the full report title and your mailing information to:
U.S. Department of Health and Human Services
Office of Disability, Aging and Long-Term Care Policy
Room 424E, H.H. Humphrey Building
200 Independence Avenue, S.W.
Washington, D.C. 20201
FAX: 202-401-7733
Email: webmaster.DALTCP@hhs.gov
RETURN TO:
Office of Disability, Aging and
Long-Term Care Policy (DALTCP) Home [http://aspe.hhs.gov/_/office_specific/daltcp.cfm]
Assistant
Secretary for Planning and Evaluation (ASPE) Home [http://aspe.hhs.gov]
U.S. Department of
Health and Human Services Home [http://www.hhs.gov]