Codes and Frequencies
VERIFY gives the response to a question about whether an individual who had not reported health insurance coverage earlier in the survey did, in fact, have any insurance coverage during the previous calendar year.
Prior to the 2001 ASEC survey, the Census Bureau identified the uninsured population as the residual population who did not report any specific type of health insurance coverage in response to a long series of questions (about Medicaid, Medicare, other government insurance programs, employment-based group coverage, and privately-purchased plans). Beginning in 2001, interviewers asked a follow-up verification question: "Does the person with no coverage reported previously have any coverage?" Those who answered "Yes" to the verification question were then asked about the type of coverage, and could report up to six types of health insurance for the previous calendar year.
For survey years prior to 2001, persons labeled uninsured may have hypothetically answered "Yes" to a verification question and subsequently would have been labeled insured. For the ASEC 2001 survey and later survey years, only persons who also answered "No" to the verification question (and are coded as "1" in VERIFY) were treated as uninsured during the previous calendar year. Persons coded as "2" or "yes, had insurance coverage" in VERIFY were, in fact, insured.
Adding the verification item dramatically reduced the estimated number of persons without health insurance coverage; about 8 percent of persons previously classified as not having health insurance (without reference to the verification question) were actually insured, based on the information in VERIFY.
The addition of the verification question introduces serious inconsistencies in estimates of insurance coverage before and after the ASEC 2001 survey. Users who wish to conduct time series analysis on the uninsured population should therefore treat persons coded as "Yes" in the VERIFY variable as uninsured if they are studying the period both prior to and after the 2001 survey. Alternatively, users can analyze the summary health insurance variables instead, which contain imputed verification information for the 1988-2000 period (see HCOVANY).
As noted above, individuals who answered "Yes" to the verification question were asked about their specific type of health insurance during the previous calendar year, and could report up to six different types of health insurance coverage. In IPUMS-CPS, these responses have been added to the variables for specific insurance types. For example, people who said they had Medicaid coverage after answering "Yes" to the verification question are added, through programming, to the pool of positive responses for the CAID variable. Thus, analyses of specific types of health insurance coverage for the period prior to and after the 2001 survey should also exclude those cases coded as "Yes" in the VERIFY variable. While the Census Bureau's analysis suggests that those with private health insurance coverage (employment-based group coverage and privately purchased plans) were most often captured by the verification question, estimates of health insurance coverage of all types could be affected. Users can produce consistent estimates of coverage for the period before and after the 2001 survey if they treat all persons coded as "Yes" in VERIFY as not covered by any health insurance.
The health insurance variables affected by this are: CAID, HIMCAID, HIMCARE, HICHAMP, CARE, CHAMPUS, CHAMPVA, MILITVA, INDIANH, GROUPOWN, GROUPDEP, PRIVOWN, PRIVDEP, OUT, COVERGH, COVERPI, KIDCAID, and KIDPRIV.
Analysts studying health insurance coverage only for the period beginning with the ASEC 2001 survey will produce more accurate estimates of insurance coverage if they treat persons coded as "Yes" in VERIFY as part of the insured population.
User Note: For a set of variables that addresses many of the comparability issues in the CPS health insurance measures, see the summary health insurance variables created by the University of Minnesota's State Health Access Data Assistance Center.
This variable is comparable over time.
- All persons who did not answer "Yes" to prior questions about coverage via specific types of insurance.
|2001 – 2018||-||-||X||-||-||-||-||-||-||-||-||-|