Ineligible for / did not purchase employer health insurance variables


I’m working with ASEC measures indicating reasons that respondents were ineligible for employer’s health insurance, or did not purchase insurance for which they were eligible. Specifically, I’m wondering whether more detail is available about the construction and coding of these measures.

For instance, -hintake7- indicates that a respondent “was eligible for employer-based health insurance, but chose not to purchase it because they were contract or temp employees and not allowed on the plan” (…). Is this respondent then indeed eligible, even though they are “not allowed on the plan”? (see also -hintake6- for employees who haven’t worked for employer long enough to be covered).

Relatedly, -hinelig5- denotes respondents who are “ineligible for employer-based health insurance if their employer offered it because it is too expensive.” I’m unclear how plan pricing would change eligibility status, and how these respondents would differ from those marked “yes” on -hintake3-.

Many thanks for any info anyone can provide.



I’ll answer each of these questions, one at a time.

(1) The universe statement for these variables may help clarify these questions a bit. For the variables HINTAKE7, HINTAKE6, and HINTAKE5 the universe statement suggest that respondents are those who who were not covered by employer-based health insurance plans who are employed and not self-employed who were eligible to purchase an employer’s health insurance plan. Therefore these people are eligible for an employer health insurance plan.

(2) I agree this does sound a bit strange. However, this is actually what the CPS survey asked respondents. See the questions and response categories under WNTAKE and WNELIG in the 2015 ASEC documentation, pp. D128-D129. Looking into this a bit myself, it looks like the variables HINELIG5 and HINTAKE3 are mutually exclusive from each other. The variable HIELIG seems to the be defining factor determining which of these two questions a respondent answered. So, I’m afraid I don’t have a great answer to this question. This paper entitled: “How did the questionnaire change in the CPS ASEC affect health insurance estimates in 2015?” may provide some insight.

After discussing this with a few folks around the office, we found this paper. It isn’t comprehensive, but it hints at how the Census Bureau uses these two variables.

Thanks for the quick and helpful response, Jeff. The paper you link is useful, too.

After digging into the CPS documentation, I think you’re absolutely right about HIELIG. It looks as if respondents are asked whether they COULD enroll in their employer’s plan if they want to. Those who say “yes” are asked about why they don’t; those who say “no” are asked why they aren’t eligible. However, perhaps because of differences in how respondents interpret “could enroll” (probably some in terms of eligibility, some in terms of other constraints, unrelated to eligibility), there is some overlap in potential response categories. That is, some respondents may be eligible, but determine that they can’t enroll (because of cost) while others say they could enroll, but don’t, for the same reason.

Finally, in case it’s a useful note for other users: based on the wording of the HIELIG quesiton in the CPS documentation, it seems that these questions refer to current health insurance status, not last year’s, as many of the other CPS health insurance measures do.