We do not regard a parent's change in coverage as a trigger event for the child's change in coverage. Rather, we think of changes in the coverage of parent and child as often having a common set of trigger events. When examined in light of changes in children's coverage, changes--or the lack thereof--in the parents' health insurance coverage can be informative with respect to the factors that may have contributed to the observed changes in children's coverage.
It is important to keep in mind how children obtain different types of health insurance coverage because it affects the relationships that we are likely to observe between children's and parents' coverage. For insurance obtained through employers, a nonworking child will be covered only as a dependent on a parent's plan. This may be less true of private insurance purchased on the open market, but what is relevant for our analysis is how SIPP captures children's coverage. For children under 15, the SIPP identifies ESI and what we are defining as other insurance only after first establishing that a responsible adult had such coverage. Therefore, if a child under 15 is reported to have ESI or other insurance, at least one parent present in the household must be reported as having the same kind of insurance. 13 Generally, this holds true for older children as well. If a child loses ESI or other insurance, the parent need not lose his or her coverage, but if the parent retains coverage the implication is that the parent either dropped dependent coverage altogether or suddenly failed to include the child among those who were reported as covered. Conversely, if a child gains ESI or other insurance, then a parent must have either gained it as well or added dependent coverage to existing individual coverage. Reporting error is also possible, but the findings we have just reviewed downplay its likely importance.
Medicaid works differently--both in practice and in how SIPP captures children's coverage. While Medicaid in 1994 was still heavily associated with Aid to Families with Dependent Children (AFDC), children in non-AFDC families could qualify under special eligibility provisions, and the proportion of children who were enrolled without their parents was growing. The SIPP asks about Medicaid coverage for children regardless of whether an adult is covered, but adults who report their own enrollment in Medicaid are asked if any children, and which, are covered. We expect to see children reported as enrolling in Medicaid without their parents and children reported as leaving Medicaid despite having no parent who was covered in the previous month.
Table 5 summarizes what we found with respect to how often parents replicate the transitions reported for their children. For 55 percent of the 23 million transitions, one or both parents made the same transition as the child (column 2). Predictably, this varies by type of transition, and it varies quite a lot. When a child moved from ESI to other insurance or vice versa, a parent made the same transition in 82 percent of the cases. Similarly, when a child made a transition--in either direction--between ESI and uninsured or between other insurance and uninsured, a parent made the same transition in 72 percent to 77 percent of the cases. When the transition involved Medicaid, however, parents were less than half as likely to make the same transition. For children who moved between Medicaid and uninsured, parents repeated the transition only 36 to 38 percent of the time. For transitions between Medicaid and ESI, parents followed their children only 22 to 26 percent of the time. Finally, for transitions involving Medicaid and other insurance, only 20 to 28 percent of the time did a parent make the same transition as the child.
|Type of Transition||Number of
|Percentage Distribution of Children Who Made a Transition|
|Total Number of Transitions||23,325,000||54.9||3.4||18.4||23.3|
|Transitions from ESI to:|
|Transitions from Uninsured to:|
|Transitions from Medicaid to:|
|Transitions from Other Insurance to:|
|SOURCE: Survey of Income and Program Participation, 1992 Panel.|
We are surprised at how often children made transitions into or out of Medicaid independently of their parents--much more than the share of children's Medicaid enrollment that was attributable to child-only provisions. And despite the 82 percent of transitions between ESI and other insurance that included the parent, the remaining 18 percent imply a higher incidence of parents dropping or adding dependent coverage than we would have anticipated.
It was possible that a parent might have made only part of the transition along with the child--specifically, the parent might have exited the initial coverage but not moved into the same final coverage. This would appear to be especially plausible for transitions from ESI to Medicaid. Our findings do not support this speculation, however. Overall, only 3 percent of the transitions involved parents who exited the original coverage along with their children but did not obtain the same new coverage. For transitions from ESI to Medicaid this fraction was only 6 percent, leaving 71 percent still unexplained.
Over all transitions, the remaining children were divided between those whose parents retained the initial coverage that the child exited and those with no parent who even shared the same initial coverage, but there are sharp differences by type of coverage before and after the transition. For children who moved from Medicaid to ESI only 7 percent of the children had parents who were themselves covered by Medicaid and retained it while 65 percent had parents with no Medicaid coverage at all. This is consistent with our expectations about the impact of child-only coverage on the likelihood of parents mirroring their children's transitions. For children who moved from ESI to Medicaid, however, the parents tended to retain ESI rather than not have it. Only 9 percent had no parent with ESI whereas 62 percent had parents with ESI who kept it. Whatever may explain this phenomenon, we recall that more than half of these transitions were reversed within four months.
For children who moved from ESI to uninsured, we have seen than nearly three-fourths of the parents did the same. But we also see that 18 percent of these children had at least one parent who remained covered by ESI. This is not as striking as the Medicaid example, but in some respects it is even more puzzling because the children in question are not replacing their coverage with free coverage but losing it altogether.
When children moved from Medicaid to ESI, Table 5 tells us that 65 percent of them had no parent covered by Medicaid. We can speculate, based on the evidence provided by the transitions from ESI to Medicaid, that many of these children must have had at least one parent covered by ESI prior to the transition. To confirm this conjecture, we compared the parents' ESI coverage before and after the child's transition. The results are summarized in Table 6, which shows the distribution of the parents' ESI coverage prior to the transition, cross-classified by the change in the parents' coverage that accompanied the child's transition from Medicaid to ESI.
Of the 1.75 million children with transitions from Medicaid to ESI and who lived with one or both parents, 74 percent had at least one parent covered by ESI prior to the transition. This included 45 percent with both parents covered by ESI, 15 percent with one parent covered and one parent not covered, and another 15 percent who were living with only one parent, who was covered by ESI. When both parents or the sole parent was already covered by ESI, there was essentially no change in the parents' coverage when the child moved from Medicaid to ESI. The only exception was an inexplicable (but negligible) 3 percent of single parents who lost ESI just as the child was gaining it. Altogether, children whose parents were already fully covered by ESI prior to the child's transition and who reported no change in their own coverage accounted for nearly 60 percent of the children who moved from Medicaid to ESI. The remaining 40 percent of children's transitions were accompanied by transitions to ESI by one or both parents. Specifically, 80 percent of the children
whose parents were not covered by ESI prior to the transition had at least one parent gain ESI, and 66 percent of the children who had one parent covered and the other parent not covered saw this other parent become covered as well.
|Child's Outcome and Parents' ESI
Coverage Prior to Child's Transition*
Children in Month
Prior to Transition
|Change in Parents' ESI Coverage|
|All Children Obtaining ESI||1,751,000||100.0||100.0||69.3||24.6||5.7||0.5|
|Both Parents Covered by ESI||787,000||44.9||100.0||100.0||0.0||0.0||0.0|
|One Covered by ESI, One Not||256,000||14.6||100.0||33.9||66.1||0.0||0.0|
|Sole Parent Covered by ESI||255,000||14.6||100.0||96.7||0.0||0.0||3.3|
|No Parent Covered by ESI||454,000||25.9||100.0||20.5||57.5||21.9||0.0|
|SOURCE: Survey of Income and Program Participation, 1992 Panel.
* Children with no parent present in the household are excluded from this table.