Effects of Trigger Events on Changes in Children's Health Insurance Coverage

Changes in children's health insurance coverage occur with far greater frequency than the modest year-to-year changes in the proportion uninsured or the proportion with different types of coverage would suggest. We present evidence that in the one-year period from July 1993 through June 1994 there were more than 23 million instances of children changing their coverage among major types of insurance or between covered and uninsured--one change for every three children. Understanding the dynamics of health insurance coverage is important in designing effective strategies to cover the uninsured, and for this reason there is a need to look at the factors that may account for the frequent changes in children's coverage. This report uses data from the 1992 Survey of Income and Program Participation (SIPP) to investigate the role of one set of factors--"trigger events" or sudden changes in the economic situation or composition of the family--in bringing about changes in children's health insurance coverage.

How Often Do Children Change Coverage?

We examined changes among three sources of coverage--employer-sponsored insurance (ESI), Medicaid, and other, primarily private insurance--plus a fourth status: uninsured. Table 1 summarizes our findings. It shows how children were distributed by major source of coverage and how many changes in coverage, or transitions, were recorded among these major types of coverage over a 12-month period. Transitions out of uninsurance and transitions out of ESI were the most common at 7.8 million and 7.2 million, respectively, although the 5.9 million transitions out of Medicaid were not much fewer. What is particularly important to note is how the numbers of changes in coverage compare to the average numbers of children who were in these states at any one time. The total transitions out of uninsurance were 87 percent of the number who were uninsured at any one time, and the transitions out of other insurance were 87 percent of the average number covered by other insurance. Transitions out of Medicaid were about 45 percent of the average enrollment while transitions out of ESI were only 17 percent of the total covered.

Transitions into each of the coverage statuses were nearly identical to the exits, explaining why we see so little change in the aggregate distribution of coverage from year to year. The uninsured and those with other insurance grew by 200 to 300 thousand over the year while Medicaid declined by about 400 thousand.

Destination statuses were distributed very differently depending on the type of coverage that children were leaving. Just over half of the children who left ESI became uninsured--more than 3.6 million. The remainder were about equally likely to enroll in Medicaid or to obtain other insurance. The 7.8 million children who left uninsurance obtained ESI or Medicaid with about equal frequency whereas the 5.9 million who left Medicaid ended up uninsured more than two times out of three.

Do Children Return to Their Original Coverage?

The nearly equal numbers of transitions into and out of each coverage status raise questions about the source of this near-equilibrium. Do children return to their original source of coverage? Indeed, many do. Children who changed their health insurance coverage often changed it again in the next four months--the interval between SIPP interviews. About 40 percent of all changes were followed by a second change within this time span, and four out of five of these (or 32 percent overall) involved a return to the original source of coverage. What may be of more importance, however, is how this phenomenon varied across types of transitions. Depending on both the original source and the destination, the frequency with which an initial change was followed by a second change ranged from 23 to 64 percent. Children leaving ESI were the most likely to have a second transition and the most likely to return to their original source of coverage, whereas children who moved into ESI were the least likely to have a second transition, doing so only 23 to 24 percent of the time.

Movements between ESI and Medicaid are of particular interest to policymakers seeking to cover the uninsured without drawing children out of employer-based coverage. It is noteworthy, then, that the children who were most likely to change their coverage again and the most likely to return to their original coverage were those who moved from ESI to Medicaid. For these children, 64 percent changed their coverage again in the next four months, and 55 percent returned to ESI. At the same time, children who moved from Medicaid to ESI were the least likely to change their coverage again in the next four months. Only 23 percent of the children who moved from Medicaid to ESI changed their coverage again in the next four months, and less than 16 percent of those who left Medicaid returned to Medicaid in four months.

To What Extent Do Parents Change Coverage with Their Children?

Parents mirrored their children's changes in coverage more than half the time. When they did not, the children split about equally between those whose parents kept the coverage that the children exited and those whose parents did not share the same coverage that their children exited. For the most part these patterns are explained by the way in which children obtain their coverage, with ESI and other insurance being obtained via a covered parent and Medicaid becoming increasingly available to children without their parents' participation. What perplexed us most was the finding that among children moving from ESI to Medicaid, two-thirds had parents who retained their ESI. Our earlier findings indicate that half of these transitions from ESI to Medicaid were reversed within four months. Nevertheless, the circumstances surrounding these transitions merit further research.

It was also noteworthy that about one-fifth of the 7.5 million children who lost ESI or Medicaid and became uninsured had parents who reportedly retained their own coverage. These transitions invite additional research as well.

What Potential Trigger Events Precede Changes in Coverage?

Events representing changes in the parents' employment status, jobs, or hours worked; family income; family headship or size; and participation in AFDC were shown to have occurred with greater frequency among children who experienced transitions in health insurance coverage than among children who did not. Depending on the type of transition, between 29 and 50 percent of transitions were accompanied by trigger events in the preceding month, and between 53 and 75 percent had trigger events in the preceding six months. The strongest association between potential trigger events and transitions appeared among children who lost ESI and became uninsured. Children who moved from ESI to Medicaid showed weaker evidence of employment or income-related events, which is consistent with the finding that two-thirds of their parents retained ESI, but we find no suggestion of what else may have helped to produce these changes in coverage. Parents' gains in employment and changes in family income appeared important in moving children out of other insurance, but this was as true of children who lost all coverage as it was of children who obtained ESI. Other than the loss of AFDC, possible trigger events were generally weakest in their influence and perhaps the most inconsistent among children who left Medicaid.

How Do Trigger Events Affect the Likelihood of Changes in Coverage?

Regression analysis of the effects of particular events on the likelihood that children who have a given type of insurance coverage (or none at all) will experience a transition to a specific other type of coverage indicates the following.

Children with ESI. Loss of employment, reduction in hours, and changes in jobs by either parent had a significant effect on children moving from ESI to uninsured--as did a marked drop in family income and a decline in family size. Only the parents' loss of employment or a parent leaving the family affected moves from ESI to Medicaid, however, while the father's reduction in hours and either a marked rise or fall in family income contributed to children leaving ESI for other insurance.

Children without Insurance. Events with a significant effect on the likelihood of uninsured children becoming covered by ESI were limited to an increase in the hours worked by either parent, a marked rise in family income, and a parent joining or rejoining the family. The mother's change in jobs or loss of employment and a parent joining or leaving the family had significant effects on the likelihood of an uninsured child gaining coverage through Medicaid. The mother's changing jobs also contributed to children becoming reinsured through other insurance, as did a marked rise or fall in family income. An increase in family size significantly reduced the likelihood of a child obtaining coverage through other insurance, however.

Children with Medicaid. The family's loss of AFDC, the father's gaining employment, and the mother increasing her hours of work to 30 or more had significant effects on the likelihood that a child would leave Medicaid either by obtaining ESI or becoming uninsured. The loss of AFDC had a stronger effect on the odds of a child becoming uninsured than obtaining ESI. The mother's changing jobs also contributed to the likelihood that a child became reinsured by obtaining ESI while the father's loss of employment and either a marked rise or fall in family income contributed significantly to children leaving Medicaid and becoming uninsured, although the mechanisms behind the effects of job loss and falling income are not obvious.

Children with Other Insurance. The mother's changing jobs or the father's increasing his hours worked had significant effects on children changing their coverage from other insurance to ESI while the father's losing employment or the family's income falling markedly contributed to children moving from other insurance to uninsured. Either parent's gaining employment also increased the odds of children moving from other insurance to uninsured, which is difficult to understand unless it represents parents dropping expensive private coverage in anticipation of ESI that will be available following a brief waiting period.

What Does This Analysis Tell Us About Why Transitions Are So Numerous?

While we did not address this macro level question explicitly, trigger events provide a mechanism that is capable of accounting for the volume of transitions--and for changes that may develop over time. The events that we examined occurred with varying frequency in the different coverage groups, and when particular events occurred the children who experienced them often experienced changes in their health insurance coverage shortly thereafter. For children with ESI, 15 to 30 percent left ESI in the next four months. For uninsured children, 35 to 45 percent became insured in the next four months. Many of the events that we examined are potentially sensitive to changes in the economy. If particular events become more frequent or less frequent, will the transitions with which they are associated be affected as well? The question is important, but to answer it we need to observe changes in the frequency of events and then assess their impact on transitions. Comparison of the late 1990s with the earlier years included in this study may provide the material with which to answer this question.