Health Affairs, 23, no. 3 (2004): 233-246 doi: 10.1377/hlthaff.23.3.233 © 2004 by Project HOPE	New Online   Obama's Success & Next Steps   3 Health Policy Options for President Obama   A Hybrid Proposal   Bipartisan Compromise

This Article Abstract Figures Only Reprint (PDF) Supplemental Exhibits Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (16) Citing Articles via Google Scholar Google Scholar Articles by Kronebusch, K. Articles by Elbel, B. Search for Related Content PubMed PubMed Citation Articles by Kronebusch, K. Articles by Elbel, B. Related Collections Access To Care Health Reform Insurance Coverage - Children Legal/Regulatory Issues Medicaid Consumer Issues

DataWatch

Simplifying Children’s Medicaid And SCHIP

Abstract

 
The states have implemented the State Children’s Health Insurance Program (SCHIP) in a variety of ways. We describe these choices and estimate the resulting enrollment impacts. Many widely adopted policies, including mail-in applications and twelvemonth continuous eligibility, have had limited impacts. Other policies that increase enrollment, including presumptive eligibility and self-declaration of income, have not been widely adopted. SCHIP programs administered as Medicaid expansions have been more successful in enrolling children than either separate SCHIP plans or combination programs. Waiting periods, premiums, and welfare reform have had important negative impacts on children’s program enrollment.

As states were implementing SCHIP, welfare reform was simultaneously being implemented. The 1996 federal welfare reform law, which replaced Aid to Families with Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF), made a number of changes. The law created state block grants; ended the entitlement status of cash assistance; and established time-limited benefits, new work requirements, and sanctions for recipients’ failure to comply with program requirements. In implementing welfare reform, the states have developed diversion programs designed to discourage enrollment in welfare, as well as new supportive services, transitional Medicaid benefits, and child care assistance designed to facilitate employment. Because of the historical and administrative connections between welfare and Medicaid, there was concern that welfare reform might adversely affect Medicaid enrollment.

Thus, important policy and administrative innovation has taken place in the states. However, little empirical research has been conducted to date on the effects of these policy choices, in part because of the complex mix of policies being pursued. We start by describing the state policy choices that were most likely to affect enrollment decisions of families, including policies concerning enrollment simplification, and other aspects of Medicaid/SCHIP administration, as well as welfare reform policies that also affect children’s insurance enrollment. To assess the effectiveness of current policy for facilitating enrollment, we use nationally representative data from the Current Population Survey (CPS). Because many of these policies were adopted at the same time, it is particularly important that this analysis control for the multiple policies that states have adopted, as well as the various demographic factors that affect program enrollment. We present our results concerning policy effectiveness and suggest some next steps for the states.

Policy Reforms

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 
Administrative model. States have used three different SCHIP organizational models: Medicaid expansion, a separate SCHIP program, or a combination (see Supplemental Exhibit 1 online).⁴ Administratively, the Medicaid expansion adopted by fifteen states and the District of Columbia was potentially the simplest: States had already designed Medicaid benefit packages and established methods to pay managed care organizations and providers. Creating a separate program (sixteen states) required state action on each of these elements, as well as decisions concerning the level of premiums and copayments. The resulting program structure—with Medic-aid and SCHIP functioning side-by-side—would also be more complex because some families cycle between Medicaid and SCHIP eligibility as incomes and family status change over time. Implementing SCHIP as a combination program (nineteen states) raised many of these same issues: There was a need to establish eligibility, benefits, contracts with health plans, and administrative methods to coordinate the programs. Moreover, many of these issues, such as the need to screen SCHIP applicants for Medicaid eligibility and to distinguish enrollees between the two programs, represent a continuing administrative burden for separate and combination programs.

View larger version (35K): [in this window] [in a new window]   EXHIBIT 1 Effects Of States’ Medicaid And State Children’s Health Insurance Program (SCHIP) Policies On Enrollment Of Children Into Medicaid And SCHIP, 2000

Enrollment simplification. Much attention has been given to policy changes to simplify or otherwise reduce the burden of the application process. Those burdens concern time, information, knowledge, and documentation. Policy changes here made the process less onerous and potentially less humiliating, by minimizing the need for visits with caseworkers in welfare or Medicaid offices.

An important component of welfare and Medicaid eligibility determinations involves examining the ownership of assets. These "asset tests" restrict the eligible pool to only those families who have a low level of assets. Dropping asset tests simplifies documentation requirements, thereby reducing the application burden on families.⁶ Currently, only one state imposes asset-related SCHIP eligibility tests, and all but six states and the District of Columbia have dropped these requirements for children’s Medicaid (Supplemental Exhibit 1).⁷

Second, states created mail-in and telephone applications, which eliminate the traditional requirement for face-to-face interviews. These innovations reduce the time burden on potential recipients and separate the Medicaid/SCHIP application process from the traditional process in welfare offices. These procedures were adopted by all states for their SCHIP-eligible recipients and today all but two states for their Medicaid programs as well.⁸

Beginning in the 1980s, "presumptive eligibility" provisions were developed to allow designated providers to determine whether a person is likely to be eligible for Medicaid and then to treat that person and receive Medicaid payment. This moves the initial stage of the application process outside the welfare office and allows uninsured children to receive needed medical care immediately. Ten states have adopted presumptive eligibility for either Medicaid or SCHIP.⁹

State enrollment procedures historically used for welfare have required that recipients provide detailed information concerning their incomes and supporting documentation, such as pay stubs. Under the typical approach to self-declaration of income, applicants provide information on their incomes for initial determinations and redeterminations, with the states using existing administrative records to verify, supplemented with random audits.¹⁰ The usual welfare-related procedures included a monthly redetermination process to verify continued eligibility. Generally, failure to provide the documentation or to complete this process led to loss of enrollment and potentially cycling on and off the program from month to month. The states have developed several modifications to this process—most prominently, the creation of twelve-month continuous eligibility, which in its most expansive form provides for eligibility that does not require monthly redeterminations and does not terminate, even with changes in family income or other characteristics that would affect eligibility.¹¹

Waiting periods and premiums. The issue of insurance substitution arose because at least some families who had become eligible for child-related coverage through the Medicaid expansions of the 1980s and 1990s appeared to have had private insurance coverage, which they dropped or did not take up when they became eligible for Medicaid. The income range for SCHIP eligibility is somewhat higher than for Medicaid and so would be expected to include families who have the option of employer-sponsored coverage for their children, as well as income that could be used to contribute to the new coverage. To reduce crowding out, policymakers have included waiting periods in SCHIP, requiring potential recipients to have lacked access to employment-related coverage for a minimum period of time before enrolling. In the thirty-three states imposing these requirements, the waiting periods range from two to twelve months (Supplemental Exhibit 1).¹²

To limit the budget impacts and to imitate private health insurance, states have also established out-of-pocket premiums for SCHIP. These apply only to separate SCHIP programs and the SCHIP part of combination programs in twenty states. Premiums or fees are often set on a sliding scale.¹³

Welfare reform. Welfare reform can affect Medicaid and SCHIP enrollment because the programs are often connected, by the use of the same or closely related administrative agencies and by public perceptions. When welfare reform was enacted in 1996, Medicaid was delinked from its historical connections with welfare, at least at the level of formal eligibility policy. Adults in families with dependent children would continue to be eligible for Medicaid based on the eligibility provisions in place in July 1996, generally without regard to how states implemented welfare reform. In addition, the 1989–90 federal mandates concerning Medicaid eligibility for children in low-income families were left in place, with children’s Medicaid eligibility protected from welfare-related time limits or state-imposed sanctions for parents’ noncompliance with welfare requirements. Nevertheless, welfare reform has been associated with declining Medicaid enrollment for adults and children.¹⁴

Several specific welfare reform policies were likely to affect program enrollment (Supplemental Exhibit 2).¹⁵ "Diversion programs" were developed to discourage potential welfare applicants from applying, often simply by explaining the time-limited nature of assistance and suggesting the pursuit of assistance from other sources. In twenty-nine states policymakers went further, establishing cash payments to those who might otherwise have applied for benefits, as long as the family does not seek formal enrollment in the program for a specified period.

View larger version (22K): [in this window] [in a new window]   EXHIBIT 2 Effects Of States’ Welfare Policies On Enrollment Of Children Into Medicaid And The State Children’s Health Insurance Program (SCHIP), 2000

Time-limited benefits were an important feature of the 1996 federal welfare reform law. The law set a timetable for the implementation of time limits, but many states established time limits sooner than required. Finally, state implementation of welfare reform has varied with respect to the stringency of administrative sanctions. The federal welfare reform law prohibits the states from applying these sanctions to children’s Medicaid enrollment. But states could apply them to adult Medicaid eligibility. In the sixteen states adopting this provision, failure to comply with work-related requirements can result in loss of not only welfare but also adult Medicaid benefits.¹⁷

Study Data And Methods

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 
Data. To determine the effects of SCHIP/Medicaid enrollment simplification and welfare reform policies on enrollment rates, we used data from the March 2001 Current Population Survey (CPS), which surveys health insurance status for 2000. We restricted the sample to all children under age eighteen with family incomes below 400 percent of poverty. The resulting data set has 25,264 observations. Because SCHIP outreach and enrollment policies may affect Medicaid enrollment, and because one administrative model for SCHIP is to expand Medicaid, it is important to examine both programs together. So, our dependent variable here is coded to reflect enrollment of a child in either Medicaid or SCHIP. We estimated logit models to determine the factors that predict Medicaid or SCHIP enrollment compared with not participating in these programs, which includes those who are uninsured as well as those with private insurance.

State-level policy characteristics. Using reported annual family income, we coded two dichotomous variables to identify children who are eligible for Medicaid either because they are under age six with family income below 133 percent of poverty or because they are age six or older with family income below 100 percent of poverty. We also coded SCHIP eligibility based on having income above the relevant Medicaid threshold but below state-set SCHIP eligibility limits. Because there may also be a spillover or communitywide impact of broadened eligibility on take-up rates, we also included the maximum income cutoff for each SCHIP program, expressed as a percentage of poverty.

In addition to directly modeling eligibility, we incorporated state policy characteristics in effect as of 30 June 2000.¹⁸ First, we included two dichotomous variables concerning the administrative form of SCHIP (Medicaid expansion or separate program; states with "combination" programs are the left-out comparison category). For enrollment simplification, we coded dichotomous variables indicating whether states have asset tests for either SCHIP or children’s Medicaid and whether they have implemented presumptive eligibility, self-declaration of income, and mail-in/telephone applications.¹⁹ For continuous eligibility provisions, there may be an interaction between the two programs, since continuous eligibility for Medicaid will lead families to continue to have Medicaid coverage even if their incomes rise into the SCHIP eligibility range. We therefore coded two dummy variables to capture the presence of program-specific continuous eligibility for Medicaid and SCHIP. We included the waiting period in months. We determined the likely premium that each family would face using their reported income and published information on premiums by state.

We measured several different aspects of welfare reform. First, we incorporated a variable indicating whether a state had time limits in place in 2000. We also included dichotomous variables indicating that a state has adopted a policy that its work-related sanctions for adult welfare recipients also affect adult Medicaid eligibility, uses cash diversion payments, or provides emergency cash assistance.

Individual and family characteristics. The starting point for modeling eligibility and program enrollment is family income, which is both the basis for administrative determinations of eligibility and the primary factor inducing people to seek enrollment. As income rises, families are more likely to have access to employment-related insurance and an increased ability to purchase private insurance, and this will reduce program participation. We measured income as "pre welfare income," which we defined as total income minus cash welfare payments, and we used annual income and income squared in our model to allow for potentially nonlinear effects. We constructed separate, individual-level eligibility variables for Medicaid and SCHIP, using each state’s income cutoffs and the child’s family income.²⁰ To control for the family-level administrative connection between welfare receipt and Medicaid coverage, we included a dichotomous variable measuring whether the family received welfare payments.

We also controlled for demographic characteristics that previous research has shown to affect the likelihood that families seek enrollment in Medicaid or SCHIP. In addition to income, these factors include age of the child, race/ethnicity (classified here as white, African American, Latino, American Indian, or Asian American), whether the child lives in a mother-only or father-only family (as opposed to a family headed by a couple), whether the family head is employed full time (more than thirty-five hours per week for at least fifty weeks of the previous year), the age and education level of the family head, and whether the family has children under age six.

Study Results

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 
To facilitate interpretation, we present our results in terms of marginal probability changes, which represent the impact of each policy on average enrollment rates. The probability of Medicaid/SCHIP enrollment for extremely poor children (family incomes less than 250 percent of poverty) is approximately 62 percent, which means that 62 percent of children in this income range are reported to be enrolled in either Medicaid or SCHIP. This declines to approximately 52 percent for incomes between 50 and 100 percent of poverty and 35 percent for those with incomes between 100 and 150 percent of poverty. Each of the policy impacts that we discuss below can be understood as additions to these baseline probabilities.

Effects of Medicaid/SCHIP policies. The presence of asset tests reduces children’s enrollment by 6.2 percentage points (Exhibit 1). For enrollment simplification, the results show that presumptive eligibility and self-declaration of income both increase enrollment. The elimination of required face-to-face interviews and the use of mail-in and telephone applications have no apparent effect on enrollment levels. Continuous eligibility provisions should lead to increased enrollment, but the results show no significant effect on combined enrollment for the two programs.

Waiting periods reduce SCHIP enrollment, through two separate effects: First, they restrict the pool of eligibles by adding the requirement that recipients must be uninsured for a specified period of time, in addition to having income that falls below the eligibility limit. Second, they create the impression of a less available program, and so even those who are eligible may become discouraged. The net result of a six-month waiting period is a fall in overall enrollment of 6.1 percentage points (Exhibit 1). SCHIP premiums also reduce enrollment. Although we measure only premiums and no other cost-sharing requirements, these results are consistent with other research on the effects of premiums on disenrollment.²¹

The choice of administrative models for the SCHIP expansion is as consequential for enrollment levels as are a number of the other measures designed to improve enrollment. States using Medicaid expansions have higher enrollment levels—an increase of 2.7 percentage points compared with combination programs and a net difference of 2.0 percentage points compared with separate SCHIP programs, which are not significantly different from the combination programs.

Effects of welfare reform. Medicaid/SCHIP enrollment in states that adopted time limits relatively early is not significantly different from that in the remaining states (Exhibit 2). However, in states that terminate welfare eligibility as the first sanction for a participant, there is a negative effect on children’s Medic-aid/SCHIP enrollment. Similarly, states with cash diversion programs also have a lower level of children’s Medicaid and SCHIP enrollment. In states that provide emergency cash assistance, however, recipients establish a connection to administrative agencies, and the result is an increased enrollment rate.

Study limitations. We note several limitations to this analysis. First, we are relying on the accuracy of the CPS. Comparisons between the CPS and administrative data reveal discrepancies, and many analysts suggest that the CPS generally underestimates Medicaid enrollment.²² Although this could affect the overall estimated enrollment levels, this bias is unlikely to be correlated with the policy and implementation characteristics that we analyzed here, and so we do not expect it to affect the estimated impacts of these policies.

Second, we lack the data to fully model the individual-level decision-making process. In particular, stigma and other perceptions about programs, and expectations about how a potential applicant will be treated by program administrative staff, are important factors in understanding enrollment decisions.²³ Since the CPS does not ascertain these individual-level perceptions, we could not incorporate them in this analysis. To the extent that these perceptions are correlated with the state policies we examine, the estimated coefficients reflect these impacts as well. But because stigma and perceptions about a program are in part a response by potential recipients to these policy characteristics themselves, the enrollment impact of these perceptions can be attributed, at least partially, to the underlying policy itself.

Third, several policy characteristics are most directly applicable to the relatively smaller number of families enrolled in SCHIP, as opposed to the much larger number of families in Medicaid. For example, SCHIP premiums will most strongly affect the decisions of those considering SCHIP enrollment and will not apply to those eligible for Medicaid. Thus, our results underestimate the impacts of premiums on these potential recipients. At the same time, many potential recipients may not understand which program they are eligible for, and they may be deterred from applying because they incorrectly perceive that they will face premiums. Although it might be possible to separate these two groups of potential recipients, we have instead simply combined them for this analysis of the impacts of premiums.

The results measure the average effectiveness of these policies across all fifty states. It is possible that the effectiveness varies across the states, with higher levels of effectiveness in some states, although we lack the data necessary to test for differential effectiveness. Among state policy characteristics, we have chosen those that appeared likely to be important influences on enrollment. We have not been able to measure some important policy characteristics—in particular, the extent of state and community-level outreach activities. A variety of approaches have been deployed to communicate to likely recipients that their children are eligible and to assist them in enrolling.²⁴ However, little systematic research has been done on these impacts, in part because of the difficulty of actually measuring across the states the size and location of these outreach activities.

Discussion

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 
The goal of providing health insurance for children is widely shared, and there is evidence that insurance enrollment improves access to health care and health outcomes. As a result of federal Medicaid mandates of 1989–90 and the creation of SCHIP in the late 1990s, the number of children eligible for state-administered health insurance programs has increased. Approximately 32.1 million children are income-eligible for Medicaid or SCHIP under current state eligibility standards (Supplemental Exhibit 3).²⁵ The federal Medicaid mandates now cover about half of the total pool of eligibles, or 15.8 million children, while the SCHIP expansions added 16.3 million.

Enrollment rates, however, are much lower than the potential eligibility. Using reported annual family income, we can identify specific children as eligible for Medicaid or SCHIP, but only about 44 percent of them are actually enrolled in the relevant programs. Enrollment rates vary among the states: Eight states have rates that are less than 35 percent, while several other states have rates greater than 60 percent.²⁶ Since these estimated enrollment rates are based on a survey sample, there is some statistical uncertainty associated with them.²⁷ Because of this, and the broader limitations of survey data discussed above, we need to be somewhat cautious about attempting to rank the states when the differences between their estimated rates are fairly small. Nonetheless, our analysis finds that the states vary greatly in the level of success they have had in implementing SCHIP and enrolling both SCHIP- and Medicaid-eligible children.

Some of the children who are not enrolled in Medicaid or SCHIP are covered by private insurance. For the nation as a whole in 2001, nearly six million children were uninsured, even though they were eligible for either Medicaid or SCHIP. Of the 32.1 million children who were eligible for either Medicaid or SCHIP, about 19 percent remained uninsured.²⁸ The differences among states are related to the extent of private insurance coverage, state policy characteristics, and program enrollment rates. Owing to a combination of these reasons, some states have achieved fairly high levels of insurance coverage, while other states have been less successful. Eligible children lacking health insurance range from as low as 4 percent to as high as 32 percent across the states (see Supplemental Exhibit 3 for state-specific estimates).²⁹

Increased challenge. The challenge of reducing these rates has now become even greater. The earlier expansions occurred during a time of relative economic prosperity and emerging government budget surpluses. Economic downturns and budget deficits place these expansions at some risk. Many states have considered cutting back the state-set maximum income limits for SCHIP eligibility. Some states have scaled back or eliminated innovative expansions directed toward low-income adults, including "family coverage" expansions designed to cover adult caregivers of children eligible for SCHIP, and expansions for childless adults. Although many states have avoided overt changes to children’s eligibility, this is not true of all.³⁰

The risk here is potentially quite large. If all states cut back their current income eligibility standard by fifty percentage points of poverty (for example, a state currently using 200 percent of poverty cutting back to 150 percent), we estimate that about 7.3 million children would lose SCHIP eligibility, or nearly one-fourth of the 32.1 million children eligible for SCHIP and Medicaid. Some of these families could purchase private insurance, but most families would be unable to afford it during a time of economic downturn. To prevent a rise in the number of uninsured children, policymakers need to maintain the safety net already in place.

States’ implementation approaches for Medicaid administrative simplification and for SCHIP have affected their levels of Medicaid enrollment. Our results show how easily certain implementation choices can reduce or even reverse the favorable impacts of administrative simplification or eligibility expansions. So, for example, the impact of eliminating asset tests is an increase in enrollment of 6.2 percentage points. If a state also adopted presumptive eligibility (an increase of 6.4 percentage points) and self-declaration of income (an increase of 3.5 percentage points), the total enrollment increase would be 16.1 percentage points. This would represent a considerable improvement over the current mean enrollment probability of 33 percent for those with incomes of 100–150 percent of poverty.

This gain can be dramatically reduced, however. Adopting a six-month crowding-out requirement reduces enrollment by 6.1 percentage points, implementing welfare reform that terminates benefits as the first sanction reduces enrollment by 5.0 percentage points, and using cash diversion approaches for welfare reform reduces enrollment by 2.1 percentage points. Together, these policy and administrative changes would lower enrollment by 13.2 percentage points, which would almost completely offset the gains described above.

Unfortunately, the most commonly adopted policies also appear to have limited, if any, enrollment impact. The majority of the states have failed to adopt policies that expand enrollment, and many have adopted SCHIP waiting periods and welfare reform policies that actually reduce enrollment. The most commonly adopted enrollment simplification has been to adopt mail-in and telephone applications, but these changes appear to have no effect on enrollment levels, when other policy and demographic characteristics are controlled for. Most of the states have dropped asset tests, a policy that significantly increases enrollment. But only fifteen states have adopted self-declaration of income for either Medicaid or SCHIP, and only ten states have adopted presumptive eligibility, both of which can improve average enrollment. At the same time, thirty-three states have adopted crowding out–related waiting periods that reduce enrollment, including fourteen states with a waiting period of six months. For welfare reform, thirty-four states provide emergency assistance, which appears to facilitate children’s insurance enrollment. But sixteen states terminate adult Medicaid as a first sanction, and twenty-nine employ cash diversion programs, both of which have a negative impact on children’s enrollment.

Choice of administrative model. One of the first implementation choices for the states concerned the administrative model for the new programs. Our results indicate an enrollment advantage for the Medicaid-expansion model, adopted by only one-third of states. It is not clear why this approach has been more successful than either separate SCHIP or combination programs. One possibility concerns the important implementation and administrative burden associated with creating a new program, either as a separate program or in combination with Medicaid. Other potential advantages for Medicaid as an administrative model include the potential for better continuity and more seamless integration of enrollment for agency staff, who are required to screen SCHIP applicants for Medicaid eligibility, as well as for recipients who potentially shift between programs when family income changes. In addition, Medicaid expansion provides recipients with the comprehensive Medicaid benefit package, rather than the sometimes more limited benefits afforded by separate SCHIP programs.³¹ Finally, outreach efforts oriented around SCHIP will automatically apply to Medicaid under a Medicaid expansion.³²

Much attention has been given to simplifying the enrollment processes. Our results are typical of administrative and policy experimentation: Some changes are effective, while others are much less so. Mail-in and telephone application procedures have had no apparent impact on enrollment, and continuous eligibility appears to have had a small and insignificant impact. Presumptive eligibility, on the other hand, has had a relatively large and favorable impact. Our models also reveal that receipt of welfare continues to have a very large impact, virtually guaranteeing Medicaid enrollment of children in such families.

Importance of personal contact. Our results reveal the importance of direct, personal administrative contact. We suggest here that welfare receipt and presumptive eligibility are effective pathways into Medicaid and SCHIP because in both cases caseworkers and eligibility workers provide potential recipients with an administrative connection to the program. Mail-in applications should be less burdensome in terms of recipients’ time. Continuous eligibility reduces recipients’ time burdens but also reduces direct contact with agency staff. Again, the administrative change is to replace direct personal contact with simplified paperwork. The assumption here is that the simplification and reduced burdens would lower enrollment barriers sufficiently so that potential recipients will apply successfully. Facing a series of complex programs, without personal assistance, recipients do not learn about the programs, fail to obtain forms, and fail to complete them. Program separation by itself does not provide the assistance that many potential recipients need.

People who receive health insurance through employment do not face administrative hurdles like those imposed on those who apply for means-tested programs; instead, workers can rely on the intervention of their employers’ human resources and personnel departments to complete these paperwork requirements.³³ Motivated by a desire to destigmatize public programs and to reduce burdens on recipients, policymakers delinked Medicaid from welfare, created SCHIP programs that were separate from Medicaid, and deployed innovative enrollment strategies. This strategy, however, may not have provided recipients with enough direct assistance to permit them to act on the expanded eligibility for their children. Many states and private organizations have been creating outreach programs that may address this gap. For this analysis, we lack information on the extent of these programs and so cannot comment on their actual effectiveness. But we believe that personal contact can play an important role in shifting from a welfare-based administrative model to the delinked approach that has been a policy goal.

Our results suggest that policymakers and program administrators who seek to increase enrollment should adopt presumptive eligibility procedures and self-declaration of income, if they haven’t already. Simply adopting mail-in application procedures will not be sufficient. A more interventionist approach seems warranted, especially since the historical administrative connections with welfare were intended to be replaced. At the same time, while federal policy requires efforts to prevent insurance substitution, states can find alternatives to simply setting minimum waiting periods, since these are associated with lower enrollment. States need also to develop new approaches to insulate children from welfare reform. Program expansions increase enrollment but are now less likely to be adopted, given budgetary pressures. These budget constraints will also tempt states to shift the enrollment burden to recipients themselves, both to save administrative costs and to reduce program expenses. During the past two decades policymakers have voiced a strong commitment to children’s health insurance. We need to take action to implement the approaches that can make this commitment a reality.

Editor's Notes

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 
Karl Kronebusch (karl.kronebusch{at}yale.edu) is an associate professor in the Division of Health Policy and Administration, Yale University, in New Haven, Connecticut. Brian Elbel is a doctoral candidate there.

NOTES

Top Editor's Notes Policy Reforms Study Data And Methods Study Results Discussion NOTES

 

1. See W. Brandon, R. Chaudry, and A. Sardell, "Launching SCHIP: The States and Children’s Health Insurance," in The New Politics of State Health Policy, ed. R.B. Hackey and D. Rochefort (Lawrence: University Press of Kansas, 2001).
2. D. Cohen Rossand L.Cox,Making It Simple: Medicaid for Children and CHIP Income Guidelines and Enrollment Procedures, Findings from a Fifty-State Survey (Menlo Park, Calif.: Henry J. Kaiser Family Foundation, October 2000); and V. Pulos and P. Denker, One Step Forward, One Step Back: Children’s Coverage after CHIP and Welfare (Washington: Families USA, October 1999).
3. D. Cutler and J. Gruber, "Does Public Insurance Crowd Out Private Insurance?" Quarterly Journal of Economics (May 1996): 391–430; L. Dubay, Expansions in Public Health Insurance and Crowdout: What the Evidence Says (Menlo Park, Calif.: Kaiser Project on Incremental Health Reform, 1999); and R. Kronick and T. Gilmer. "Insuring Low Income Adults: Does Public Coverage Crowd Out Private?" Health Affairs 21, no. 1 (2002): 225–239.[Abstract/Free Full Text]
4. Complete tabular results, Supplemental Exhibit 1, are available at content.healthaffairs.org/cgi/content/full/23/3/233/DC1.
5. Ibid.
6. V.Smith, E. Ellis, and C. Chang, Eliminating the Medicaid Asset Test for Families: A Review of State Experiences (Menlo Park, Calif.: Kaiser Family Foundation, April 2001).
7. See Supplemental Exhibit 1, content.healthaffairs.org/cgi/content/full/23/3/233/DC1.
8. Ibid.
9. Ibid.
10. L.Cox, Allowing Families to Self-Report Income: A Promising Strategy for Simplifying Enrollment in Children’s Health Coverage Programs (Washington: Center for Budget and Policy Priorities, 28 December 2001).
11. Pulos and Denker, One Step Forward, One Step Back.
12. See Supplemental Exhibit 1, content.healthaffairs.org/cgi/content/full/23/3/233/DC1.
13. Ibid. We present in Supplemental Exhibit 1 the level of these premiums at 151 percent of the federal poverty level. SCHIP insurance benefits may also require out-of-pocket cost sharing as well (data not shown).
14. Pulos and Denker, One Step Forward, One Step Back;R. Klein and C. Fish-Parcham, Losing Health Insurance: The Unintended Consequences of Welfare Reform, Pub. no. 99-103 (Washington: Families USA Foundation, 1999); L. Ku and B. Bruen, The Continuing Decline in Medicaid Coverage, New Federalism: Issues and Options for States no. A-37 (Washington: Urban Institute, 1999); W. Chavkin, D. Romero, and P.H. Wise, "State Welfare Reform Policies and Declines in Health Insurance," American Journal of Public Health 90, no. 6 (2000): 900–908;[Abstract/Free Full Text] and K. Kronebusch, "Medicaid for Children: Federal Mandates, Welfare Reform, and Policy Backsliding," Health Affairs 20, no. 1 (2001): 97–111.[Abstract/Free Full Text]
15. See Supplemental Exhibit 2, content.healthaffairs.org/cgi/content/full/23/3/233/DC2.
16. Ibid.
17. Ibid.
18. Policy characteristics were identified from the following sources: C. Pernice et al., Charting SCHIP: Report of the Second National survey of the State Children’s Health Insurance Program (Portland, Maine: National Academy of State Health Policy, July 2001); Cohen Ross and Cox, Making It Simple; State Policy Documentation Project, "Temporary Assistance for Needy Families (TANF)," 6 June 2001, www.spdp.org/tanf.htm (20 February 2004); and Centers for Medicare and Medicaid Services, "State Medicaid Plans and Plan Amendments," 26 June 2003, www.cms.gov/medicaid/stateplans (20 February 2004).
19. Since all states but one have adopted mail-in applications for SCHIP, this refers to whether they have adopted mail-in applications for Medicaid. Only thirty-nine states had adopted this procedure in 2000.
20. This ignores the use of various disregards by the states for earned income, child care, and so on. Incorporating these disregards would increase the number of children in the sample who are simulated to be eligible. State eligibility limits are generally based on monthly income. The Current Population Survey, however, reports annual income. We coded individual-level eligibility based on having annual income below twelve times the relevant monthly income limits. For states implementing SCHIP via a Medicaid expansion, we coded the expansion group as Medicaid eligible, which corresponds to the insurance status reported in the CPS.
21. See E. Shenkman et al., "Disenrollment and Re-enrollment Patterns in SCHIP," Health Care Financing Review 23, no. 3 (2002): 47–63.[Medline]
22. There is also some uncertainty about how people respond to health insurance survey questions.See K. Swartz, "Interpreting the Estimates from Four National Surveys of the Number of People without Health Insurance," Journal of Economic and Social Measurement 14, no. 3 (1986): 219–232; and J.C. Ham and L. Shore-Sheppard, "The Effect of Medicaid Expansions for Low-Income Children on Medicaid Participation and Insurance Coverage: Evidence from the SIPP," Journal of Business and Economic Statistics (forthcoming).
23. M.Perry et al., Medicaid and Children: Overcoming Barriers to Enrollment. Findings from a National Survey (Menlo Park, Calif.: Kaiser Family Foundation, 2000); and J. Stuber and K. Kronebusch, "Stigma, Enrollment Barriers, and Policy: Determinants of Participation in TANF and Medicaid," JournalofPolicyAnalysisandManagement (forthcoming).
24. See, for example, the projects funded by the Robert Wood Johnson Foundation under its Covering Kids initiative, described at www.coveringkids.org (20 February 2004).
25. See Supplemental Exhibit 3, content.healthaffairs.org/cgi/content/full/23/3/233/DC3.
26. Ibid.
27. The level of uncertainty is related to the sample size for each state. For the percentages in Supplemental Exhibit 3, the 95 percent confidence intervals range from plus/minus seven percentage points for Wyoming to plus/minus 1.8 percentage points for California. For comparisons between states, differences of more than five percentage points are generally statistically significant. The 95 percent confidence interval for the U.S. estimate is plus/minus 0.6 percentage points.
28. See Supplemental Exhibit 3, content.healthaffairs.org/cgi/content/full/23/3/233/DC3.
29. Ibid.
30. M.Nathanson and L. Ku, Proposed State Medicaid Cuts Would Jeopardize Health Insurance Coverage for 1.7 Million People: An Update (Washington: Center on Budget and Policy Priorities, 31 March 2003).
31. C.Mann, Why Not Medicaid? Using Child Health Funds to Expand Coverage through the Medicaid Program (Washington: Center for Budget and Policy Priorities, 19 November 1997).
32. We thank an anonymous peer reviewer for suggesting this.
33. D.K.Remler, J.E. Rachlin, and S.A. Glied, "What Can the Take-Up of Other Programs Teach Us about How to Improve Take-Up of Health Insurance Programs?" NBER Working Paper no. 8185 (Cambridge, Mass.: National Bureau of Economic Research, March 2001).

                      What's this?

This article has been cited by other articles:

				J. E. DeVoe, C. Tillotson, and L. S. Wallace Uninsured Children and Adolescents With Insured Parents JAMA, October 22, 2008; 300(16): 1904 - 1913. [Abstract] [Full Text] [PDF]

				J. E. DeVoe, A. Baez, H. Angier, L. Krois, C. Edlund, and P. A. Carney Insurance + Access != Health Care: Typology of Barriers to Health Care Access for Low-Income Families Ann. Fam. Med, November 1, 2007; 5(6): 511 - 518. [Abstract] [Full Text] [PDF]

				M. Schlesinger and J. S. Hacker Secret Weapon: The "New" Medicare as a Route to Health Security Journal of Health Politics Policy and Law, April 1, 2007; 32(2): 247 - 291. [Abstract] [PDF]

				S. E. Rushing and M. Genel Our Porous Safety Net for Children Arch Pediatr Adolesc Med, November 1, 2005; 159(11): 1080 - 1081. [Full Text] [PDF]

				L. A. Blewett, M. Davern, and H. Rodin Covering Kids: Variation In Health Insurance Coverage Trends By State, 1996-2002 Health Aff., November 1, 2004; 23(6): 170 - 180. [Abstract] [Full Text] [PDF]