A Children's Health Insurance Factsheet from Families USA
January 1998

What is Crowd Out and Why Should Children's Health Advocates Care?

"Crowd out" refers to the substitution of public programs for private arrangements. It occurs to some extent in every public program. In the context of health insurance, it can occur if either employers or employees drop private insurance in favor of coverage under a public insurance program. Crowd out has become a concern in the expansion of public insurance programs to higher income families who are more likely to have access to private insurance options than do the poor. The new children's health insurance program, Title XXI of the Social Security Act, requires states to describe how new or expanded public health insurance for children will avoid substitution for private employer coverage .

This is an important issue for those who want to assure that public dollars target uninsured children. Unfortunately, some of the methods for avoiding crowd out exclude uninsured children from eligibility for health insurance under Title XXI. Advocates need to learn about this issue in order to make the case for avoiding crowd out without harming children.

WHAT DOES THE RESEARCH SHOW ABOUT THE EXTENT OF CROWD OUT?

Lower income families don't have employer-based coverage to lose.

Efforts to avoid crowd out should focus primarily on higher income families because lower income families are less likely to have private coverage to lose. Concerns about substitution arose when Medicaid eligibility levels were expanded to cover higher income populations with greater access to employer-based insurance than the traditional Medicaid population. Nationally, fewer than 12 percent of children in families with incomes below the poverty level have employer-based insurance, and only 36 percent of children in families with incomes between 100 and 150 percent of poverty have employer-based insurance. The percentage of children with employer based insurance rises with income. According to a recent study of state experiences with insurance expansion, insurance programs that target populations up to 400 percent of poverty generally devote more attention to crowd out than programs that cap eligibility at 150 to 200 percent of poverty. Eight-nine percent of children in families with income over 400 percent of poverty have employer-based insurance.

National studies disagree on the extent of the crowd out problem.

Employer-based insurance coverage is eroding for many reasons -- including rising health costs, changes in the labor market, and competitive pressures on employers -- that have nothing to do with Medicaid eligibility rules. At the same time, the Medicaid safety net for children has been expanding to cover children who have been disproportionately affected by the erosion of employer-based coverage. This makes it difficult for researchers to separate out the effect of Medicaid expansion from the underlying trend in employer coverage.

Several studies have tried to determine the extent to which substitution occurred when Medicaid expanded coverage for pregnant women and children up to 133 percent of poverty (and, in some states, up to 185 percent of poverty for pregnant women and infants). One study found that for every two people who enrolled in Medicaid, one person dropped private insurance. However, other researchers have criticized the methodology of this study and have come to different conclusions. Recent studies have shown that either there was no crowd out over the expansion period or that, at most, for every five people who enrolled in Medicaid, only one person dropped private coverage.

State studies have not found evidence of significant crowd out.

Officials in states that offered expanded coverage to children prior to the passage of Title XXI, report that crowd out has not been a significant problem. Available evidence from large scale children's health insurance programs in Florida and Minnesota suggest crowd out was not a significant problem in those states. The Florida Healthy Kids program subsidizes insurance premiums for school children in families with incomes at or under 185 percent of the federal poverty level. A recent study found that only two percent of enrollees had employer coverage any time during the year before enrolling in Florida Healthy Kids. Also, the analysis of reasons for disenrollment from Healthy Kids showed that when parents had an opportunity to select employer-based coverage, they did so. A study of MinnesotaCare, a program that subsidizes premiums for families under 275 percent of poverty, found no evidence that the program resulted in significant erosion of coverage in the private market. Further, only three percent of enrollees shifted from employer-based insurance to MinnesotaCare.

[Return to top] 

WHAT HAVE STATES DONE TO AVOID CROWD OUT THAT MAY HARM CHILDREN?

States have used three main techniques to try to discourage crowd out:
· Eligibility restrictions,
· Employer regulations and incentives, and
· Indirect effects of program design.

Eligibility restrictions can either target families least likely to have private coverage options, like lower income families, or deny coverage to families who have past or current insurance or access to insurance. Employer regulations can prohibit employers from dropping or changing coverage; incentives can subsidize the employer cost of expanding insurance. Finally, decisions about the benefit package, the family share of costs of coverage, and the ease of application may all affect the likelihood of substitution.

Restrictions on eligibility harm uninsured children.

In order to deter crowd out, Title XXI restricts eligibility to children who are not currently insured. This restriction applies both to separate state programs and Medicaid expansions qualifying for the Title XXI enhanced matching rate. (States can, of course, cover insured and underinsured children under Medicaid at the regular matching rate). In addition, states establishing a separate state program also have the option of imposing additional eligibility restrictions based on past insurance status or access to insurance.

Cures for crowd out may be worse than the disease. States may be considering denying coverage to children who are currently uninsured, but have been insured under group plans in the past. Thousands of uninsured children with no option to retain prior insurance status may be excluded by such policies. Further, even when employees do drop private coverage to enroll in a public program, the program is probably providing better coverage for children and needed financial relief for working poor families.

States may also be considering the exclusion of children who have access to group coverage or have had such access in the past even though the children have never actually been insured by the group plan. This is unfair to families if the available insurance benefits are not comprehensive, or the employee share of premium costs is high. One study showed that 30 percent of employees earning less than $14,000 annually in firms that offer insurance face annual contributions of $2400 or more to obtain family coverage. Eligibility rules that deny public coverage to employees with access to employer-based insurance also send the wrong message to employers by favoring those who don't offer insurance over those who do.

[Return to top] 

WHAT ARE THE STATES DOING TO AVOID CROWD OUT THAT HARMS THE LEAST NUMBER OF CHILDREN?

In light of the uncertainty about the extent of crowd out, what is the best approach for states to take? Title XXI requires states to exclude from eligibility children who are insured at the time of application. Generally, states should also ask about the prior insurance status of applicants, and be prepared to take appropriate corrective action if it appears that crowd out is occurring to any significant extent. If states determine that further measures are needed to prevent substitution, they should try to avoid exclusions of uninsured children with no other options for coverage.
· If a state does adopt eligibility restrictions, it should target them narrowly to those families voluntarily dropping employer-based insurance coverage because of the state program.
· States should also recognize that some features of a separate state program that they may plan to implement anyway, like a premium scale based on income, could also serve to deter crowd out.
· Finally, states can try to encourage employers to retain or expand group coverage for dependents.

States with experience running large scale children's health insurance programs are not imposing additional eligibility restrictions under Title XXI.

Table 1 shows how states filing Title XXI plans to date have addressed the issue. Those states with experience running large scale state-funded children's health insurance programs have proposed Title XXI plans that avoid crowd out without the use of additional eligibility restrictions. Based on its experience running Child Health Plus, New York's Title XXI plan does not anticipate crowd out to be a problem. New York proposes to deter crowd out through premium charges for families with incomes over 160 percent of poverty, and continuation of existing programs to assist small employers in providing coverage. (Exempting lower income families from premium costs is important to avoid a large drop off in participation of just those families least likely to have other insurance options). Florida and Pennsylvania filed Title XXI plans seeking federal matching funds for their currently state-funded children's insurance programs. Neither state has added any eligibility restrictions beyond the exclusion of currently insured children. A recent study of states that had expanded health insurance for children and families prior to the passage of Title XXI, found that only five of 16 programs had established a waiting period for children.

Table 1
Procedures identified in selected plans for avoiding substitution of coverage (§4.4.3 of Title XXI application)

Narrowly target eligibility restrictions.

If a state does require that a child be uninsured for some time prior to application, it is important that the waiting period not be unreasonably long and that the restrictions narrowly target children in families who have voluntarily dropped employer coverage. Generally, states should target only voluntary termination of employer-based coverage as triggering a waiting period, and recognize exceptions where a waiting period, even then, would be inequitable. Because lower income families are much less likely to have employer-based insurance, states may want to exempt families below certain income levels. Minnesota, for example, which covers families with incomes up to 275 percent of poverty, exempts children in families with incomes under 150 percent of poverty from its four-month waiting period for MinnesotaCare.

Of seven states that have so far either filed Title XXI plans or passed legislation to implement Title XXI, four have established eligibility criteria based on past insurance status. California authorizes exclusion of children who had employer-based insurance in the three months prior to application, and permits exclusion for up to six months if the state finds that a substantial share of program funds are going to families dropping employer coverage. Colorado excludes children who had coverage during the last three months under an employer plan if the employer paid at least 50 percent of the cost. Missouri will deny benefits if the parent dropped coverage within the last six months. Connecticut looks back six months for prior employer coverage, and authorizes a 12-month period if necessary to deter crowd out.

Missouri's plan does not address exemptions, but California, Colorado, and Connecticut all attempt to exempt children who had prior insurance but lost it for reasons unrelated to crowd out. See Table 2 for a list of the exemptions used in one or more of the California, Colorado and Connecticut laws.

Examples of reasons for insurance loss that will not preclude eligibility

• Loss of non-group insurance (i.e. individual insurance or Medicaid)
• Loss of group insurance for which employer paid less than 50 percent of premium cost
• Loss of employment for reasons other than voluntary termination
• Death of a parent
• Change to a new employer such that coverage is unavailable
• Change of address such that coverage is unavailable
• Discontinuance of benefits to all employees
• Expiration of a coverage period under COBRA
• Self-employment
• Termination of health benefits due to long-term disability
• Termination of dependent coverage due to economic hardship
• Substantial reduction in either lifetime medical benefits or a category of benefits

Avoid burdensome application procedures.

In order to avoid burdensome application procedures that can deter participation, states should try to keep anti-crowd out measures simple. A recent study of 16 states that had expanded health insurance found that most states determined insurance status by requesting a self-declaration from the applicant and did not attempt verification of insurance status.

State efforts to avoid crowd by imposing eligibility restrictions based on past insurance status or access to insurance are difficult and expensive for states to administer. A recent study estimated that even if crowd out occurred at the rate of 20 percent, it amounted to less than one percent of total Medicaid expenditures. This suggests that the administrative expense of identifying and verifying past insurance status may be greater than the expense of any crowd out that is otherwise likely to occur.

States can learn from the experience of those states that have already expanded insurance coverage for children of the working poor. For example, in Florida, the Healthy Kids Corporation originally restricted eligibility to children who were uninsured for the prior six months. However, Florida found verification of prior insurance status to be so cumbersome to administer and unfair to low-income families struggling to pay high premiums that they discontinued the requirement. Nonetheless, the Florida program has found little evidence of crowd out.

Give employers a reason to retain or expand insurance coverage

States can also address crowd out with laws regulating insurance or offering employers incentives to retain or expand coverage. California's new law amends its existing unfair labor practices act to prohibit an employer from changing the employee-employer share-of-cost ratio based upon the employee's wage base or job classification in order that the employee enroll in a Title XXI program. It also prohibits employers from making any modification of coverage for employees and their dependents in order that they enroll in the Title XXI program. Further, the California law prohibits an employer from encouraging employees to drop group coverage in favor of the Title XXI program. Wisconsin passed legislation requiring employers who offer insurance to some of their employees to offer insurance to all. Rhode Island prohibits an employer from discriminating in benefits against employees eligible for public health insurance or offering an incentive to only such employees to drop employer-based coverage. In drafting such laws, states must take care to avoid federal preemption. A federal law called ERISA governs employee benefit plans, and preempts state laws seeking to directly regulate employer plans. ERISA does not preempt states from regulating the sale of insurance, however state regulation will not reach the substantial number of self-insured group health plans.

States have also used purchasing cooperatives, tax credits and other incentives for employers to expand private coverage. New York's Title XXI plan refers to its Small Business Health Insurance Partnership Program, which offers financial assistance to small businesses and sole proprietors to purchase insurance, as one way it deters crowd out. Another approach is for the public program to subsidize the employee share of costs for group coverage. Provided that it is "cost effective," this can be done with a family coverage waiver under Title XXI, and is authorized as a state option under the Medicaid program. However, because children must be uninsured to qualify for Title XXI, employer buy-in programs do not reduce crowd out, but may make it easier for families to transition into private coverage.

[Return to top] 

Conclusion

Title XXI requires that state plans describe how new or expanded public insurance programs will avoid substitution for group coverage. However, the only thing states must do to satisfy this requirement is to limit benefits to children who are currently uninsured. States setting up separate state programs may choose to adopt additional measures to avoid crowd out. However, states should be cautious in adopting program design features that will restrict eligibility to needy uninsured children until more is known about crowd out and what to do about it.

[Return to top] 

Other resources on crowd out:

The Center for Studying Health System Change, Medicaid Eligibility Policy and the Crowding-Out Effect, Issue Brief No. 3, October 1996, Washington, DC. www.hschange.com

The Center on Budget and Policy Priorities, Strategies for Responding to Concerns about Crowd Out, October 17, 1997, Washington, D.C. www.cbpp.org.

Children's Defense Fund, Fears that Employer Coverage Will Fall if Uninsured Children Are Helped Are Exaggerated, October 14, 1997, Washington, D.C. www.childrensdefense.org

Chollet, D., Birnbaum, M. Sherman, M., Deterring Crowd-Out in Public Insurance Programs: StatePolicies and Experience, Alpha Center, October 1997, Washington, DC. www.ac.org

Merlis, M., Employer Coverage and the Children's Health Insurance Program under the Balanced Budget Act of 1997: Options for States, The Institute for Health Policy Solutions, August 27, 1997, Washington, D.C. 202-857-0810.

Also, The Lewin Group is preparing a study on crowd out for the Assistant Secretary of Planning and Evaluation at HHS.

1. §2102(b)(3)(C) of Title XXI of the Social Security Act, Pub. L. 105-33.
2. Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 1996 Current Po;ulation Survey, EBRI Issue Brief. November 1996.
3. Chollet, D., Birnbaum, M. Sherman, M., Deterring Crowd-Out in Public Insurance Programs: State Policies and Experience, (Alpha Center, October 1997).
4. U.S. General Accounting Office, Employment-Based Health Insurance: Costs Increase and Family Coverage Decreases, (GAO/HEHS-97-35, February 1997).
5. Cutler, David M., and Jonathan Gruber, "Medicaid and Private Insurance: Evidence and Implications." 16 Health Affairs (January/February 1997): 194-200.
6. Yazici, Esel, Medicaid Expansions and the Crowding Out of Private Health Insurance, paper presented at the 18th Annual Research Conference of the Association for Public Policy Analysis and Management, Pittsburgh, Pa., Nov. 2, 1996.
7. Dubay, Lisa C., and Genevieve M. Kenney, "The Effects of Medicaid Expansions on Insurance Coverage of Children." The Future of Children (Spring 1996): 152-160.
8. Chollet, D., op. cit.
9. Shenkman, Elizabeth et al., "The Florida Healthy Kids Program: Are There Indications of Crowd Out?" (Institute for Child Health Policy, September 1997).
10. Call, K.T, Lurie, N., et al., "Who Is Still Uninsured in Minnesota?" 278 JAMA 1191, October 8, 1997.
11. Curtis, R. and Page, A. "Improving Health Care Coverage for Low-Income Children and Pregnant Women: Public and Employer-Financed Coverage Relations," Table 5, (Institute for Health Policy Solutions, December 1996).
12. Dallek, G., Cost-Sharing and Low-Income People, Families USA, October 1997.
13. Chollet, D., op. cit.
14. Chollet, D. op. cit.
15. Holahan, J., "Crowding Out: How Big A Problem?" 16 Health Affairs 204 (January/February 1997).
16. Gauthier, A and Schroedel, S., State Initiatives in Health Care Reform: Expanding Children's Coverage: Lessons from State Initiatives in Health Care Reform (Alpha Center, May 1997) p.30.
17. Shenkman, op. cit.
18. Section 1906 of the Social Security Act as amended by §4741 of the Balanced Budget Act of 1997 (Medicaid); §2105(c)(3) of Title XXI of the Social Security Act. For a discussion of Medicaid employer buy-in programs, see Curtis op. cit.

[Return to top]