This is the accessible text file for GAO report number GAO-09-252 
entitled 'State Children's Health Insurance Program: CMS Should Improve 
Efforts to Assess whether SCHIP Is Substituting for Private Insurance' 
which was released on March 23, 2009. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to the Chairman, Committee on Finance, U.S. Senate: 

United States Government Accountability Office: 
GAO: 

February 2009: 

State Children's Health Insurance Program: 

CMS Should Improve Efforts to Assess whether SCHIP Is Substituting for 
Private Insurance: 

GAO-09-252: 

GAO Highlights: 

Highlights of GAO-09-252, a report to the Chairman, Committee on 
Finance, U.S. Senate. 

Why GAO Did This Study: 

Congress created the State Children’s Health Insurance Program (SCHIP) 
to reduce the number of uninsured children in low-income families that 
do not qualify for Medicaid. States have flexibility in structuring 
their SCHIP programs, and their income eligibility limits vary. 
Concerns have been raised that individuals might substitute SCHIP for 
private health insurance—known as crowd-out. GAO was asked to examine 
the Centers for Medicare & Medicaid Services’ (CMS) and states’ efforts 
to minimize crowd-out and determine whether it should be a concern. GAO 
examined (1) CMS’s guidance to states for minimizing crowd-out and 
assessment of whether it should be a concern and (2) states’ policies 
to minimize crowd-out and how they assess whether it should be a 
concern. To do the work, GAO reviewed federal laws and guidance, 
examined state annual reports, and interviewed CMS officials. GAO also 
interviewed SCHIP officials from nine states. 

What GAO Found: 

CMS provided guidance to states about activities to minimize crowd-out 
in SCHIP, and the information it collected was of limited use in 
assessing the extent to which crowd-out should be a concern. Along with 
this guidance, CMS instituted specific requirements for certain program 
designs it identified as being at greater risk of crowd-out, including 
programs with higher income eligibility thresholds. CMS said that among 
other sources, it used states’ SCHIP annual reports to assess the 
occurrence of crowd-out, and on this basis it believed that crowd-out 
was occurring. Yet each of the approaches CMS used was limited in 
providing information about the occurrence of crowd-out and thus the 
extent to which it should be a concern. CMS did not collect certain 
indicators of the potential for crowd-out in SCHIP annual reports, such 
as the extent to which private health insurance was available and 
affordable to families. States’ responses to CMS were inconsistent: 
GAO’s review of annual reports for 2007 found that less than half of 
the 50 states and the District of Columbia provided a percentage in 
response to CMS’s question on the percentage of applicants who dropped 
private health insurance to enroll in SCHIP. 

In general, states implemented similar types of policies in their 
activities to minimize crowd-out, but not all states collected 
information adequate to assess whether crowd-out should be a concern. 
The majority of states used policies such as waiting periods—a required 
period of uninsurance before an applicant can enroll in SCHIP—to try to 
reduce incentives for dropping private health insurance. All 39 states 
with waiting periods offered exemptions for involuntary loss of private 
health insurance. These exemptions were mostly related to whether 
insurance was available rather than affordable. Not all states 
collected information that was adequate to assess whether crowd-out 
should be a concern. For example, while all 50 states and the District 
of Columbia asked SCHIP applicants if they were currently insured, 24 
states asked applicants if they had access to private health insurance, 
which is important to understanding the potential for crowd-out. Of the 
9 states we interviewed, 5 states measured the occurrence of crowd-out, 
but they all used different methodologies to develop their estimates; 
the remaining 4 states did not measure crowd-out. None of the officials 
in the 9 states viewed crowd-out as a concern, with most basing this 
assessment on a variety of factors, including the lack of available and 
affordable private health insurance for the SCHIP population in their 
state. 

Overall, CMS concurred with the report’s findings and recommendation, 
but raised concerns regarding the difficulty of measuring crowd-out, 
particularly assessing the affordability of private coverage. While GAO 
agrees that measuring crowd-out is complicated, the actions GAO 
recommends are an essential first step to better assessing whether 
concerns about crowd-out are warranted. 

What GAO Recommends: 

To improve information on whether crowd-out should be a concern in 
SCHIP, GAO recommends that the Acting Administrator of CMS ensure that 
states (1) collect and report consistent information on the extent to 
which SCHIP applicants have private insurance available to them and (2) 
take appropriate steps to determine whether available private health 
insurance is affordable for SCHIP applicants. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-09-252]. For more 
information, contact James C. Cosgrove at (202) 512-7114 or 
cosgrovej@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

CMS Provided Guidance to States on Crowd-Out; Information It Collected 
Was of Limited Use in Assessing the Extent to Which Crowd-Out Should Be 
a Concern: 

States Used Similar Types of Policies to Minimize Crowd-Out, but Not 
All Collected Adequate Information to Assess whether Crowd-Out Should 
Be a Concern: 

Conclusions: 

Recommendation for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Comments from the Centers for Medicare & Medicaid Services: 

Appendix II: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: SCHIP Crowd-Out Estimates, by Study Type: 

Table 2: Examples of Program Designs for Which CMS Provided Specific 
Guidance to States: 

Table 3: States' Responses to Annual Report Question on the Incidence 
of Crowd-Out: 

Table 4: States' Exemptions to Waiting Periods and the Conditions They 
Addressed: 

Table 5: States' Efforts to Verify Insurance Status: 

Figures: 

Figure 1: Estimates of the Percentage of Employed Households That Are 
Offered Individual or Family Health Insurance through Their Employers, 
by Percentage of FPL: 

Figure 2: Estimated Average Annual Premiums Households Paid for 
Individual or Family Health Insurance, by FPL: 

Figure 3: Estimated Percentage of Household Income Spent on Individual 
or Family Health Insurance, by FPL: 

Figure 4: States' Policies to Minimize Crowd-Out: 

Figure 5: States' Waiting Period Lengths: 

Abbreviations: 

CMS: Centers for Medicare & Medicaid Services: 

COBRA: Consolidated Omnibus Budget Reconciliation Act of 1985: 

FPL: federal poverty level: 

MEPS: Medical Expenditure Panel Survey: 

SCHIP: State Children's Health Insurance Program: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

February 20, 2009: 

The Honorable Max Baucus: 
Chairman: 
Committee on Finance: 
United States Senate: 

Dear Mr. Chairman: 

Congress created the State Children's Health Insurance Program (SCHIP) 
in August 1997 with the goal of significantly reducing the number of 
low-income uninsured children. SCHIP offers health insurance for 
children in low-income families that do not qualify for Medicaid and 
are not enrolled under a group health plan or under other health 
insurance.[Footnote 1] The program enrolled over 7 million children at 
some point during fiscal year 2007. States that participate in SCHIP 
receive a larger share of federal funds--an enhanced federal matching 
rate--than what is offered under Medicaid.[Footnote 2] States have 
considerable flexibility in structuring their SCHIP programs and may 
use one of three basic options: (1) expanding Medicaid, (2) 
establishing a separate child health program, or (3) combining these 
two approaches. States also vary in the income eligibility limits that 
they have established for SCHIP. States' inclusion of higher income 
eligibility levels--some at or exceeding three times the federal 
poverty level (FPL)[Footnote 3]--has increased concerns about the 
possible substitution of public health insurance under SCHIP for 
private insurance--a phenomenon known as crowd-out.[Footnote 4] Under 
the SCHIP program, states must monitor crowd-out. In certain 
circumstances, states must also describe their policies to minimize 
crowd-out in their state plan for SCHIP, which must then be reviewed 
and approved by the Centers for Medicare & Medicaid Services (CMS), the 
agency responsible for administering SCHIP. 

Efforts to assess the potential for crowd-out vary in their 
methodological approaches, definitions of crowd-out, and estimates of 
the extent to which it occurs.[Footnote 5] For example, some studies 
measured crowd-out by including any decline in private health insurance 
within the population, while other studies did not count certain 
instances where insurance was not available to applicants--such as 
individuals who lost private health insurance because of job loss and 
took up SCHIP--in their measurement. Additionally, researchers varied 
in how they separated changes in private health insurance because of 
SCHIP from other confounding factors, such as changes in the economy. 
As a result, crowd-out estimates--the percentage of SCHIP enrollees who 
left private health insurance to enroll in SCHIP--have been as low as 0 
percent and as high as 60 percent nationally.[Footnote 6] 

Uncertainty surrounding the extent to which crowd-out occurs has led to 
debates about how to ensure that children have access to health 
insurance while making certain that SCHIP does not crowd-out private 
health insurance. SCHIP was reauthorized on February 4, 2009, and 
concerns about crowd-out remain.[Footnote 7] 

To help Congress understand crowd-out, you asked us to examine CMS's 
and states' efforts to minimize crowd-out and how they assess whether 
crowd-out should be a concern. 

In this report, we examine (1) CMS's guidance to states aimed at 
minimizing crowd-out and the agency's efforts to assess whether crowd- 
out should be a concern and (2) states' policies to minimize crowd-out 
in their SCHIP programs, including how states assess whether crowd-out 
should be a concern. 

To determine CMS's guidance to states and the agency's efforts to 
assess whether crowd-out should be a concern, we reviewed relevant 
SCHIP legislation and CMS guidance, interviewed CMS officials, and 
reviewed 2007 SCHIP annual reports from all 51 states.[Footnote 8] To 
determine states' policies to minimize crowd-out and their efforts to 
assess whether crowd-out should be a concern, we reviewed state 
documents from all 51 states, including SCHIP annual reports, state 
SCHIP plans, and SCHIP applications from 2007. We sent the information 
from our review to SCHIP officials from all 51 states for verification. 
[Footnote 9] We also conducted a literature review to determine the 
effect of activities to minimize crowd-out. To further understand CMS's 
guidance, states' activities to minimize crowd-out, and assessments of 
whether crowd-out should be a concern, we selected a sample of 9 states 
(California, Georgia, Missouri, New Jersey, New Mexico, New York, North 
Dakota, Oklahoma, and Oregon) and interviewed SCHIP officials from 
those states. To select the 9 states, we considered SCHIP program 
structure, SCHIP eligibility level, total SCHIP program enrollment, and 
geographic variation. Together, the states in our sample represented 
approximately 40 percent of all SCHIP enrollees in 2007. 

For our analysis, we considered an individual's access to private 
health insurance to consist of two main factors--availability and 
affordability. In this report, private insurance is said to be 
available to individuals if their employers offered health insurance 
and if these individuals and their families were eligible for this 
benefit.[Footnote 10] Affordability refers to the capacity of low- 
income families to purchase available private health insurance. 
[Footnote 11] For purposes of this report, we did not further define 
levels of availability or affordability.[Footnote 12] 

We defined crowd-out as the substitution of SCHIP for private health 
insurance. Crowd-out can occur when employers drop, decrease, or decide 
not to offer private health insurance to employees because SCHIP is 
available. Crowd-out can also occur when individuals either drop 
private health insurance or do not enroll in private health insurance 
because SCHIP is available. We did not include in our definition of 
crowd-out the substitution of SCHIP for Medicaid. We included 
information as reported by states but did not conduct interviews with 
employers or insurers to corroborate it. We conducted this performance 
audit from July 2008 through February 2009 in accordance with generally 
accepted government auditing standards. Those standards require that we 
plan and perform the audit to obtain sufficient, appropriate evidence 
to provide a reasonable basis for our findings and conclusions based on 
our audit objectives. We believe that the evidence obtained provides a 
reasonable basis for our findings and conclusions based on our audit 
objectives. 

Results in Brief: 

CMS provided guidance to states regarding activities to minimize crowd- 
out in SCHIP, and the information it collected was of limited use in 
assessing whether crowd-out should be a concern. In issuing guidance to 
states, CMS instituted specific requirements for program designs that 
the agency identified as being at greater risk of crowd-out, including 
programs with higher income eligibility thresholds. For example, the 
agency required separate child health programs at all eligibility 
thresholds to monitor crowd-out, and it required states with SCHIP 
income eligibility levels from 201 to 250 percent of the FPL to monitor 
crowd-out and take action if levels of crowd-out reached an agreed-upon 
threshold. CMS reported reviewing states' SCHIP annual reports, 
analyzing national trends in public and private health insurance, and 
commissioning studies on SCHIP, and on this basis believed crowd-out 
was occurring. However, each of the approaches CMS used provided 
limited information on the occurrence of crowd-out and thus on the 
extent to which it should be a concern. CMS did not collect certain 
indicators that could improve assessments of the potential for crowd- 
out in the SCHIP annual report, such as the extent to which private 
health insurance was available and affordable to SCHIP-eligible 
children. Moreover, states did not provide CMS with consistent 
responses in their annual reports. For example, our review of SCHIP 
annual reports for 2007 found that less than half of the 51 states 
provided a percentage in response to CMS's question on the percentage 
of applicants who dropped private health insurance to enroll in SCHIP. 

In general, states implemented similar types of policies in their 
activities to minimize crowd-out, but not all states collected 
information adequate to assess whether crowd-out should be a concern. 
The majority of states implemented policies, such as waiting periods-- 
a required period of uninsurance before an applicant can enroll in 
SCHIP--and premiums, with the aim of reducing incentives for 
individuals to drop private health insurance. All 39 states with 
waiting periods offered exemptions designed to account for instances 
where a child involuntarily lost private health insurance. These 
exemptions were mostly related to the availability of insurance rather 
than the affordability of it. Officials from our sample of 9 states 
generally believed that the policies they implemented were effective in 
minimizing crowd-out, but they provided little support for this 
assessment. Not all states collected information that was adequate to 
assess whether crowd-out should be a concern. For example, while all 51 
states asked SCHIP applicants if they were currently insured, 24 states 
asked applicants if they had access to private health insurance through 
their employers, which is important to understanding the potential for 
crowd-out. Within our sample of 9 states, 5 states measured the 
occurrence of crowd-out, but they developed their estimates 
differently; the remaining 4 states did not measure crowd-out. None of 
the officials from the sample states viewed crowd-out as a concern, 
with most basing this assessment on a variety of indicators, including 
the lack of available and affordable private health insurance for the 
SCHIP population in their state. 

To improve information on whether crowd-out should be a concern in 
SCHIP, we recommend that the Acting Administrator of CMS ensure that 
states (1) collect and report consistent information on the extent to 
which SCHIP applicants have private insurance available to them and (2) 
take appropriate steps to determine whether available private health 
insurance is affordable for SCHIP applicants. 

Overall, CMS concurred with the report's findings, conclusions, and 
recommendation; however, it identified several items of concern that it 
addressed in written comments. CMS stated that the issue of crowd-out 
is complicated and that trying to measure it is difficult because of 
variations in definition and methodological approaches. In particular, 
CMS raised concerns about collecting and assessing information to 
determine whether health insurance was affordable for those who had 
access to private coverage. We agree that measuring crowd-out-- 
particularly in terms of assessing affordability of coverage--is 
complicated. However, as noted in our report, some states already make 
such designations in their assessments of an individual's access to 
private health insurance. 

Background: 

SCHIP is a federal-state program that, in general, allows states to 
provide health insurance to children in families with incomes of up to 
200 percent of FPL, or 50 percentage points above states' Medicaid 
eligibility limits that were in place as of March 31, 1997.[Footnote 
13] States submit plans for their use of SCHIP funds to CMS. The agency 
reviews and approves these plans and monitors their implementation. 
Within broad federal guidelines, states have considerable flexibility 
in designing their SCHIP programs in terms of establishing eligibility 
guidelines, the scope of benefits, and administrative procedures. A 
number of states have made use of this flexibility by expanding SCHIP 
eligibility to children in families with income levels as high as 350 
percent of FPL.[Footnote 14] 

States' choices in structuring their SCHIP programs have important 
programmatic and financial implications. States have three basic 
options for structuring their SCHIP programs: (1) a Medicaid expansion 
program, (2) a separate child health program, or (3) a combination 
program that includes both a Medicaid expansion and a separate child 
health program. A Medicaid expansion program allows a state to expand 
eligibility levels within the state's existing Medicaid program and 
requires the state to follow Medicaid rules, including those on 
eligibility determination, benefits, and cost sharing.[Footnote 15] 
States that operate Medicaid expansions continue to receive federal 
funds at the regular Medicaid matching rate after they have exhausted 
their SCHIP funds.[Footnote 16] In contrast, a separate child health 
program may depart from Medicaid rules, including introducing limited 
cost sharing and creating waiting lists for enrollment. A separate 
child health program only receives a defined SCHIP allotment from the 
federal government, and the state can limit its own annual contribution 
and or enrollment once funds for SCHIP are exhausted. Even programs 
with the same structure do not always operate in the same way. For 
example, Medicaid expansion programs can operate under a section 1115 
demonstration waiver, which allows states to implement policies that do 
not follow traditional Medicaid rules.[Footnote 17] Similarly, separate 
child health programs can operate as "Medicaid look-alike" programs, 
generally following Medicaid rules but maintaining limited funding. 

Federal law requires states to implement policies to minimize the 
potential for crowd-out. The SCHIP statute defines a "targeted low- 
income child" as one from a family that does not qualify for Medicaid 
and is not covered under a group health plan or under other health 
insurance.[Footnote 18] To help minimize crowd-out, states are required 
to implement policies that would discourage a family from dropping 
other private health insurance, and states must coordinate eligibility 
screening with other health insurance programs, such as Medicaid. 
[Footnote 19] 

CMS has provided general guidance to the states regarding activities to 
minimize and monitor crowd-out. In 1998 and 2001, CMS outlined the 
types of activities states would need to implement to minimize and 
monitor crowd-out, and gave states flexibility in choosing specific 
activities. In August 2007, CMS issued additional guidance for states 
that wished to expand eligibility to children in families with 
effective income levels at or above 250 percent of FPL. 

Most research efforts describe crowd-out as the movement of individuals 
from private to public health insurance.[Footnote 20] However, among 
these research efforts there is no universally accepted method to 
measure the extent of crowd-out, and as a result, estimates vary 
widely. One reason for this variation is differences in study type. 
Broadly, researchers have used population-based, enrollee-based, and 
applicant-based studies to measure crowd-out. Population-based studies 
measure crowd-out by estimating any decline in private health insurance 
within a population. Enrollee-based studies estimate the number of 
SCHIP enrollees who had insurance within a specified time frame, 
accounting for specific losses of private health insurance because of 
job loss or other circumstances. Applicant-based studies use state 
application data to identify the number of applicants declined for 
having current or prior health insurance. (See table 1.) Instead of 
providing a measure of crowd-out, applicant-based studies estimate the 
amount of crowd-out averted because of a state's eligibility 
determination process. 

Table 1: SCHIP Crowd-Out Estimates, by Study Type: 

Characteristic: Area of focus; 
Population-based studies[A]: National examination of trends in public 
and private health insurance; 
Enrollee-based studies[B]: Multistate or state-specific examination of 
SCHIP recipients; 
Applicant-based studies[C]: State-specific review of applicant data. 

Characteristic: Crowd-out estimate; 
Population-based studies[A]: 0 to 60 percent; 
Enrollee-based studies[B]: 0.7 to 15 percent; 
Applicant-based studies[C]: 0 to 17 percent. 

Characteristic: Description of crowd-out estimate; 
Population-based studies[A]: Compares change in private health 
insurance relative to change in public insurance; 
Enrollee-based studies[B]: Estimates the percentage of children 
enrolled in SCHIP who had private health insurance within a specified 
time frame; 
Applicant-based studies[C]: Estimates the percentage of families that 
applied for SCHIP but were declined because of existing or prior 
private health insurance. 

Characteristic: Sources of data; 
Population-based studies[A]: Current Population Survey, Medical 
Expenditure Panel Survey, and Survey of Income and Program 
Participation[D]; 
Enrollee-based studies[B]: Telephone and mail surveys; 
Applicant-based studies[C]: State administrative data from 
applications. 

Source: GAO analysis of research, as of November 2008. 

Note: GAO's analysis is similar to that conducted by Mathematica Policy 
Research, Inc. See S.S. Limpa-Amara, A. Merrill, and M. Rosenbach. Last 
accessed on December 4, 2008, at: [hyperlink, http://www.mathematica-
mpr.com/publications/redirect_pubsdb.asp?strSite=pdfs/SCHIPsubstitution.
pdf]. 

[A] For an example of this type of study, see J. Gruber and K. Simon, 
"Crowd-Out Ten Years Later: Have Recent Public Insurance Expansions 
Crowded Out Private Health Insurance?" 

[B] For an example of this type of study, see L.P. Shone, et al., 
"Crowd-Out in the State Children's Health Insurance Program (SCHIP): 
Incidence, Enrollee Characteristics and Experiences, and Potential 
Impact on New York's SCHIP," Health Services Research, vol. 43, no. 1, 
Part II (February 2008). 

[C] For an example of this type of study, see S.S. Limpa-Amara, A. 
Merrill, and M. Rosenbach. 

[D] The population-based studies reviewed used data from a variety of 
surveys that differed according to sample size, study type (for 
example, longitudinal versus cross-sectional), population 
characteristics (for example, applying different income levels), and 
insurance information collected. 

[End of table] 

Assessments of the potential for crowd-out must take into account an 
understanding of the extent to which private health insurance is 
available and affordable to low-income families that qualify for SCHIP. 
With regard to the availability of private health insurance, the extent 
to which private firms offer health insurance to their employees varies 
by state. According to the 2006 Medical Expenditure Panel Survey 
(MEPS), 55.8 percent of private sector firms offered insurance--either 
individual or family health insurance--to their employees. Across the 
states, the percentage of private sector firms offering individual or 
family health insurance to their employees ranged from 89.6 percent of 
firms in Hawaii to 40.1 percent of firms in Montana.[Footnote 21] 
Smaller firms were less likely to offer individual or family health 
insurance to their employees than larger firms. For example, 35.1 
percent of firms with fewer than 10 employees offered individual or 
family health insurance, while 98.4 percent of firms with 1,000 or more 
employees offered individual or family health insurance to their 
employees. There is some evidence suggesting that the availability of 
private health insurance for families is declining. For example, a 
recent study found that from 2003 through 2007, the percentage of low- 
income employees offered individual insurance through their employers 
did not change, but the percentage of low-income employees offered 
family health insurance decreased from 71.1 percent to 63.6 percent. 
[Footnote 22] 

With regard to affordability, MEPS looks at the amount that individuals 
paid annually in order to obtain private health insurance.[Footnote 23] 
The national average annual premium--which includes both employee and 
employer contributions--for a single employee was $4,118 and for a 
family was $11,381. Premiums also showed some variation by state, 
ranging from $3,549 in Hawaii to $4,663 in Maine for an individual, 
while premiums for family health insurance ranged from $9,426 in Hawaii 
to $12,686 in New Hampshire. 

When viewed by household income level, households at higher income 
levels were more likely to be offered health insurance--either 
individual or family health insurance--through their employers. For 
example, employed households with incomes less than 300 percent of FPL 
were less likely to be offered health insurance than households earning 
more than 300 percent of FPL. (See figure 1.) 

Figure 1: Estimates of the Percentage of Employed Households That Are 
Offered Individual or Family Health Insurance through Their Employers, 
by Percentage of FPL: 

[Refer to PDF for image: horizontal bar graph] 

Percentage of FPL: Below 227; 
Percentage of households that are offered insurance: 58.3%. 

Percentage of FPL: Below 300; 
Percentage of households that are offered insurance: 64.5%. 

Percentage of FPL: Below 400; 
Percentage of households that are offered insurance: 69.3%. 

Percentage of FPL: 400 and above; 
Percentage of households that are offered insurance: 92.4%. 

Source: GAO analysis of the 2006 Medical Expenditure Panel Survey. 

Notes: Offers of health insurance include offers of both individual and 
family health insurance. The standard error rate for the estimates in 
this figure is within +/-3 percentage points. 

[End of figure] 

The 227 percent of FPL represents the national average income level for 
the SCHIP population in 2006, or about $45,400 for a household or 
family of four. This average was weighted by the FPL and enrollment 
levels of each state. 

In addition to employers offering health insurance, the costs of such 
insurance play an important role in the extent to which individuals 
accept such insurance. Based on an analysis of MEPS, lower-income 
households paid less for insurance premiums--either individual or 
family health insurance--than did households with higher incomes (see 
figure 2). However, the estimated premiums paid by lower-income 
households constituted a larger percentage of the total income for 
these households (see figure 3).[Footnote 24] 

Figure 2: Estimated Average Annual Premiums Households Paid for 
Individual or Family Health Insurance, by FPL: 

[Refer to PDF for image: vertical bar graph] 

Percentage of FPL: less than 227; 
Estimated average premium: $2,118. 

Percentage of FPL: 227-299; 
Estimated average premium: $2,671. 

Percentage of FPL: 300-399; 
Estimated average premium: $3,416. 

Percentage of FPL: 400 or more; 
Estimated average premium: $3,354. 

Source: GAO analysis of the 2006 Medical Expenditure Panel Survey. 

Notes: The estimated average annual premiums paid do not include other 
out-of-pocket costs that insured individuals could face, such as co- 
payments and deductibles. These percentage of income estimates include 
spending for both individual and family health insurance premiums. The 
standard error rate for the estimates in this figure is within +/-$608. 

The 227 percent of FPL represents the national average income level for 
the SCHIP population in 2006, or about $45,400 for a household or 
family of four. This average was weighted by the FPL and enrollment 
levels of each state. 

[End of figure] 

Figure 3: Estimated Percentage of Household Income Spent on Individual 
or Family Health Insurance, by FPL: 

[Refer to PDF for image: vertical bar graph] 

Percentage of FPL: less than 227; 
Percentage of income: 7.3%. 

Percentage of FPL: 227-299; 
Percentage of income: 5.2%. 

Percentage of FPL: 300-399; 
Percentage of income: 5.2%. 

Percentage of FPL: 400 or more; 
Percentage of income: 2.9%. 

Source: GAO analysis of the 2006 Medical Expenditure Panel Survey. 

Notes: The estimated average annual premiums paid do not include other 
out-of-pocket costs that insured individuals could face, such as co- 
payments and deductibles. These percentage of income estimates include 
spending for both individual and family health insurance premiums. The 
standard error rate for the estimates in this figure is within +/-1.1 
percentage points. 

The 227 percent of FPL represents the national average income level for 
the SCHIP population in 2006, or about $45,400 for a household or 
family of four. This average was weighted by the FPL and enrollment 
levels of each state. 

[End of figure] 

CMS Provided Guidance to States on Crowd-Out; Information It Collected 
Was of Limited Use in Assessing the Extent to Which Crowd-Out Should Be 
a Concern: 

CMS provided guidance to states regarding activities to minimize crowd- 
out in SCHIP, and the information it collected was of limited use in 
assessing the extent to which crowd-out should be a concern. In issuing 
guidance to states, CMS instituted specific requirements for program 
designs the agency identified as being at greater risk of crowd-out, 
including programs with higher income eligibility thresholds. CMS 
officials told us that they reviewed states' SCHIP annual reports, 
analyzed national trends in public and private health insurance, and 
commissioned studies on SCHIP, and on this basis believed crowd-out was 
occurring. However, each of the approaches CMS used was limited in 
providing information on the occurrence of crowd-out and the extent to 
which it should be a concern. In particular, CMS did not collect 
certain indicators in the SCHIP annual report--such as whether 
applicants' employers made private health insurance for families 
available and at what cost to the applicant--that could help show the 
potential for crowd-out. Moreover, information CMS did collect was not 
provided consistently by states. 

CMS Established Specific Requirements for Certain SCHIP Program 
Designs: 

CMS established specific requirements for SCHIP program designs it 
identified as being at a greater risk for crowd-out. In issuing the 
SCHIP final rule, CMS outlined broad regulatory requirements regarding 
substitution of private health insurance with SCHIP and stated that it 
planned to incorporate additional flexibility into its review of state 
plans in this area.[Footnote 25] CMS could not apply eligibility- 
related crowd-out prevention requirements to Medicaid expansion 
programs except those operating under an 1115 demonstration waiver. 
[Footnote 26] CMS did outline specific requirements for separate child 
health programs with higher income eligibility levels, explaining that 
there is a greater likelihood of crowd-out as incomes increase. For 
example, the agency required separate child health programs at all 
eligibility thresholds to monitor crowd-out, and it required states 
with SCHIP eligibility thresholds from 201 to 250 percent of FPL to 
implement policies to minimize crowd-out should an "unacceptable level" 
of crowd-out be detected. CMS officials told us that they did not 
define an "unacceptable level" for all states, but rather negotiated 
with states individually. CMS viewed waiting periods as a policy to 
minimize crowd-out; therefore, states with waiting periods were not 
required to monitor for an "unacceptable level" of crowd-out, and only 
three states currently do so.[Footnote 27] CMS also required that 
states institute a number of requirements for premium assistance 
programs, including a waiting period.[Footnote 28] (Table 2 provides 
examples of program designs for which CMS provided specific guidance.) 

Table 2: Examples of Program Designs for Which CMS Provided Specific 
Guidance to States: 

SCHIP design: Medicaid expansion programs; 
Examples of CMS guidance: 
* Medicaid expansion programs operating under 1115 demonstration 
waivers were subject to CMS review.[A] 

SCHIP design: Separate child health programs; 
Examples of CMS guidance: 
* Required separate child heath programs at all eligibility thresholds 
to monitor crowd-out; 
* Required states with SCHIP eligibility thresholds from 201 percent to 
250 percent of FPL to implement policies to minimize crowd-out should 
an "unacceptable level" of crowd-out be detected[B]; 
* Required programs above 250 percent of FPL to have policies in place 
to minimize crowd-out regardless of what any monitoring efforts showed; 
* Required states with premium assistance programs that subsidized 
private health insurance offered by employers to implement waiting 
periods of at least 6 months but no more than 12 months.[C] 

Source: GAO analysis of CMS guidance as of December 2008. 

[A] CMS did not apply crowd-out requirements to all Medicaid expansion 
programs. 

[B] CMS did not define an "unacceptable level" for all states, but 
rather negotiated with states individually. 

[C] Premium assistance programs use Medicaid funds, SCHIP funds, or 
both to subsidize the purchase of private health insurance. 

[End of table] 

In a letter issued on August 17, 2007, CMS provided specific guidance 
on crowd-out for states operating separate child health programs above 
250 percent of FPL, but implementation of the requirements in this 
guidance has been suspended.[Footnote 29] In this letter, CMS outlined 
specific activities states with separate child health programs should 
use to minimize crowd-out if they wish to cover children with effective 
family incomes above 250 percent of FPL. Among other things, these 
activities included a 12-month waiting period,[Footnote 30] a cost- 
sharing requirement,[Footnote 31] and efforts to prevent employers from 
changing health insurance offerings that would encourage a shift to 
public programs such as SCHIP. The letter caused states to raise a 
number of concerns, including those about the potential effect on 
current enrollees,[Footnote 32] and we have reported concerns regarding 
the issuance of this guidance.[Footnote 33] In a memorandum dated 
February 4, 2009, for the Secretary of Health and Human Services, the 
President directed that the August 17 letter be withdrawn immediately. 
[Footnote 34] 

Approaches CMS Used Did Not Address All Information Important to 
Assessing the Extent to Which Crowd-Out Should Be a Concern: 

CMS reported using a variety of approaches to assess the occurrence of 
crowd-out in SCHIP, including reviewing states' SCHIP annual reports, 
national estimates, and CMS-commissioned studies. From this 
information, CMS officials said they believed that crowd-out was 
occurring and that the potential for crowd-out was greater at higher 
income levels. However, the approaches CMS used provided limited 
information on the occurrence of crowd-out and thus the extent to which 
it should be a concern. 

The questions on crowd-out that CMS asked states to answer for their 
annual reports did not collect certain indicators of the potential for 
crowd-out, such as the extent to which SCHIP applicants were offered 
private health insurance for their families through their employers. In 
the annual reports that states must submit, CMS requires states to 
report the "incidence of substitution" in their state by calculating 
the percentage of applicants who drop private health insurance to 
enroll in SCHIP; however, CMS did not specify how states should 
calculate this incidence.[Footnote 35] While this question measures the 
number of SCHIP applicants who were enrolled in private health 
insurance and decided to take up SCHIP, it does not include SCHIP 
applicants who had private health insurance available through their 
employers but never enrolled. It also does not provide information on 
changes in insurance status, such as the extent to which SCHIP 
recipients gain access to private health insurance but remain enrolled 
in SCHIP. In addition, CMS does not specifically require states to 
provide information in their annual reports on the affordability of 
private health insurance available to SCHIP applicants.[Footnote 36] 

CMS has recently updated the requirements for the information that 
states must provide on crowd-out in their 2008 annual reports. For 
example, states will be expected to include more information on SCHIP 
applicants' insurance status. Whereas states have previously been 
required to report only the percentage of applicants who have other 
insurance, CMS plans to require states to provide the percentage of 
applicants who have Medicaid and the percentage of applicants who have 
other insurance.[Footnote 37] In addition, CMS plans to require that 
states with waiting periods report on the percentage of applicants who 
meet exemptions to the waiting period. While these questions are useful 
from an eligibility standpoint, they still do not account for 
applicants who were offered private health insurance for their families 
through their employers but did not take it up.[Footnote 38] These 
questions also do not address the extent to which available private 
health insurance is affordable to the SCHIP population. 

CMS's information on the occurrence of crowd-out is also limited 
because states did not consistently provide the information that CMS 
requested in their annual reports (see table 3). We reviewed SCHIP 
annual reports from 2007 for state responses on the percentage of 
applicants who dropped private health insurance to enroll in SCHIP and 
found that less than half of the 51 states provided a percentage in 
response to CMS's question.[Footnote 39] Of those states providing a 
percentage, 7 states answered the question directly, with 1 reporting 
no crowd-out and the remaining 6 reporting an incidence of crowd-out of 
1 percent or less.[Footnote 40] Two of the states reporting an 
incidence of less than 1 percent also reported this number for the 
percentage of applicants who had insurance at the time of application. 
Forty-one states provided a response related to what they knew about 
the insurance status of applicants,[Footnote 41] provided descriptive 
responses, or reported that data were not available. Three states did 
not answer this question in their annual reports. 

CMS officials told us that in cases where states had missing or 
incomplete responses to questions in the annual report, the agency 
allowed states to address these questions in the next year's 
report.[Footnote 42] CMS officials also noted that states used 
different data sources when analyzing crowd-out. For example, they said 
that some states used surveys to measure crowd-out, while others relied 
on data collected through application questions. 

Table 3: States' Responses to Annual Report Question on the Incidence 
of Crowd-Out: 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: Answered question 
directly; 
Number of states: 7. 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: Provided the percentage 
denied for having health insurance within a certain number of months 
before applying; 
Number of states: 8. 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: Provided the percentage 
who had health insurance within a certain number of months before 
applying; 
Number of states: 6. 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: Provided a descriptive 
response; 
Number of states: 8. 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: Responded that data were 
not available; 
Number of states: 19. 

Response to question asking for the percentage of applicants who drop 
private health insurance to enroll in SCHIP: No response; 
Number of states: 3. 

Source: GAO analysis of 51 states' annual reports for 2007. 

[End of table] 

In addition to reviewing state annual reports, CMS officials told us 
that they used national data to assess the occurrence of crowd-out. 
Specifically, they said that they used the Current Population Survey 
and CMS enrollment data to track changes in public and private health 
insurance over time. While this type of analysis can account for broad 
trends in changes in insurance usage, it cannot account for the reasons 
for these changes. For example, it cannot isolate whether an increase 
in public insurance resulted from crowd-out or an unrelated decline in 
the availability of private health insurance. 

CMS officials told us that they commissioned two studies to look at the 
issue of crowd-out in SCHIP. However, the results of these studies did 
not provide a clear sense of whether crowd-out should be a concern. A 
2003 study provided information on states' early experiences with SCHIP 
based on state annual reports from 2000.[Footnote 43] The study 
reported that states' estimates of crowd-out in SCHIP ranged from 0 to 
20 percent, but cautioned that state data must be interpreted carefully 
because of a number of factors, including the limited experience upon 
which states based their estimates. A 2007 study assessed the SCHIP 
program,[Footnote 44] and included a background paper on crowd-out. 
[Footnote 45] The background paper synthesized evidence on the 
occurrence of crowd-out in SCHIP and concluded that while the evidence 
suggested that crowd-out occurred, the magnitude of the occurrence 
ranged widely--from 0.7 to 56 percent--depending on how crowd-out was 
defined and measured. The study commented on the strengths and 
weaknesses of various ways to define and measure crowd-out, but it did 
not indicate a preference for any specific methodology. 

States Used Similar Types of Policies to Minimize Crowd-Out, but Not 
All Collected Adequate Information to Assess whether Crowd-Out Should 
Be a Concern: 

In general, states implemented similar types of policies in their 
activities to minimize crowd-out, but not all states collected 
information adequate to assess whether crowd-out should be a concern. 
States' policies largely sought to deter individuals from dropping 
private health insurance. Officials we interviewed in the nine sample 
states generally believed that their policies were effective in 
minimizing crowd-out. However, little evidence existed to confirm this 
belief. Less than half of states investigate whether applicants had 
access to private health insurance, which is key to understanding the 
extent to which crowd-out should be a concern. 

States Used Similar Types of Policies to Minimize Crowd-Out, but Little 
Is Known about Their Effect: 

Forty-seven states used one or more policies to minimize crowd-out. 
These policies deter individuals from dropping private health insurance 
in order to take up SCHIP. Our analysis found that waiting periods-- 
required periods of uninsurance before applicants can enroll in SCHIP-
-were the most common policy states had in place to minimize crowd-out; 
premiums and other types of cost sharing were also frequently used (see 
figure 4).[Footnote 46] As expected, these policies were more common in 
separate child health and combination programs than in Medicaid 
expansion programs operating with 1115 demonstration waivers.[Footnote 
47] A waiting period is meant to discourage individuals from dropping 
private health insurance. Premiums and other types of cost sharing are 
used in SCHIP to narrow the cost difference between SCHIP and private 
health insurance, thereby reducing the likelihood that lower out-of- 
pocket costs for SCHIP will attract individuals who already have other 
health insurance. Ten states reported having premium assistance 
programs, where states used SCHIP funds to pay premiums for private 
health insurance sponsored by employers. In contrast to CMS's view that 
premium assistance programs could raise concerns about crowd-out, 5 
states reported that such programs were a policy to prevent crowd-out. 
One state reported instituting a policy directly aimed at affecting 
employer behavior. This state made it an unfair labor practice for 
insurance companies or employers to encourage families to enroll in 
SCHIP, for example, through changing benefit offerings, when the 
families already have private health insurance. 

Figure 4: States' Policies to Minimize Crowd-Out: 

[Refer to PDF for image: vertical bar graph] 

Policy: Waiting period; 
Number of states: 39. 

Policy: Premium; 
Number of states: 35. 

Policy: Other cost sharing; 
Number of states: 30. 

Policy: Premium assistance program; 
Number of states: 10. 

Source: GAO analysis of state information as of July 31, 2008. 

Notes: The total number of states with policies to minimize crowd-out 
is 51; some states used more than one policy. Other cost sharing 
includes co-payments or deductibles, though only 3 states had 
deductibles. 

Four states did not respond to our request to verify information; 
therefore, we used information from states' 2007 annual reports. 

[End of figure] 

Thirty-nine states had waiting periods, and they varied in length. The 
most common waiting period length was 3 or 6 months (see figure 5). At 
least 2 of the 39 states reduced their waiting periods after concluding 
that crowd-out was not a significant problem and that the original 
waiting period length was unnecessarily long.[Footnote 48] 

Figure 5: States' Waiting Period Lengths: 

[Refer to PDF for image: vertical bar graph] 

Length of waiting period: 1 month; 
Number of states: 3. 

Length of waiting period: 2 months; 
Number of states: 1. 

Length of waiting period: 3 months; 
Number of states: 12. 

Length of waiting period: 4 months; 
Number of states: 3. 

Length of waiting period: 6 months; 
Number of states: 15. 

Length of waiting period: 12 months; 
Number of states: 2. 

Source: GAO analysis of state information as of July 31, 2008. 

Notes: This figure shows the waiting periods imposed by 36 states. 
While the total number of states with waiting periods is 39, 3 states 
used more than one waiting period. These 3 states implemented multiple 
waiting periods to be applied in specific situations. For example, one 
state applied different waiting periods to different levels of 
applicant income. The length of these multiple waiting periods ranged 
from 3 to 12 months and are not included in this figure. 

Four states did not respond to our request to verify information; 
therefore, we used information from states' 2007 annual reports. 

[End of figure] 

All 39 states with a waiting period included exemptions designed to 
account for instances where a child involuntarily lost private health 
insurance. These exemptions were mostly related to the availability 
rather than the affordability of insurance (see table 4). Exemptions 
varied among states, but the most common exemptions were for a change 
in job status that led to a loss of private health insurance, a change 
in family structure, and exhaustion of health insurance provided 
through the Consolidated Omnibus Budget Reconciliation Act of 1985 
(COBRA).[Footnote 49] CMS does not require states to consider issues of 
affordability in developing exemptions, and just over 10 states 
provided an exemption specifically for economic hardship, that is, if 
private health insurance makes up too much of a family's income. 

Table 4: States' Exemptions to Waiting Periods and the Conditions They 
Addressed: 

Condition: Availability of insurance; 
Exemptions: Change in job status (loss of job or reduction of hours 
resulting in loss of benefits); 
Number of states: 29. 

Condition: Availability of insurance; 
Exemptions: Loss of COBRA; 
Number of states: Condition: 20. 

Condition: Availability of insurance; 
Exemptions: Termination/reduction in group health insurance; 
Number of states: Condition: 26. 

Condition: Availability of insurance; 
Exemptions: Special health care needs or termination of health 
insurance for a disability; 
Number of states: Condition: 9. 

Condition: Availability of insurance; 
Exemptions: Exhaustion of prior benefits; 
Number of states: Condition: 7. 

Condition: Availability of insurance; 
Exemptions: Address change that is not covered by private health 
insurance sponsored by an employer; 
Number of states: 9. 

Condition: Affordability of insurance; 
Exemptions: Economic hardship (income threshold specified); 
Number of states: 11. 

Condition: Availability and affordability of insurance; 
Exemptions: Change in family circumstance (e.g., death or divorce); 
Number of states: 20. 

Condition: Availability and affordability of insurance; 
Exemptions: Self-employed/individual health insurance; 
Number of states: 7. 

Source: GAO analysis of state information as of July 31, 2008. 

Note: All 39 states that had waiting periods had exemptions to them. 
Additionally, states can have more than one exemption. 

[End of table] 

SCHIP officials in the nine sample states generally believed that their 
policies to minimize crowd-out were effective but did not cite any 
specific data to support their assessment. Officials from these nine 
states generally reported that waiting periods, outreach efforts to 
inform applicants about program requirements, and premiums and other 
cost-sharing mechanisms were successful policies to minimize crowd-out. 
Officials in four of the states also noted that these policies may have 
unintended consequences, such as limiting families' access to 
insurance. 

Among the nine states in our sample, seven states reported no immediate 
plans to change their policies to minimize crowd-out, while two states 
were considering reducing the requirements in place regarding waiting 
periods. Officials from one of the two states said that they are 
considering reducing the state's waiting period. Officials from the 
other state wanted to add an economic hardship exemption to the state's 
waiting period. 

Our analysis of previous research found few studies that estimated the 
effect of different policies on minimizing crowd-out, but those that 
did primarily focused on waiting periods and cost sharing.[Footnote 50] 
These studies concluded that waiting periods and cost sharing can have 
negative effects on individuals' participation in SCHIP.[Footnote 51] 
These studies also suggested that policies to minimize crowd-out may 
deter SCHIP enrollment by eligible uninsured children at a faster rate 
than they deter use by individuals who have private health insurance. 
[Footnote 52] However, little is known about how variation in policies, 
such as the length of a state's waiting period, affects efforts to 
minimize crowd-out. 

Not All States Collected Information Adequate to Assess Whether Crowd- 
Out Should Be a Concern: 

All 51 states monitored crowd-out by asking applicants questions about 
whether they had private health insurance, but fewer states asked 
whether applicants were offered health insurance through their 
employers--a key piece of information in understanding whether crowd- 
out should be a concern. 

All 51 states asked applicants if they were currently insured, and 44 
of the 51 states asked applicants whether they had been insured in the 
past. Twenty-four states asked applicants about their access to private 
insurance. Of these 24 states, 15 states asked applicants if they had 
access to any private health insurance and 9 states asked applicants if 
they had access to specific types of private health insurance (such as 
through state or school employment). The remaining 27 states did not 
ask questions to determine whether applicants had access to private 
health insurance through their employers. 

The majority of states made efforts to verify applicants' responses to 
their questions about insurance. Twenty-five states used a database to 
verify applicants' enrollment in private health insurance, 14 states 
verified applicants' responses with employers, and 7 states used both. 
States conducted verification of insurance most frequently at the time 
of application, although about half of the states verified insurance 
status at eligibility renewal or at other regular intervals. (See table 
5.) 

Table 5: States' Efforts to Verify Insurance Status: 

Effort: Employer verification only; 
At any point in time: 9; 
At application: 7; 
At eligibility renewal: 8; 
At regular intervals: 3. 

Effort: Database verification only; 
At any point in time: 20; 
At application: 18; 
At eligibility renewal: 13; 
At regular intervals: 17. 

Effort: Both employer and database verification; 
At any point in time: 9; 
At application: 7; 
At eligibility renewal: 5; 
At regular intervals: 5. 

Effort: No verification; 
At any point in time: 13; 
At application: 19; 
At eligibility renewal: 25; 
At regular intervals: 26. 

Source: GAO analysis of state information as of July 31, 2008. 

Note: States were considered to verify insurance status at any point in 
time if they verified insurance status at application, eligibility 
renewal, or other regular intervals. States may have verified insurance 
status at more than one point in time. 

[End of table] 

Twenty-nine states used databases at some point to verify individuals' 
enrollment in private health insurance; however, there were differences 
in the databases used. The database sources states used included (1) 
databases from private insurance companies, (2) databases run by third- 
party administrators, and (3) databases that were maintained by the 
states.[Footnote 53] The databases also varied in the scope of 
insurance information they provided. For example, some databases 
included information on the majority of the population with private 
health insurance in a state, while other databases were more limited, 
such as those for state employees.[Footnote 54] While all of the 
databases verify insurance status, they do not determine whether 
individuals have private health insurance available through their 
employers but did not make use of it for their families. 

Another way states often verified insurance status was by checking with 
individuals' employers--which could allow a state to determine whether 
the individuals had private insurance available to them. We identified 
18 states that used various approaches to verify insurance status with 
employers at some point. Generally, the states asked whether (1) the 
individual was insured, (2) the individual had access to employer 
insurance, and (3) the individual paid out-of-pocket expenses for 
private health insurance.[Footnote 55] Other differences included the 
number of individuals states verified insurance status for and how 
employer information was collected. For example, at least 2 states only 
verified insurance status with employers if they found evidence that an 
individual might have private health insurance available through an 
employer, such as a pay stub showing that wages were deducted for 
health insurance premiums. Another state reported that it verified the 
availability of private health insurance through (1) an annual survey 
sent to all employers in the state and (2) an individual questionnaire 
sent to the employers of specific applicants and enrollees. 

The nine states whose officials we interviewed varied in the extent to 
which they measured crowd-out, with five states providing some estimate 
of crowd-out and four states reporting that they could not estimate it 
at all. The five states that did measure crowd-out developed their 
estimates differently. Officials from one of the five states based 
their estimate on the number of applicants who did not meet the 
eligibility criteria for SCHIP because they were enrolled in other 
insurance. However, this estimate measures how successful the state was 
in avoiding the inappropriate use of SCHIP rather than crowd-out. 
Officials from the remaining four states based their estimates on the 
number of individuals who dropped insurance for reasons that--by their 
states' definitions--constituted crowd-out. For example, one state 
included in its estimate of crowd-out circumstances in which families 
had private health insurance available but dropped it because it was 
unaffordable, while another state explicitly excluded this reason from 
its estimate. 

None of the officials in our sample of nine states viewed crowd-out as 
a concern, with most basing this assessment on a variety of indicators, 
including the availability and affordability of private health 
insurance for the SCHIP population in their state. Six states told us 
that private health insurance was not readily available and three other 
states believed that where insurance was available it was not 
affordable. Another state referred to the small number of SCHIP 
enrollees at its highest eligibility level--from 250 to 300 percent of 
FPL--as indicating a low potential for crowd-out. 

Conclusions: 

Understanding the potential for crowd-out is a complex task that begins 
with understanding the extent to which private health insurance is 
available to children in families whose incomes make them eligible for 
SCHIP. SCHIP is designed to offer health insurance to eligible children 
who would otherwise be uninsured, not to replace private health 
insurance. For crowd-out to occur, private health insurance must be 
available to low-income families who qualify for SCHIP, and these 
families must find such insurance affordable. CMS and the states differ 
in their views on whether crowd-out is a concern for SCHIP, yet the 
information on which both base their views is limited in providing a 
basis for assessing the occurrence of crowd-out. 

CMS's guidance to and reporting requirements for states provide limited 
information on the extent to which crowd-out should be a concern. While 
precise measurements of crowd-out are difficult, certain indicators 
could help improve assessments of the extent to which crowd-out should-
-or should not--be a concern. In particular, asking SCHIP applicants 
who work whether they are offered private health insurance for their 
families through their employers could provide an initial assessment of 
the extent to which private health insurance is available to these 
individuals and thus better assess whether concerns about crowd-out are 
warranted. For low-income working families, the affordability of 
available private health insurance is also important in determining 
whether crowd-out should be considered a concern. However, CMS does not 
currently have this information, as states are not required to collect 
and report it. Information about the availability and affordability of 
private health insurance could help ensure the best use of SCHIP funds 
and help determine the future funding needs of this important program. 

Recommendation for Executive Action: 

To improve information on whether crowd-out should be a concern in 
SCHIP, we recommend that the Acting Administrator of CMS refine CMS 
policies and guidance to better collect consistent information on the 
extent to which applicants have access to available and affordable 
private health insurance for their children eligible for SCHIP. Such 
actions should include ensuring that states: 

* collect and report consistent information on the extent to which 
SCHIP applicants have private insurance available to them and: 

* take appropriate steps to determine whether available private health 
insurance is affordable for SCHIP applicants. 

Agency Comments and Our Evaluation: 

CMS reviewed a draft of this report and provided written comments, 
which are reprinted in appendix I. In addition to comments on our 
recommendation, CMS provided us with technical comments that we 
incorporated where appropriate. 

Overall, CMS concurred with the report's findings, conclusions, and 
recommendation. CMS commented that the issue of crowd-out is 
complicated and that assessing or measuring it is difficult because of 
variations in definition and methodologies for assessing its 
occurrence. CMS agreed that the availability and affordability of 
health insurance are relevant issues in reviewing crowd-out. They also 
agreed with our recommendation, stating that information on the 
availability and affordability of health insurance from states would be 
extremely helpful in evaluating the potential for crowd-out. 

CMS also listed a number of concerns in response to our recommendation 
about the collection and submission of additional information related 
to the availability and affordability of private health insurance. CMS 
stated that the reliability of any information states collect from 
applicants about the availability or affordability of private coverage 
would be suspect if it were self-reported. However, our work found that 
much of the information currently collected by states and submitted to 
CMS on applicants' insurance status, which is used for eligibility 
determinations, is also self-reported by applicants. Moreover, our work 
shows that the majority of states take steps to verify this self- 
reported information. Thus, the use of self-reported data is not 
unusual and collecting such data could help CMS make an initial 
assessment of the extent to which private health insurance is available 
and affordable. With these data, CMS could better assess whether 
concerns about crowd-out are warranted. 

CMS also noted that there is currently no national definition of 
affordability and that it does not believe it would be appropriate for 
the agency to develop one because of the diversity of states' economies 
and health insurance markets, as well as the flexibility that is a key 
tenet of SCHIP. Further, CMS commented that its existing definition of 
a targeted low-income child does not specify that children only receive 
SCHIP if private health insurance is not affordable. We agree that CMS 
should not be responsible for devising a national standard of what is 
considered affordable within a state. We believe that states are in the 
best position to make such a determination, and as noted in our report, 
some states already make such designations in their assessments of an 
individual's access to private health insurance. 

Finally, CMS commented that because of the administrative and reporting 
burdens that would be associated with additional data reporting, the 
agency would need to provide states with opportunities to discuss the 
changes needed to collect and report this information. We agree that 
such steps may be necessary. We also note that CMS recently updated 
questions in its SCHIP annual report template to help ensure that 
states are collecting and reporting necessary information related to 
crowd-out. Thus, the administrative efforts associated with additional 
information collection and reporting are well within the scope of CMS's 
responsibility. 

As agreed with your office, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies to the Acting 
Administrator of the Centers for Medicare & Medicaid Services, 
committees, and others. This report also will be available at no charge 
on the GAO Web site at [hyperlink, http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-7114 or cosgrovej@gao.gov. Contact points for 
our Offices of Congressional Relations and Public Affairs may be found 
on the last page of this report. GAO staff who made major contributions 
to this report are listed in appendix II. 

Sincerely yours, 

Signed by: 

James C. Cosgrove: 
Director, Health Care: 

[End of section] 

Appendix I: Comments from the Centers for Medicare & Medicaid Services: 

Department Of Health & Human Services: 
Office Of The Secretary
Assistant Secretary for Legislation: 
Washington, DC 20201: 

January 29, 2009: 

James C. Cosgrove: 
Director, Health Care: 
U.S. Government Accountability Office: 
441 G Street N.W. 
Washington, DC 20548: 

Dear Mr. Cosgrove: 

Enclosed are comments on the U.S. Government Accountability Office's 
(GAO) report entitled: "State Children's Health Insurance Program: CMS 
Should Improve Efforts to Assess Whether SCHIP Is Substituting for 
Private Insurance (GAO-09-252). 

The Department appreciates the opportunity to review this report before 
its publication. 

Sincerely, 

Signed by: 

Barbara Pisaro Clark: 
Acting Assistant Secretary for Legislation: 

Attachment: 

Department Of Health & Human Services:	
Centers for Medicare & Medicaid Services: 
Administrator: 
Washington, DC 20201: 

Date: January 29, 2009: 

T0: James Cosgrove: 
Director, Health Care: 
Government Accountability Office: 

From: [Signed by] Charlene Frizzera: 
Acting Administrator: 

Subject: Government Accountability Office (GAO) Draft Report: State 
Children's Health Insurance Program (SCHIP): CMS Should Improve Efforts 
to Assess Whether SCHIP Is Substituting for Private Insurance
(GAO-09-252): 

Thank you for the opportunity to review the GAO draft report entitled, 
"State Children's Health Insurance Program (SCHIP): CMS Should Improve 
Efforts to Assess Whether SCRIP Is Substituting for Private Insurance" 
(GAO-09-252). 

The GAO was asked to examine the Centers for Medicare and Medicaid 
Services' (CMS) and States' efforts to minimize crowd-out and determine 
whether it should be a concern. GAO examined: 1) CMS' guidance to 
States for minimizing crowd-out and assessment of whether it should be 
a concern: and 2) States' policies to minimize crowd-out and how they 
assess whether it should be a concern. GAO reviewed Federal laws and 
guidance, examined State annual reports, and interviewed CMS officials. 
GAO also interviewed SCHIP State officials from a select sample of nine 
States (California, Georgia, Missouri, North Dakota. New Jersey, New 
Mexico, New York, Oklahoma, and Oregon). 

GAO Recommendations: 

To improve information on whether crowd-out should be a concern in 
SCHIP, GAO recommends that the Acting Administrator of CMS ensure that 
States: 1) collect and report consistent information on the extent to 
which SCRIP applicants have private insurance available to them; and 2) 
take appropriate steps to determine whether available private insurance 
is affordable for SCHIP applicants. 

CMS Response: 

We concur with comment on these recommendations. The issue of crowd-out 
is complicated and trying to assess or measure it is difficult because 
of variations in both its definition as well as in the methodologies 
for assessing its occurrence. We agree that individual decisions to 
forego enrollment when private coverage is available, as well as the 
affordability of coverage, are relevant issues in the overall review of 
whether public coverage is substituting for private, and having States 
provide this information, as recommended by the GAO, would be extremely 
helpful in evaluating the potential for substitution of coverage. 

However, there are limitations in the conclusions that could be drawn 
since reliable data collection methods have not been developed to 
determine whether families had access to private coverage that they 
have declined, either before enrolling their children in SCHIP or while 
the child is enrolled. It would also be difficult to ascertain and make
conclusions in each instance, when someone had access to private 
coverage and declined it, and whether that would be attributable to 
crowd-out. 

We have concerns with requiring the additional collection and 
submission of information related to access and affordability. While 
States may ask applicants if they have access to private coverage, or 
if that coverage is affordable, the reliability of this data would be 
suspect if it were self-reported. The only way to ensure that it is 
reliable would be to contact the applicant's employer. Additionally, 
there is no national definition of "affordability." The existing 
definition of a targeted low-income child does not specify that 
children can only receive SCHIP if the coverage is not "affordable" in 
the private sector. We do not believe it would be appropriate for CMS 
to develop a national affordability standard due to the diverse 
characteristics of the economic and health insurance markets across 
States, as well as the flexibility that is a key tenet of SCHIP. 

In light of the new administrative and reporting burdens that this data 
reporting would place on States, CMS would need to provide States with 
opportunities to discuss the feasibility of defining and measuring 
access to employer-sponsored insurance and affordability of insurance. 
Discussions and planning sessions would also be required on the 
feasibility of modifying applications, administrative procedures, and 
data systems to capture and report this information. We emphasize that 
issuance of any type of new reporting standards would need to comply 
with all applicable procedural requirements of the Administrative 
Procedure Act and Paperwork Reduction Act, including providing 
stakeholders with the opportunity for public comment. 

We have also provided technical comments in an attachment that we hope 
will be helpful in refining your draft. Please contact Ms. Kathleen 
Farrell at kathleen.farrell@cms.hhs.gov if we can assist in any other 
way. CMS appreciates the opportunity to comment on this draft report 
and we look forward to working with the GAO on this and other issues. 

[End of section] 

Appendix II: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

James C. Cosgrove, (202) 512-7114 or cosgrovej@gao.gov: 

Acknowledgments: 

In addition to the contact named above, Carolyn Yocom, Assistant 
Director; Kristin Bradley; William Crafton; Kevin Milne; Elizabeth T. 
Morrison; Rachel Moskowitz; and Samantha Poppe made key contributions 
to this report. 

[End of section] 

Footnotes: 

[1] SCHIP defines low-income in terms of the federal poverty level 
(FPL), which is a measure used to establish eligibility for certain 
federal assistance programs, including SCHIP. SCHIP defines low-income 
children as those in families earning up to 200 percent of FPL, or 50 
percentage points above a state's existing Medicaid eligibility 
standard as of March 31, 1997. Therefore, states vary in their 
definition of low-income families. For example, Delaware's SCHIP 
program covers children up to 200 percent of FPL, while Washington's 
covers children up to 250 percent of FPL. 

[2] Medicaid program expenditures are shared between states and the 
federal government with each state's share determined by a formula that 
compares a state's per capita income to the national average. Federal 
matching rates for SCHIP are "enhanced"--they are established under a 
formula that takes 70 percent of a state's Medicaid matching rate and 
adds 30 percentage points, with an overall federal share that may not 
exceed 85 percent. Thus, a state with a 50 percent Medicaid match 
receives a 65 percent match under SCHIP. 

[3] The FPL is updated annually to reflect changes in the cost of 
living and varies according to family size. For example, 200 percent of 
FPL for a family of four was $42,400 in 2008. 

[4] In general, the issue of substitution of SCHIP for private health 
insurance arises primarily with regard to group health plans, which are 
employer-sponsored plans that provide health care benefits to 
employees. See 42 C.F.R. § 457.10. We use the term private health 
insurance to denote health care benefits provided by group health 
plans. 

[5] See, for example, J. Gruber and K. Simon, "Crowd Out Ten Years 
Later: Have Recent Public Insurance Expansions Crowded Out Private 
Health Insurance?" National Bureau of Economic Research Working Paper 
12858, January 2007; and T. T. Alteras, Understanding the Dynamics of 
"Crowd-out": Defining Public/Private Coverage Substitution for Policy 
and Research Working Paper 12858, (Washington, D.C.: Academy for Health 
Services Research and Health Policy, June 2001). 

[6] Precise crowd-out estimates vary according to their definition and 
chosen parameters. For example, some studies estimate crowd-out as the 
change in the number of individuals with private health insurance 
relative to the change in those publicly insured. A crowd-out estimate 
of 30 percent, for instance, would mean that 30 out of 100 children 
enrolling in SCHIP dropped private health insurance to enroll in the 
program. For a summary of studies on SCHIP crowd-out, see S.S. Limpa- 
Amara, A. Merrill, and M. Rosenbach, SCHIP at 10: A Synthesis of the 
Evidence on Substitution of SCHIP for Other Coverage, Final Report 
(Cambridge, Mass.: Mathematica Policy Research, Inc., September 2007). 

[7] See Children's Health Insurance Program Reauthorization Act of 
2009, Pub. L. No. 111-3, _Stat. _. In general, the amendments made by 
this Act are effective April 2009. 

[8] In this report, we use the term state to refer to the 50 states and 
the District of Columbia. 

[9] We received complete verification from 46 of 51 states; 3 states 
did not respond and 2 states provided incomplete responses. We 
contacted states by phone and sent information by e-mail about GAO's 
review and how to verify information. We used information identified 
through our review of state documents if a state did not respond or 
provided an incomplete response. 

[10] See Social Security Act § 2102(b)(3); 42 C.F.R. §§ 457.800- 
.810(2008). 

[11] If available insurance is not affordable, families may decline 
such insurance regardless of SCHIP, and by definition, crowd-out would 
not occur. 

[12] For example, other considerations that could apply to assessments 
of crowd-out include analyzing the value or type of insurance offered, 
for example, whether available insurance is creditable. 

[13] See Pub. L. No. 105-33, §§ 4901, et. seq., 111 Stat. 251, 552, 
codified at 42 U.S.C. § 1397aa(a), 1397jj. 

[14] As of July 31, 2008, 21 states reported providing SCHIP for 
children in families with incomes above 200 percent of FPL. The state 
with the highest FPL is New Jersey, which provides SCHIP for children 
in families earning up to 350 percent of FPL, or $74,200 annually for a 
family of four. 

[15] Federal regulations do not allow SCHIP programs operating as 
Medicaid expansion programs to apply eligibility-related substitution 
prevention provisions, because such eligibility conditions are 
inconsistent with the Medicaid statute. Examples include provisions 
related to establishing waiting periods of uninsurance for individuals 
who are enrolled in private health insurance. 

[16] A Medicaid expansion creates an entitlement by requiring the 
states to continue providing services to eligible children even when 
states exceed their SCHIP allotments. At that point, states with 
Medicaid expansion programs continue to receive federal matching funds-
-but at the Medicaid matching rate rather than the enhanced match that 
SCHIP provides. 

[17] Section 1115 of the Social Security Act allows waivers to many of 
the statutory requirements of Medicaid or SCHIP in the case of 
experimental, pilot, or demonstration projects that promote program 
objectives. For example, SCHIP programs operating under a section 1115 
waiver may have eligibility-related substitution prevention provisions. 

[18] Pub. L. No. 105-33, § 4901, 111 Stat. at 569. 

[19] See 42 U.S.C. §§ 2105(c)(3), 2102(b)(3). 

[20] L.A. Blewett and K.T. Call, "Revisiting Crowd-Out," Robert Wood 
Johnson Foundation Synthesis Project, September 2007, 1; A. Sommers, et 
al., "Substitution of SCHIP For Private Coverage: Results from a 2002 
Evaluation in Ten States," Health Affairs, vol. 26, no. 2 (2007), 529; 
S.S. Limpa-Amara, A. Merrill, and M. Rosenbach, SCHIP at 10: A 
Synthesis of the Evidence on Substitution of SCHIP for Other Coverage, 
Final Report, and M. Rosenbach, et al., National Evaluation of the 
State Children's Health Insurance Program: A Decade of Expanding 
Coverage and Improving Access, Final Report (Washington, D.C.: 
Mathematica Policy Research, Inc., September 2007), 34. 

[21] MEPS is a set of large-scale surveys of families and individuals, 
their medical providers, and employers across the United States. MEPS 
collects data on the specific health services that Americans use, how 
frequently they use them, the cost of these services, and how they are 
paid for, as well as data on the cost, scope, and breadth of health 
insurance held by and available to U.S. workers. 

[22] P. Cunningham, S. Artiga, and K. Schwartz, The Fraying Link 
Between Work and Health Insurance: Trends in Employer-Sponsored 
Insurance for Employees, 2000-2007, (Washington, D.C.: Henry J. Kaiser 
Family Foundation, Kaiser Commission on Medicaid and the Uninsured, 
November 2008). 

[23] Considerations of the affordability of private insurance could 
also be based on the share of income one spends on health insurance, as 
well as on such factors as income level and cost of living. 

[24] The estimated premiums paid do not include other out-of-pocket 
costs that insured individuals could face, such as co-payments and 
deductibles. Additionally, the extent to which private health insurance 
policies were equivalent in terms of benefits, out-of-pocket costs, or 
other features is not known. 

[25] See 66 Fed. Reg. 2601-04. For example, CMS officials told us that 
they decided against requiring all states to implement a waiting 
period, recognizing that specific circumstances could affect which 
activities related to minimizing crowd-out were most appropriate for a 
given state. Instead, CMS suggested waiting periods as an example of an 
activity states could use to minimize crowd-out and invited states to 
propose other options. 

[26] Federal law does not allow SCHIP programs operating as Medicaid 
expansion programs to apply eligibility-related substitution prevention 
provisions, because such eligibility conditions are inconsistent with 
the Medicaid statute. See 66 Fed. Reg. 2605. However, section 1115 of 
the Social Security Act allows waivers of certain statutory 
requirements of Medicaid or SCHIP for demonstration projects that 
promote program objectives, including allowing crowd-out provisions 
where appropriate. 

[27] Two of these three states will implement a waiting period if they 
reach "unacceptable levels" of 8 percent and 15 percent, respectively, 
while the third state plans to extend the length of its waiting period 
should an unacceptable level be attained. CMS officials said that none 
of these states has reached its specified threshold. 

[28] Premium assistance programs use Medicaid funds, SCHIP funds, or 
both to subsidize the purchase of private health insurance. CMS's 
stated concern for instituting special provisions for these programs 
was that families currently enrolled in private insurance may be more 
likely to seek public funding if they could use it to pay for the cost 
of that insurance. 

[29] CMS also indicated that this guidance would apply to Medicaid 
expansion programs operating under 1115 demonstration waivers. 

[30] In a clarifying letter on May 7, 2008, CMS noted that while a 12- 
month waiting period was the standard against which all states would be 
evaluated, it would consider alternative proposals from states and the 
justifications for them. 

[31] The letter said that states were expected to impose cost sharing 
that could not be below the cost sharing required by competing private 
health insurance plans by more than 1 percent of a family's income 
unless such cost sharing would exceed 5 percent of a family's income, 
which is the maximum allowed in the SCHIP statute. 

[32] In its May 7, 2008 letter, CMS stated that any changes made in 
response to the August 17, 2007, letter need not be applied to those 
enrolled in the program prior to the effective date of the letter, as 
long as they remain continuously enrolled in the program. However, 
because of issues of retention in the SCHIP program, concerns remain 
that prior enrollees who leave the SCHIP program will be subject to the 
August 17 requirements if they try to reenroll. 

[33] In particular, in response to a request from Senators Rockefeller 
and Snowe for an opinion on whether the agency complied with the 
Congressional Review Act in issuing the August 17, 2007 letter, we 
determined that CMS should have submitted this letter to Congress and 
the Comptroller General, as required by the Congressional Review Act, 
Pub. L. No. 104-121 § 251, 110 Stat. 847, 868-74, codified at 5 U.S.C. 
§§801 - 808, before the letter could take effect. See GAO, 
Applicability of the Congressional Review Act to Letter on State 
Children's Health Insurance Program, B-316048 (Washington, D.C.: Apr. 
17, 2008). 

[34] See 74 Fed. Reg. 6347 (Feb. 6, 2009). 

[35] CMS has recently updated the annual report that states must 
complete for 2008 to include the specific calculation states should 
make to arrive at this percentage: ((# applicants who drop coverage/ 
total # applicants) * 100). However, it remains unclear how states 
should count the number of applicants who drop coverage--whether states 
should include all applicants who dropped private insurance coverage 
regardless of whether they were granted SCHIP eligibility, or if they 
should count only those applicants who dropped private insurance and 
then became eligible for SCHIP after having served a waiting period, 
having obtained an exemption, or being otherwise deemed eligible for 
SCHIP. 

[36] CMS officials did note that they ask open-ended questions in the 
annual report that could lead to information on the affordability of 
private health insurance. For example, they told us that the annual 
report asks states to explain factors that may account for increases or 
decreases in the number of uninsured children in the state. 

[37] CMS does allow states to provide a combined percentage if they are 
unable calculate separate percentages. 

[38] CMS guidance and regulations do not address explicitly whether it 
has authority to consider this population as an element of crowd-out. 

[39] In 2007, 14 of the 51 states had Medicaid expansion programs, 
which are only subject to crowd-out requirements if the state requests 
this authority under an 1115 demonstration waiver. As a result, states 
with Medicaid expansion programs may not have collected the same data 
as states subject to crowd-out requirements. 

[40] Of these seven states, just three states gave an indication of how 
they arrived at this percentage, with one state noting that it used 
data from a 2005 survey to answer the question. 

[41] In particular, eight states provided the percentage of applicants 
denied SCHIP for having health insurance, while six states just 
reported the percentage of applicants who were currently enrolled in 
health insurance. This distinction is important because the percentage 
denied SCHIP likely excludes those who had other insurance but were 
allowed to enroll in SCHIP because of state-allowed exemptions, such as 
job loss. 

[42] CMS officials explained that resubmitting a response requires a 
lot of administrative work on the part of states. Instead, CMS will 
talk with states about why an answer is incomplete, offer assistance in 
collecting data, and ask states to include a complete answer in the 
annual report for the following year. CMS officials told us that they 
are in the process of developing a follow-up letter to send to states 
within 60 days of submitting their 2008 annual reports. This letter is 
intended to provide feedback and technical assistance to states and 
will target specific areas of the annual reports, including the section 
on crowd-out. 

[43] M. Rosenbach, et al., Implementation of the State Children's 
Health Insurance Program: Synthesis of State Evaluations (Cambridge, 
Mass.: Mathematica Policy Research, Inc., March 2003). 

[44] M. Rosenbach, et al., National Evaluation of the State Children's 
Health Insurance Program: A Decade of Expanding Coverage and Improving 
Access (Cambridge, Mass.: Mathematica Policy Research, Inc., September 
2007). 

[45] S.S. Limpa-Amara, A. Merrill, and M. Rosenbach, SCHIP at 10: A 
Synthesis of the Evidence on Substitution of SCHIP for Other Coverage. 

[46] Premiums are payments required of enrollees for health insurance 
for a given period of time. Other types of cost sharing include 
coinsurance and deductibles. 

[47] While combination programs include both a Medicaid expansion and a 
separate child heath component, our analysis of these programs showed a 
similar trend. Waiting periods, premiums, and other cost sharing were 
more prevalent in the separate child health component than in the 
Medicaid expansion component. In general, Medicaid expansions are 
expected to follow Medicaid rules and thus generally do not impose 
waiting periods and are limited in the amount of premiums and cost 
sharing that they impose. 

[48] These two states were New Jersey, which lowered its waiting period 
from 12 months to 3 months, and California, which lowered its waiting 
period from 6 months to 3 months. 

[49] COBRA, in general, requires employers that offer group health 
plans and have 20 or more employees to make health insurance available 
for a limited period of time for employees and their dependents who 
have lost health insurance because of certain events, including the 
loss of employment or the loss of eligibility as a dependent on a 
parent's plan. 

[50] An analysis of 23 articles found 5 studies that estimated the 
effectiveness of waiting periods--C. Bansak and S. Raphael, "The 
Effects of State Policy Design Features on Take-Up and Crowd-Out Rates 
for the State Children's Health Insurance Program," Journal of Policy 
Analysis and Management, vol. 26, no. 1 (2006); K. Kronebusch and B. 
Elbel, "Simplifying Children's Medicaid and SCHIP," Health Affairs, 
vol. 23, no. 3 (2004); J. Gruber and K. Simon, "Crowd-Out Ten Years 
Later: Have Recent Public Insurance Expansions Crowded Out Private 
Health Insurance?" A.T. Lo Sasso and T.C. Buchmueller, "The Effect of 
the State Children's Health Insurance Program on Health Insurance 
Coverage," Journal of Health Economics, vol. 23 (2004); and B. Wolfe 
and S. Scrivner, "The Devil May Be in the Details: How the 
Characteristics of SCHIP Programs Affect Take-Up," Journal of Policy 
Analysis and Management, vol. 24, no. 3 (2005)--and three articles--J. 
Gruber and K. Simon, "Crowd-Out Ten Years Later: Have Recent Public 
Insurance Expansions Crowded Out Private Health Insurance?"; J. Hadley, 
et al., "Insurance Premiums and Insurance Coverage of Near-Poor 
Children," Inquiry, vol. 43, no. 4 (Winter 2006/2007); B. Wolfe and S. 
Scrivner, "The Devil May Be in the Details: How the Characteristics of 
SCHIP Programs Affect Take-Up"--that estimated the effectiveness of 
cost sharing. 

[51] Four of the five studies that estimated the effectiveness of 
waiting periods and the three studies that estimated the effectiveness 
of cost sharing reported that these policies have a negative and 
significant impact on SCHIP enrollment. See B. Wolfe and S. Scrivner, 
"The Devil May Be in the Details: How the Characteristics of SCHIP 
Programs Affect Take-Up"; A.T. Lo Sasso and T.C. Buchmueller, "The 
Effect of the State Children's Health Insurance Program on Health 
Insurance Coverage"; J. Gruber and K. Simon, "Crowd-Out Ten Years 
Later: Have Recent Public Insurance Expansions Crowded Out Private 
Health Insurance?"; J. Hadley, et al., "Insurance Premiums and 
Insurance Coverage of Near-Poor Children"; C. Bansak and S. Raphael, 
"The Effects of State Policy Design Features on Take-Up and Crowd-Out 
Rates for the State Children's Health Insurance Program"; and 
Kronebusch and B. Elbel, "Simplifying Children's Medicaid and SCHIP." 

[52] See L.A. Blewett and K.T. Call, "Revisiting Crowd-Out"; A.T. Lo 
Sasso and T.C. Buchmueller, "The Effect of the State Children's Health 
Insurance Program on Health Insurance Coverage"; B. Wolfe and S. 
Scrivner, "The Devil May Be in the Details: How the Characteristics of 
SCHIP Programs Affect Take-Up"; and J. Hadley, et al., "Insurance 
Premiums and Insurance Coverage of Near-Poor Children." 

[53] Some states have contracted with a third-party administrator, an 
outside organization, to perform the administrative task of identifying 
SCHIP applicants and enrollees with private health insurance. For 
example, a number of states reported contracting with an organization, 
Health Management Systems, that compares SCHIP applicant and enrollee 
information to private health insurance plan enrollment files to help 
identify individuals. 

[54] One official from our sample of nine states indicated that third- 
party databases were not always up to date. For example, this official 
noted that while the third-party administrator may update its database 
quarterly, participating insurance companies may only provide updated 
information annually. 

[55] Over half of the 18 states asked whether the individual was 
insured; states asked the remaining two questions less frequently. 

[End of section] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "E-mail Updates." 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s Web site, 
[hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional 
information. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: