Manajemen | Fakultas Ekonomi Universitas Maritim Raja Ali Haji 073500104000000208
Journal of Business & Economic Statistics
ISSN: 0735-0015 (Print) 1537-2707 (Online) Journal homepage: http://www.tandfonline.com/loi/ubes20
The Measurement of Medicaid Coverage in the
SIPP
David Card, Andrew K.G Hildreth & Lara D Shore-Sheppard
To cite this article: David Card, Andrew K.G Hildreth & Lara D Shore-Sheppard (2004) The
Measurement of Medicaid Coverage in the SIPP, Journal of Business & Economic Statistics,
22:4, 410-420, DOI: 10.1198/073500104000000208
To link to this article: http://dx.doi.org/10.1198/073500104000000208
Published online: 01 Jan 2012.
Submit your article to this journal
Article views: 58
View related articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ubes20
Download by: [Universitas Maritim Raja Ali Haji]
Date: 13 January 2016, At: 00:33
The Measurement of Medicaid Coverage in
the SIPP: Evidence From a Comparison of
Matched Records
David C ARD
University of California of Berkeley, Berkeley, CA 94720 (card@econ.berkeley.edu )
Andrew K. G. H ILDRETH
California Census Research Data Center, University of California Berkeley, Berkeley, CA 94720
Lara D. S HORE -S HEPPARD
Downloaded by [Universitas Maritim Raja Ali Haji] at 00:33 13 January 2016
Williams College, Williamstown, MA 01267
This article studies the accuracy of reported Medicaid coverage in the Survey of Income and Program
Participation (SIPP) using administrative records from the State of California. Overall, we estimate that
the SIPP underestimates Medicaid coverage in the California population by about 10%. The probability
of correctly reporting coverage for those actually covered by Medicaid is around 85%, and is even higher
for low-income children. The probability that people who are not covered by Medicaid incorrectly report
that they are covered is about 1.3% for the population as a whole, but is higher (up to 7%) for low-income
children.
KEY WORDS: Attenuation bias; Misclassification.
1. INTRODUCTION
One of the most widely debated policy concerns in the United
States is the adequacy of health insurance coverage for lowincome children and adults. The Medicaid program was established in 1965 to provide health insurance for female-headed
families on public assistance and for the aged, blind, and disabled. Over the past two decades the program has gradually
expanded to cover low-income families that are not participating in other welfare programs (Gruber 2002). Despite these expansions, data from the Current Population Survey (CPS) show
that about 25% of poor children lacked health insurance coverage in the mid-1990s (U.S. Department of Commerce, Bureau
of the Census 2000). In the wake of recent federal and statelevel welfare reforms, there have been renewed efforts to maintain and expand Medicaid coverage. Nevertheless, CPS data
show that the fraction of the U.S. population with measured
health insurance coverage fell slightly over the 1990s.
Although analysts agree that expansions in the potential
availability of Medicaid have not led to equivalent increases
in measured coverage (Shore-Sheppard 1999; Gruber 2002),
there is less consensus on the reasons for this phenomenon. One
simple explanation is that people underreport their true Medicaid status in surveys such as the CPS or the Survey of Income and Program Participation (SIPP), the two key sources of
data on health insurance coverage in the United States. Indeed,
comparisons between administrative data and CPS estimates
of the number of Medicaid recipients show a growing divergence in the 1990s. For example, administrative data from the
“Form 2082” reporting system show a 29% rise in the number of people covered by Medicaid at any point during the year
from 1992 to 1998 (U.S. Department of Health and Human
Services, Center for Medicaid and Medicare Services 2002),
whereas survey data from the March CPS show only a 7% rise.
Even if underreporting by people who have coverage is partially
offset by “false-positive” responses among nonrecipients, measurement errors in Medicaid coverage can lead to significant
understatement of the takeup rate for the program, potentially
explaining some of the puzzling results in the literature.
In this article we present new evidence on the accuracy of
Medicaid coverage responses in the SIPP. Unlike the March
CPS, which asks individuals whether they were covered by
Medicaid at any time in the previous year, the SIPP asks
questions about coverage on a month-by-month basis. We use
a unique dataset formed by merging survey information from
the 1990–1993 SIPP panels with administrative data on Medicaid coverage from California’s Medi-Cal Eligibility File (MEF).
The combined sample contains actual and reported Medicaid
coverage status for 20,000 individuals and 640,000 personmonths. We construct estimates of net and gross error rates
in reported coverage for the overall population and for various
subgroups that can be used by researchers to gauge the potential
biases in statistical analyses that use the SIPP data.
The article is organized as follows. Section 2 provides a brief
overview of the Medicaid program. In Section 3 describes the
SIPP survey and presents various data on measured Medicaid
participation patterns in the California sample. Section 4 describes the administrative data and summarizes the matching
process. It also presents information on the characteristics of
the matched sample versus the overall California population.
Section 5 contains our main results, including cross-tabulations
of reported Medicaid status in the SIPP survey and the MEF
for the overall matched sample, and various subsamples. It also
summarizes the implications of our findings for studies that
use reported Medicaid coverage as either a dependent variable or an explanatory variable. Finally, Section 6 reviews our
main conclusions.
410
© 2004 American Statistical Association
Journal of Business & Economic Statistics
October 2004, Vol. 22, No. 4
DOI 10.1198/073500104000000208
Card, Hildreth, and Shore-Sheppard: Measuring Medicaid Coverage in the SIPP
Downloaded by [Universitas Maritim Raja Ali Haji] at 00:33 13 January 2016
2. THE CALIFORNIA MEDICAID PROGRAM
Medicaid is a joint state–federal program that pays for medical services for low-income individuals, including the elderly,
blind, and disabled recipients of SSI; the “medically needy”
(people who have recently incurred large medical expenses);
and people in low income families. Historically, the latter group
was made up exclusively of participants in the Aid to Families with Dependent Children (AFDC) program. Starting in the
mid-1980s, however, a series of federal law changes expanded
Medicaid eligibility to families with incomes above the AFDC
threshold and others that did not meet the family composition
rules of AFDC. The 1989 Omnibus Budget Reconciliation Act
(OBRA) mandated that states offer Medicaid coverage to pregnant women and children up to age 6 with family incomes
below 133% of the federal poverty threshold. OBRA 1990 further expanded coverage to children born after September 30,
1983 and living in families with incomes below the poverty
line. Other legislative changes in the late 1980s and early 1990s
allowed states to expand Medicaid coverage beyond these minimum mandates. California, for example, raised the family income limit for pregnant women and infants to 200% of the
federal poverty line.
During the 1990s, enrollment patterns in the California Medicaid program—known as Medi-Cal—closely tracked national
trends. Between 1991 and 1998, the state accounted for a
steady 16% of average monthly Medicaid enrollment in the
United States. Further, the ratio of per-capita expenditures in
California to the nation as a whole remained relatively constant. In light of this stability and the size and diversity of the
411
California population, we believe that the state provides an excellent testing ground for evaluating the quality of Medicaid
coverage responses.
Table 1 reports the various Medi-Cal eligibility categories
in effect in California as of late 1995, along with estimates of
the number of people covered under each category. Despite the
coverage expansions in the late 1980s and early 1990s, more
than 75% of individuals covered by Medi-Cal in 1995 were
adults or children enrolled in AFDC or SSI. The majority of this
group—about 60% of total Medi-Cal enrollees—were AFDC
recipients. Another 10% were medically needy adults and children, 5% were refugees and undocumented aliens, and 5% were
medically indigent adults and children. Only about 3% of MediCal enrollees in 1995 were women or children who were receiving coverage as a result of the poverty-related expansions.
In part this reflects the generous AFDC limits in California,
where the fraction of children in families with incomes above
the AFDC threshold but below the poverty line is lower than in
most other states.
Given the high fraction of Medi-Cal enrollees whose coverage is linked to welfare participation, it is not surprising that
changes in Medi-Cal enrollment are strongly related to changes
in the welfare caseload. The recession of the early 1990s led
to a rise in California’s welfare rolls and increases in MediCal enrollment. Since 1996, welfare rolls and Medicaid enrollment have both declined in California, with evidence that most
of the drop in Medi-Cal has been attributable to the drop in
the number of families receiving cash assistance (Broaddus and
Guyer 2000).
Table 1. Medi-Cal Eligibility Criteria and Caseload as of 1995
Eligibility basis
Categorically eligible AFDC/SSI recipients:
Families with dependent children in AFDC
Aged, blind, and disabled in SSI/SSP
Women and children in low-income families:
Pregnant women with family income
ISSN: 0735-0015 (Print) 1537-2707 (Online) Journal homepage: http://www.tandfonline.com/loi/ubes20
The Measurement of Medicaid Coverage in the
SIPP
David Card, Andrew K.G Hildreth & Lara D Shore-Sheppard
To cite this article: David Card, Andrew K.G Hildreth & Lara D Shore-Sheppard (2004) The
Measurement of Medicaid Coverage in the SIPP, Journal of Business & Economic Statistics,
22:4, 410-420, DOI: 10.1198/073500104000000208
To link to this article: http://dx.doi.org/10.1198/073500104000000208
Published online: 01 Jan 2012.
Submit your article to this journal
Article views: 58
View related articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ubes20
Download by: [Universitas Maritim Raja Ali Haji]
Date: 13 January 2016, At: 00:33
The Measurement of Medicaid Coverage in
the SIPP: Evidence From a Comparison of
Matched Records
David C ARD
University of California of Berkeley, Berkeley, CA 94720 (card@econ.berkeley.edu )
Andrew K. G. H ILDRETH
California Census Research Data Center, University of California Berkeley, Berkeley, CA 94720
Lara D. S HORE -S HEPPARD
Downloaded by [Universitas Maritim Raja Ali Haji] at 00:33 13 January 2016
Williams College, Williamstown, MA 01267
This article studies the accuracy of reported Medicaid coverage in the Survey of Income and Program
Participation (SIPP) using administrative records from the State of California. Overall, we estimate that
the SIPP underestimates Medicaid coverage in the California population by about 10%. The probability
of correctly reporting coverage for those actually covered by Medicaid is around 85%, and is even higher
for low-income children. The probability that people who are not covered by Medicaid incorrectly report
that they are covered is about 1.3% for the population as a whole, but is higher (up to 7%) for low-income
children.
KEY WORDS: Attenuation bias; Misclassification.
1. INTRODUCTION
One of the most widely debated policy concerns in the United
States is the adequacy of health insurance coverage for lowincome children and adults. The Medicaid program was established in 1965 to provide health insurance for female-headed
families on public assistance and for the aged, blind, and disabled. Over the past two decades the program has gradually
expanded to cover low-income families that are not participating in other welfare programs (Gruber 2002). Despite these expansions, data from the Current Population Survey (CPS) show
that about 25% of poor children lacked health insurance coverage in the mid-1990s (U.S. Department of Commerce, Bureau
of the Census 2000). In the wake of recent federal and statelevel welfare reforms, there have been renewed efforts to maintain and expand Medicaid coverage. Nevertheless, CPS data
show that the fraction of the U.S. population with measured
health insurance coverage fell slightly over the 1990s.
Although analysts agree that expansions in the potential
availability of Medicaid have not led to equivalent increases
in measured coverage (Shore-Sheppard 1999; Gruber 2002),
there is less consensus on the reasons for this phenomenon. One
simple explanation is that people underreport their true Medicaid status in surveys such as the CPS or the Survey of Income and Program Participation (SIPP), the two key sources of
data on health insurance coverage in the United States. Indeed,
comparisons between administrative data and CPS estimates
of the number of Medicaid recipients show a growing divergence in the 1990s. For example, administrative data from the
“Form 2082” reporting system show a 29% rise in the number of people covered by Medicaid at any point during the year
from 1992 to 1998 (U.S. Department of Health and Human
Services, Center for Medicaid and Medicare Services 2002),
whereas survey data from the March CPS show only a 7% rise.
Even if underreporting by people who have coverage is partially
offset by “false-positive” responses among nonrecipients, measurement errors in Medicaid coverage can lead to significant
understatement of the takeup rate for the program, potentially
explaining some of the puzzling results in the literature.
In this article we present new evidence on the accuracy of
Medicaid coverage responses in the SIPP. Unlike the March
CPS, which asks individuals whether they were covered by
Medicaid at any time in the previous year, the SIPP asks
questions about coverage on a month-by-month basis. We use
a unique dataset formed by merging survey information from
the 1990–1993 SIPP panels with administrative data on Medicaid coverage from California’s Medi-Cal Eligibility File (MEF).
The combined sample contains actual and reported Medicaid
coverage status for 20,000 individuals and 640,000 personmonths. We construct estimates of net and gross error rates
in reported coverage for the overall population and for various
subgroups that can be used by researchers to gauge the potential
biases in statistical analyses that use the SIPP data.
The article is organized as follows. Section 2 provides a brief
overview of the Medicaid program. In Section 3 describes the
SIPP survey and presents various data on measured Medicaid
participation patterns in the California sample. Section 4 describes the administrative data and summarizes the matching
process. It also presents information on the characteristics of
the matched sample versus the overall California population.
Section 5 contains our main results, including cross-tabulations
of reported Medicaid status in the SIPP survey and the MEF
for the overall matched sample, and various subsamples. It also
summarizes the implications of our findings for studies that
use reported Medicaid coverage as either a dependent variable or an explanatory variable. Finally, Section 6 reviews our
main conclusions.
410
© 2004 American Statistical Association
Journal of Business & Economic Statistics
October 2004, Vol. 22, No. 4
DOI 10.1198/073500104000000208
Card, Hildreth, and Shore-Sheppard: Measuring Medicaid Coverage in the SIPP
Downloaded by [Universitas Maritim Raja Ali Haji] at 00:33 13 January 2016
2. THE CALIFORNIA MEDICAID PROGRAM
Medicaid is a joint state–federal program that pays for medical services for low-income individuals, including the elderly,
blind, and disabled recipients of SSI; the “medically needy”
(people who have recently incurred large medical expenses);
and people in low income families. Historically, the latter group
was made up exclusively of participants in the Aid to Families with Dependent Children (AFDC) program. Starting in the
mid-1980s, however, a series of federal law changes expanded
Medicaid eligibility to families with incomes above the AFDC
threshold and others that did not meet the family composition
rules of AFDC. The 1989 Omnibus Budget Reconciliation Act
(OBRA) mandated that states offer Medicaid coverage to pregnant women and children up to age 6 with family incomes
below 133% of the federal poverty threshold. OBRA 1990 further expanded coverage to children born after September 30,
1983 and living in families with incomes below the poverty
line. Other legislative changes in the late 1980s and early 1990s
allowed states to expand Medicaid coverage beyond these minimum mandates. California, for example, raised the family income limit for pregnant women and infants to 200% of the
federal poverty line.
During the 1990s, enrollment patterns in the California Medicaid program—known as Medi-Cal—closely tracked national
trends. Between 1991 and 1998, the state accounted for a
steady 16% of average monthly Medicaid enrollment in the
United States. Further, the ratio of per-capita expenditures in
California to the nation as a whole remained relatively constant. In light of this stability and the size and diversity of the
411
California population, we believe that the state provides an excellent testing ground for evaluating the quality of Medicaid
coverage responses.
Table 1 reports the various Medi-Cal eligibility categories
in effect in California as of late 1995, along with estimates of
the number of people covered under each category. Despite the
coverage expansions in the late 1980s and early 1990s, more
than 75% of individuals covered by Medi-Cal in 1995 were
adults or children enrolled in AFDC or SSI. The majority of this
group—about 60% of total Medi-Cal enrollees—were AFDC
recipients. Another 10% were medically needy adults and children, 5% were refugees and undocumented aliens, and 5% were
medically indigent adults and children. Only about 3% of MediCal enrollees in 1995 were women or children who were receiving coverage as a result of the poverty-related expansions.
In part this reflects the generous AFDC limits in California,
where the fraction of children in families with incomes above
the AFDC threshold but below the poverty line is lower than in
most other states.
Given the high fraction of Medi-Cal enrollees whose coverage is linked to welfare participation, it is not surprising that
changes in Medi-Cal enrollment are strongly related to changes
in the welfare caseload. The recession of the early 1990s led
to a rise in California’s welfare rolls and increases in MediCal enrollment. Since 1996, welfare rolls and Medicaid enrollment have both declined in California, with evidence that most
of the drop in Medi-Cal has been attributable to the drop in
the number of families receiving cash assistance (Broaddus and
Guyer 2000).
Table 1. Medi-Cal Eligibility Criteria and Caseload as of 1995
Eligibility basis
Categorically eligible AFDC/SSI recipients:
Families with dependent children in AFDC
Aged, blind, and disabled in SSI/SSP
Women and children in low-income families:
Pregnant women with family income