KEY FINDINGS
- Medicaid and CHIP have succeeded in reaching the target population of uninsured children and have contributed greatly to the reduction in uninsurance among low-income children from 25 percent in 1997 to 13 percent in 2012.
- During the same period, uninsurance rates rose among adults, who were less likely to qualify for Medicaid and CHIP.
- All racial and income groups experienced gains in coverage, but the gains have been particularly striking among Hispanic children.
- Participation in Medicaid and CHIP among eligible children increased nationwide from 82 percent in 2008 to 88 percent in 2012; by 2012, 21 states had achieved participation rates of 90 percent or higher while just 5 states had rates of 80 percent or lower.
- The number of children eligible for Medicaid or CHIP yet uninsured fell from 4.9 to 3.7 million between 2008 and 2012, and 68 percent of all remaining uninsured children are eligible for Medicaid or CHIP.
Previous research has documented substantial declines in uninsurance among low-income children following CHIP’s implementation. Those declines stand in contrast, sometimes sharply, with uninsurance trends for low-income parents and other groups not eligible for the program (Rosenbach et al. 2007; Choi, Sommers, and McWilliams 2011; Howell and Kenney 2012; Blavin et al. 2012a). Studies also show that CHIP expansions have contributed to a reduction in racial and ethnic disparities in coverage among low-income children (Shone et al. 2005; Currie et al. 2008; Choi et al. 2011; Blavin et al. 2012b; Coyer and Kenney 2013; Kenney, Coyer, and Anderson 2013).
In this chapter, we present coverage trends since CHIP’s enactment, including trends in the proportion of children without health insurance. We use a consistent time series of data from the Current Population Survey Annual Social and Economic Supplement (CPS-ASEC), the most widely cited source of information about health insurance coverage; the CPS-ASEC covers the 15-year period—1997 through 2012— since CHIP’s enactment. We also include in this chapter an analysis of data from the ACS for 2008 and 2012 to show changes over time as well as variation across states and key subpopulations in the rate of Medicaid and CHIP participation among eligible children. Overall, the findings suggest that Medicaid and CHIP have succeeded in reaching the target population of uninsured children and have contributed greatly to the reduction in uninsurance among low-income children from 25 percent in 1997 to 13 percent in 2012.
-
Children’s Health Insurance Coverage In The CHIP Era, 1997–2012
-
Medicaid and CHIP coverage offset declines in employment-based coverage, fueling a substantial decline in uninsurance among children.
Between 1997 and 2012, most children (all incomes) had coverage from a parent’s employer, although the proportion with such coverage dropped from 63 to 55 percent over the period (Figure III.1).26Meanwhile, Medicaid and CHIP coverage among all children increased from 20 to 35 percent over the same period.27 Increased public coverage more than offset the loss of employer-sponsored coverage so that the percentage of all children who were uninsured fell by 6 percentage points (from 15 to 9 percent) despite recession conditions that separated many families from their connection to employer-sponsored coverage and left families with fewer resources to purchase coverage on their own.
Figure III.1. Percentage with Medicaid/CHIP, Employer-Sponsored Insurance, and Uninsured: All Children, 1997–2012
Source: CPS-ASEC.
Notes: Children are ages 0 to 18.
Given that both CHIP and Medicaid are means-tested programs, their impact is even more visible when low-income children (in families with income below 200 percent of the FPL) are the focus, as in Figure III.2. Among low-income children, Medicaid and CHIP coverage exceeded employer-sponsored coverage throughout the period, rising from 41 percent in 1997 to 63 percent in 2012. The proportion of low-income children who were uninsured fell from 25 percent in 1997 to 13 percent in 2012.
Figure III.2. Percentage with Medicaid/CHIP, Employer-Sponsored Insurance, and Uninsured: Low-Income Children, 1997–2012
Source: CPS-ASEC.
Notes: Children are ages 0 to 18. Low income is below 200 percent of the FPL.
Although public coverage rates are consistently highest among children with the lowest incomes, the gains in public coverage among children in families with incomes in the range targeted by CHIP—between 100 and 300 percent of the FPL—were even greater (Figure III.3). Over the 15-year period since CHIP’s inception, public coverage rates for children in families with income between 100 and 200 percent of the FPL increased by 26 percentage points; rates for children in families with income between 200 and 300 percent of the FPL increased by 18 percentage points; and rates for children with income in the Medicaid range of under 100 percent of the FPL increased by 15 percentage points.
Figure III.3. Percentage of Children Covered by Medicaid/CHIP, by Poverty Level: 1997–2012
Source: CPS-ASEC.
Note: Children are ages 0 to 18.
The coverage gains for low-income children were not matched by similar gains for low-income adults, pointing to the importance of public coverage in driving the decline in uninsurance among children.
Throughout the 15-year period, uninsured rates were substantially higher among low-income adults than among children (Figure III.4). Uninsured rates were consistently highest among adults without children, who were less likely to be eligible for public coverage. Uninsurance among low-income adults without children remained fairly constant during the period, while rates for adult parents increased from 33 to 38 percent. In contrast, rates for children declined steadily, falling from 25 to 13 percent over the 15-year period.
Figure III.4. Percentage Uninsured: Low-Income Children and Adults, 1997–2012
Source: CPS-ASEC.
Notes: Children are ages 0 to 18. Low income is below 200 percent of the FPL.
Public coverage gains were similar across groups of children defined by race and ethnicity, helping to narrow disparities in uninsured rates, especially for Hispanic children.
The coverage trends for low-income children show similar patterns across racial and ethnic groups. Medicaid and CHIP coverage increased (Figure III.5), and uninsurance fell (Figure III.6) for Hispanics, non-Hispanic whites, non-Hispanic blacks, and other groups of children (which includes Asian-Americans, Native Hawaiians and Other Pacific Islanders, and American Indians and Alaska Natives). The trends among low-income Hispanic children are particularly striking: the uninsured rate was cut in half, falling from 34 percent in 1997 to 17 percent in 2012. The improvement was driven by the increase in Medicaid and CHIP coverage, from 42 percent in 1997 to 65 percent in 2012.
Figure III.5. Percentage of Low-Income Children Covered by Medicaid/CHIP Coverage, by Race and Ethnicity, 1997–2012
Source: CPS-ASEC.
Notes: Children are ages 0 to 18. Low income is below 200 percent of the FPL. Hispanic includes all races. Other includes Asian-American, Native-Hawaiian and Other Pacific Islander, and American Indian and Alaska Native. Non-Hispanic respondents indicating more than one race are assigned to a primary race based on a hierarchy originally developed for ASPE’s TRIM3 microsimulation model. Prior to March 2003, individuals could only report one race.
By 2012, the uninsured rate had fallen by 10 percentage points for non-Hispanic black and non-Hispanic white children and by 17 percentage points for Hispanic children (Figure III.5). Disparities in coverage for Hispanic children also declined sharply; the differential between non-Hispanic white and Hispanic children narrowed from a 13 percentage points in 1997 to 5 percentage points in 2012.
Figure III.6. Percentage of Low-Income Children Uninsured, by Race and Ethnicity, 1997–2012
Source: CPS-ASEC.
Notes: Children are ages 0 to 18. Low income is below 200 percent of the FPL. Hispanic includes all races. Other includes Asian-American, Native-Hawaiian and Other Pacific Islander, and American Indian and Alaska Native. Non-Hispanic respondents indicating more than one race are assigned to a primary race.
26 Interpretation of CPS-ASEC health insurance data is subject to several caveats. Research matching CPS-ASEC responses with Medicaid and CHIP administrative data shows that significant percentages of respondents in all age groups who are enrolled in Medicaid or CHIP do not report such coverage on the survey. Consequently, coverage reported through the survey is lower than totals in administrative data, and uninsurance estimates from survey data are inflated. Introduction of verification questions in the survey, in which respondents who said no when asked about all coverage types were asked to confirm that they were uninsured, increased reported coverage rates beginning with the data for 1999, as did retroactive improvements in procedures for imputing responses among those who did not provide answers to the health insurance questions. A portion of the increase in coverage since 1997 is thus attributable to changes in CPS-ASEC methods. See U.S. Census Bureau (2008) and U.S. Census Bureau (2011).
27 The CPS-ASEC asks separate questions about Medicaid and CHIP coverage. Many analysts, however, believe that respondents do not always distinguish accurately between the two programs. It is particularly difficult to do so in states where the programs have the same name.
-
-
Participation in Medicaid and CHIP, 2008 and 2012
-
Medicaid and CHIP participation rates for children increased substantially overall and in most states and for all major subpopulations examined.
Nationally, the estimated number of uninsured children and the proportion of children uninsured fell between 2008 and 201228, from 7.2 to 5.4 million and from 9.2 to 7.0 percent, respectively, according to data from the ACS (Table III.1).29 Decreases in uninsured rates were widespread among children, with statistically significant declines in 33 states. As a result, the distribution of uninsured rates narrowed across states, although a differential of 14.5 percentage points persisted in the uninsured rates among children; in 2012, Nevada had an uninsured rate of 15.8 percent among children compared to a rate of 1.3 percent among children in Massachusetts. In 2008, a differential of 18 percentage points existed between Nevada and Massachusetts, the states with the highest and lowest uninsured rates for children, respectively.
Table III.1. Medicaid/CHIP Eligibility, Participation, and Uninsurance of Children (Ages 0 to 18) by State, 2008 and 2012
Uninsured Medicaid/CHIP Eligibles Medicaid/CHIP Participation 2008 2012 2008 2012 2008 2012 Rate (%) Rate (%) Number (1,000s) Number (1,000s) Rate (%) Rate (%) Nation 9.2 7.0 ++ 41,548 46,025 81.7 88.1 ++ Alabama 7.6** 4.0**++ 651 855 85.4** 92.6**++ Alaska 12.1** 12.8** 86 93 70.4** 81.7**++ Arizona 14.8** 12.5**++ 909 976 76.3** 81.8**++ Arkansas 8.3 5.5**++ 444 452 87.8** 93.9**++ California 10.0** 7.8**++ 5,687 6,010 81.4 87.0**++ Colorado 12.9** 8.1**++ 519 681 69.3** 85.0**++ Connecticut 5.0** 3.6**++ 389 419 85.8** 93.0**++ Delaware 8.0 3.6**++ 101 110 81.2 93.9**++ District of Columbia 2.8** 2.7** 78 74 95.6** 97.1** Florida 16.6** 10.6**++ 2,082 2,385 69.8** 85.5**++ Georgia 10.7** 8.6**++ 1,538 1,644 81.0 85.8**++ Hawaii 2.9** 2.9** 200 230 91.5** 92.6** Idaho 12.6** 7.6 ++ 216 237 73.6** 86.3 ++ Illinois 5.4** 4.0**++ 2,004 2,012 88.0** 93.8**++ Indiana 9.1 7.6**++ 981 1,032 78.5** 84.4**++ Iowa 4.7** 4.1** 545 587 85.9** 89.8 + Kansas 7.8** 6.8 342 431 81.4 86.4++ Kentucky 5.7** 5.9** 579 613 89.5** 90.2** Louisiana 7.4** 5.5**++ 747 773 88.3** 92.5**++ Maine 5.4** 4.7** 136 149 91.0** 94.0** Maryland 5.2** 3.8**++ 704 749 86.3** 91.9**++ Massachusetts 1.6** 1.3** 685 723 95.0** 97.4**++ Michigan 4.9** 4.2**++ 1,230 1,299 89.6** 92.2**++ Minnesota 5.6** 5.5** 646 687 81.3 85.3**++ Mississippi 12.0** 7.2 ++ 514 537 81.4 90.3**++ Missouri 6.6** 7.1 968 1,002 85.3** 85.5** Montana 14.1** 11.2** 114 150 67.9** 81.0**++ Nebraska 6.9** 5.4** + 205 238 80.8 88.4 + Nevada 20.0** 15.8**++ 318 392 56.1** 70.6**++ New Hampshire 5.0** 4.1** 146 149 85.5 89.7 New Jersey 6.8** 5.0**++ 1,089 1,159 82.4 88.7++ New Mexico 12.8** 8.2++ 344 389 81.6 89.3++ New York 5.3** 3.9**++ 2,432 3,214 89.2** 92.4**++ North Carolina 9.5 7.1++ 1,181 1,368 84.6** 89.6**++ North Dakota 7.3 7.4 47 44 75.9 84.5 Ohio 6.7** 5.4**++ 1,366 1,494 83.3** 89.5++ Oklahoma 11.8** 10.0**++ 542 579 81.2 85.8**++ Oregon 11.7** 5.4**++ 402 621 74.9** 90.2**++ Pennsylvania 5.7** 4.8**++ 1,757 1,828 86.1** 89.4**++ Rhode Island 5.3** 5.7** 111 120 85.1 90.4 + South Carolina 10.9** 7.9*++ 596 686 79.4** 87.5++ South Dakota 8.4 3.9**++ 106 102 83.2 92.1**++ Tennessee 6.7** 5.6**+ 945 1019 86.3** 90.3**++ Texas 16.3** 12.2**++ 3,756 4,142 74.6** 84.3**++ Utah 12.0** 9.4** 362 446 65.8** 95.8**++ Vermont 3.7** 2.9** 81 84 93.5** 95.2** Virginia 7.2** 5.5**++ 752 824 80.0** 87.5++ Washington 7.7** 5.4**++ 831 1,025 82.5 89.4*++ West Virginia 6.1** 4.1**++ 241 288 89.3** 91.1** Wisconsin 4.7** 4.7** 784 839 86.2** 88.7 Wyoming 8.9 9.3* 58 65 76.4 81.5+ Source: Analysis of the Urban Institute's Health Policy Center's ACS Medicaid/CHIP Simulation Model based on data from the Integrated Public Use Microdata Series (IPUMS) from 2008 and 2012.
Notes: See text for definitions of eligibility, participation, and uninsurance. Since eligibility estimates reflect potential eligibility based on meeting the income, asset, and immigration requirements, they include some children with ESI coverage who do not necessarily qualify for Medicaid/CHIP coverage. Eligibility estimates do not take into account waiting periods which vary by state.
**(*) indicates estimate is statistically different from national estimate at the 0.05 (0.1) level.
++(+) indicates 2012 estimate is statistically different from 2008 estimate at the 0.05 (0.1) levelIncreased take-up of Medicaid and CHIP coverage among eligible children was associated with a decrease in the number of uninsured children eligible for but not enrolled in Medicaid or CHIP (Table III.I). The decline in the number of uninsured children who were eligible for Medicaid or CHIP but not enrolled occurred even as states expanded eligibility to additional groups of children over that period and as more children became eligible for public coverage due to the economic downturn. Of the ten states that saw the largest decreases in uninsured rates for children between 2008 to 2012, seven saw the largest increases in Medicaid and CHIP participation rates for children over the same period (Colorado, Delaware, Florida, Idaho, Nevada, Oregon, and Texas) and seven were one of the ten states in 2008 with the highest uninsured rates (Colorado, Florida, Idaho, Mississippi, New Mexico, Nevada, and Texas).
Between 2008 and 2012, participation in Medicaid and CHIP rose nationally among eligible children, with statistically significant increases in 37 states.30 Nationwide, Medicaid and CHIP participation rates among children increased from 82 percent in 2008 to 88 percent in 2012; by 2012, 21 states had participation rates of 90 percent or higher and just two states had rates of 80 percent or lower (Table III.1).31 The 21 states that had participation rates above 90 percent draw from all four regions and include states that differ in terms of the demographic and socioeconomic composition of the children who are targeted by Medicaid and CHIP. Over the four-year period, many states implemented changes in their enrollment and re-enrollment processes and adopted new outreach strategies aimed at increasing the take-up and retention of Medicaid and CHIP coverage among eligible children (Heberlein et al. 2013; Hoag et al. 2013). Participation rates increased in Medicaid programs as well as in separate CHIP and Medicaid expansion CHIP programs (data not shown).32
Medicaid and CHIP participation rates increased across many subpopulations but still vary across groups.
Medicaid and CHIP participation rates increased among children across subgroups defined by income, age, race and ethnicity, immigration status, and functional status (Table III.2).33 In 2012, participation rates exceeded 85 percent across most subgroups, with the exception of American Indian children (78 percent), children with income above 138 percent of the FPL (80 to 81 percent depending on the income group), children ages 13 to 18 (83 percent), noncitizen children (83 percent), and citizen children with no parents in the household (83 percent). Participation rates exceeded 90 percent for children under age six (92 percent), children with income below 138% percent of the FPL (90 percent), black non-Hispanic children (92 percent), non-Hispanic children with multiple or other race (90 percent), and children with functional limitations (94 percent). Despite these gains, participation rates remained lower for some groups of children--for adolescents compared to younger children; for non-citizen compared to citizen children, and for children without functional limitations compared to children with functional limitations. For example, children ages 13 to 18 had participation rates that were nearly ten percentage points lower than children under age six.
Table III.2. Medicaid/CHIP Participation Rates for Children by Subgroup, 2008 and 2012
Participation Rates 2008 2012 Nation 81.7% 88.1% Income Less than 138 percent of FPL 84.4%** 90.1%**++ Between 138 and 200 percent of FPL 74.2%** 81.2%**++ Greater than 200 percent of FPL 71.6%** 79.8%**++ Age 0 to 5 85.6%** 91.5%**++ 5 to 12 82.4%** 88.9%**++ 13 to 18 75.5%** 82.8%**++ Sex Male 81.6% 88.0%++ Female 81.7% 88.2%++ Race/Ethnicity Hispanic 78.8%** 87.2%**++ White, non-Hispanic 81.4% 87.0%**++ Black, non-Hispanic 86.8%** 92.2%**++ Asian, non-Hispanic 79.2%** 85.8%**++ American Indian, non-Hispanic 68.4%** 78.4%**++ Other/multiple, non-Hispanic 86.4%** 90.3%**++ Citizenship Citizen with no citizen parents 78.5%** 88.8%**++ Citizen with at least one citizen parent 83.3%** 88.8%**++ Noncitizen 78.5%** 82.9%**++ Citizen with no parents in household 75.5%** 82.8%**++ Functional Limitation Yes 91.0%** 94.2%**++ No 78.2%** 85.6%**++ Source: Analysis of the Urban Institute's Health Policy Center's ACS Medicaid/CHIP Simulation Model based on data from the Integrated Public Use Microdata Series (IPUMS) from 2008 and 2012.
Notes: See text for definitions of eligibility, participation, and uninsurance.
**(*) indicates estimate is statistically different from national estimate at the 0.05 (0.1) level.
++(+) indicates 2012 estimate is statistically different from 2008 estimate at the 0.05 (0.1) level.
The number of uninsured children eligible for public coverage has been declining but most remaining uninsured children are eligible for Medicaid or CHIP.
Between 2008 and 2012, the number of children eligible for Medicaid or CHIP but uninsured fell by about 1.2 million to 3.7 million and the estimated number of uninsured children fell from 7.2 to 5.4 million (Figure III.7). Over that period, more than a third of the states expanded coverage to new groups of children.34 Altogether, an additional 4.5 million children became eligible for Medicaid or CHIP between 2008 and 2012 as a consequence of a combination of the following: expansions of eligibility to new groups of children, increases in the total number of children, and shifts in the income distribution that made more children eligible for public coverage.
Figure III.7. Estimated Number of Uninsured Children (Ages 0 to18), 2008 and 2012 (in millions)
Source: Urban Institute’s Health Policy Center’s ACS Medicaid/CHIP Eligibility Simulation Model based on data from the Integrated Public Use Microdata Series (IPUMS) from 2008 and 2012.
Notes: Estimates reflect an adjustment for the misreporting of coverage on the ACS (see Footnote 28).
As of 2012, an estimated 68 percent of uninsured children were eligible for Medicaid or CHIP but not enrolled in either program (Figure III.8). The remaining 32 percent of uninsured children were not eligible for Medicaid or CHIP coverage because of their immigration status (7 percent of all uninsured children, and 22 percent of the uninsured children who are not eligible for Medicaid or CHIP) or because their income levels exceeded Medicaid and CHIP eligibility levels (25 percent of all uninsured children, and 78 percent of the uninsured children who are not eligible for Medicaid or CHIP).
Figure III.8. Profile of Medicaid and CHIP Eligibility Among Uninsured Children, 2012
Source: Analysis of the Urban Institute’s Health Policy Center’s ACS Medicaid/CHIP Eligibility Simulation Model based on data from the Integrated Public Use Microdata Series (IPUMS).
Notes: Estimates reflect an adjustment for the misreporting of coverage on the ACS.
The high participation rates achieved in a large and growing number of states and for many subgroups of children suggest that there is the potential for additional increases in Medicaid and CHIP coverage among the remaining 3.7 million eligible but uninsured children, particularly among the states and groups that are lagging behind. As discussed later in Chapter IX, the vast majority of low-income parents reported that they would enroll their uninsured child in Medicaid or CHIP if told their child was eligible, but many of these families did not know that their child was eligible or how to apply for coverage, or they thought the enrollment and renewal processes were difficult to navigate. Moreover, evidence presented in Chapter VI on the extent of churning in public coverage and gaps in coverage between Medicaid and separate CHIP programs suggests that state policy choices can have substantial effects on how successful states are at enrolling and retaining children in Medicaid and CHIP and transferring them seamlessly between programs.
28 We analyze trends between 2008 and 2012 because 2008 is the first year that the American Community Survey included questions on health insurance coverage and 2012 was the most recent year that was available when these analyses were being performed.
29 Reported estimates of uninsurance from the ACS are lower than the CPS estimates of uninsurance presented earlier in this chapter (Figure III.1). Even though the ACS coverage estimates released by the Census Bureau are generally considered reliable and align fairly well with those from other surveys, the estimates presented here reflect a set of logical coverage edits that are applied if other information collected in the ACS implies that coverage for a sample case likely has been misclassified (Lynch et al. 2011). The edits bring the ACS estimates closer to distributions reported in other national surveys, such as the National Health Interview Survey (NHIS), and bring the Medicaid/CHIP coverage estimates from the ACS more in line with administrative totals. For more details, see Kenney et al. (2011).
30 Participation rates are the ratio of eligible children enrolled in Medicaid or CHIP to that number plus eligible children not enrolled in Medicaid or CHIP. We exclude the small number of children with both Medicaid/CHIP and employer-provided/union-based, military, or private nongroup coverage; we also exclude the children with Medicaid/CHIP coverage without a known eligibility pathway since we cannot include them in a consistent way.
31 Although six states showed a slight increase in uninsured rates for children over this period (AK, KY, MO, ND, RI and WY), the increases were not statistically significant.
32 There is substantial error in the measurement of program type based on the information available on the ACS which is why we do not provide participation rates for children who are eligible for different types of programs.
33 The Indian Health Service (IHS) is not typically counted as health insurance coverage because of limitations in the scope of available services and the geographic reach of IHS facilities. For most states, the participation rates do not change in a meaningful way when IHS was considered a source of health insurance coverage; however, in six states—Alaska, Montana, New Mexico, North Dakota, Oklahoma, and South Dakota—the participation rate increased by more than 2 percentage points when IHS was reclassified as insurance coverage when using data from 2009, with a particularly noticeable impact in Alaska. The other estimate sensitive to how IHS was treated was the participation rate among American Indian/Alaska Native children, which increased from 74.5 to 91.8 percent when the IHS was classified as health insurance coverage.
34 By 2011, 25 states and the District of Columbia had eligibility levels at or above 250 percent of the FPL; of those, 17 had thresholds of 300 percent of the FPL or higher, and nearly half of states covered lawfully residing immigrant children who had been in the country fewer than five years. See Heberlein et al. (2013) for more information.
-
View full report
"rpt_CHIPevaluation.pdf" (pdf, 1.01Mb)
Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®