Children enrolling through ELE appear to be similar to other Medicaid and CHIP enrollees in most respects except for one: we observe sizable differences in the distribution of age by enrollment pathway, suggesting that ELE may have been successful at enrolling some traditionally hard-to-reach older children. Additionally, there are smaller but notable differences in family size in the two programs that provided this data. And, finally, in the two Iowa programs, we find ELE enrollees have substantially lower incomes than non-ELE enrollees. Table III.1 compares ELE and non-ELE children in terms of demographic characteristics.
For every state, a larger proportion of ELE enrollees than non-ELE enrollees are teenagers (13 to 18 years old) than non-ELE enrollees (Figure III.16). This difference is greatest in Louisiana, mostly driven by the fact that a sizable proportion of non-ELE enrollees are under age one, while less than 5 percent of ELE enrollees are. However, even if infants are excluded from the analysis, the average age of ELE enrollees in Louisiana is higher than that of non-ELE enrollees. Other states show smaller, but still sizeable differences between the two age groups. For example, in Iowa Medicaid, 31 percent of ELE enrollees are ages 13-18, compared to 23 percent of non-ELE enrollees; and there is a similar 8-percentage point difference by group among new enrollees in New Jersey Medicaid (35 to 27 percent). Given that teenagers are traditionally the most likely age group of children to be uninsured, this finding may speak to the promise of using ELE as a means of reaching and enrolling older children.
Table III.1. Characteristics of ELE and Non-ELE New Enrollees (Percentages)
|Alabama Medicaid||Iowa Medicaid||Iowa’s CHIP Program||Louisiana Medicaid||New Jersey Medicaid||New Jersey CHIP (Medicaid and Separate CHIP Programs)|
|< 1 year||4||36||3||8||0||0||2||52||5||10||2||2|
|< 100% FPL||NA||NA||88||80||3||5||NA||NA||8||7||0||0|
|>100 to 133% FPL||NA||NA||11||17||5||6||NA||NA||91||89||2||10|
|>133 to 185% FPL||NA||NA||0||2||73||53||NA||NA||1||5||57||44|
|>185 to 200% FPL||NA||NA||0||0||8||13||NA||NA||0||0||13||19|
|>200 to 300% FPL||NA||NA||0||1||11||23||NA||NA||0||0||24||24|
Sample: Children newly enrolled in coverage; data for Iowa’s separate CHIP program and New Jersey Medicaid and CHIP are for June 2009–December 2011; data for Louisiana Medicaid and Medicaid expansion CHIP are for February 2010–November 2011.
Note: Iowa’s separate CHIP program income and family size numbers included some (fewer than 50) individuals counted twice.
CHIP = Children’s Health Insurance Program; ELE = Express Lane Eligibility; FPL = Federal poverty level; NA = not available.
Figure III.16. Percentage of New Enrollees Ages 13–18, ELE and Non-ELE
Source: Mathematica analysis of Maximizing Enrollment grantee state data, 2012 for Alabama and Louisiana, and Mathematica analysis of state-reported data, May 2008 – December 2011 for New Jersey, June 2009 – February 2012 for Iowa Medicaid, and June 2009 – December 2011 for Iowa’s separate CHIP program.
CHIP = Children’s Health Insurance Program; ELE = Express Lane Eligibility.
In Iowa’s separate CHIP and Medicaid programs, children enrolling through ELE are from families with lower incomes. In the separate CHIP program, 81 percent of ELE enrollees are below 185 percent of the FPL, compared to 64 percent of non-ELE enrollees. In Iowa Medicaid, enrollees below 100 percent of the FPL make up 88 percent of all new ELE enrollees, while the corresponding percentage for the non-ELE group is 80. In Iowa’s separate CHIP ELE program, because ELE enrollees are by definition applicants who have first applied to Medicaid, they are more likely to be closer to the Medicaid income eligibility threshold than those entering through standard enrollment pathways. Iowa Medicaid’s ELE program has a different program design (matching to SNAP data). Here, it may be the case that ELE mailings are more successful at reaching potential enrollees below 100 percent of the FPL than standard forms of outreach. However, so few children have been enrolled through the ELE pathway in Iowa Medicaid that it is difficult to draw any solid conclusions from these observed differences.
ELE-enrolled children in Iowa’s separate CHIP program and New Jersey are more likely to be from families comprising two or three people, and less likely to be from families comprising four or more people, compared to non-ELE enrolled children. However, the observed differences are small (5 to 6 percentage points) and information on family size is only available from two of the four states, making it difficult to conclude much at this early stage.
Prior Public Coverage
Because of data availability, we can only review data on prior public coverage in three ELE programs: Alabama, Iowa Medicaid, and Louisiana. In both Iowa Medicaid and Louisiana, we find that children who enrolled via ELE were less likely than non-ELE enrollees to have a spell of public coverage in the 12-months prior to enrollment (Figure III.17). For example, ELE enrollees in Louisiana were 12 to 15 percentage points less likely than children who enrolled through standard processes to have a prior spell of public coverage in the 12 months before enrolling.
We find the opposite relationship in Alabama, with ELE enrollees more likely to have a recent spell of public coverage than non-ELE enrollees, with a difference of approximately 10 percentage points. However, whereas Iowa Medicaid and Louisiana used ELE as a mechanism to identify and enroll eligible but uninsured children, Alabama’s ELE program does not include an outreach component.42 Because families of children still need to apply for coverage as they would if ELE were not in place, regardless of whether state workers use TANF and SNAP income determinations in establishing Medicaid eligibility, we would not necessarily expect to find children in the ELE group more likely to be disconnected from public coverage than the non-ELE children.
Figure III.17. Public Coverage in 12 Months Prior to Enrollment, ELE and Non-ELE, Alabama, Iowa Medicaid, and Louisiana
Mathematica analysis of Maximizing Enrollment grantee State data, 2012 for Alabama and Louisiana, and Mathematica analysis of state-reported data, June 2009 – February 2012 for Iowa Medicaid.
ELE = Express Lane Eligibility
42 Alabama has plans to expand ELE to auto-enroll children who apply to TANF or SNAP and are found to be eligible for and not already enrolled in Medicaid (a process similar to Louisiana’s ELE program). However, this phase has not yet been implemented.