CHIPRA Mandated Evaluation of Express Lane Eligibility: First Year Findings. Findings

As of April 1, 2012, CMS had approved nine states (Alabama, Colorado, Georgia, Iowa, Louisiana, Maryland, New Jersey, Oregon, and South Carolina) to implement ELE. Five of the states (Alabama, Colorado, Louisiana, Maryland, and South Carolina) are approved for ELE in their Medicaid programs, while four (Georgia, Iowa, New Jersey, and Oregon) are approved for ELE in both Medicaid and their separate CHIP programs. Even among this small number of states, the striking variation in ELE designs reflects the flexibility that CHIPRA afforded. For example, states’ partner agencies range from those administering other public benefits—such as the National School Lunch Program (NSLP) (New Jersey and Colorado), SNAP (Alabama, Iowa, Louisiana, Oregon, and South Carolina), Temporary Assistance for Needy Families (TANF) (Alabama and South Carolina), and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) (Georgia)—to those responsible for aspects of state revenue, including the Division of Taxation (New Jersey) and the Office of the Comptroller (Maryland). Tables 1 and 2 summarize the key features and aims of the ELE programs in the six states included in the descriptive analysis of the costs and new enrollment associated with ELE implementation.  

Table 1. Features and Implementation Status in Six States with Approved ELE Programs as of December 2010

Source: Mathematica interviews with state staff conducted between January and May 2012 and analysis of state administrative data provided by states between January and June 2012.

^a New Jersey also has an NSLP ELE program, which has been implemented in all counties as of October 2011. Because it was not in place as of December 2010, the NSLP partnership is excluded from this initial analysis.

^b As of December 2011, Oregon was also piloting an NSLP ELE program in four school districts.

^c Implementation dates are those recognized by CMS as the program’s “Effective Date” unless available information suggests that a different date more accurately reflects the timing of meaningful changes. Implementation dates for Alabama, Iowa Medicaid, New Jersey, and Oregon are the CMS effective dates. The date for Iowa CHIP was given by state staff in an interview. For Louisiana, the date reflects when children were first enrolled via ELE. For Maryland, the 2008 date reflects the first outreach mailings on the basis of tax return information, which is the event considered by state staff to be the start of ELE in that state; this date also is referenced in Idala et al. (2011). Maryland’s process was not recognized as ELE by CMS until 2010, when the state started using the mailing process to establish in-state residency.

^d Iowa’s separate CHIP ELE process does not include renewals. However, children whose eligibility is redetermined by Medicaid, resulting in an ineligibility finding, may be ELE-referred to CHIP, with income-eligibility for CHIP established based on Medicaid’s income findings. This constitutes an ELE transfer, rather than a renewal.

^e Enrollment counts were computed from state administrative data and reflect the number of enrollments resulting from ELE application processes. Renewal counts reflect estimates provided by state staff during interviews of the approximate number of renewals resulting from ELE renewal processes.

^f We requested enrollment data from states for the period beginning one year prior to the ELE program effective date. Because CMS recognizes the effective date for Iowa CHIP’s ELE process as June 2010, we present data on enrollments from June 2009 through December 2011. The number of enrollees since the ELE process began in 2004 is likely much higher.

^g As of November 2011.

CHIP = Children’s Health Insurance Program; CMS = Centers for Medicare & Medicaid Services; ELE = Express Lane Eligibility; NSLP = National School Lunch Program; SNAP = Supplemental Nutrition Assistance Program; TANF = Temporary Assistance for Needy Families.

Table 2. Aims of ELE in Six States with Approved ELE Programs as of December 2010

Source: Mathematica interviews with state staff conducted between January and May 2012.

^a Since April 2010, self-declaration of income has been accepted for most ELE and non-ELE children, if income cannot be verified through databases accessible to state eligibility staff. This is not the case for children of self-employed parents, for whom income verification is required. Thus, ELE reduces the documentation burden for children with self-employed parents whose income cannot be verified via other databases but can be verified via SNAP or TANF databases.

CHIP = Children’s Health Insurance Program; ELE = Express Lane Eligibility; SNAP = Supplemental Nutrition Assistance Program; TANF = Temporary Assistance for Needy Families.

Across the seven ELE programs implemented as of December 2010, Medicaid and CHIP staff cited a number of potential benefits to applicants from their ELE programs. In five of the seven ELE programs (all except Alabama and Maryland), ELE required that applicants submit fewer documents in support of their ELE applications compared to standard applications, and required fewer interactions with state staff (Table 3). In addition, in six programs (all but Maryland), ELE applications were processed more quickly than standard applications, which resulted in expedited eligibility determinations for families. These expedited processes shortened the typical time from application to enrollment for an eligible child from roughly three or four weeks to one week or less.

Some states also noted that ELE has improved the application experience of non-ELE applicants. For example, in New Jersey, the use of tax information in ELE has led to more extensive use of tax information in standard application processing. Specifically, people applying via the standard route are permitted to attest to income and are not required to provide pay stubs if their attestation matches tax return data. Also, in Alabama and Louisiana, state staff suggested that the time saved by diverting some applications to ELE routes has meant that standard processing times for non-ELE applications are quicker than they would be otherwise.

Net administrative costs or savings differed widely across the six early states adopting ELE (Figure 1) (although these findings should be reviewed with caution given methodological challenges computing administrative costs and savings, including recall bias, incomplete data, and other process changes in the study period). For example, the ELE programs in Alabama and Louisiana yielded large net annual savings, driven by the staff time saved through processing large numbers of ELE applications and renewals in a more efficient manner than possible through traditional means. By contrast, in Maryland, ELE did not generate annual savings (for example, in staff time) that could offset new mailing costs associated with the policy, resulting in a net cost to the state from the policy. And, in New Jersey, targeted mailings cost the state about $250,000 per year, but no savings from ELE accrues to the state; any savings generated from more efficient ELE application processing are absorbed by a third-party administrator. Finally, ELE programs in Oregon and Iowa Medicaid were essentially “cost neutral” from an administrative perspective, as added mailing costs and the time spent processing unsuccessful ELE applications offset the time savings from processing successful ELE applications more efficiently than by traditional means.

Notably, in no case did program staff describe ELE as directly resulting in the need to hire additional staff or the ability to eliminate staff positions. Any time savings resulting from ELE were instead used to address other program needs. For example, Louisiana saved about 69,000 staff hours per year from ELE—the equivalent of 33 full-time positions—but it used the savings to make up for staffing reductions caused by state budget cuts. Staff described ELE as enabling Medicaid to stay on top of its workload in the face of reduced staffing. Similarly, Alabama noted that staff time saved through ELE may have enabled staff to process the traditional applications more quickly than would otherwise have been possible.

Table 3. Perceived Benefits of ELE for Applicants

Source: Mathematica analysis of interviews with state staff. Information about documentation requirements reflects interviews with state staff supplemented by a review of Medicaid and CHIP websites and application materials.

^a ELE enrollees to the separate CHIP program are enrolled roughly five days after the referral from Medicaid; however, coverage is retroactive to the Medicaid filing date.

CHIP = Children’s Health Insurance Program; ELE = Express Lane Eligibility.

Figure 1. Net Annual Administrative Savings and (Costs) from ELE

Source: Mathematica analysis of Interviews and follow-up correspondence with state staff between January and June 2012.

ELE = Express Lane Eligibility.

Except for Louisiana, where we observe a noticeable spike in both overall new enrollments and ELE-based new enrollments around the adoption of the policy, we see little descriptive evidence from the new enrollment analysis that ELE increased the number of children added to coverage (although among children who were added, 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). We likewise see little evidence of differences in retention between ELE and non-ELE new enrollees, again except for Louisiana, where roughly half of all ELE new enrollees exited coverage within 12 months due to a temporary policy requiring them to use their Medicaid card within 12 months. However, because the descriptive analysis can provide no estimate of the counterfactual—that is, what enrollment in ELE states would have been without the policy—it may significantly underestimate the effects of the policy on coverage. Moreover, because the descriptive analysis focused only on new enrollments tied directly to ELE, it could not account for enrollment gains arising indirectly from the policy, nor could it fully assess those affected by ELE renewal processes. These limitations are relevant for several states, including Alabama and Louisiana (both of which have adopted ELE for renewals) and New Jersey (where ELE has led to more extensive use of tax information, a “spillover” that cannot be identified in our data).

In contrast to these descriptive findings, the multivariate analysis, which combines aggregate total enrollment data for both ELE and non-ELE states, finds significant evidence of an ELE effect on total enrollment. Drawing on a range of models, the analysis finds that ELE increased total Medicaid enrollment by an estimated 5.5 percent and Medicaid and CHIP enrollment combined by an estimated 4.2 percent. Findings for Medicaid were notably robust, showing little variation in magnitude and remaining statistically significant across many sensitivity tests. However, as with all impact studies using regression methods to estimate the counterfactual, our multivariate findings might be biased from unobserved factors that differ between ELE and non-ELE states over time.

Taken together, these first-year findings suggest that ELE can be an effective tool for enrolling children into coverage. However, because evidence that ELE significantly increased enrollment is stronger in the multivariate findings than in the descriptive findings, it will be critical to revisit both analyses in the evaluation’s second year, as more post-ELE enrollment data become available. The case studies will further support these second-year analyses, permitting a more thorough understanding of other simplifications and the trends in enrollment under ELE.

In our discussions with administrators, several themes emerged about what factors facilitated ELE process improvements. Among them, states should think carefully about their selection of a viable Express Lane agency partner: it is a critical first step in the ELE process. First-year findings do not suggest that partnering with a certain type of agency—such as SNAP, TANF, Medicaid, or state tax agency—makes cost savings or enrollment gains more likely. Rather, it is likely tied to the processes for using partner agency information and whether that agency has data on large numbers of eligible-but-uninsured children or enrolled children up for renewal. States might consider “test” data matches with potential partners to understand how ELE might function in practice. For example, besides partnering with SNAP and TANF, Alabama considered an ELE partnership for its CHIP program with the state’s child care subsidy program because the income eligibility levels for both programs are similar. Through a test data match, Alabama administrators found that few eligible children were identified through the child care subsidy program, and that, given its older information systems, the costs of the systems changes that would be needed far outweighed the potential gains to coverage.

State administrators also said it was easier to implement ELE processes when they could build upon existing relationships, because of the familiarity with agency staff and their operations, as well as the existence of data use agreements that could easily be modified to accommodate ELE. Of the seven ELE programs studied as part of the qualitative analysis, six chose an Express Lane partner with which they had a prior data-sharing relationship.

Finally, in states where the ELE process uses partner agencies’ findings as a means to target application mailings, we find far fewer ELE-linked enrollments. These more tepid findings suggest that this outreach-focused approach to ELE, which requires parents to respond to mailings, may offer less promise as a means for enrolling large numbers of new children.

The findings presented in this report should be considered tentative; because ELE is so new and so varied, in its potential uses, and its implementation has been limited to a handful of States, it is too soon to draw conclusions about its effects on administrative costs or enrollment. In many instances, we need the second-year evaluation activities to assess the robustness of findings from this first year and to further elucidate the reasons behind, and meaning of, most of them. In addition, although this report looks at ELE’s implementation, ELE might have differential longer-term effects. For example, as ELE processes mature, the costs and savings that accrue to States could change. The second-year evaluation activities will provide a longer post-implementation period and offer more extensive data to assess both short- and longer-term policy effects. We also will be better able to distinguish between inherent features of ELE and issues that arise based on particular State choices about how to implement this new option.

Future analyses will both extend and assess the robustness of the first-year findings on administrative costs and enrollment by including additional states in these analyses and focusing on a longer period post-ELE implementation. First-year findings regarding retention rates are particularly limited and also mixed; for example in Iowa Medicaid, retention rates were higher for those entering through ELE, but in Louisiana and Iowa’s separate CHIP program, retention rates were lower. In Alabama, there was no discernible difference in retention between ELE and those who entered through the standard route. It is likely that, in Iowa’s separate CHIP program, the cost-sharing policy in CHIP is affecting retention rates. Because Louisiana changed its policy regarding ELE renewal, moving from affirmative consent to an opt-in policy for data matching, we are eager to study the effect of this state policy change on retention rates there, and will be able to do so in the second year.

In addition, at the Federal level, future analyses will examine how the CHIPRA performance bonuses may have influenced ELE adoption and whether and how this policy might be modified in the future. Congress specified that states that implemented at least five out of eight simplifications (one of which was ELE) and that increased Medicaid enrollment by a specified threshold could qualify for bonus funds. These funds represent a significant Federal investment: more than $500 million has been awarded in the first three years (2009 through 2011) to 23 states. All 9 ELE states are among these 23 states; 3 of them (Georgia, Maryland, and South Carolina) needed ELE to meet the 5 of 8 threshold (the other states would have met the 5 of 8 policy criteria without having ELE in place). The case studies will help us better understand whether the availability of funds acted as an incentive to implement ELE, and further assess whether this new investment in states that implemented ELE is warranted or needs adjustment, given enrollment outcomes using ELE methods.

A further important question for future evaluation work regards ELE’s potential value following implementation of the Affordable Care Act in 2014. States have a compressed timeline with which to prepare to enroll, by Congressional Budget Office estimates, approximately 9 million subsidized individuals through the Affordable Insurance Exchanges and 7 million individuals into Medicaid or CHIP as of 2014, rising to 23 and 10 million, respectively, by 2016 (Congressional Budget Office 2012). Asked about the potential for ELE to benefit states in meetings these targets, administrators who had implemented ELE felt it was too soon to answer this question, but that ELE programs and experiences likely would be useful in the context of Affordable Care Act implementation.

Finally, as the research on this project unfolds, we expect to learn more about whether ELE supports or compromises program integrity. Program integrity involves the incorrect application of a program’s eligibility rules to a particular household. As this is an important policy concern, states will report their error rates to CMS, and these will be included in the final Report to Congress on ELE submitted in September 2013.