In December 2017, the Republican Congress, working with the Trump administration, repealed the tax penalties enforcing the Affordable Care Act’s (ACA) individual mandate, effective in 2019. Although the degree of the mandate’s efficacy is uncertain, its repeal is sure to lead to additional Americans going without coverage, exacerbating the instability that now affects the individual insurance markets of many states.

In this context, it is incumbent on federal and state policymakers to enact replacement policies for the ACA’s individual mandate. More fundamentally, the individual market’s core dysfunction remains in place: Many young and healthy consumers remain outside the market, increasing average risk levels and premiums, and leaving millions of people needlessly uninsured. Those enrollment choices involve the perception and the reality of unaffordable coverage, but they also reflect burdens of enrollment and plan selection that have an outsize impact on program participation.

@musiena via Twenty20

Similar dynamics extend beyond the individual market, leaving many people uninsured despite qualifying for Medicaid or employer-sponsored insurance. Most people who remain uninsured today qualify for insurance, often at very low cost to themselves.

One promising strategy to address these challenges involves the use of automatic, default enrollment into insurance. An automatic enrollment program can improve participation by creating options that require little or no premium payment and that require very little effort from the consumer—apart from providing consent, perhaps through failing to opt out.

We propose an approach aimed at making enrollment into insurance as automatic as possible. This will be a complex undertaking. Nonetheless, once it is up and running, we believe this approach can dramatically improve enrollment into insurance, and thus help to stabilize the market and make it more attractive for all consumers.

Past Use Of Automatic Enrollment

401(k) Enrollment

Making participation rather than non-participation the default option has greatly increased take-up in contexts outside health coverage, illustrating the surprisingly significant impact of lifting small procedural barriers to enrollment. One classic example involves 401(k) retirement savings accounts. In companies where new employees must complete a form to establish such accounts, roughly one-third enroll within six months. By contrast, in firms where new employees are automatically enrolled unless they complete opt-out forms, 90 percent join.

Medicare Part B

Health programs have also used automatic enrollment to achieve high take-up levels. Perhaps the best known example involves Medicare Part B, which historically achieved 96 percent participation levels. People turning 65 are automatically enrolled by default, unless they object. Part B premium payments are withheld from Social Security checks.

Medicare Part D

Within six months of the Medicare Part D prescription drug benefit becoming available for enrollment in January 2006, the Centers for Medicare and Medicaid Services (CMS) achieved remarkable success, enrolling 74 percent of eligible seniors in the low-income subsidy (LIS) component of the program. Only 14 percent of eligible seniors completed applications for enrollment, however. The others were auto-enrolled based on data matches with state Medicaid programs and the federal Supplemental Security Income (SSI) program. Medicare beneficiaries continue to qualify automatically for LIS assistance based on their receipt of Medicaid or SSI during the previous year.

Louisiana Express Lane Eligibility

In 2010, Louisiana implemented Express Lane Eligibility (ELE), an option that lets states base Medicaid on the eligibility determinations of other need-based programs. Unlike other states that reached many fewer children because they required parents to complete forms requesting coverage, Louisiana used largely default-enrollment methods to provide children with Medicaid when their families participated in the Supplemental Nutrition Assistance Program (SNAP).

Only 1 percent of families whose children received SNAP but not Medicaid opted out of ELE. The remainder were sent Medicaid cards, which were automatically activated upon first use. Nearly 30,000 children received health coverage, substantially cutting the state’s already low percentage of uninsured children. After initial enrollment, 83 percent of ELE children used Medicaid to access care within a year—only slightly below the 88 percent of children who enrolled in Medicaid through other channels.

When information technology problems forced Louisiana to change its enrollment method to require parents to check an opt-in box on the SNAP application form, ELE enrollment fell by 62 percent.

Lessons Learned

Based on this prior experience, several features appear essential to the effective use of default or automatic enrollment:

• Eligibility criteria are structured to fit available data, so additional information or other action from the individual is not required before coverage begins.
• Either default enrollees are not required to make payments or the administrative entity doing the enrollment can make payments on the consumer’s behalf (e.g., employers’ paycheck withholding of workers’ 401(k) contributions and the Social Security Administration’s withholding of Medicare Part B premium payments from social security checks).
• Default choices are believed to match the preferences of most affected consumers, with consumer gains significantly exceeding costs.

Maryland Tests The Possibilities Of Default Enrollment In The Individual Market

Meeting these criteria in the context of individual-market coverage is not easy, but policymakers in Maryland have been pushing the boundaries of what is possible under current federal law. Legislation introduced in Maryland, the “Protect Maryland Health Care Act of 2018,” would use the state’s income tax system to replace federal enforcement of the ACA’s individual mandate. Rather than simply impose a penalty, the legislation would encourage the uninsured to convert their penalties into “down payments” to buy health insurance, whenever possible.

Roughly 100,000 Maryland adults who were uninsured in 2016 could obtain Exchange coverage at zero additional premium beyond the applicable premium tax credit (PTC) plus the payment owed because of coverage gaps the previous year, according to a Families USA analysis of 2016 data from the American Community Survey and premium information from Maryland’s Exchange. More than two-thirds of these 100,000 consumers could purchase gold plans with deductibles of $1,500 or less. Roughly two-thirds are adults under age 45, and 39 percent are under age 35, suggesting that their addition to the individual market could lower overall risk levels and unsubsidized premiums. However, several legal and policy obstacles prevented legislators from proposing full automatic enrollment:

• State mandate enforcement means that consumers who were uninsured the previous year identify themselves on state income tax returns. However, many who were uninsured the previous year have coverage by the time they file tax returns. The Maryland legislation therefore requires formerly uninsured tax filers to indicate whether they remain uninsured before tax-based enrollment begins.
• Tax return information provides most of the information needed to determine an uninsured taxpayer’s eligibility for PTCs. But the state revenue agency cannot disclose return information to the Exchange without legal authorization. The Maryland legislation thus requires uninsured taxpayers to authorize disclosure of relevant return information before the Exchange moves forward.
• State tax returns do not provide all of the information needed to determine PTC eligibility. The Exchange can obtain some of the missing data elements based on matches from third-party data sources, authorization for which is provided by uninsured consumers on their tax returns. But other items may need to be furnished by the consumer, either on the tax return or through later provision of information to the Exchange.
• Advance PTCs are needed for taxpayers to enroll in coverage at zero additional premium cost. However, PTC claimants must file later federal tax returns that reconcile advance payments with annual circumstances shown on the return. Excess advance PTC claims can thus lead to federal tax debts. The Maryland proposal accordingly requires consumers, before enrollment, to be informed of and acknowledge the need for tax reconciliation, the risk of year-end federal tax debts, and the obligation to notify the Exchange of changes in mid-year household circumstances.

The Legislature was unable to complete action on this complex, groundbreaking proposal during the brief 2018 legislative session, but it is a leading item on the agenda of a bipartisan legislative working group preparing legislation for action in early 2019. More broadly, Maryland’s initial efforts illustrate what states can do under current federal law, as well as operational barriers that could be lowered through federal policy intervention.

A Federal Framework With Multiple Options For States And Employers

In addition to other steps to stabilize markets and increase coverage, the federal government could establish an overall framework for states to experiment with more frictionless enrollment into insurance. Flexibility is key, given the many challenges ahead and the need to garner lessons learned that can inform future policy. Such a federal framework could address the individual market, employer coverage, and Medicaid/CHIP.

The Individual Market: Increased State Flexibility

Several possible changes to federal law would increase state capacity to use default enrollment to provide individual market coverage to uninsured residents, including young and healthy adults whose participation would improve risk pools and lower unsubsidized premiums. First, states could be given the option to base eligibility for federal financial assistance on prior-year tax returns, eliminating the risks that auto-enrollment would otherwise create with later reconciliation. Other programs use similar eligibility methods to simplify enrollment, including federal grants and loans for post-secondary-education, tax rebates in the 2008 stimulus bill, and means-tested variations in Medicare Part B and Part D premiums. In these programs, consumers can obtain additional assistance by demonstrating a recent drop in income, but increases to current-year income affect future assistance levels without requiring subsidy repayment.

Second, federal law could increase state flexibility to base eligibility determinations on reliable sources of objective data. For example, lawmakers could more clearly authorize states to determine eligibility for PTCs (premium tax credits) (and Medicaid) based entirely on third-party data sources, without obtaining affirmative attestations from consumers. Congress could authorize the IRS to share federal income tax data with states to help administer auto-enrollment programs, provided that affected tax filers give consent. With other sources of reliable and relevant data, federal legislation could permit their use to determine eligibility for subsidized coverage, so long as affected consumers receive clear notice and a chance to opt out of data sharing. The Medicaid statute already permits such an opt-out approach to obtaining information needed to verify eligibility, with guardrails to protect privacy and data security.

Third, Congress could give states more access to data about health coverage. This would help states target the uninsured for auto-enrollment and prevent public funds from being wasted to cover consumers who already have insurance. For example:

• Federal law could give states access to existing data sources that identify people with coverage. Such sources include Medicare coordination-of-benefits records, information returns that carriers and sponsors provide under Internal Revenue Code Section 6055, coverage data provided by private contractors to HHS for purposes of verifying eligibility for special enrollment periods, and third-party-liability data sources for Medicaid. Currently, such data are available for specified purposes that do not include helping states focus enrollment efforts on people who are uninsured.
• The federal government could create an exemption from ERISA allowing states to compel self-insured plans to provide coverage information to the state.
• The federal government could incorporate information about recipients of employer-based coverage into the National Database of New Hires (NDNH) and make NDNH accessible both for determining eligibility for subsidies and helping state auto-enrollment programs.
• Federal agencies could give states access to information about who receives coverage through federal employment, Medicare, other states’ Medicaid and CHIP programs, or the federally operated healthcare.gov website.

Fourth, the federally facilitated marketplace that serves most of the country could create options for data exchange and plan choice for states interested in collaborating around default-enrollment strategies.

Finally, states could receive the option to let consumers of all ages use PTCs to auto-enroll into catastrophic plans, letting more people receive coverage that costs no more than PTC amounts. Currently, PTCs may not be used with such plans, and they are generally limited to people under age 30. Other plans could potentially be offered at even lower net cost to the consumer, bringing entirely PTC-funded coverage within reach of additional uninsured.

Enrollment in high-deductible plans improves the individual market’s overall risk pool and shields enrollees from catastrophic costs, but such plans provide much more limited access to care than is typically furnished by group coverage, and many enrollees may not perceive their high-deductible plans as valuable. Reasonable people can disagree about the desirability of high-deductible plans—including bronze coverage—but several steps could mitigate concerns about their use for default enrollment.

As a starting point, states can improve such plans’ short-term usefulness by encouraging or requiring high-deductible plans used for default enrollment to offer significant coverage of pre-deductible services, including generic medications and visits with primary care providers. In addition, federal law could permit carriers to let default enrollees quickly move up to silver coverage by combining PTCs with additional household premium payments. This might require revising current federal limits on changing plans outside open and special enrollment periods. Moreover, default enrollment into high-deductible plans, based on zero additional premium cost beyond PTC amounts (plus payments, if any, required by the state because of coverage gaps the previous year), could be limited to consumers who: (1) were offered more generous plans and chose not to enroll; (2) are not offered more generous plans at zero additional premium cost; and (3) can opt out of the default high-deductible option, either before or within a defined period following auto-enrollment.

The Individual Market: A Supportive Federal Structure

Beyond increasing state flexibility, federal lawmakers could create a basic structure to support state auto-enrollment efforts. Short-term federal grants, perhaps with modest state matching requirements, could fund state policy planning and initial development of information technology infrastructure. The federal government could also fund independent studies that inform future policy choices about whether and, if so, how to pursue auto-enrollment.

To qualify for these federal resources, states could be asked to submit an outline describing key elements of the proposed auto-enrollment strategies, such as:

• Venue For Auto-Enrollment. Will the state use its income-tax system as the place to identify the uninsured and provide them with coverage? Will health care providers begin auto-enrollment for uninsured patients (perhaps leveraging systems hospitals already use to retroactively enroll uninsured patients into Medicaid)? Will auto-enrollment efforts begin when drivers’ licenses are renewed or when workers are laid-off from jobs that provide employer-based coverage? Will the state proactively identify the uninsured through analysis of coverage data and initiate auto-enrollment based on such identification?
• What data will the state access to determine eligibility for assistance? How will the state prevent default enrollment from reaching consumers who already have insurance? Will the state create new databases to supplement existing data sources? How will data privacy and security be protected? What notice will the state provide consumers? What opt-ins, opt-outs, and defaults will apply? How will consumers be able to review and correct their data, securely and easily?
• Financial Assistance. What financial assistance will pay for coverage? How will consumers be warned about or shielded from risks associated with claiming assistance, such as the risk of federal tax liability for advance PTC claims that turn out to be excessive? Will states impose a tax payment on consumers who fail to maintain coverage, and thus provide added incentive for insurance enrollment, or will they rely on automatic enrollment to boost coverage? If states impose a penalty for going without coverage, will they allow payments of those penalties to be applied toward enrollment in insurance?
• Health Plan Selection. What health plans will be used for auto-enrollment? If more than one plan is available, how will the default plan be chosen? How will the state promote consumers’ receipt of coverage they value? Will consumers have options to make plan choices before auto-enrollment or to change or drop coverage after auto-enrollment?

As federal and state policymakers decide on their approach to auto-enrollment, several trade-offs are important to evaluate:

• Enrollment Versus Privacy. Requiring consumers to affirmatively waive privacy protections before a state accesses data establishing eligibility for assistance, for example, may protect privacy at the expense of coverage. Further analysis may be needed to assess the actual preferences of affected consumers and determine which default settings make consumers better off “as judged by themselves.”
• High-Deductible Coverage. Default enrollment in the individual market reaches more people, all else equal, if such enrollment can put people into high-deductible plans. Some observers are concerned about the limited access to care such coverage provides, but few would argue that going completely uninsured is better.

Employer Coverage

While most working Americans and their families readily sign up for employer-sponsored insurance, some do not. According to a study published by the Kaiser Family Foundation, 3.7 million people were uninsured in 2016 despite offers of employer-sponsored coverage that disqualified them from subsidized health insurance in the ACA exchanges. In a separate study, based on survey data from employer plans, Kaiser found that 78 percent of workers whose employers offered coverage accepted those offers.

Federal law could make clear that employers, under certain conditions, have the flexibility to automatically enroll their employees into coverage. Employers electing this option would need to ask workers in advance if they have other coverage. Those who state that they are uninsured could be enrolled into the employer plan unless they opt out, with worker premium shares withheld from paychecks.

Employers could limit these default options to plans with minimal worker premium costs. Workers would need the right to decline coverage after their employers notify them of enrollment. This option would mirror some employers’ current practice of automatically enrolling their employees into pension or retirement savings arrangements unless workers opt out.

Medicaid And CHIP

Federal policymakers could take several steps to let states automate and otherwise streamline enrollment into Medicaid and CHIP when reliable sources of data show that consumers qualify. States could receive the flexibility to enroll such consumers by default, without requiring an affirmative request for coverage. If a state picks this option, consumers would need to receive notice and a chance to opt out before coverage begins. As with Medicaid’s current procedures for ex parte/administrative renewal, the notice could inform beneficiaries of their obligation to notify the state if household circumstances change in ways that may affect eligibility.

Congress could also give states the flexibility to use Express Lane Eligibility with adults as well as children. This would let states qualify families for health coverage when SNAP or other benefit programs have already found them to meet income and other eligibility requirements, notwithstanding small technical details about how different programs define households and measure income. When children and adults qualify based on ELE, states could be given the flexibility to enroll them by default, so long as they do not opt out. The expedient originally used by Louisiana—where eligible families showed affirmative consent by using a Medicaid card to seek care—would no longer be required if a state used opt-out rather than opt-in enrollment procedures for people who qualify through ELE.

What A State’s Automatic Enrollment Program Might Look Like In Practice

With the above changes to federal law, here are examples of how a state might facilitate automatic enrollment.

Tax-Based Enrollment

The tax-based component of a state auto-enrollment plan might unfold as follows:

• Tax Season. The state could require residents to disclose health insurance status during the prior calendar year as part of filing state income tax returns. By checking a single box, tax filers could authorize a state health agency to obtain all necessary information to determine eligibility for free or low-cost insurance and to enroll uninsured household members into coverage through Medicaid, CHIP, or private marketplace coverage. The state has the option to impose taxes on those who were uninsured the prior year.
• Immediate Data Cross-Check And Eligibility Determination. For tax filers who check the authorization box, the state would access coverage data and use information from the tax return as well as other third-party sources to identify uninsured tax filers and determine their eligibility for Medicaid, CHIP, and PTCs. Until the state has built a system for compiling and vetting coverage data, tax filers could be asked to indicate on their return whether they are uninsured at the time the return is filed.
• Rapid Enrollment. Those found eligible for Medicaid and CHIP would be enrolled. Those who filed tax returns by April 15 and qualify for PTCs might have a brief special enrollment period in which they could choose from any available QHPs (qualified health plans). Those who fail to enroll during that period and who are offered QHPs costing no more than PTCs (plus tax payments for prior-year coverage gaps, in a state imposing such payments) would be auto-enrolled.
  To increase the number of residents with auto-enrollment options, the state might make catastrophic-level plans available for PTC use by auto-enrollees of all ages. To lessen consumer risks and reduce the need for affirmative consent, the state could base PTC eligibility on prior-year tax returns, without any need for later reconciliation on federal income tax returns. Coverage would begin at the earliest practical date, given standard rules about coverage-effective dates.
• November-December Open Enrollment After Tax Time. Individuals who were auto-enrolled into a QHP could change plans during open enrollment. If they did not make a selection, they would remain enrolled in the same QHP, following normal default renewal rules.

This cycle would repeat itself each year, with refinements over time making the process more seamless, less costly, and with fewer mistakes and errors.

What If The State Does Not Use Or Wants To Supplement Tax-Based Enrollment?

If the state has no income tax or hesitates to use tax filing for this purpose, it could develop automatic enrollment systems that find uninsured residents through other means. For example, a state could:

• Proactively reach out to residents who are reliably identified as uninsured based on a state-developed comprehensive, regularly updated coverage database;
• Leverage existing state functions that “touch” numerous residents, such as automobile registration or drivers’ license renewal, by asking about health insurance and requiring all uninsured consumers to say whether they want their contact information forwarded to the state health agency to see if they qualify for free or low-cost health insurance; and
• Creating systems through which health care providers can easily and quickly enroll their uninsured patients into coverage, perhaps building on successful state initiatives to let hospitals and clinics act as their patients’ authorized representatives in signing up for health coverage. Newborn children could be a particular focus of hospital-based enrollment, given the surprisingly large percentage who lack coverage—5 percent of all children under age 1, a much higher proportion than for any other age group.

Even a state that uses its income tax system as a pathway to enrollment could supplement that effort with one or more of these other channels.

Employment-Based Enrollment

Auto-enrollment into employer-sponsored insurance could accompany any of the above methods for identifying uninsured state residents and signing them up for coverage. With hiring that can take place at any time, employers could offer automatic enrollment of new employees throughout the calendar year.

Looking Forward

We are convinced that auto-enrollment could greatly reduce the number of uninsured while lowering individual-market premiums substantially by increasing the participation among the young and healthy uninsured. Among other gains, such measures could more than offset the effects of ending the ACA’s individual mandate enforcement and help remedy some of the core dysfunctions plaguing the individual market. We hope that future analysis and state-level implementation will test these hypotheses and assess the magnitude of the effects we foresee.

The operational challenges of implementing auto-enrollment strategies would be significant. Numerous details, such as those involving immigration status and citizenship, would be critically important to design with great care. Nevertheless, state health officials have a track record of remarkable creativity, persistence, and effectiveness, given the right tools. We encourage federal policymakers to give such tools to their state colleagues so that the whole country can learn from state innovation in this promising and important arena.