The ACA Expands Comprehensive Coverage on Campuses

In 2010, when President Obama signed the Patient Protection and affordable Care Act (ACA) into law,¹ 19.2 percent of higher education students had no health insurance at all.^{Reference Bauer-Wolf2} And many who did have insurance had quite limited coverage, including through campus-sponsored plans. By late 2016, as Donald Trump prepared to take office with a promise to repeal and replace the ACA, the percentage of uninsured and underinsured students had dropped dramatically. At that point, 8.7 percent of students, roughly 1.7 million people, lacked insurance, and for the rest, their coverage options were generally — although not universally — more robust. The uninsured rate had been slashed by more than half for this population of about 20 million people attending community colleges, 4-year institutions, and graduate programs (hereinafter “students”).³

This dramatic change is daylighted in a nationwide study commissioned by the Lookout Mountain Group (LMG) and grounded in census data. Previous studies of student health insurance had considered merely attendees under age 25, or only those at private, residential undergraduate colleges. The LMG study is broader, encompassing private and public institutions, and students of all ages who are seeking associate, bachelors, or graduate degrees. It also stratifies the data in a variety of ways, including by state and by student race, highlighting persistent regional and racial coverage disparities.⁴

Credit for the expansion of campus coverage rests with all the ACA's four M's of Access: the mandate for individuals; the mandate for large employers; the marketplace for subsidized coverage; and, especially it seems, the Medicaid expansion. Thus, campuses provide a fertile ground to consider the ACA's complex, sometimes contradictory, and often controversial impacts on access to health care.

The ACA's individual mandate — rendered toothless and perhaps unconstitutional by a 2017 federal tax law — requires most citizens to maintain qualified health insurance unless exempted (for financial or other reasons) or pay a tax penalty, which is set at $0 starting in 2019.⁵ The individual mandate has been perhaps the most unpopular aspect of the law.⁶ Prior to its adoption, though, many students were subject to a form of it and required to maintain some level of health insurance as a condition of enrollment. In some instances, this was a state-law requirement and in others an institutional decision.

In 2000, for example, the University of California (“UC”) Board of Regents voted to require undergraduate coverage throughout the UC system after a retention study showed that unexpected medical bills were a leading cause of students dropping out.^{Reference Reed7} The Board of Trustees of the California State University (“Cal State”) system, by contrast, declined to adopt that requirement for Cal State students, who tend to come from lower-income families than do UC students.^{Reference Zelman8} This is consistent with a key reason school administrators gave for not favoring this type of mandate in the 2000s: coverage, even limited coverage, could be expensive, and adding that cost on top of tuition might prevent students from going to school, or going to school full-time.

Prior to the ACA, even campuses without a student insurance mandate frequently offered a student-specific plan, though often with limited coverage. Nearly all included lifetime or annual limits. A federal study found that annual limits in 2008 student plans ranged from $15,000 – $250,000, with the median being $50,000.⁹ Many excluded pre-existing conditions (PECs), pregnancy, and serious mental illness; other restrictions included low dollar-limits for prescription medications or outpatient care. Some had moral turpitude exclusions that could, for example, disallow coverage for injury resulting from the use of alcohol.¹⁰ And, of course, a student who became too sick or injured to attend school would typically lose attendance-linked coverage in short order.

The ACA's mandate as to “qualified” coverage caused these plans to improve, but also, at some institutions, to disappear. A late amendment to the ACA allowed schools to continue offering student-specific plans. Aside from this limitation on the pool of purchasers, the ACA's standards for individual insurance apply to the vast majority of these plans.¹¹ Thus, among other requirements, they must be free of PEC consideration, impose no lifetime or annual limits, include a defined set of ten “essential health benefits,” and cover preventive services without copay or coinsurance. This preventive services mandate has been further defined to include such things as cancer screenings, routine vacci-nations, and, controversially for some higher education institutions, coverage of FDA-approved contraceptives.

This contraceptive coverage requirement is one of the reasons some schools cited for dropping their plans altogether. Brigham Young University, for example, dropped its student plan in 2014, noting among other ACA-related issues that birth control coverage would be inconsistent with its expectations of unmarried student behavior.^{Reference Chen12} Other schools stopped selling plans citing the ACA's expansion of other coverage options. The University of Washington and Washington State University, among others, dropped their general student plans in 2014, reasoning that students would find better choices among the ACA's expanded coverage options, and thus shrink the on-campus market, perhaps to unsustainably low levels.^{Reference Blankinship13}

Among those possibly better options is coverage by an employer. The ACA's employer mandate requires large employers either to offer statutorily adequate, affordable insurance to their employees who work at least 30 hours a week, or instead to pay a federal tax penalty.¹⁴ Coverage as an employee is tricky for students. Attending school, particularly full-time, can preclude holding down a job that comes with benefits. Most off-campus part-time or on-campus work-study jobs would not offer health insurance. Teaching or research assistant positions (sometimes offered as part of a graduate student package) are not necessarily considered employment relationships for purposes of insurance coverage.^{Reference Falasco, Jackson and Harris15} A married student might have spousal coverage available; but the ACA does not require employers to offer spousal coverage and, where that coverage is offered, does not require subsidization or otherwise set affordability standards.

Large employers are, though, required to cover children of employees and to do so up to age 26, though, again, without subsidization or affordability standards.¹⁶ This to-age-26 provision polls as among the ACA's most popular.¹⁷ And it was one of the first to go into effect, in mid-2010. Indeed, a federal government report on the uninsured noted that in 2010 only one of its surveyed age groups saw a decline in the percentage without coverage. That group was those 18-24, whose uninsured rate dropped from 29.3% in 2009 to 27.2% in 2010. The report attributed this drop to the to-age-26 provision.^{Reference DeNavas-Walt18}

Particularly on campuses that tend to enroll students from higher income families (where at least one parent works for a large employer), many of the undergraduates and some graduate students will be eligible for coverage on a parent's plan. This would be less true at schools whose enrollment skews older or from lower-income families, where parents are less likely to have employer-provided insurance and less likely to be able to afford the cost of adding a child to a policy. In addition to costs, other issues with adding a student to a parent's plan include network limitations (health providers located near the employer, not the campus) and privacy concerns (with a students' health care bills presumptively being sent to their parents). Ultimately, the LMG study concluded that the to-age-26 provision did not significantly reduce the overall uninsured student rate.¹⁹ Perhaps this is because employer plans are increasingly likely to have high deductibles, narrow networks, and limited subsidies for dependent coverage. A good campus plan could be a better, more affordable option for many.

The ACA's marketplaces (also known as exchanges) provide another option for access to comprehensive health insurance, with no pre-existing condition consideration, regulated benefits, and, crucially, premium subsidies for low- and moderate-income purchasers (as well as cost-sharing reductions for some).²⁰ There was never any doubt that these plans would attract people who were older or sicker than average, particularly if eligible for a premium subsidy. Stability of the marketplaces and affordability (particularly for the unsubsidized), though, depends on attracting the young and healthy in significant numbers as well. A little noted issue is that the to-age-26 provision and the availability of student-only plans drew off young people who might otherwise have chosen a marketplace plan.

Attracting young adults, including those on campuses, was a focus of early advertising campaigns and enrollment efforts. An outreach campaign on Cal State campuses noted that those students were “the low hanging fruit of the uninsured population,” being geographically concentrated, tech savvy, educationally minded, and with their financial data at hand, most having recently filled out financial aid forms.²¹ Chris Walla of the band Death Cab for Cutie visited campuses in Washington State promoting its marketplace where he and his wife, both self-employed, had purchased an unsubsidized plan.^{Reference Hallmark22} Among other efforts targeting young people generally was President Obama's widely viewed appearance on the comedic talk show “Between Two Ferns.”

The LMG study concluded that the marketplaces and the ACA's standards for individually purchased insurance significantly boosted rates of comprehensive coverage on campuses.²³ Indeed, one of the earliest enrollees when the marketplaces first (barely) opened in October 2013 was a 30-year-old 3rd-year law student previously covered by a skimpy high-deductible plan.^{Reference Mahoney24} Student concerns about marketplace coverage reflect those of many purchasers — narrow networks of providers and high out-of-pocket charges. In common with their non-student age mates, they are highly price-sensitive. The Cal State campaign concluded that concerns about affordability rather than perceptions of invincibility were a chief barrier to student purchases on the marketplaces.

Finally, the Medicaid expansion began to go into effect in 2014, though not uniformly around the country. As written, the ACA required states to expand their Medicaid programs from coverage of only categories of low-income people (including children and adults with qualifying disabilities) to coverage of all otherwise uninsured citizens (and authorized immigrants) whose incomes fall below 138 percent of the federal poverty level (FPL). As revised by the Supreme Court's 2012 decision, though, this expansion requirement became an option for the states.²⁵ By 2016, more than half of states had adopted Medicaid expansion; holdouts included populous Texas and Florida, many southern states and several northern ones.²⁶ Pursuant to the ACA framework, the federal government picks up at least 90 percent of the cost for the “expansion population,” an amount that is significantly higher than the federal contribution for the “categorically eligible.” The expansion population comprises low-income adults who do not have access to employer-provided coverage and who are not categorically eligible (because they are not pregnant, disabled, or over age 65). This describes a lot of students.

Indeed, the LMG study suggests that of all the ACA's access provisions, the Medicaid expansion appears to be the most critical factor in reducing the percentage of uninsured students. State variation in adoption of the expansion helped isolate this factor.²⁷ In some states more than 97 percent of students had coverage in 2016, while in others less than 86 percent did.²⁸ This coverage difference is highly correlated with whether the state had adopted Medicaid expansion by the time of the study, including adoption in 2016. Since the time of the study, several more states have adopted Medicaid expansion. In addition to regional variation in coverage, the LMG study also highlights persistent racial variation. Black and Hispanic students were about twice as likely to lack insurance in 2016 as were white students.

Of all the confusing panoply of coverage options that led to the dramatically reduced percentage of uninsured and underinsured students, Medicaid is the most novel. Student-specific plans, employer-sponsored coverage, and individually purchased plans were aspects of campus coverage prior to the ACA, albeit in different forms. Medicaid, though, had only a limited presence. Prior to 2014 (and through at least 2019 in many states, especially in the South) that coverage would be an option only for those low-income students who were pregnant, disabled or (in some states) the parent of a young child. What might it mean that for the foreseeable future this historically stigmatized program is likely to include millions of the most educationally privileged? The following section considers this question.

Students Exemplify the Possibilities and Pitfalls of Medicaid Universality

Writing for the majority in the Supreme Court decision that made the expansion a state option, Chief Justice John Roberts opined that under the ACA's framework as enacted, Medicaid becomes “no longer a program to care for the neediest among us, but rather an element of a comprehensive national plan to provide universal health insurance coverage.”²⁹ Responding to this statement, Justice Ruth Bader Ginsburg's dissent on this aspect of the decision asks rhetorically, “What makes that so? Single adults earning no more than $14,856 per year — 133% of the current federal poverty level — surely rank among the Nation's poor.”³⁰ Both justices' statements are true. To view “the Nation's poor” as within the “neediest among us” as to health care coverage is a conceptual and structural shift from past American practice.

This shift, writes Nicole Huberfeld, reflects a vision of Medicaid universality, a movement “from fragmentation and exclusivity to universality and inclusivity.”^{Reference Huberfeld31} Medicaid universality is grounded in “non-judgmental principles of unification and equalization” and reflects the reality that lack of access to good health care makes anyone vulnerable, regardless of any other aspects of privilege that person might have.^{Reference Huberfeld and Roberts32} In many ways, students — and especially those in graduate professional programs such as law, medicine, and business — are exemplars of the potential of Medicaid universality.

Privilege and vulnerability intersect in this regard on campuses. The ability to pursue higher education is an aspect of privilege. Not only do economic and noneconomic factors limit attendance, but the time to study is itself something of a luxury; and degree-attainment is generally associated with higher social status and higher income. Despite this privilege, all students, even those in the most elite graduate programs, are of course vulnerable to illness and injury, and many enroll with chronic conditions. Justice Ginsburg and her husband were Harvard law students when he was diagnosed with prostate cancer. ^{Reference Lepore33} Justice Sonia Sotomayor's enrollment as an undergraduate at Princeton did not cure her diabetes.^{Reference Sotomayor34} On campuses throughout the country students need medical care, sometimes unexpectedly and expensively.

Inclusion of the educationally privileged as Medicaid beneficiaries highlights several key concerns — public support, provider engagement, and program mainstreaming — central to the successful transition of a program originally targeted at “the neediest among us” into one with more of a universal focus. Medicaid's early advocates hoped that the program would bring the care of the poor from the charity wards into the mainstream of American medicine.^{Reference Stevens and Stevens35} However, the program's categorical coverage and low-payment rates, among other issues, perpetuated a separation, particularly as to physician services.

Medicaid universality theory posits that expansion should facilitate mainstreaming. “Because they are no longer labeled worthy or unworthy of medical assistance, Medicaid patients will not be limited to the obstetrics unit, long term care, or pediatricians' offices. The infiltration of Medicaid patients throughout the health care sector will facilitate integration for the Medicaid population.”³⁶ One concern with this potential is that mainstreaming not come at the expense of those traditionally covered by Medicaid. Low-income, disabled, and elderly people, for example, need and are entitled to broader health benefits (such as nursing home care and non-emergency medical transportation). The fact that the expansion population has health care needs more typical of the general adult population should not lead to the diminishment of services to those with broader health care needs.

Student inclusion could accelerate Medicaid's involvement in the mainstream of medical care. Consider graduate students in the health sciences professions. Might having medical students covered by Medicaid act as a destigmatizing force? Although a tiny percentage of the overall Medicaid population, medical students and other health professions' graduate students could have an outside influence while on campus and later while in practice.

Student participation in Medicaid also highlights some of its access barriers. One is mobility, which is not unique to students. Eligibility is tied to state residency, but students often matriculate at schools away from their family home. If a student is in a state solely to attend school, and does not intend to remain, it might be hard to meet Medicaid residency requirements near their school. Furthermore, there can be education-related reasons why students do not want to change their residency, including maintaining in-state status for subsequent graduate school application, and where scholarship or other educational benefits are tied to state residency. And as with parental employer-based coverage, if a student is covered by Medicaid in their home state but attending school in another, or in-state but distant from their home county, limited services, and perhaps only emergency coverage, might be available nearby. Non-students enrolled in Medicaid can face these mobility issues when job options or family needs augur for a move, particularly to a non-expansion state.

A related access barrier is continuity of coverage. Students tend to live by the academic calendar, with, for example, a summer's gap in attendance, a semester externship, or a quarter break for work. States account for this, or don't, in varying ways. While other enrollees are less likely to be tied to an academic calendar, seasonal work, income variations, and program bureaucracy present continuity challenges of a similar nature.

A key challenge to mainstreaming is the need to expand provider participation. While participation rates have always varied substantially between specialties and localities, most physicians have not accepted Medicaid patients in their private practices or have strictly capped the number they will accept.^{Reference Sparer37} With the high coverage rates among children and pregnant women, pediatricians and obstetricians are among the specialties historically more likely to accept Medicaid. Medicaid's low reimbursement rates are certainly one reason for physician hesitancy, as the ACA recognized by temporarily (for two years) raising payments for primary care and some specialty care to Medicare rates, with the federal government covering the difference.³⁸ Administrative hassles and the challenges of treating often-complex patients are other reasons for physician hesitancy.

Universality theory suggests that the expansion should encourage more physician participation, and student involvement even more so. Might, for example, community college enrollees — with their wide age-range and dispersal throughout many localities — help expand the pool of physicians who accept their coverage? One reason that students might be particularly desirable patients among the general expansion population is their likelihood (perceived or actual) of graduating into a job with employer-provided insurance and its typically much higher physician payments. The extent to which students face difficulties finding Medicaid-accepting physicians depends on a variety of factors including their school's location and whether it has an associated medical school or health system.

An intriguing state innovation further integrates Medicaid into coverage mechanisms for students, though in a way that arguably maintains a separation from the program. Under this innovation, Medicaid covers the premiums for student health-insurance plans for eligible students, and in effect serves as secondary coverage.³⁹ In August 2014, Cornell University worked with New York to pilot one of the first such post-expansion arrangements. One goal was to provide better coverage for the students (including access to more physicians) at the same or lower cost to the government than if they had been fully part of the state's Medicaid program. About two hundred Cornell students were covered the first year, and the numbers have grown since.^{Reference Andrews and Whitman40}

This type of coverage arrangement is now a standard option in Massachusetts and is under consideration in other states.⁴¹ It is conceptually akin to premium assistance arrangements that predated the ACA whereby, for example, the Medicaid program in some states would cover the additional premium charge for employer-sponsored insurance to add an eligible child to a low-income parent's employer plan and would provide additional wrap-around services. The idea is to support consistent, mainstream medical coverage, at least as good as Medicaid, at the same or less cost to the government.^{Reference Alker, Miskell, Musumeci and Rodowitz42} So, while this premium support model is not novel, its broad intersection with higher education students is.

One question is how these types of arrangements and other intersections with Medicaid-supported medical care will influence student connection to the larger program. If 2nd-year law students have difficulty finding Medicaid-accepting mental health providers, will they ally with advocates for medically underserved communities? Or will their access to campus resources shield them from the realities of limited physician availability off-campus? Similarly, students' skills at navigating bureaucratic, computer, and transportation systems put them in a far different position than many disabled Medicaid enrollees for whom those logistical factors can be serious hurdles. And, of course, given the vagaries of program names — “Apple Health,” “Badger-Care,” “Healthy Michigan” and the like — which mostly do not include the word “Medicaid,” students (and others) might not be fully aware they are covered by Medicaid. And where coverage is via financial support for a student plan, ignorance of the Medicaid connection might be more pronounced.

Finally, what might student inclusion mean for public support of Medicaid? If students and their families are at least satisfied with their own public program coverage or grateful for its availability to their fellow students, that might translate into electorally visible support. There is precedent, though inexact, for this within Medicaid's history. When states began adding severely disabled children to their Medicaid programs regardless of parental income, a powerful advocacy network developed that continues to challenge attempted cuts to that coverage. The limited evidence for students' impact here is explored in the final section of this article.

Another possibility is that student inclusion faces political blowback. Covered students might be viewed as exemplars of “voluntary poverty,” and thus unde-serving of government-financed coverage. This narrative would fit within the ongoing, polarized debate in many parts of the country over whether to cover working-age, non-disabled, low-income citizens at all. Said one governor, in a common refrain against expanded Medicaid: “able-bodied adults should be self-reliant.”^{Reference Leonard43} This view animates aspects of the Trump-era health reform retrenchment discussed in the following section.

Trump-Era Retrenchment Threatens Comprehensive Student Coverage

The multifaceted ACA repeal efforts of the Trump era have undercut — though as of mid-2019 not yet destroyed — the law's major access provisions. Complete legislative repeal is off the table with a politically divided Congress. Complete judicial invalidation is a possibility, though perhaps unlikely.^{Reference Ober-lander44} Partial executive branch abrogation of the ACA, though, persists on a multitude of fronts.^{Reference Sanford45}

While many aspects of the Trump administration's health reform retrenchment have the potential to impact student coverage, two initiatives are particularly noteworthy: Medicaid eligibility restrictions and cheaper, skimpier plan offerings. Both are being challenged in court. And both, if allowed to proceed, rely on individual state action for implementation, thus ensuring variable consequences across the country.

With the Trump administration's strong encouragement, several states have applied for Medicaid waivers that would, among other limitations, condition coverage for many people on ongoing documentation of work or “community engagement activities.” As explained in federal guidance to state Medicaid directors, this new policy invites Section 1115 waivers to test these and other eligibility-limiting conditions on “non-elderly, non-pregnant adult Medicaid beneficiaries who are eligible for Medicaid on a basis other than disability.”⁴⁶ This encompasses the bulk of the expansion population, including students. In non-expansion states, this requirement would apply mostly to mothers of minor children. These types of waivers are a distinct move away from Medicaid universality, back towards a stigmatizing division of the poor into the deserving and undeserving, and incorporating bureaucratic hurdles that could hinder even the eligible.

Kentucky's waiver was the first to be approved by the federal government. Kentucky had expanded its Medicaid program in 2014 under Governor Steven Beshear, a Democrat, and saw a drop in its uninsured rate from 16.3 percent in 2013 to 5.1 percent in 2016. One of the most dramatic uninsured reductions in the country, this drop was attributable largely to the expansion. In 2015, Matt Bevin, a Republican, campaigned on a promise to repeal the expansion and won the governor's seat.^{Reference Goodnough47}

At Governor Bevin's direction, the state submitted a Section 1115 waiver request that would, among other provisions, require most non-disabled Medicaid beneficiaries ages 19 to 64 to provide documentation every month that their work or recognized community engagement activities totaled at least 80 hours. Those who did not meet that documentation requirement would lose their coverage and be locked out of regaining coverage for several months. Pregnant women, the medically frail, and primary caretakers of minor children or disabled adults could obtain an exemption from the requirement, as could full-time students. Part-time students could count their schooling as part of “community engagement,” in addition to paid employment, volunteer work, and specified types of caregiving.

Kentucky estimated that about a quarter of its Medicaid population would be subject to these requirements, and that, within 5 years of the new rule, 95,000 fewer Kentuckians would be covered by Medicaid, many of whom would satisfy the new coverage conditions but would fail to provide appropriate documentation.⁴⁸ CMS approved the waiver, and the new requirements were set to phase in beginning July 2018.

Shortly before then, a federal district court blocked Kentucky's plan, finding that the administration's approval was arbitrary and capricious because it did not adequately consider whether the plan would “help the state furnish medical assistance to its citizens, a central objective of Medicaid.”^{Reference Watson49} Pointing to Kentucky's estimate (which the plaintiffs argued was low), the court concluded that “[a]t bottom, the record shows that 95,000 people would lose Medicaid coverage, and yet the secretary [of Health and Human Services] paid no attention to that deprivation.”⁵⁰

The administration has vowed to press on with granting these waivers, and litigation will inevitably continue. By mid-2019, waiver requests of this type were pending as to six states, and had been approved for seven, though two of those (Kentucky's and Arkansas's) had been judicially set aside and implementation was on hold in a few of the others.⁵¹

The state plans differ in their details, but all likely would result in less Medicaid coverage for students, particularly those who attend part-time. One issue is whether student status can confer an exemption or instead would be counted towards “community engagement” hours, with its added documentation burden. Under Kentucky's waiver, full-time attendance warrants an exemption; under Indiana's, which was also approved, part-time attendance may also satisfy exemption criteria under specified circumstances. However, under New Hampshire's approved waiver, school attendance would be considered as part of the required 100 community engagement hours per month to qualify for Medicaid coverage.⁵²

Another issue is a potential preference for employment-focused education. Kansas, for example, lists as activities that could be counted towards the monthly hours' requirement “vocational education” (which includes study towards a two-year degree), “job skills training directly related to employment” and “education related to employment.”⁵³ In the same vein, Maine's application states that to count towards the required hours, “[t]he goal of education must be employment.”⁵⁴ Exactly what these details mean in practical terms might not be clear until implementation is well under way, if that happens. Students who juggle work or family obligations are among those likely to be dissuaded from Medicaid coverage where these waivers are in effect. Indeed, a few of the plain-tiffs in the Kentucky case are full- or part-time students, including a law student who has children and a student-husband with fluctuating work hours.⁵⁵

The impact on student coverage nationwide is difficult to predict. This is partly because if waivers of the type approved by the administration survive judicial review, expanded Medicaid might be available in more states, as these coverage restrictions might make expansion more politically and financially palatable. However, students would likely find it more onerous to obtain and maintain coverage given the documentation requirements. Of course, Medicaid enrollees without students' ready computer access face potentially greater documentation challenges.

Students are also likely to be impacted by the administration's multi-faceted promotion of inexpensive, limited-coverage plans. Rules effective beginning in the fall of 2018, for example, could vastly expand access to short-term, limited duration plans (STLDs).^{Reference Keith56} Absent state regulation, STLD plans do not need to comply with key ACA requirements. Instead, STLDs can exclude coverage for maternity, prescription medications, and mental health treatment (among other “essential benefits”), can have lifetime and annual limits, and can consider a person's health status in determining whether to offer a plan, its benefits, and its terms.

Under the Obama-era regulations, STLD plans could last for only three months and could not be renewed. Under the Trump-era regulations, STLD plans can last for 364 days and can be renewed — at the insurer's option — two times (for a total term of just under three years). In addition, nothing in the rules prevents a person from thereafter buying a new STLD plan, making these skimpier plans a potentially ongoing source of coverage. (Though, of course, pregnancy, disease, or injury might mean that a new STLD plan is unavailable, unaffordable, or unhelpful.)

Some states have prohibited or limited the sale of STLD plans. And the 2018 federal rules that vastly expanded their potential availability have been challenged in court.⁵⁷ If STLDs and other newly authorized non-ACA compliant plans survive the legal challenges and become widely available, healthy students are likely to seek them out. This is particularly likely where expanded Medicaid is not an option. Given their coverage limitations, STLD plan premiums likely would be much less expensive than a student plan or an unsubsidized marketplace plan (and less than many subsidized marketplace plans). If students switch to these cheaper, skimpier plans in any significant numbers, the costs for health insurers and the insureds remaining on the comprehensive, ACA-compliant plans are sure to rise, perhaps to unsustainable levels.

Thus, the student coverage experience reflects, in microcosm, concerns about the sustainability of the health insurance system for individually insured and for low-income people throughout the country. The pending Medicaid restrictions and limited plan options impact students in ways that are both unique and of relevance to the broader population.

Campus Coverage Experience Portends Support for Medicare Commonality

The student coverage experience also bears on emerging proposals to expand Medicare. Whether as an offering on the individual marketplaces, a choice for those over age 55, the default for all, or some other variant, Medicare as a common option is poised to be a key question for the next wave of health reform proposals. The 2020 election seems certain to feature debate about various Medicare commonality ideas, government-grounded coverage generally, and, yet again, the ACA.

The student coverage experience could impact this debate in at least two ways: students' Medicaid- and marketplace-coverage opinions could influence their own voting behavior; and their coverage decisions could highlight or exacerbate problems that spur other voters. Overall, the student coverage experience and its reverberating effects portend support for nascent versions of Medicare commonality.

There is scant evidence regarding student opinions about Medicaid or marketplace coverage for themselves or their classmates. If most are at least satisfied with Medicaid (or wish it were an available option in their state), that would presumably translate to a generally favorable view of government-grounded coverage. And if unhappy with it, to a generally unfavorable view. The same is likely true as to marketplace coverage, although perhaps with less force because — despite being federally regulated for all and federally subsidized for most — it is more identifiably private coverage.

National polling does show increasing support over the past several years for the government's role in this arena, with particularly high levels of support among the student-age population.^{Reference Kiley58} While public opinion is interesting, it is voting that really matters for legislative change. Americans between the ages of 18 and 29 vote at much lower rates than do those in older age groups.⁵⁹ Furthermore, voting rates among low-income people have historically been low, and not nudged upwards by government programs intended to benefit them. In fact, there is some evidence that enrollment in stigma-tized government programs, such as cash welfare, correlates with decreased voting rates.^{Reference Clinton and Sances60}

Medicaid expansion, though, seems different. According to three recent studies, where Medicaid is expanded more people register to vote and more people do vote, at least in the election immediately following the expansion.^{Reference Baicker, Finkelstein, Haselswerdt and Sanger-Katz61} How they vote is unknown. The expansion population is younger and more likely to be single than the general adult population, though, and conventional wisdom suggests that both circumstances correlate now with stronger support for Democratic candidates.

None of the three studies conclusively identifies reasons for the increased electoral participation. There are at least a few plausible theories why this low-income program might have differing voting impacts than previous low-income programs. First, considering targeted impacts, Medicaid enrollment is associated with increased depression treatment and improved financial stability, both of which could make participation easier.⁶² Poor health, particularly as related to depression or mobility limitations, is associated with decreased electoral participation.⁶³ If enroll-ees, because of new health coverage, are healthier and more financially secure, they might be more likely to vote. In addition, their registration to vote might have been facilitated by access to registration materials at the time of Medicaid enrollment, whether online or in person.

Second, recognizing generalized reactions, this embrace of a key aspect of the ACA could inspire registration and voting by opponents (as well as proponents) who are not Medicaid enrollees. Kentucky provides an example. Under a Democratic governor it expanded its Medicaid program in 2014 and saw a dramatic drop in its uninsured rate; the following year, a Republican campaigned on a promise to repeal the expansion and won the governor's office.⁶⁴ Although central aspects of the ACA are popular, the law overall remains highly polarizing and it has been a key issue in several recent elections.

The idea that it is non-enrollees who are boosting the voting rates is undercut somewhat by the one study of the three mentioned above that focused narrowly on Medicaid enrollees' registration and voting patterns. That study, grounded in Oregon's pre-ACA lottery system for low-income adult coverage, found that enrollees were more likely to register and to vote, at least in the national election immediately after their enrollment.⁶⁵

Third, theorizing about motivational forces, enrollment might boost enrollees' appreciation of government programs and thus their interest in electoral politics. This jibes with older studies of universal, nonstigmatized federal programs such as Social Security and the G.I. Bill, both of which increased voting rates.⁶⁶ If students tend to view Medicaid favorably, and through a lens of universality, they might well be receptive to policy proposals that would expand government-grounded health care, and to vote accordingly. In addition, their electoral impacts could be magnified because higher education and higher income have historically correlated with higher voting rates. Furthermore, over the past couple of national elections, including the 2018 mid-terms, voter turn-out was notably up among millennials generally and students specifically.^{Reference Thomas67}

Beyond student perceptions, student actions might lend credence to arguments for Medicare commonality. As described in the prior section, the Trump administration in 2018 authorized and began promoting the sale of inexpensive plans with limited coverage and pre-existing condition exclusions. It is difficult to predict how widely available these plans will be, particularly given the ongoing legal challenges and the efforts of some states to limit their availability.

What is predictable, though, is that where cheap, skimpy plans are available, budget-conscious, healthy students are likely to seek them out. The individual marketplaces are already attracting fewer people in their 20s and 30s than expected, thus skewing the risk pool towards those who are older and more expensive to insure. In 2015, for example, per capita health care costs for people between 55 and 64 years of age were $9,707 compared with $4,442 for those between 26 and 44, and $2,915 for those between 19 and 25.^{Reference Bodenheimer68}

If the cheaper, skimpy plans cause students and others in their age cohort to further abandon the individual marketplaces in significant numbers, premium costs for comprehensive, ACA-compliant individual insurance plans could rise substantially. Purchasers of any age who receive a premium subsidy would be largely shielded from this cost increase (though the federal government, which pays the subsidy, would not be). Subsidies are available for those who make less than four times the poverty level; in 2019 this equates to about $50,000 for an individual or $100,000 for a family of four. Purchasers who make more than this amount, though, will bear the full brunt of premium increases. And those in their 50's and early 60's will be particularly impacted, as older purchasers in most states can be charged up to three times as much as younger purchasers for the same insurance product.

Older individual marketplace purchasers, whether subsidized or not, are already a target demographic for many calls to make Medicare a more commonly available option. Under one line of thinking, if middle-aged people could buy into or otherwise be absorbed into Medicare, their own premiums should go down, as should those of younger people who remain in the individual coverage pool. Of course, this could have its own problematic repercussions. Without middle-aged people, the resulting smaller market might be less desirable to insurers, thus raising premiums and limiting coverage options.^{Reference Blumberg, Holahan and Sanger-Katz69} Making Medicare a more common option (or the common option) involves multiple tradeoffs. Solving problems and not merely redistributing them will continue to be a challenge in the ongoing American quest to expand access to health care, improve its quality, and reduce its cost.