Medicaid

The 2010s have been a momentous decade for Medicaid. With enrollment of over seventy-two million people (19% of the country’s population), Medicaid is the nation’s largest public health insurance program,[footnote]Ctrs. for Medicare & Medicaid Servs., July 2019 Medicaid & CHIP Enrollment Data Highlights, Medicaid.gov, https://www.medicaid.gov/medicaid/program information/medicaid-and-chip-enrollment-data/report-highlights/index.html [https://perma.cc/ZV8L-FUXQ] (last visited Oct. 4, 2019).[/footnote] and it is the primary or sole source of health insurance for vulnerable groups such as low-income children and pregnant women, adults with disabilities, and people in need of long-term care.[footnote]See 42 U.S.C. § 1396a(a)(10)(A) (Supp. V 2017) (defining mandatory populations for Medicaid coverage).[/footnote] Since 2014, the pendulum of Medicaid policy has swung from an unprecedented expansion of coverage under the Affordable Care Act (ACA), toward more recent federal regulations and state policy innovations that are instead predicted to limit uptake of benefits. These developments have raised questions about the purposes and scope of the Medicaid program, which are central to ongoing litigation over state Medicaid waivers. Under § 1115 of the Social Security Act, states can seek approval from the Secretary of Health and Human Services (HHS) to implement demonstration projects in Medicaid programming.[footnote]Social Security Act § 1115, 42 U.S.C. § 1315 (2012) (explaining requirements for waivers of state plans based on “experimental, pilot, or demonstration project[s]”).[/footnote] States have previously used this waiver pathway to expand coverage and benefits, to incorporate incentives for healthy behaviors, to charge copayments and premiums, and to make delivery system modifications.[footnote]See Elizabeth Hinton, MaryBeth Musumeci, Robin Rudowitz, Larisa Antonisse & Cornelia Hall, Kaiser Family Found., Section 1115 Medicaid Demonstration Waivers: The Current Landscape of Approved and Pending Waivers app. A (2019), https://www.kff.org/report-section/section-1115-medicaid-demonstration-waivers-the-current-landscape-of-approved-and-pending-waivers-appendices [https://perma.cc/BNC3-T3BA]; Laura Hermer, On the Expansion of “Welfare” and “Health” Under Medicaid, 9 St. Louis U. J. Health L. & Pol’y 235, 237–41 (2016).[/footnote] Beginning in 2018, however, HHS has approved numerous state demonstration projects that make eligibility for some beneficiaries contingent on work requirements: the completion of monthly quotas of work, education, or volunteering hours to stay on Medicaid.[footnote]The first such waiver was in Kentucky; HHS subsequently approved work requirement waivers in Arizona, Arkansas, Indiana, Kentucky, Michigan, New Hampshire, Ohio, and Utah. See Medicaid Waiver Tracker: Approved and Pending Section 1115 Waivers by State, Kaiser Family Found. (Sept. 18, 2019), https://www.kff.org/medicaid/issue-brief/medicaid-waiver-tracker-approved-and-pending-section-1115-waivers-by-state/#Table2 [https://perma.cc/D8ST-4SC2] [hereinafter Kaiser Family Found., Medicaid Waiver Tracker].[/footnote] Although work requirements have long been a prerequisite for accessing federal cash welfare[footnote]See Heather Hahn, David Kassabian & Sheila Zedlewski, Urban Institute, TANF Work Requirements and State Strategies to Fulfill Them 2 (2012), https://www.acf.hhs.gov/sites/default/files/opre/work_requirements_0.pdf [https://perma.cc/ET6Q-E9QT] (describing work requirements in the Temporary Assistance for Needy Families program, which were established at the program’s inception in 1996 and modified by subsequent legislation including the Deficit Reduction Act of 2005 and the American Recovery and Reinvestment Act of 2009).[/footnote] and nutritional assistance,[footnote]See Able-Bodied Adults Without Dependents (ABAWDs), Food & Nutrition Serv., U.S. Dep’t of Agric., https://www.fns.usda.gov/snap/ABAWD [https://perma.cc/KX6X-2VVZ] (last visited Oct. 4, 2019) (describing work requirements for able-bodied adults without dependents using the Supplementary Nutritional Assistance Program).[/footnote] their use in Medicaid is a historic first.[footnote]Lola Fadulu, Why States Want Certain Americans to Work for Medicaid, Atlantic (Apr. 12, 2019), https://www.theatlantic.com/health/arcarch/2019/04/medicaid-work-requirements-seema-verma-cms/587026/ [https://perma.cc/7CW4-AZJJ]; Hinton et al., supra note 4, at 1 (“In January 2018, CMS posted new guidance to allow state Section 1115 waiver proposals to condition Medicaid on meeting a work requirement and subsequently has approved the first waivers of that type in the history of the Medicaid program.”).[/footnote] Program beneficiaries have challenged HHS’s approvals, alleging in part that work requirement waivers are unlikely to advance the purposes of the Medicaid program.[footnote]See Philbrick v. Azar, No. 19-773 (JEB), 2019 WL 3414376, at *1 (D.D.C. July 29, 2019); Stewart v. Azar (Stewart II), 366 F. Supp. 3d 125, 145 (D.D.C. 2019); Gresham v. Azar, 363 F. Supp. 3d 165 (D.D.C. 2019); Stewart v. Azar (Stewart I), 313 F. Supp. 3d 237 (D.D.C. 2018).[/footnote] As a result, these cases squarely ask the court to identify Medicaid’s purposes. Although the statute specifies that the program aims to “furnish medical assistance,” HHS has proffered other goals, including the promotion of public health. Thus far, the District of D.C. has relied on the narrower interpretation to strike down waivers in Kentucky,[footnote]See Stewart II, 366 F. Supp. 3d at 125; Stewart I, 313 F. Supp. 3d at 237.[/footnote] Arkansas,[footnote]See Gresham, 363 F. Supp. 3d at 179.[/footnote] and New Hampshire,[footnote]See Philbrick, 2019 WL 3414376, at *11.[/footnote] finding that the Secretary did not adequately consider whether work requirement waivers are likely to advance the standalone goal of furnishing medical assistance. These decisions have explicitly found that promoting public health is not a freestanding purpose of Medicaid programming.[footnote]See, e.g., id. (reiterating that the court “has found that health is not a freestanding objective of the Medicaid Act”); Stewart II, 366 F. Supp. 3d at 145 (“[T]he Court finds that health is not a freestanding objective of the statute.”); Gresham, 363 F. Supp. 3d at 179 (citing the findings from Stewart I that “the agency’s ‘focus on health is no substitute for considering Medicaid’s central concern: covering health costs’” and that “HHS has offered no argument here that calls those conclusions into question”).[/footnote] The stakes of this interpretive choice are greater than the fate of work requirements. An enormous body of literature in public health documents the influence of social determinants of health: the broad conditions of poverty, social inclusion, housing, nutrition, neighborhood safety, discrimination, instability, economic opportunity, and education that shape health outcomes.[footnote]See Sandro Galea, Melissa Tracy, Katherine J. Hoggatt, Charles DiMaggio & Adam Karpati, Estimated Deaths Attributable to Social Factors in the United States, 101 Am. J. Pub. Health 1456, 1456, 1461–62 (2011) (calculating relative risks, prevalence, and estimated deaths resulting from factors such as low educational attainment levels, racial segregation, low social support, individual-level poverty, income inequality, and area-level inequality); U.S. Burden of Disease Collaborators, The State of US Health, 1990–2016: Burden of Diseases, Injuries, and Risk Factors Among US States, 319 JAMA 1444, 1444, 1451 (2018) (reporting trends on leading risk factors for death and disease); Samantha Artiga & Elizabeth Hinton, Beyond Health Care: The Role of Social Determinants in Promoting Health and Health Equity, Kaiser Family Found. (May 10, 2018), https://www.kff.org/disparities-policy/issue-brief/beyond-health-care-the-role-of-social-determinants-in-promoting-health-and-health-equity/ [https://perma.cc/Y8XG-RZH5] (providing an overview of social factors that contribute to disparate health outcomes and surveying existing federal, state, and nongovernmental initiatives that seek to address them).[/footnote] Indeed, these risk and protective factors may account for a far greater share of our morbidity and mortality than access to medical care.[footnote]See Steven A. Schroeder, We Can Do Better—Improving the Health of the American People, 357 New Eng. J. Med. 1221, 1221 (2007) (arguing that “the pathways to better health do not generally depend on better health care” but rather on addressing nonbehavioral determinants of health).[/footnote] In recent years, states have begun to test pathways to use federal and state Medicaid funds to intervene in these determinants, and they have often done so by waiver. Using mechanisms such as § 1115 programming and incentives for Medicaid managed care organizations, states have found creative ways to address risk and protective factors beyond the doctor’s office, including paying for improvements in housing, nutrition, and linkages to social services.[footnote]See infra Part I.[/footnote] These programs may be more tenuous, however, if health promotion is excluded from the reading of Medicaid’s purposes. On this interpretation, state Medicaid waivers may hold only if they advance the provision of medical assistance, which may trade off against using program resources to address upstream determinants of health. This Piece argues that a narrow construction of Medicaid’s purposes may have unanticipated consequences for future Medicaid waivers, inadvertently limiting the long-term capacity of Medicaid programming to address social and structural influences on health. The promise of public insurance programs like Medicaid for intervening in these determinants is already bounded by available funds, political will, and the limits of federal versus state power in a jointly administered program. But judicial precedents disallowing health promotion as a goal of Medicaid programming may compound these barriers, further complicating efforts to leverage Medicaid to alleviate structural risks. This Piece considers tradeoffs in setting the scope of Medicaid and other public health insurance programs. Using public health insurance to purchase health, and not just medical assistance, may require a fundamental reconceptualization of programs like Medicaid and Medicare. This Piece begins with a brief overview of social determinants of health and ways in which states have begun to address these influences through Medicaid waiver programming. Subsequent sections will consider work requirement waivers and ongoing litigation parsing the purposes of the Medicaid statute. Concluding remarks will consider whether public health insurance can and should serve as a vehicle for addressing social determinants of health. I. USING MEDICAID TO ADDRESS SOCIAL DETERMINANTS OF HEALTH Over the past half-century, public health scholars have amassed extensive documentation of social and structural determinants of health.[footnote]See generally World Health Org., Social Determinants of Health: The Solid Facts (Richard Wilkinson & Michael Marmot eds., 2d ed. 2003); Michael Marmot, Sharon Friel, Ruth Bell, Tanja A.J. Houweling & Sebastian Taylor, Closing the Gap in a Generation: Health Equity Through Action on the Social Determinants of Health, 372 Lancet 1661 (2008); Michael Marmot, Social Determinants of Health Inequalities, 365 Lancet 1099 (2005); Sandra Putnam & Sandro Galea, Epidemiology and the Macrosocial Determinants of Health, 29 J. Pub. Health Pol’y 275 (2008).[/footnote] These determinants are many but include factors such as affordable high-quality housing; neighborhood factors such as violence, transportation, and walkability; social experiences such as discrimination, segregation, stress, and social support; socioeconomic status; education; employment opportunities; access to healthy food; and environmental hazards such as poor air quality and unsafe water.[footnote]See Paula Braveman, Susan Egerter & David R. Williams, The Social Determinants of Health: Coming of Age, 32 Ann. Rev. Pub. Health 381, 384–89 (2011); Galea et al., supra note 14, at 1462–63; Artiga & Hinton, supra note 14, at 2.[/footnote] Together, these determinants are powerful influences on morbidity and mortality. A much-quoted analysis has estimated that only 10% of health status is attributable to healthcare access, while 20% arises from social and environmental exposures, 30% from genetics, and 40% from behavior.[footnote]Schroeder, supra note 15, at 1222.[/footnote] When we consider that behavior is also in part determined by exposure to social and structural risk factors,[footnote]See Paula Braveman & Laura Gottlieb, The Social Determinants of Health: It’s Time to Consider the Causes of the Causes, 129 Pub. Health Rep. 19, 22–24 (2014) (discussing research showing that exposure to social factors and stressors can influence behaviors that in turn influence health outcomes).[/footnote] the role of social determinants looms even larger. The failure to address many of these health determinants can help to explain what Elizabeth Bradley and Lauren Taylor have called the “American health care paradox”: How is it that we spend so much more on care but purchase health outcomes that are mediocre or worse than outcomes in comparably wealthy countries?[footnote]Elizabeth H. Bradley & Lauren A. Taylor, The American Health Care Paradox: Why Spending More Is Getting Us Less 2–3 (2013) (describing the “spend more, get less” paradox long studied in U.S. health policy). See generally Nat’l Research Council & Inst. of Med., U.S. Health in International Perspective: Shorter Lives, Poorer Health (Stephen H. Woolf & Laudan Aron eds., 2013) (comparing U.S. health outcomes against other high-income countries and finding a pattern of higher mortality and inferior health in nine categories).[/footnote] As these authors and others have deduced, the paradox in large part derives from our failure to pay for “services that address the broader determinants of health.”[footnote]Bradley & Taylor, supra note 21, at 3.[/footnote] These include improvements in “housing, nutrition, education, the environment and unemployment support,”[footnote]Id. at 12.[/footnote] among others. These services fall outside traditional conceptions of healthcare, and thus outside definitions of covered benefits for private and public health insurance. As Bradley and Taylor have argued: [Americans] continue to pay top dollar for hospitals, physicians, medications, and diagnostic testing yet skimp in broad areas that are central to health, such as housing, clean water, safe food, education, and other social services. It may even be that Americans are spending large sums for healthcare to compensate for what they are not paying in social services—and the trade-off is not good for the country’s health.[footnote]Id. at 15–16.[/footnote] Expenditures on social determinants of health are not covered benefits for most healthcare payors; indeed, they fall well outside what is traditionally considered to be medical care. For both private policies and public insurance programs in the United States, coverage of services depends largely on whether the care is deemed medically necessary.[footnote]See Kristen Underhill, Paying for Prevention: Challenges to Health Insurance Coverage for Biomedical HIV Prevention in the United States, 38 Am. J.L. & Med. 607, 647–49 (2012) (providing background on insurers’ reliance on medical necessity determinations).[/footnote] Contract terms limiting health insurance coverage to “medically necessary” care arose in private insurance plans after World War II, and Medicare took up the language in 1965.[footnote]Id. at 647.[/footnote] Medicare currently excludes coverage for any care that is “not reasonable and necessary” for diagnosis, treatment, prevention, and palliation.[footnote]42 U.S.C. § 1395y(a)(1)(A)–(E) (Supp. V 2017).[/footnote] Medicaid benefits are defined more specifically by statute[footnote]Id. § 1396d(a) (defining “medical assistance”).[/footnote] and set forth in detailed regulations by HHS,[footnote]42 C.F.R. §§ 440.1–440.185 (2018) (defining mandatory Medicaid benefits).[/footnote] including categories such as inpatient hospital care, outpatient services, laboratory and imaging services, and “other diagnostic, screening, preventive, and rehabilitative services.”[footnote]See id. § 440.130; 42 U.S.C. § 1396d(a)(1)–(3), (13); see also Mandatory & Optional Medicaid Benefits, Medicaid.gov, https://www.medicaid.gov/medicaid/benefits/list-of-benefits/index.html [https://perma.cc/J6EL-S534] (last visited Oct. 6, 2019) (listing Medicaid benefits).[/footnote] States must cover mandatory benefits but may also receive federal matching funds for optional benefits such as dental care and prescription coverage.[footnote]See Robin Rudowitz, Kendal Orgera & Elizabeth Hinton, Kaiser Family Found., Medicaid Financing: The Basics 1 (2019), http://files.kff.org/attachment/Issue-Brief-Medicaid-Financing-The-Basics [https://perma.cc/X62A-QLE8] (providing an overview of Medicaid financing).[/footnote] For adults covered under the ACA Medicaid expansion, states must cover ten categories of care known as “essential health benefits”;[footnote]See Affordable Care Act, Pub. L. No. 111-148, § 2001(a)(2), 124 Stat. 271, 271–72 (2010); see also id. § 1302, 124 Stat. at 163–68 (defining essential health benefits).[/footnote] these are the same categories of care that must be covered by private plans eligible for ACA tax subsidies,[footnote]See Ian Spatz & Michael Kolber, The Future of Essential Health Benefits, Health Affairs (Feb. 14, 2017), https://www.healthaffairs.org/do/10.1377/hblog20170214.058765/full/ [https://perma.cc/E3MG-LPU5] (describing how private plans interact with the ACA’s essential health benefits).[/footnote] and they include care such as ambulatory patient services, emergency care, hospitalization, maternity and newborn care, mental health and sub-stance use disorder services, and preventive and wellness services.[footnote]42 U.S.C. § 18022(b)(1)(A)–(J) (Supp. V 2017).[/footnote] Given the definitions that structure mandatory benefits, essential health benefits, and medical necessity clauses, there may be little interpretive room to extend health insurance coverage to services such as housing subsidies, educational supports, and environmental improvements. But in recent years, states have begun experimenting with Medicaid waiver programming to pay for interventions that improve social determinants of health. States have considered or used multiple mechanisms to achieve these goals, including § 1115 experimental waivers, a subset of § 1115 programs known as Delivery System Reform Incentive Payment (DSRIP) programs, and incentives for Medicaid managed care organizations.[footnote]States may also address some social determinants of health for disabled populations in § 1915(c) waivers, which pay for home supports to enable in-home care for patients who would otherwise be institutionalized. See Home & Community-Based Services 1915(c), Medicaid.gov, https://www.medicaid.gov/medicaid/hcbs/authorities/1915-c/index.html [https://perma.cc/CL49-APXE] (last visited Oct. 6, 2019); see also Deborah Bachrach, Jocelyn Guyer & Ariel Levin, Milbank Mem’l Fund, Medicaid Coverage of Social Interventions: A Road Map for States 7–8 (2016), https://www.milbank.org/wp-content/uploads/2016/09/MMF-NYS-Health-Issue-Brief-FINAL.pdf [https://perma.cc/K48G-4M4Q] [hereinafter Bachrach et al., Road Map for States] (discussing § 1915(c) waivers). Section 1945(c)(1) also provides some leeway for mitigating social risks. The statute authorizes payments for health homes for people with chronic disease, which should include referrals to “community/social supports.” Medicaid.gov, Health Homes (1945 of SSA/ Section 2703 of ACA) Frequently Asked Questions Series II 4–5, 12 (2015), https://www.medicaid.gov/state-resource-center/medicaid-state-technical-assistance/health-home-information-resource-center/downloads/health-homes-section-2703-faq.pdf [https://perma.cc/CCM3-4HTY].[/footnote] A. § 1115 Waivers Under § 1115 of the Social Security Act, states can seek relief from certain federal requirements for a period of up to five years to implement experiments, pilots, or demonstrations in their Medicaid programming.[footnote]42 U.S.C. § 1315(a), (e)(6).[/footnote] Waivers qualify for approval if the experimental program is “likely to assist in promoting the objectives” of the statute, with an additional agency requirement of being budget-neutral for the federal government.[footnote]42 U.S.C. § 1315; see also Sidney D. Watson, Out of the Black Box and into the Light: Using Section 1115 Medicaid Waivers to Implement the Affordable Care Act’s Medicaid Expansion, 15 Yale J. Health Pol’y L. & Ethics 213, 214 (2015) [hereinafter Watson, Black Box] (discussing Section 1115 waivers). Under current budget neutrality rules, the waiver must not increase costs to the federal government, relative to what the federal government would spend in ordinary Medicaid programming. See Letter from Timothy B. Hill, Acting Dir., Ctrs. for Medicare & Medicaid Servs. to State Medicaid Dirs. 1–2 (Aug. 22, 2018), https://www.medicaid.gov/federal-policy-guidance/downloads/smd18009.pdf [https://perma.cc/7GJN-AHJ4] [hereinafter Hill, Budget Neutrality Policy Letter].[/footnote] Although the language of § 1115 suggests time-limited and incremental experiments, these waivers have increasingly functioned as programmatic waivers, with “experimental” programs lasting decades.[footnote]See U.S. Gov’t Accountability Office, GAO-18-249, Medicaid Demonstrations: Evaluations Yielded Limited Results, Underscoring Need for Changes to Federal Policies and Procedures 12 (2018), https://www.gao.gov/assets/690/689506.pdf [https://perma.cc/DH2X-4L3B] (noting that some states have operated portions of their Medicaid program as demonstrations for decades); Watson, Black Box, supra note 37, at 215 (“Some waivers have continued for decades with no public evaluation of their impact on Medicaid access, cost, or quality.”).[/footnote] The section of the Medicaid statute that outlines mandatory benefits is not part of the Secretary’s § 1115 waiver authority,[footnote]See Watson, Black Box, supra note 37, at 225 (“[R]equests to reduce benefits implicate statutory provisions in Section 1937, and are therefore outside the Secretary’s Section 1115 authority to waive provisions in Section 190.”).[/footnote] so waiver programs cannot cover less than the mandatory set of services.[footnote]See id.[/footnote] But commentators have suggested that states can use § 1115 waivers to obtain Medicaid matching funds to cover more.[footnote]See, e.g., Barak D. Richman, Behavioral Economics and Health Policy: Understanding Medicaid’s Failure, 90 Cornell L. Rev. 705, 762–63 (2005) (“Through Section 1115 waivers, states can move away from simple insurance expansions and may instead pursue social policies that appreciate the behavioral component of health care consumption.”).[/footnote] Should states seek to use Medicaid funds to reduce social risks to health—such as, for example, by investing in neighborhood safety programs in areas with a high proportion of residents on Medicaid, or by subsidizing the purchase and storage of healthy food—approving this coverage by waiver is likely within the Secretary’s discretion. Waiver authority is broad and judged on an abuse of discretion standard, whereby waiver approvals are found to be arbitrary and capricious if “the agency ‘entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.’”[footnote]Stewart II, 366 F. Supp. 3d 125, 135 (D.D.C. 2019) (quoting Motor Vehicle Mfrs. Ass’n, Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)).[/footnote] To justify the grant of a waiver, HHS must “examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.”[footnote]Id. (internal quotation marks omitted) (quoting State Farm, 463 U.S. at 43).[/footnote] These standards should be straightforward for § 1115 waivers seeking to finance interventions into social determinants of health. Epidemiological studies have documented the pathways between social risk factors and health outcomes, and evidence for interventions that address these risks is growing.[footnote]See, e.g., Clare Bambra, Marsha Gibson, Amanda Sowden, Kath Wright, Margaret Whitehead & Mark Petticrew, Tackling the Wider Social Determinants of Health and Health Inequalities: Evidence from Systematic Reviews, 64 J. Epidemiology Community Health 284, 285–90 (2010) (discussing research on interventions related to the social determinants of health); David R. Williams, Manuela V. Costa, Adebola O. Odunlami & Selina A. Mohammed, Moving Upstream: How Interventions that Address the Social Determinants of Health Can Improve Health and Reduce Disparities, 14 J. Pub. Health Mgmt. Prac. S8, S8–S14 (2008) (same).[/footnote] Testing these interventions in Medicaid would be supportable by relevant data and seems to be a sound use of waiver authority. There are few other legal barriers to using § 1115 waivers to advance social interventions. Medicaid benefits are set forth at some length by statute, which may tend to imply that Congress deliberately excluded nonenumerated benefits from coverage;[footnote]This is known as the expressio unius canon of statutory interpretation, whereby a statute that expresses one thing is read to imply that other things are excluded. See Richard A. Posner, Statutory Interpretation—In the Classroom and the Courtroom, 50 U. Chi. L. Rev. 800, 805 & n.25 (1983).[/footnote] one covered category, however, consists of “any other medical care, and any other type of remedial care recognized under State law, specified by the Secretary.”[footnote]42 U.S.C. § 1396d(a)(29) (Supp. V 2017).[/footnote] If the Secretary wished to approve Medicaid waiver financing for innovative social interventions, HHS could make a colorable claim that waiver authority and the Secretary’s own judgments about covered care are sufficient. Indeed, the statutory provision allowing for § 1115 waivers notes that the “costs of such project[s] which would not otherwise be a permissible use of funds . . . shall to the extent and for the period prescribed by the Secretary, be regarded as a permissible use of funds under such part.”[footnote]42 U.S.C. § 1315(a)(2)(B) (2012).[/footnote] The stumbling block for such waivers is likely to be budget neutrality, but assuming some structural interventions are indeed cost-effective, § 1115 waivers attempting to alleviate structural risk are legally feasible if they align with CMS’s priorities. To date, states have not made full use of § 1115 waiver proposals to address social determinants of health, despite calls to do so.[footnote]See, e.g., Richman, supra note 41, at 762–64 (encouraging the use of Section 1115 waivers by states); Bachrach et al., Road Map for States, supra note 35, at 7–8.[/footnote] Possible innovations under such waivers, however, are highly attractive. For example, surveys show a high prevalence of food insecurity in the Medicaid population, and SNAP benefits do not extend to everyone eligible for Medicaid.[footnote]See, e.g., Am. Heart Ass’n, Policy Position Statement: Expanding Access to Healthy Food for Medicaid Beneficiaries 2, 5 (2019), https://www.heart.org/-/media/files/about-us/policy-research/policy-positions/access-to-healthy-food/expanding-access-to-healthy-food-for-medicaid-beneficiaries–policy-statement-2019.pdf [https://perma.cc/EJ4S-LHGT]; see also Deborah Bachrach, Jocelyn Guyer, Sarah Meier, John Meerschaert & Shelly Brandel, Enabling Sustainable Investment in Social Interventions: A Review of Medicaid Managed Care Rate-Setting Tools 3 (2018), https://www.commonwealthfund.org/sites/default/files/documents/___media_files_publications_fund_report_2018_jan_bachrach_investment_social_interventions_medicaid_rate_setting.pdf [https://perma.cc/MM94-UCAV] [hereinafter Bachrach et al., Enabling Sustainable Investment] (noting that although about 70% of adults eligible for Medicaid in expansion states are below the federal poverty level, only about half receive SNAP benefits).[/footnote] States might respond by seeking § 1115 approval to spend on the subsidization and storage of healthy food.[footnote]See Bachrach et al., Enabling Sustainable Investment, supra note 49, at 8 (outlining states’ options for using § 1115 waivers to use Medicaid strategies to invest in social interventions that might offer help with, among other things, food assistance).[/footnote] Medicaid is an important support for people experiencing homelessness or unsafe housing; perhaps states might adopt “housing first” models that spend on supportive services, or they may seek to defray the costs of mold remediation or lead paint abatement.[footnote]See Amanda Cassidy, Medicaid and Permanent Supportive Housing 3 (2016), https://www.healthaffairs.org/do/10.1377/hpb20161014.734003/full/healthpolicybrief_164.pdf (on file with the Columbia Law Review) (identifying “a variety of strategies on how states can use Medicaid funding to support individuals who are or have been homeless and emphasize options for covering housing-related services under Medicaid”).[/footnote] At the neighborhood level, Medicaid might also pay for housing inspectors or remediation teams. For individuals living with energy insecurity, and who are exposed to extremes of heat and cold in their homes—exacerbating problems of asthma, stress, and poor mental health[footnote]See Diana Hernández, Understanding ‘Energy Insecurity’ and Why It Matters to Health, Soc. Sci. & Med., Oct. 2016, at 6 (reporting study results demonstrating the negative health consequences of home energy insecurity).[/footnote]—another possible use of § 1115 programming could spend Medicaid funds to supplement assistance from the Low Income Home Energy Assistance Program (LIHEAP).[footnote]LIHEAP Fact Sheet, Office of Cmty. Servs. (Nov. 16, 2018), https://www.acf.hhs.gov/ocs/programs/liheap/about [https://perma.cc/83SF-27LX] (noting that the objective of LIHEAP is “[t]o assist households with low incomes, particularly those with the lowest incomes that pay a high proportion of household income for home energy, primarily in meeting their immediate home energy needs”).[/footnote] Coverage of GEDs, childcare assistance, and transportation subsidies may also pay off in long-term health improvements. Oregon, for example, implements Medicaid through contracts to organizations that function as both insurers and providers, and the program serves the goal of “addressing the social determinants of health” by directly covering “community-level interventions” and services such as housing.[footnote]See Bachrach et al., Enabling Sustainable Investment, supra note 49, at 9.[/footnote] This style of waiver might also move the focus from individuals to communities, funding interventions at the community level rather than services for individuals. Consider the Michigan effort to address harms inflicted by the lead contamination crisis in Flint. Michigan recently sought a § 1115 waiver to expand Medicaid eligibility for people who were exposed to the Flint water supply, up to 400% of the federal poverty level (FPL).[footnote]Michigan’s Medicaid Section 1115 Waiver to Address Effects of Lead Exposure in Flint, Kaiser Family Found. (Mar. 7, 2016), https://www.kff.org/medicaid/fact-sheet/michigans-medicaid-section-1115-waiver-to-address-effects-of-lead-exposure-in-flint [https://perma.cc/6A8T-MHEF].[/footnote] But instead of paying for medical care, it may have been a better use of funds to pay for earlier action that would have mitigated or prevented the contamination. In cities or communities with high proportions of Medicaid beneficiaries, Medicaid funds could purchase outsize returns in public health by detecting and reducing water contamination by lead, hazardous chemicals, or bacteria.[footnote]See Bradley & Taylor, supra note 21, at 15 (noting that approximately 15% of the U.S. population in 2009 was exposed to contaminated drinking water).[/footnote] Similarly, funds could be directed to other neighborhood-level hazards, such as violence prevention or the construction of safe spaces for physical activity and social interaction.[footnote]See, e.g., Diana Hernández, Affording Housing at the Expense of Health: Exploring the Housing and Neighborhood Strategies of Poor Families, 37 J. Fam. Issues 921, 922 (2016) (discussing violence-avoidance and coping strategies employed by low-income families who confront housing and neighborhood hardships); Karyn A. Tappe, Karen Glanz, James F. Sallis, Chuan Zhou & Brian E. Saelens, Children’s Physical Activity and Parents’ Perception of the Neighborhood Environment: Neighborhood Impact on Kids Study, Int’l J. Behav. Nutrition & Physical Activity, Mar. 2013, at 7 (reporting a positive correlation between neighborhood safety from crime and children’s use of public recreation spaces).[/footnote] One objection to waivers in this style is that they lead to positive externalities that cannot be recaptured by the Medicaid program. For example, intervening at the neighborhood level to promote social inclusion and prevent violence may have benefits for residents who are not Medicaid beneficiaries. Assuming budget neutrality, however, positive externalities need not disqualify waivers from approval. Other provisions of the Medicaid statute, such as the requirement that state Medicaid programs pay hospitals Disproportionate Share Hospital (DSH) funds for uncompensated care burdens,[footnote]See Medicaid Disproportionate Share Hospital (DSH) Payments, Medicaid.gov, https://www.medicaid.gov/medicaid/finance/dsh/index.html [https://perma.cc/QG7G-PBY3] (last visited Oct. 4, 2019) (“Federal law requires that state Medicaid programs make [DSH] payments to qualifying hospitals that serve a large number of Medicaid and uninsured individuals.” (emphasis added)).[/footnote] or the section providing healthcare providers with high proportions of Medicaid caseloads with incentives to adopt electronic health records,[footnote]42 U.S.C. § 1903(a)(3)(F), (t) (Supp. V 2017); see also Ctrs. for Medicare & Medicaid Servs., Medicaid Electronic Health Record Incentive Payments for Eligible Professionals 1 (2013), https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Downloads/MLN_MedicaidEHRProgram_TipSheet_EP.pdf [https://perma.cc/DB7D-47MA] (describing the Medicaid Electronic Health Record (EHR) Incentive Program, which “provides incentive payments for Medicaid eligible professionals (EPs) who adopt, implement, upgrade, or meaningfully use certified EHR technology”); Federal Match Rates for Medicaid Administrative Activities, Medicaid & CHIP Payment & Access Comm’n, https://www.macpac.gov/federal-match-rates-for-medicaid-administrative-activities/ [https://perma.cc/8M5Q-XFRT] (last visited Oct. 4, 2019) (compiling federal compensation schedules for various Medicaid activities).[/footnote] have similar benefits for third parties. B. DSRIP Waivers Unlike direct programming—whereby states directly finance investments in social determinants of health—DSRIP waivers work by giving states money to incentivize activity by healthcare providers, typically hospitals.[footnote]See Michael K. Gusmano & Frank J. Thompson, Medicaid Delivery System Reform Incentive Payments: Where Do We Stand? Health Aff.: Blog (Sept. 28, 2018), https://www.healthaffairs.org/do/10.1377/hblog20180920.103967/full [https://perma.cc/DYD2-J37G] [hereinafter Gusmano & Thompson, Where Do We Stand?].[/footnote] DSRIP programs are a subset of § 1115 waivers that have been in use since California first proposed one in 2010.[footnote]Alexandra Gates, Robin Rudowitz & Jocelyn Guyer, An Overview of Delivery System Reform Incentive Payment (DSRIP) Waivers 3 (2014), http://files.kff.org/attachment/an-overview-of-dsrip [https://perma.cc/P8JD-TDGZ].[/footnote] The goal of DSRIP programs is to hold out Medicaid funds to incentivize providers to make improvements in healthcare delivery[footnote]Felicia Heider, Tina Kartika & Jill Rosenthal, Nat’l Acad. for State Health Policy, Exploration of the Evolving Federal and State Promise of Delivery System Reform Incentive Payment (DSRIP) and Similar Programs 5 (2017), https://www.macpac.gov/wp-content/uploads/2018/03/Exploration-of-the-Evolving-Promise-of-DSRIP-and-Similar-Programs.pdf [https://perma.cc/6YFS-CWV5].[/footnote] and population health;[footnote]See Gates et al., supra note 61, at 3 (“At the highest level, DSRIP waivers are designed to advance the ‘Triple Aim’ of improving the health of the population, enhancing the experience and outcomes of the patient and reducing the per capita cost of care.”).[/footnote] about a dozen states have now implemented a program as part of a § 1115 waiver.[footnote]See Medicaid & CHIP Payment & Access Comm’n, Delivery System Reform Incentive Payment Programs 1 (2018), https://www.macpac.gov/wp-content/uploads/2018/03/Delivery-System-Reform-Incentive-Payment-Programs.pdf [https://perma.cc/MGG7-KVAE] [hereinafter MACPAC, Delivery System Reform]; Gusmano & Thompson, Where Do We Stand?, supra note 60.[/footnote] Payments to providers under these programs are tied to specific milestones, which can include health outcomes.[footnote]Medicaid & CHIP Payment & Access Comm’n, Delivery System Reform, supra note 64, at 5.[/footnote] States like New Jersey and Kansas, for example, have offered providers incentives to reduce diabetes complications, post-surgical complications, and emergency department visits.[footnote]See Michael K. Gusmano & Frank J. Thompson, An Examination of Medicaid Delivery System Reform Incentive Payment Initiatives Under Way in Six States, 34 Health Aff. 1162, 1165–66 (2015).[/footnote] In California, DSRIP programming provides incentives for hospitals to improve health outcomes, with a particular focus on sepsis infections and conditions including HIV and asthma.[footnote]See Gates et al., supra note 61, at 6.[/footnote] New York’s program incentivizes providers to address substance use prevention and mental health, along with projects focusing on HIV, cardiovascular health, perinatal health, diabetes, palliative care, asthma, renal care, and care for women and children.[footnote]Id. at 7.[/footnote] New York’s program also involves the work of a governmental task force, the Medicaid Redesign Team, which coordinated the use of state Medicaid funds to provide supportive housing for 4,500 beneficiaries at high risk for homelessness and poor health outcomes.[footnote]See Kelly M. Doran, Elizabeth J. Misa & Nirav R. Shah, Housing as Health Care—New York’s Boundary-Crossing Experiment, 369 New Eng. J. Med. 2374, 2374–75 (2013).[/footnote] In 2012 and 2013 respectively, the state budget allocated $75 million and $86 million in state (not federal) Medicaid funds to purchase housing for beneficiaries with especially high healthcare costs.[footnote]Id. at 2375.[/footnote] As the project leaders wrote in the New England Journal of Medicine, “We envision a Medicaid system in which spending on social determinants of health is not only allowable, but recognized as a best practice.”[footnote]Id. at 2376.[/footnote] CMS has not approved a new DSRIP program since the change of administration and in 2018 directed Texas to prepare a phase-out plan to develop delivery system reforms without DSRIP funding.[footnote]See Eliot Fishman, Sinsi Hernández-Cancio & Ellen Albritton, Families USA, Financial Support for Safety Net and Small Community Providers to Participate in Delivery System Reform: Medicaid-Based Options for States 2 (2018) (on file with the Columbia Law Review); Tex. Health & Human Servs., Draft Delivery System Reform Incentive Payment (DSRIP) Transition Plan 2–3 (2019), https://hhs.texas.gov/sites/default/files/documents/laws-regulations/policies-rules/1115-waiver/waiver-renewal/draft-dsrip-transition-plan-cms.pdf [https://perma.cc/EB7H-99HX] (describing the mandate to develop a transition plan to phase out DSRIP funding); Letter from Angela D. Garner, Dir., Div. of Sys. Reform Demonstrations, Ctrs. for Medicare & Medicaid Servs., to Stephanie Muth, Assoc. Comm’r, Medicaid/CHIP, Tex. Health & Human Servs. Comm’n 41 (Feb. 14, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/tx/tx-healthcare-transformation-ca.pdf [https://perma.cc/47MC-YB9E] (specifying the end of DSRIP funding).[/footnote] But the viability of population health efforts under the DSRIP program is an intriguing proof of concept for waivers that seek to address social determinants of health, and future administrations may revive the DSRIP tool for this purpose. C. Medicaid Managed Care Organization (MCO) Incentives Like DSRIP waivers, managed care contracts offer another means of promoting health indirectly, namely by incentivizing managed care organizations to take on projects that address social determinants.[footnote]See Bachrach et al., Enabling Sustainable Investment, supra note 49, at 6 (“States can elect to use incentive payments to reward plans that perform well on quality metrics related to social issues and/or that make use of value-based payments.”); see also Bachrach et al., Road Map for States, supra note 35, at 8 (“MCOs may determine to cover additional social services—i.e., those not covered under the MCO contract—in order to reduce the cost and improve the quality of care.”).[/footnote] Many states operate Medicaid through managed care organizations—insurance companies that are paid a per-patient (capitated) rate to provide coverage for Medicaid benefits.[footnote]Bachrach et al., Road Map for States, supra note 35, at 8.[/footnote] Managed care organizations have incentives under capitated rates to cover any care that is cost effective, which might include services such as supportive housing for beneficiaries most at risk.[footnote]See id. at 8, 11–13 (describing the use of Medicaid MCO contracts to incentivize improvements in housing for covered beneficiaries and summarizing examples in various states); see also Bachrach et al., Enabling Sustainable Investment, supra note 49, at 7 (describing Arizona’s practice of making state housing grants available to regional behavioral health authorities, and requiring those regional authorities to reinvest 6% of their profits in community projects like food banks or housing).[/footnote] As Deborah Bachrach and colleagues have noted, “While states cannot direct plans to invest in non-Medicaid social supports, they can indirectly encourage such investments by linking incentive and withhold payments to outcomes that can be improved by offering social supports.”[footnote]Bachrach et al., Enabling Sustainable Investment, supra note 49, at 11.[/footnote] A number of states—including at least Colorado, Michigan, Maine, and California—have made use of this strategy, often encouraging managed care contractors to provide case management services that connect beneficiaries to social supports and help to navigate public assistance for housing, energy, and other needs.[footnote]Bachrach et al., Road Map for States, supra note 35, at 9–10.[/footnote] In one example, a Medicaid managed care organization in Minnesota achieved a 9% reduction in emergency department visits by connecting patients with housing specialists.[footnote]Dawn E. Alley, Chisara N. Asomugha, Patrick H. Conway & Darshak M. Sanghavi, Accountable Health Communities—Addressing Social Needs Through Medicare and Medicaid, 374 New Eng. J. Med. 8, 8 (2016).[/footnote] These three pathways—§ 1115 waivers, DSRIP programming, and managed care incentives—are all means through which states are already exploring uses of the Medicaid program to address social determinants of health. But ongoing litigation shaping the purposes of the Medicaid statute may affect states’ enthusiasm for these types of creative programs. This litigation arises from the use of work requirement waivers, to which this Piece now turns. II. WORK REQUIREMENT WAIVERS The prior Part described ways in which states have considered using Medicaid to address social determinants of health. In approving recent § 1115 waivers, CMS has characterized work as a determinant of health and has conceptualized work requirement waivers as an effort to shape labor force participation. This section describes the course of work requirement waivers that led to the emerging set of federal court decisions interpreting Medicaid’s statutory purpose. Over the past two administrations, federal Medicaid policy has vacillated between two poles. In 2010, a crucial provision of the Affordable Care Act sought to extend coverage to adults with incomes below 138% of the federal poverty level, including adults lacking traditional plus-factors such as disability or pregnancy.[footnote]Patient Protection and Affordable Care Act, Pub. L. No. 111-148, § 2001(a), 124 Stat. 119, 271 (2010) (codified as amended at 42 U.S.C. § 1396a(a) (Supp. V 2017)) (expanding Medicaid eligibility to 133% FPL); 42 U.S.C. § 1396a(a)(10)(A)(i)(VIII) (expanding Medicaid eligibility to adults under 65 years of age who are not pregnant). The text of the ACA sets the minimum FPL at 133%, but due to its new methodology of calculating income, the effective minimum eligibility threshold was actually raised to 138%. ACA Frequently Asked Questions, Am. Pub. Health Assoc., https://www.apha.org/topics-and-issues/health-reform/aca-frequently-asked-questions [https://perma.cc/V5YD-Y495] (last visited Oct. 5, 2019). Prior to this expansion, Medicaid coverage had been reserved for those considered the “deserving” poor—women, children, blind people, disabled people, and poor elderly people. Nicole Huberfeld, Elizabeth Weeks Leonard & Kevin Outterson, Plunging into Endless Difficulties: Medicaid and Coercion in National Federation of Independent Business v. Sebelius, 93 B.U. L. Rev. 1, 13, 16 (2013) (denoting categories of people historically considered to be “deserving poor”). The concept of deserving poor also had a racial dimension, which factored into local control of Medicaid and other programs intended to provide relief for the poor. See Huberfeld et al., supra, at 13 n.62; see also Jamila Michener, Fragmented Democracy: Medicaid, Federalism, and Unequal Politics 34–36 (2018). The ACA abolished these categories, extending coverage on the basis of income alone without requiring a plus-factor for eligibility. See Huberfeld et al., supra, at 25.[/footnote] Congress enforced this directive the same way it had enforced all prior Medicaid expansions: States that refused to expand the program would risk losing all Medicaid funds.[footnote]See Huberfeld et al., supra note 79, at 5–6 (“[T]he Court held that an existing statute, on the books for almost eight decades, constitutionally could not be applied to withhold states’ Medicaid funding for failing to implement the Medicaid expansion.”); see also id. at 21 (describing prior expansions of Medicaid, all enforced by the same provision that risked “all Medicaid funding for non-cooperating states”).[/footnote] Two years later, the Supreme Court’s decision in NFIB v. Sebelius eliminated this penalty as coercive, rendering the expansion optional.[footnote]567 U.S. 519, 585–86 (2012).[/footnote] With thirty-three states opting in since funds became available in 2014, the optional expansion yielded an increase of approximately twelve to thirteen million people insured through Medicaid.[footnote]Medicaid Enrollment Changes Following the ACA, Medicaid & CHIP Payment & Access Comm’n, https://www.macpac.gov/subtopic/medicaid-enrollment-changes-following-the-aca [https://perma.cc/4XL4-6K54] (last visited Oct. 6, 2019) (finding an increase in Medicaid enrollment of 13.1 million people in expansion states); see also Bowen Garrett & Anuj Gangopadhyaya, Urban Inst., Who Gained Health Insurance Coverage Under the ACA, and Where Do They Live? 7 (2016) (finding that over nineteen million nonelderly people gained insurance from 2010 to 2015 and that about twelve million of these were in Medicaid expansion states).[/footnote] Medicaid thus accounts for a majority of the people who newly gained insurance under the ACA.[footnote]Rachel Garfield, Kendal Orgera & Anthony Damico, Kaiser Family Found., The Uninsured and the ACA: A Primer 7 (2019), http://files.kff.org/attachment/The-Uninsured-and-the-ACA-A-Primer-Key-Facts-about-Health-Insurance-and-the-Uninsured-amidst-Changes-to-the-Affordable-Care-Act [https://perma.cc/Q75U-Q76K] (finding that nearly twenty million people gained insurance coverage after the ACA).[/footnote] This expansion, in turn, has led to improved access to healthcare, increased self-reported health, increased financial security, and financial gains for hospitals and clinics.[footnote]Lara Antonisse, Rachel Garfield, Robin Rudowitz & Madeline Guth, Kaiser Family Found., The Effects of Medicaid Expansion Under the ACA: Updated Findings from a Literature Review 1 (2019), http://files.kff.org/attachment/Issue-brief-The-Effects-of-Medicaid-Expansion-under-the-ACA-Findings-from-a-Literature-Review [https://perma.cc/PKM3-RRZU]; Olena Mazurenko, Casey P. Balio, Rajender Agarwal, Aaron E. Carroll & Nir Menachemi, The Effects of Medicaid Expansion Under the ACA: A Systematic Review, 37 Health Aff. 944, 946–48 (2018).[/footnote] Some commentators have even characterized expansion as a step toward Medicaid “universality”[footnote]Nicole Huberfeld, The Universality of Medicaid at Fifty, 15 Yale J. Health Pol’y L. & Ethics 67, 68 (2015).[/footnote]—a shift away from formalized exclusions in healthcare on the basis of factors such as ability to pay, and toward a norm of inclusion verging on “outright solidarity” across eligible populations.[footnote]Id. at 69. Despite this shift toward inclusivity, undocumented migrants continue to be excluded from Medicaid even in expansion states due to federal regulations restricting eligibility. Id. at 68 n.7.[/footnote] Some states also capitalized on Medicaid expansion to increase flexibility via § 1115 waiver programming, securing new abilities to impose premiums,[footnote]See Watson, Black Box, supra note 37, at 213.[/footnote] cost sharing,[footnote]Id.[/footnote] and stick-based personal responsibility conditions for accessing enhanced or optional benefits.[footnote]See, e.g., Susan McKernan, Elizabeth Momany, Aparna Ingleshwar, Padmaja Ayyagari, Astha Singhal, Dan Shane, Andrew Ghattas & Peter Damiano, Univ. of Iowa Pub. Policy Ctr., Access, Utilization, and Cost Outcomes: Iowa Dental Wellness Plan Evaluation 2014–2016, at 6 (2017), http://ppc.uiowa.edu/sites/default/files/dwp_outcomes_report.pdf [https://perma.cc/PBB5-L67K] (explaining a dental plan that rendered members who sought routine dental exams at regular intervals eligible for “additional covered services”); Tami Gurley-Calvez, Genevieve M. Kenney, Kosali I. Simon & Douglas Wissoker, Impacts on Emergency Department Visits from Personal Responsibility Provisions: Evidence from West Virginia’s Medicaid Redesign, 51 Health Serv. Res. 1424, 1426–27 (2016) (describing a health plan that would provide participants with “more comprehensive benefits” if they agreed to “certain program rules” designed to reduce costs and improve health behavior); Brad Wright, Natoshia M. Askelson, Monica Ahrens, Elizabeth Momany, Suzanne Bentler & Peter Damiano, Completion of Requirements in Iowa’s Medicaid Expansion Premium Disincentive Program, 2014–2015, 108 Am. J. Pub. Health 219, 219 (2018) (evaluating rates of member compliance with Iowa’s Medicaid expansion premium disincentive program).[/footnote] In 2017, however, the change in federal administration brought a change in emphasis for Medicaid policy. Federal regulation of Medicaid has mirrored a shift in broader administration goals, visible in new “public charge” regulations (that is, rules that consider Medicaid uptake as potentially disqualifying for citizenship and immigration determinations),[footnote]Inadmissibility on Public Charge Grounds, 84 Fed. Reg. 41,292, 41,292 (Aug. 14, 2019) (to be codified at 8 C.F.R. pts. 103, 212, 213, 214, 245, 248).[/footnote] proposed rules that would eliminate federal oversight of state provider reimbursement levels,[footnote]See Methods for Assuring Access to Covered Medicaid Services-Rescission, 84 Fed. Reg. 33,732, 33,732 (proposed July 15, 2019) (to be codified at 42 C.F.R. pt. 447); Sara Rosenbaum, The Latest Installment in the Saga of the Medicaid Equal Access Guarantee, Health Aff.: Blog (Aug. 23, 2019), https://www.healthaffairs.org/do/10.1377/hblog20190820.245433/full [https://perma.cc/RMG7-AHEU].[/footnote] proposed rollbacks of nondiscrimination rules that had initially sought to protect patients on the basis of gender identity and sexual orientation,[footnote]Nondiscrimination in Health and Health Education Programs or Activities, 84 Fed. Reg. 27,846, 27,847–49 (proposed June 14, 2019) (to be codified at 42 C.F.R. pts. 438, 440, 460).[/footnote] and the issuance of experimental state waivers that allow states to enforce new conditions on eligibility for adult beneficiaries deemed able-bodied.[footnote]See infra note 109 and accompanying text.[/footnote] In late 2017, CMS issued new goals for Medicaid waivers, including “improv[ing] access to high-quality, person-centered services that produce positive health outcomes,” “ensur[ing] Medicaid’s sustainability,” and “address[ing] certain health determinants that promote upward mobility, greater independence, and improved quality of life.”[footnote]Hinton et al., supra note 4, at 9; see also About Section 1115 Demonstrations, Medicaid.gov, https://www.medicaid.gov/medicaid/section-1115-demo/about-1115/index.html [https://perma.cc/MZ28-48RV] (last visited Oct. 6, 2019).[/footnote] The latter set of CMS decisions culminated in the first work requirements[footnote]The term “community engagement requirements” is preferred by CMS and many states. See, e.g., 1115 Community Engagement Initiative, Medicaid.gov, https://www.medicaid.gov/medicaid/section-1115-demo/community-engagement/index.html [https://perma.cc/K7PT-MSEZ] (last visited Oct. 6, 2019). This terminology denotes that beneficiaries can fulfill requirements by not only paid work but also volunteering, education, job search and training activities, caretaking, and participating in substance use treatment. I here use “work requirements” in order to participate in conversation with other legal scholars writing about these programs and to echo the discussion of these programs in public debate.[/footnote] internal to the Medicaid program, a break from the decisions of prior administrations. Work requirements have long been part of cash welfare (Temporary Assistance to Needy Families) and nutritional assistance programming (Supplemental Nutrition Assistance Programs), both of which are governed by the Personal Responsibility and Work Opportunity Reconciliation Act of 1996.[footnote]See Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Pub. L. No. 104-193, § 824, 110 Stat. 2105, 2323 (adding work requirements to SNAP programming); id. § 402, 110 Stat. 2105, 2262 (codified as amended at 8 U.S.C. § 1612 (2018)) (adding work requirements to cash welfare programming).[/footnote] Both programs require work effort as a condition of participation for adults who are able-bodied, and sanctions for noncompliance extend to program exclusion.[footnote]See supra notes 6–7 and accompanying text.[/footnote] Before Medicaid expansion, however, similar efforts to include work requirements in Medicaid programming were rejected by CMS.[footnote]See Laura D. Hermer, What to Expect When You’re Expecting . . . TANF-Style Medicaid Waivers, 27 Annals Health L. 37, 44 & n.41 (2018) [hereinafter Hermer, What to Expect]; Letter from Andrew Slavitt, Acting Adm’r, Ctrs. for Medicare & Medicaid Servs., to Thomas Betlach, Dir., Ariz. Health Care Cost Containment Sys. (Sept. 30, 2016), https://www.azahcccs.gov/Resources/Downloads/1115Waiver/LetterToState09302016.pdf [https://perma.cc/EKL9-38RH] (disallowing work requirements in Arizona); Letter from Marilyn Tavenner, Acting Adm’r, Ctrs. for Medicare & Medicaid Servs., to Michael Hales, Deputy Dir., Utah Dep’t of Health (Apr. 6, 2012), https://le.utah.gov/interim/2012/pdf/00002139.pdf [https://perma.cc/8SJ4-LLC4] (disallowing community service requirements in Utah); Letter from Marilyn Tavenner, Adm’r, Ctrs. for Medicare & Medicaid Servs., to Beverly Mackereth, Sec’y, Pa. Dep’t of Pub. Welfare (Aug. 18, 2014), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/pa/Healthy-Pennsylvania-Private-Coverage-Option-Demonstration/pa-healthy-ca.pdf [https://perma.cc/7B8G-2875] (disallowing work requirements in Pennsylvania).[/footnote] This was in part because the populations that were previously eligible for Medicaid—children, elderly people, disabled people, and pregnant women—have been considered “deserving” regardless of labor force participation.[footnote]Huberfeld et al., supra note 79, at 13.[/footnote] In states that expanded Medicaid under the ACA, however, the program now included a new population of adult beneficiaries, including “non-elderly, non-disabled, low-income single adults or couples without children,”[footnote]Id. at 25.[/footnote] who have not historically been considered deserving without working.[footnote]The expectation that work effort would identify “deserving” poor dates back to the design of Elizabethan poor laws. See id. at 13; see also Hermer, What to Expect, supra note 98, at 41.[/footnote] In March of 2017, CMS Administrator Seema Verma and then–U.S. Department of Health and Human Services Secretary Tom Price issued a letter to state governors promising new flexibilities in state waivers, including the approval of “innovations that build on the human dignity that comes with training, employment, and independence.”[footnote]Letter from Thomas E. Price, Sec’y, U.S. Dep’t of Health & Human Servs., & Seema Verma, Adm’r, Ctrs. for Medicare & Medicaid Servs., to State Governors 2 (Mar. 14, 2017), https://www.hhs.gov/sites/default/files/sec-price-admin-verma-ltr.pdf [https://perma.cc/YQ86-MLR6] [hereinafter Letter to State Governors].[/footnote] In January of 2018, CMS issued guidelines for state proposals involving community engagement requirements for beneficiaries considered able-bodied.[footnote]Letter from Brian Neale, Deputy Adm’r & Dir., Ctrs. for Medicare & Medicaid Servs., to State Medicaid Dirs. (Jan. 11, 2018), https://www.medicaid.gov/federal-policy-guidance/downloads/smd18002.pdf [https://perma.cc/YR5S-K2LX].[/footnote] These guidelines outline CMS’s hypothesis that mandating individual work effort will improve health and alleviate poverty, and they specify that CMS will not allow federal Medicaid funds to be spent on “work supports” such as transportation or childcare for beneficiaries during working hours.[footnote]Id. at 7.[/footnote] The Administration’s interest in work effort as a condition of benefits participation extends not only to Medicaid but also to other federal benefits like federal housing assistance, as President Trump has emphasized via executive order.[footnote]Exec. Order No. 13,828, Reducing Poverty in America by Promoting Opportunity and Economic Mobility, 83 Fed. Reg. 15,941 (Apr. 13, 2018).[/footnote] Seventeen states have now sought CMS approval for work requirement waivers, of which nine have been granted, with required quotas ranging from 80 to 100 hours per month as a condition of maintaining Medicaid eligibility.[footnote]Kaiser Family Found., Medicaid Waiver Tracker, supra note 5.[/footnote] Interest in work requirements has been keen in states characterized by popular, legislative, or executive resistance to the ACA, including nonexpansion states[footnote]See David A. Super, A Hiatus in Soft-Power Administrative Law: The Case of Medicaid Eligibility Waivers, 65 UCLA L. Rev. 1590, 1600 (2018).[/footnote] and several states that had previously used waivers to expand Medicaid in exchange for concessions from CMS.[footnote]See Watson, Black Box, supra note 37, at 213; Sidney D. Watson, Premiums and Section 1115 Waivers: What Cost Medicaid Expansion?, 9 St. Louis U. J. Health L. & Pol’y 265, 266 (2016) (describing expansion by § 1115 waiver in Michigan, Arkansas, Indiana, and Montana). All four states have since sought work requirement waivers. See Kaiser Family Found., Medicaid Waiver Tracker, supra note 5.[/footnote] State applications and CMS approval letters concerning work requirements have articulated a range of purposes for these conditions. These have included alleviating poverty, building dignity, promoting state flexibility and tailored programming, advancing national economic interests, reverting Medicaid to its pre-ACA intentions, moving public beneficiaries into private insurance, controlling costs, promoting program sustainability, and “reserving” Medicaid benefits for people who are not considered able-bodied.[footnote]For Arizona, see Letter from Andrea J. Casart, Dir., Div. of Medicaid Expansion Demonstrations, U.S. Dep’t of Health & Human Servs., to Jami Snyder, Dir., Ariz. Health Care Cost Containment Sys. (Sept. 13, 2019), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/az/az-hccc-ca.pdf [https://perma.cc/6YS2-7939]. For Arkansas, see Letter from Seema Verma, Adm’r, Ctrs. for Medicare & Medicaid Servs., U.S. Dep’t of Health & Human Servs., to Asa Hutchinson, Governor of Ark., & Cindy Gillespie, Dir., Ark. Dep’t of Human Servs. (Mar. 5, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ar/ar-works-ca.pdf [https://perma.cc/GWB5-B6E9]. For Indiana, see Letter from Demetrios Kouzoukas, Principal Deputy Adm’r, Ctrs. for Medicare & Medicaid Servs., to Allison Taylor, Medicaid Dir., Ind. Family & Soc. Serv. Admin. (Feb. 1, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/in/in-healthy-indiana-plan-support-20-ca.pdf [https://perma.cc/V4C6-FBXZ]. For Kentucky, see Letter from Paul Mango, Chief Principal Deputy Adm’r & Chief of Staff, Ctrs. for Medicare & Medicaid Servs., to Carol H. Steckel, Comm’r, Dep’t for Medicaid Servs., Commonwealth of Ky. (Nov. 20, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ky/ky-health-ca.pdf [https://perma.cc/5ZH4-EJUF]. For Michigan, see Letter from Seema Verma, Adm’r, Ctrs. for Medicare & Medicaid Servs., to Rick Snyder, Governor of Mich., & Kathy Stiffler, Acting Dir., Dep’t of Health & Human Servs., State of Mich. (Dec. 21, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/mi/mi-healthy-michigan-ca.pdf [https://perma.cc/7FME-FVFC]. For New Hampshire, see Letter from Mary C. Mayhew, Deputy Adm’r & Dir., Ctrs. for Medicare & Medicaid Servs., to Henry D. Lipman, Medicaid Dir., N.H. Dep’t of Health & Human Servs. (Nov. 30, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/nh/nh-granite-advantage-health-care-program-ca.pdf [https://perma.cc/KN8R-SYUK]. For Ohio, see Letter from Andrea J. Casart, Dir., Div. of Medicaid Expansion Demonstrations, U.S. Dep’t of Health & Human Servs., to Maureen Corcoran, Dir., Ohio Dep’t of Medicaid (May 23, 2019), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/oh/oh-work-requirement-community-engagement-ca.pdf [https://perma.cc/TVU7-7ZPK]. For Utah, see Letter from Seema Verma, Adm’r, Ctrs. for Medicare & Medicaid Servs., to Nathan Checketts, Dir., Utah Dep’t of Health (Mar. 29, 2019), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ut/ut-primary-care-network-ca.pdf [https://perma.cc/G5Z5-KYMH]. For Wisconsin, see Letter from Seema Verma, Adm’r, Ctrs. for Medicare & Medicaid Servs., to Casey Himebauch, Deputy Medicaid Dir. & Adm’r, Div. of Medicaid Servs., Wis. Dep’t of Health Servs. (Oct. 31, 2018), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/wi/wi-badgercare-reform-ca.pdf [https://perma.cc/4DDD-WVYU].[/footnote] The goal of greatest interest for this Piece is health promotion. All of the state applications and CMS approval letters have explicitly argued that work requirements will improve beneficiaries’ health—that is, the requirements will motivate participation in work or volunteering, which will result in employment, which in turn will promote health. In their 2017 letter to state governors, Administrator Verma and Secretary Price argued that “[t]he best way to improve the long-term health of low-income Americans is to empower them with skills and employment.”[footnote]Letter to State Governors, supra note 102, at 2.[/footnote] States have echoed the position that employment brings a “sense of accomplishment, personal satisfaction, self-reliance, social interaction, and integration”—all of which, states posit, lead to better physical health.[footnote]State of Kan., KanCare 2.0: Section 1115 Demonstration Renewal Application 9 (2017), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ks/ks-kancare-pa3.pdf [https://perma.cc/4GT5-LLF6].[/footnote] Some states have also used the language of “health determinants,” in keeping with CMS’s emphasis on “certain health determinants” in waiver programming.[footnote]See supra note 94 and accompanying text.[/footnote] Across these waivers, states have framed employment, or the lack thereof, as a social determinant of health, and they seek to remedy unemployment by holding out Medicaid as an incentive to motivate individual behavior. Commentators have disagreed with the behavioral model behind Medicaid work requirements, and some have expressed skepticism about states’ and CMS’s intentions in adopting work requirements as a condition of participation.[footnote]See, e.g., Hannah Katch, Jennifer Wagner & Aviva Aron-Dine, Ctr. on Budget & Policy Priorities, Taking Medicaid Coverage Away from People Not Meeting Work Requirements Will Reduce Low-Income Families’ Access to Care and Worsen Health Outcomes (2018), https://www.cbpp.org/sites/default/files/atoms/files/2-8-18health2.pdf [https://perma.cc/R9E2-MTWV]; Laura D. Hermer, Medicaid: Welfare Program of Last Resort, or Safety Net?, 44 Mitchell Hamline L. Rev. 1203, 1224 (2018); Nicole Huberfeld, Can Work Be Required in the Medicaid Program?, 375 New Eng. J. Med. 788, 789–90 (2018); Super, supra note 107, at 1600; Nicholas Bagley & Eli Savit, Michigan’s Discriminatory Work Requirements, N.Y. Times (May 8, 2018), https://www.nytimes.com/2018/05/08/opinion/michigan-medicaid-work-requirement.html [https://perma.cc/3YJX-KJZY]; Sara Rosenbaum, Experimenting on the Health of the Poor: Inside Stewart v. Azar, Health Aff.: Blog (Feb. 5, 2018), https://www.healthaffairs.org/do/10.1377/hblog20180204.524941/full [https://perma.cc/GD2K-BTQC].[/footnote] In recent months, observers have also cited the experience of Arkansas as cause for concern. Arkansas was the first to implement a Medicaid work requirement, which was in force for approximately nine months until the Gresham v. Azar decision vacated the program.[footnote]363 F. Supp. 3d 165 (D.D.C. 2019).[/footnote] Between June and December 2018, almost 17,000 people were notified that they had lost their coverage (estimated to be about 12% of the Medicaid population); a statewide representative survey found significant decreases in Medicaid coverage, increases in people who reported being uninsured, no change in employment, and widespread confusion about the policy.[footnote]See Benjamin D. Sommers, Anna L. Goldman, Robert J. Blendon, E. John Orav & Arnold M. Epstein, Medicaid Work Requirements: Results from the First Year in Arkansas, 381 New Eng. J. Med. 1073, 1079–80 (2019).[/footnote] Work requirement waivers offer an opportunity to consider broader questions about what public health insurance can and should seek to accomplish. In the ensuing litigation, federal courts must now weigh in on the purposes of the Medicaid program, with implications that may shape future Medicaid programming by both progressive and conservative states and federal administrations. III. PARSING THE PURPOSES OF MEDICAID: STEWART, GRESHAM, AND PHILBRICK Shortly after CMS approved the first work requirement waiver in Kentucky, a group of Medicaid beneficiaries sued to enjoin the program.[footnote]See Class Action Complaint for Declaratory and Injunctive Relief at 3–4, Stewart I, 313 F. Supp. 3d 237 (D.D.C. 2018) (No. 1:18-cv-00152), 2018 WL 525491.[/footnote] Similar suits have challenged waivers in Arkansas,[footnote]See First Amended Complaint for Declaratory and Injunctive Relief at 3–4, Gresham v. Azar, 363 F. Supp. 3d 165 (D.D.C. 2019) (No. 1:18-cv-1900 (JEB)), 2018 WL 8265789.[/footnote] New Hampshire,[footnote]See Class Action Complaint for Declaratory and Injunctive Relief at 3–4, Philbrick v. Azar, No. 1:19-cv-00773 (D.D.C. Mar. 20, 2019), 2019 WL 3414376.[/footnote] and Indiana.[footnote]See Shari Rudavsky, Lawsuit Is Filed Against Indiana’s Medicaid Work Requirements, Indianapolis Star (Sept. 25, 2019), https://www.indystar.com/story/news/health/2019/09/25/medicaid-indiana-work-requirements-healthy-indiana-plan-lawsuit-filed-hhs-trump-administration-cms/2429229001 [https://perma.cc/447S-GKYH].[/footnote] Each complaint named the Secretary of HHS, arguing that the Secretary’s approval of each waiver was an abuse of discretion under the Social Security Act. Judge James E. Boasberg in the District of D.C. has now decided cases concerning all three states, reasoning on substantially identical grounds at summary judgment: The Secretary’s waiver approvals were indeed arbitrary and capricious, the court ruled, because the Secretary had not adequately “considered” the extent to which the programs would advance the Medicaid objective of furnishing medical assistance.[footnote]Philbrick, 2019 WL 3414376, at *1; Stewart II, 366 F. Supp. 3d 125, 131 (D.D.C. 2019); Gresham, 363 F. Supp. at 169 (D.D.C. 2019); Stewart I, 313 F. Supp. 3d at 259.[/footnote] These opinions squarely interrogate the purposes of the Medicaid statute, and they answer the novel question whether statutory goals included the promotion of health alongside the purchase of healthcare. In each of these decisions, the court has interpreted two provisions of the Social Security Act. The first, 42 U.S.C. § 1315, sets out conditions for experimental waivers: HHS may approve state experiments, pilots, or demonstration programs that “in [the Secretary’s] judgment . . . are likely to assist in promoting the [Act’s] objectives.”[footnote]Stewart I, 313 F. Supp. 3d at 244–45 (quoting 42 U.S.C. § 1315(a) (2012)). Budget neutrality is an agency requirement, not a formal statutory requirement. There are additional procedural requirements for waivers, including public notice and comment, hearings, an evaluation, and a time limitation on waiver periods. See Watson, Black Box, supra note 37, at 215.[/footnote] To identify these objectives, the court has followed prior precedents in consulting 42 U.S.C. § 1396-1, which authorizes the appropriation of federal Medicaid funds “[f]or the purpose of enabling each State . . . to furnish (1) medical assistance . . . [to] individuals[ ] whose income and resources are insufficient to meet the costs of necessary medical services, and (2) rehabilitation and other services to help such families and individuals attain or retain capability for independence or self-care.”[footnote]42 U.S.C. § 1396-1 (Supp. V 2017).[/footnote] The meaning of “furnishing medical assistance” is defined elsewhere in the statute as “payment of all or part of the cost” of care,[footnote]See id. § 1396d(a); Stewart I, 313 F. Supp. 3d at 260.[/footnote] and precedents in the D.C. Circuit and the Supreme Court have referred to the goal of Medicaid as providing federal money to states that purchase medical care on behalf of needy residents.[footnote]See Alexander v. Choate, 469 U.S. 287, 289 (1985); Harris v. McRae, 448 U.S. 297, 301 (1980); Adena Reg’l Med. Ctr. v. Leavitt, 527 F.3d 176, 180 (D.C. Cir. 2008); Stewart I, 313 F. Supp. 3d at 261.[/footnote] The beneficiaries in the ACA expansion population—the adults new to Medicaid—indisputably qualify as these needy residents. Although the Medicaid expansion is optional, states that elect to expand have chosen to include individuals below 138% FPL as a mandatory population.[footnote]See Philbrick, 2019 WL 3414376, at *2; Stewart II, 366 F. Supp. 3d at 132; Gresham, 363 F. Supp. 3d at 170; Stewart I, 313 F. Supp. 3d at 261 (“Under the ACA, states can choose to expand their Medicaid coverage to include this new, low-income group. Should a state choose to do so, those individuals become part of its mandatory population.” (citations omitted)).[/footnote] “As amended,” the court in Stewart I explained, “one objective of Medicaid thus became ‘furnishing medical assistance’ for this new group of low-income individuals.”[footnote]Stewart I, 313 F. Supp. 3d at 261.[/footnote] The court then identified furnishing medical assistance as “a central objective” of the Medicaid Act;[footnote]Id. at 243, 273 (emphasis added); see also Philbrick, 2019 WL 3414376, at *7–8; Stewart II, 366 F. Supp. 3d at 138; Gresham, 363 F. Supp. 3d at 176.[/footnote] although the court in each case has left the door open for CMS to read in additional purposes, it is clear that approving any waiver that is unlikely to advance the “central” objective of furnishing medical assistance is beyond the Secretary’s discretion.[footnote]See Stewart II, 366 F. Supp. 3d at 139; Stewart I, 313 F. Supp. 3d at 272.[/footnote] Given the goal of paying for care, the court in each work requirement waiver decision has identified the same problem: Although the administrative records showed that reduced Medicaid coverage was a likely consequence of each work requirement waiver, there was insufficient evidence that the Secretary had considered how those losses would affect the objective of furnishing medical assistance for beneficiaries.[footnote]See Philbrick, 2019 WL 3414376, at *8; Stewart II, 366 F. Supp. 3d at 138; Gresham, 363 F. Supp. 3d at 177; Stewart I, 313 F. Supp. 3d at 262.[/footnote] Instead, CMS interpreted the Medicaid statute as implying other objectives, and the court found that the agency had focused on these objectives as alternatives[footnote]See Philbrick, 2019 WL 3414376, at *8; Gresham, 363 F. Supp. 3d at 176; Stewart I, 313 F. Supp. 3d at 268.[/footnote] or supplements[footnote]See Philbrick, 2019 WL 3414376, at *11; Stewart II, 366 F. Supp. 3d at 139.[/footnote] to the goal of furnishing medical assistance. CMS’s proffered objectives for the Medicaid program have varied somewhat across decisions.[footnote]In Stewart I, the Secretary offered other objectives such as “address[ing] behavioral and social factors that influence health outcomes,” “incentiviz[ing] beneficiaries to engage in their own health care,” and “facilitat[ing] smoother beneficiary transition to commercial coverage.” Stewart I, 313 F. Supp. 3d at 261–62.[/footnote] But by the Stewart II, Gresham, and Philbrick decisions, CMS had restated the other objectives as threefold, including “advanc[ing] the health and wellness needs” of beneficiaries,[footnote]Id. at 266.[/footnote] promoting beneficiaries’ “financial independence,”[footnote]Stewart II, 366 F. Supp. 3d at 139.[/footnote] and “ensur[ing] the fiscal sustainability of the Medicaid program.”[footnote]Philbrick, 2019 WL 3414376, at *4; Gresham, 363 F. Supp. 3d at 180.[/footnote] Although CMS argued in each case that health promotion is a secondary or even overriding purpose of Medicaid,[footnote]See Philbrick, 2019 WL 3414376, at *11; Gresham, 363 F. Supp. 3d at 179; Stewart I, 313 F. Supp. 3d at 144.[/footnote] the district court firmly decided that CMS may not “conflate[]” the general goal of improving health with the specific goal of furnishing medical assistance: “[F]ocus[ing] on health is no substitute for considering Medicaid’s central concern: covering health costs. While improving public health and health outcomes might be one consequence of ‘furnishing . . . medical assistance,’ the Secretary cannot choose his own means to that end.”[footnote]Stewart I, 313 F. Supp. 3d at 266.[/footnote] The court emphasized Congress’s focus on the high costs of care at the time of the ACA—it is, after all, the Affordable Care Act—and noted that insofar as health promotion is a goal, it may only be pursued through Congress’s chosen pathway of purchasing healthcare for the low-income population.[footnote]Id. at 267 (“Had Congress maintained a singular focus on promoting health, it easily could have said as much, but the text and structure of Medicaid shows its desire to provide health coverage to those groups. After all, the Act does not convert states into hospitals, forcing them to provide [] medical services.”).[/footnote] Indeed, the court went further to say that recognizing a separate health promotion rationale for Medicaid could have “bizarre results” in waiver approvals: [I]magine that the Secretary could exercise his waiver authority solely to promote health, rather than cover healthcare costs. Nothing could stop him from conditioning Medicaid coverage on consuming more broccoli (at least on an experimental basis). Or, as Plaintiffs suggest, he might force all recipients to enroll in pilates classes or take certain nutritional supplements. . . . The penalty for non-compliance? No more Medicaid.[footnote]Id. at 267–68 (citation omitted).[/footnote] In light of the specific waiver terms at issue in Stewart, Gresham, and Philbrick, it is easy to see why the court’s imagination ran toward other waiver terms that would impose behavioral conditions on program eligibility—all three cases concerned programs that made working, studying, or volunteering a condition of program participation.[footnote]See Philbrick, 2019 WL 3414376, at *1; Stewart II, 366 F. Supp. 3d 125, 130 (D.D.C. 2019); Gresham, 363 F. Supp. 3d at 172; Stewart I, 313 F. Supp. 3d at 243.[/footnote] But at least in this series of cases, the court has not considered other forms of innovation that may be undertaken by states willing and able to spend more. Moreover, the court has done little to imagine other waiver designs, including possible progressive efforts to improve social determinants of health. Oral arguments before the D.C. Circuit in Stewart and Gresham focused heavily on the purposes of the Medicaid statute,[footnote] Oral Argument, Gresham v. Azar, No. 19-5094 (D.C. Cir. Oct. 11, 2019), Stewart v. Azar, No. 19-5095 (D.C. Cir. Oct. 11, 2019), https://www.cadc.uscourts.gov/recordings/recordings2019.nsf/2FEFECFB798E422C85258490005718BA/$file/19-5095.mp3 [https://perma.cc/WDU8-NDL2].[/footnote] with questions from all three panel judges suggesting that coverage is the “critical” or “principal” program goal.[footnote]See Amy Goldstein, Appeals Panel Expresses Skepticism About Medicaid Work Requirements, Wash. Post (Oct. 11, 2019), https://www.washingtonpost.com/health/appeals-panel-expresses-skepticism-about-medicaid-work-requirements/2019/10/11/a8357c4e-eb8a-11e9-9c6d-436a0df4f31d_story.html (on file with the Columbia Law Review).[/footnote] Although goals such as health promotion may be “laudable,” in the words of Judge David Sentelle, there were questions about whether these objectives should “outweigh” the goal of coverage.[footnote]Id.[/footnote] Judge Cornelia Pillard echoed the district court’s concern about other behavioral conditions on coverage, posing a hypothetical waiver that would condition Medicaid coverage on watching only minimal hours of television.[footnote]Oral Argument, supra note 141, at 26:50–27:10.[/footnote] Conditions on eligibility, again, were top of the judges’ minds when imagining the types of waivers that might be justified in health promotion terms. The D.C. Circuit has yet to rule, and its ruling may shed light on whether health promotion may serve as an additional program goal. Under Stewart, Gresham, and Philbrick, CMS could, assuming budget neutrality, continue to approve waivers designed to promote health by alleviating structural risk factors. But there may be new pitfalls for these types of programs. Imagine, for example, that a state sought a waiver to reduce the risk of lead poisoning in families with Medicaid coverage, such as by using funds to finance improvements to housing and water supply for Medicaid beneficiaries. Depending on cost and design, a waiver like this would fall into one of three categories. First, a waiver seeking to reduce lead poisoning may increase the furnishing of medical assistance. If so, approval of such a waiver by the Secretary of HHS would easily pass muster under Stewart. And indeed, where beneficiaries receive other types of support, they may be better able to access medical assistance. Purchasing transportation and childcare assistance specifically for attending medical care appointments, for example, is justifiable on these grounds, and Medicaid already covers nonemergency medical transportation in many states.[footnote]Medicaid & CHIP Payment & Access Comm’n, Medicaid Coverage of Non-Emergency Transportation 1 (2019), https://www.macpac.gov/wp-content/uploads/2019/05/Medicaid-Coverage-of-Non-Emergency-Medical-Transportation.pdf [https://perma.cc/TPB6-2UTU].[/footnote] Perhaps families less vulnerable to lead poisoning are better able to secure other types of care. But this argument at best advances an indirect benefit of what these waivers would actually aim to do, which is to promote the health of beneficiaries by means other than the purchase of medical care. Second, the hypothetical lead poisoning waiver could have no effect on furnishing medical assistance. If so, the waiver may still be approvable—but it may be difficult. In Stewart, Gresham, and Philbrick, the administrative records for the waivers projected coverage losses; perhaps the Secretary could approve waivers that increase spending on social determinants in the name of health promotion so long as they have a neutral impact on coverage. The statutory standard for approval, however, is that Medicaid waivers must be “likely to advance” the purposes of the statute. If these purposes must always—and perhaps must only—include the furnishing of medical assistance, Medicaid waivers will lose some of the flexibility needed to support innovation from both sides of the political spectrum. Appellate decisions in the Stewart, Gresham, and Philbrick cases may shed light on whether programs that promote health with neutral impacts on coverage may continue. Third, the imagined lead poisoning waiver could reduce the furnishing of medical assistance. This may occur if, for example, the state chose to refrain from coverage for optional benefits, or if the program put in place protocols to limit overtreatment[footnote]Overtreatment refers to “waste that comes from subjecting patients to care that, according to sound science and the patients’ own preferences, cannot possibly help them,” and it is estimated to cost between $158 billion and $226 billion per year. Donald M. Berwick & Andrew D. Hackbarth, Eliminating Waste in U.S. Health Care, 307 JAMA 1513, 1514 (2012).[/footnote] or coverage for “low-value care.”[footnote]See William H. Shrank, Teresa L. Rogstad & Natasha Parekh, Waste in the US Health Care System: Estimated Costs and Potential for Savings, 322 JAMA 1501, 1501–03 (2019).[/footnote] Under Stewart, a waiver that undermines the furnishing of medical assistance will not stand, even if it results in net improvements in public health outcomes. Perhaps this is as it should be—if Congress wishes to limit lead poisoning, it could do so by means other than Medicaid waiver, and a decision to let this waiver stand would indeed allow the health promotion rationale to outweigh the coverage goal. But public health problems often affect communities with comparatively less political power. Lead poisoning from paint and water is an ideal example—there has been vanishingly little political action on lead poisoning, in large part because the people most at risk are low-income children of color. Where political will and resources are not forthcoming, waivers would provide an alternative pathway to accomplish public health goals, and investing in social determinants such as safe housing may in fact purchase more health than some investments in care. Other forms of § 1115 waivers might redraw coverage lines, eliminating coverage for some treatments or benefits in order to ensure access to care with greater impact. Since 1994, for example, Oregon has held a § 1115 waiver to cover Medicaid benefits according to a Prioritized List,[footnote]Ctrs. for Medicare & Medicaid Servs., Oregon Demonstration Factsheet 4–5 (Jan. 12, 2017), https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/or/or-health-plan2-fs.pdf [https://perma.cc/85KJ-V4JM].[/footnote] where a line is drawn between covered and noncovered benefits based on health benefit and category of care.[footnote]Prioritized List Overview, Or. Health Auth., https://www.oregon.gov/oha/HPA/DSI-HERC/Pages/Prioritized-List-Overview.aspx [https://perma.cc/8H6F-XW5K] (last updated Jan. 2, 2018).[/footnote] This approach necessarily eliminates the possibility of coverage for some treatments (for example, routine foot care for those not at risk of amputation; infertility treatments; repair of uncomplicated hernias[footnote]Id.[/footnote]), but it ensures that funds spent on furnishing medical assistance are allocated to maximize the program’s impact on health. A Secretary reviewing this type of waiver may decide that it has no impact on furnishing medical assistance, or perhaps that it increases assistance if it adds services that were not already covered (such as bone marrow transplants, which were excluded before the waiver’s adoption[footnote]See Philip A. Perry & Timothy Hotze, Oregon’s Experiment with Prioritizing Public Health Care Services, 13 Virtual Mentor 241, 241 (2011), https://journalofethics.ama-assn.org/sites/journalofethics.ama-assn.org/files/2018-06/pfor1-1104.pdf [https://perma.cc/R75T-V57R].[/footnote]). But if it is possible to interpret this waiver as reducing medical assistance for those in need of the excluded benefits, the health promotion rationale is an important additional justification. Commentators have long hoped that states may consider invoking § 1115 waivers to address structural determinants of health.[footnote]See supra note 41 and accompanying text.[/footnote] As Part I noted, some states have come closer than others in doing so. But under the Stewart line of cases, a narrow purpose of Medicaid may limit not only stick-based waiver terms such as work requirements as a condition of eligibility but also carrot-based waivers and the direct purchase of services that alleviate structural risk factors.[footnote]Those concerned about all behavioral conditions on Medicaid eligibility (for example, work requirements, premiums, wellness activities) may find this an acceptable outcome. But rather than chilling all waivers (even uniformly), CMS and the courts could use other steps to narrow approval to only meritorious proposals, including rigorous evaluations of experimental programs and increased CMS willingness to deny renewal or insist on modifications. Courts could also place more emphasis on whether the Secretary has permissibly found that a program is “likely to advance” program goals on the basis of the administrative record, rather than narrowing the program goals themselves. Under the current standard, the Secretary must “examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.” Stewart I, 313 F. Supp. 3d 237, 249 (D.D.C. 2018) (internal quotation marks omitted) (quoting Motor Vehicle Mfrs. Ass’n, Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)).[/footnote] States attentive to the Stewart cases may be reluctant to design waivers primarily geared toward health promotion, even if the latter goals may be more efficient and impactful uses of funds. IV. THE FUTURE ROLE OF PUBLIC HEALTH INSURANCE Given the spectacularly high costs of U.S. healthcare,[footnote]See Bradley & Taylor, supra note 21, at 1–20; Elisabeth Rosenthal, An American Sickness: How Healthcare Became Big Business and How You Can Take It Back 1–8 (2018); Elisabeth Rosenthal, Paying Till It Hurts, N.Y. Times: Health, https://www.nytimes.com/interactive/2014/health/paying-till-it-hurts.html [https://perma.cc/K5SZ-D25T] (last visited Oct. 6, 2019).[/footnote] it is perhaps enough for public insurance programs like Medicaid and Medicare to purchase healthcare services. There are always more benefits to add, more technologies to cover, and more people who could qualify for coverage for states with the desire to spend more in state funds. But to focus exclusively on purchasing healthcare services misses a novel opportunity to use Medicaid § 1115 waivers to intervene productively in social determinants—which may lead to greater improvements in health than merely paying for care. Chilling health promotion programming in Medicaid raises concerns, in part because the country’s largest public insurance program could be a powerful vehicle for improving public health. Medicaid is vast in scope, covering “traditional” populations in all fifty states and expansion adults in thirty-three.[footnote]See supra notes 1, 82 and accompanying text. An additional two million otherwise-eligible adults would be covered had their state chosen to expand coverage. See Rachel Garfield, Kendal Orgera & Anthony Damico, Kaiser Family Found., The Coverage Gap: Uninsured Poor Adults in States that Do Not Expand Medicaid 2 (2019), http://files.kff.org/attachment/Issue-Brief-The-Coverage-Gap-Uninsured-Poor-Adults-in-States-that-Do-Not-Expand-Medicaid [https://perma.cc/MW8A-VUJM].[/footnote] Although state § 1115 waivers can at most extend statewide, recent experiences with state waivers have revealed rapid diffusion across states, where new states look to prior waivers as having near-precedential effects on what innovations should be approved. The Trump Administration has entrenched this dynamic by promising to expedite future approvals of any § 1115 waiver terms that have been approved in the past. This may not be ideal for a mechanism intended to promote innovative experimentation, but it does lend itself well to the diffusion of programmatic terms that prove to be effective. States that propose health promotion waivers—and that find such waivers to be successful, ideally through methodologically rigorous independent evaluations[footnote]See Kristen Underhill, Atheendar Venkataramani & Kevin G. Volpp, Fulfilling States’ Duty to Evaluate Medicaid Waivers, 379 New Eng. J. Med. 1985, 1986–87 (2018) (suggesting the use of randomized, controlled trials to strengthen approved Section 1115 waiver evaluations, which are “essential in order for such experiments to yield useful lessons for Medicaid program design”).[/footnote]—can provide models for other jurisdictions where populations face similar risk factors for ill health. Congress could also use state waiver experiences as models for national legislation or modifications to Medicaid terms. Medicaid could also be a valuable vehicle for health promotion for several reasons. First, Medicaid receives open-ended federal funding instead of block grants or state grants with per-capita caps,[footnote]See Rachel Sachs & Nicole Huberfeld, The Problematic Law and Policy of Medicaid Block Grants, Health Aff.: Blog (July 24, 2019), https://www.healthaffairs.org/do/10.1377/hblog20190722.62519/full/ [https://perma.cc/3GMC-QE2Q].[/footnote] giving the program more financial flexibility than insurance programs with more limited funding, such as the State Children’s Health Insurance Program. Medicaid also explicitly contains mechanisms to encourage experimentation, including § 1115 waivers and § 1332 waivers, which leverage not only Medicaid funds but also federal tax subsidies under the Affordable Care Act.[footnote]Section 1332 waivers provide states flexibility to waive out of certain Affordable Care Act requirements for a five-year period and reallocate the funds that would have been provided for activities such as maintaining the marketplaces or tax subsidies for private insurance purchases. See 42 U.S.C. § 18052 (Supp. V 2017); Tracking Section 1332 State Innovation Waivers, Kaiser Family Found. (Aug. 29, 2019), https://www.kff.org/health-reform/fact-sheet/tracking-section-1332-state-innovation-waivers [https://perma.cc/8YNE-3Q2C]. Thus far, states have generally used these waivers for reinsurance mechanisms to stabilize ACA markets.[/footnote] This explicit call for experimentation not only encourages states to develop more innovative public health approaches but also allows states to diverge from one another to focus on individually pressing health problems.[footnote]For example, twenty-eight states have now received approval for waivers targeting the opioid epidemic, with innovations such as expanding access to medication-assisted treatment and lifting federal restrictions on the size of inpatient facilities for mental and behavioral health. See, e.g., 1115 Substance Use Disorder Demonstrations, Medicaid.gov, https://www.medicaid.gov/state-resource-center/innovation-accelerator-program/program-areas/reducing-substance-use-disorders/1115-sud-demonstrations/index.html [https://perma.cc/4G3K-QEBS] (last visited Oct. 6, 2019); Section 1115 Waivers for Substance Use Disorder Treatment, Medicaid & CHIP Payment & Access Comm’n, https://www.macpac.gov/subtopic/section-1115-waivers-for-substance-use-disorder-treatment [https://perma.cc/7YV4-NCF8] (last visited Oct. 6, 2019).[/footnote] Finally, Medicaid experimentation provides an opportunity to obtain federal matching funds for public health approaches that have not yet attracted political will in Congress. Although there are federal programs that aim to address some of the determinants described above—SNAP, TANF, the Earned Income Tax Credit, LIHEAP, federal housing assistance, and federal student loans—many low-income individuals eligible for Medicaid struggle with problems of food insecurity,[footnote]See New Research Collaboration Yields Insights into Food Insecurity Among Medicaid Beneficiaries, Root Cause Coal. (Oct. 9, 2018), https://www.rootcausecoalition.org/e_news/new-research-collaboration-yields-insights-into-food-insecurity-among-medicaid-beneficiaries/ [https://perma.cc/4TWM-WQKD].[/footnote] energy insecurity,[footnote]See Daniel R. Taylor, Bruce A. Bernstein, Eileen Carroll, Elizabeth Oquendo, Linda Peyton & Lee M. Pachter, Keeping the Heat on for Children’s Health: A Successful Medical-Legal Partnership Initiative to Prevent Utility Shutoffs in Vulnerable Children, 26 J. Health Care for Poor & Underserved 676, 682–83 (2015) (working primarily in a population eligible for Medicaid and finding high levels of energy insecurity).[/footnote] homelessness and poor housing,[footnote]See Alley et al., supra note 78, at 10; Joshua Bamberger, Reducing Homelessness by Embracing Housing as a Medicaid Benefit, 176 JAMA Internal Med. 1051, 1051 (2016).[/footnote] and other disadvantages linked to poverty. It may be politically challenging to expand on any of these programs, or to develop new federal programs with separate line items. But if state Medicaid programs can demonstrate possible advantages in paying to alleviate broader risk factors for poor health, these waivers may be practicable means of receiving federal assistance in achieving public health goals without requiring Congressional modifications.[footnote]State experimentation in this area could eventually build momentum for changes to federal Medicaid policy; in healthcare, the use of Massachusetts as a model for the ACA private market reforms is one such example of the federal government looking to states for inspiration. See Sharon K. Long, Karen Stockley & Heather Dahlen, Massachusetts Health Reforms: Uninsurance Remains Low, Self-Reported Health Status Improves as State Prepares to Tackle Costs, 31 Health Aff. 444, 449–50 (2012).[/footnote] Despite these potential advantages, however, states have done little to date to pursue the types of public health promotion waivers suggested above. This reluctance may stem from multiple sources including a lack of state funds, disagreement over how to prioritize and approach public health goals, uncertainty about how best to intervene in social and structural risk factors, the budget-neutrality requirement at CMS for § 1115 waivers,[footnote]This has been interpreted as requiring each component of a § 1115 waiver to demonstrate budget neutrality for the federal government. See Hill, Budget Neutrality Policy Letter, supra note 37, at 2. This means that even if savings accrue across an entire set of waiver terms, individual waiver terms that are not budget neutral may be difficult to justify.[/footnote] and possible uncertainty about the authority of HHS to approve waivers for novel types of expenditure.[footnote]See supra notes 38–40.[/footnote] None of these sources of uncertainty would be fatal to social determinants waivers, including those pursued under § 1115 or DSRIP programming. But the Stewart line of cases may add a nationwide roadblock to such waivers, casting doubt on the extent to which waivers justified in health-promotion terms may be eligible for approval without also expanding the direct purchase of medical assistance. Besides Medicaid, other public assistance programs aim to “furnish medical assistance,” and we might ask similar questions about whether the federal government could seek to pay for social and structural-level interventions in Medicare, the Indian Health Services, TRICARE and the VA, and other federal benefits programs. To date, with the exception of some covered benefits in Medicare Advantage plans (for example, transportation, meal services, and connection to social services[footnote]See CMS Finalizes Medicare Advantage and Part D Payment and Policy Updates to Maximize Competition and Coverage, Ctrs. for Medicare & Medicaid Servs. (Apr. 1, 2019), https://www.cms.gov/newsroom/press-releases/cms-finalizes-medicare-advantage-and-part-d-payment-and-policy-updates-maximize-competition-and [https://perma.cc/44UJ-4R3F]; Paige Minemyer, Experts: Offering Medicare Advantage Supplemental Benefits Is “A Long Play,” FierceHealthcare (Sept. 10, 2019), https://www.fiercehealthcare.com/payer/experts-offering-medicare-advantage-supplemental-benefits-a-long-play [https://perma.cc/H82G-EC2C].[/footnote]), Medicare has done little to mobilize federal funds for broader risk factors. The 2020 election has galvanized discussions about Medicare for All (M4A), with multiple permutations of how Medicare might be expanded to accommodate all U.S. residents as a single-payer plan, or perhaps cover only those not covered by employer-sponsored insurance, or perhaps offer a public option or early buy-in.[footnote]See, e.g., Abby Goodnough & Trip Gabriel, ‘Medicare for All’ vs. ‘Public Option’: The 2020 Field Is Split, Our Survey Shows, N.Y. Times (June 23, 2019), https://www.nytimes.com/2019/06/23/us/politics/2020-democrats-medicare-for-all-public-option.html [https://perma.cc/SD37-KV97] (summarizing plan ideas among Democratic presidential candidates); Dylan Scott, On Health Care, There Are 3 Types of Democrats Running for President, Vox (July 30, 2019), https://www.vox.com/2019/7/30/20747974/democratic-debate-health-care-medicare-for-all [https://perma.cc/RC2X-FQJD]; see also Linda J. Blumberg, John Holahan, Matthew Buettgens, Anuj Gangopadhyaya, Bowen Garrett, Adele Shartzer, Michael Simpson, Robin Wang, Melissa M. Favreault & Diane Arnos, Comparing Health Insurance Reform Options: From “Building on the ACA” to Single Payer, Commonwealth Fund (Oct. 16, 2019), https://www.commonwealthfund.org/publications/issue-briefs/2019/oct/comparing-health-insurance-reform-options-building-on-aca-to-single-payer [https://perma.cc/7E6X-VYTF].[/footnote] All of these plans, however, focus on ways of purchasing care from providers and manufacturers, and they have not turned the discussion to intervening in upstream social determinants of health. Like Medicaid, Medicare has an open-ended budget and statutory avenues for innovation; more expansive thinking about Medicare and social determinants could yield better returns to health than paying for care alone. To be sure, purchasing care alone is daunting in the most expensive healthcare market in the world. But to date, the United States has lacked the political will and the creativity to address the social and structural-level risk and protective factors that shape our health outcomes—and that put us near the bottom of our peer group in international health rankings.[footnote]See Karen Davis, Kristof Stremikis, David Squires & Cathy Schoen, Commonwealth Fund, Mirror, Mirror on the Wall: How the Performance of the U.S. Health Care System Compares Internationally 7–10 (2014), https://www.commonwealthfund.org/sites/default/files/documents/___media_files_publications_fund_report_2014_jun_1755_davis_mirror_mirror_2014.pdf [https://perma.cc/Q5M7-THJB]; Bradley Sawyer & Daniel McDermott, Kaiser Family Found., How Does the Quality of the U.S. Healthcare System Compare to Other Countries?, Peterson-Kaiser Health Sys. Tracker (Mar. 28, 2019), https://www.healthsystemtracker.org/chart-collection/quality-u-s-healthcare-system-compare-countries/#item-start [https://perma.cc/24N2-MSVC].[/footnote] Reframing public health insurance as a means of addressing social determinants of health may sound like a tall order, but states have begun to make inroads into using Medicaid for this purpose, and the lessons learned in Medicaid may be useful for other public benefits programs as well. To cultivate these efforts, we may need to take a more inclusive view of the purposes and capacity of public health insurance.[footnote]This thinking would run directly counter to traditionally conservative efforts to limit the size and scope of such programs, such as by per capita caps and block grants in Medicaid. See Sachs & Huberfeld, supra note 157 (describing pressures to transform Medicaid into a block grant system). But justifying work requirement waivers in terms of health promotion and social determinants has committed CMS to a position in favor of leveraging Medicaid for reducing social and structural risks to health. Future common ground on some of these structural risks may be possible.[/footnote] CONCLUSION In light of Medicaid’s scope, open-ended budgetary allocation, and built-in flexibilities for state waiver approaches, the time is ripe for more ambitious Medicaid § 1115 waivers that target social determinants of health. But just as such waivers are gaining traction, they may be impeded by decisions that exclude health promotion from acceptable Medicaid purposes. The Stewart line of cases has important implications on several fronts, and although many commentators have focused on the implications for work requirement waivers, these cases may have broader consequences for efforts to use Medicaid as a means of public health promotion. Medicaid has unique advantages as an avenue for addressing social determinants of health, and truncating Medicaid’s purposes may affect not only waivers that set conditions on eligibility but also waivers that reallocate resources to tackle upstream causes of poor health. Public health scientists are continuing to build evidence that demonstrates how social and structural factors influence health. Where there are neither the funds nor the will to address these problems in new legislation or federal programming, Medicaid waivers could provide a promising vehicle for action. But these waivers may become less appealing, and less tenable, in light of judicial decisions that narrow the purposes of the program.