Restrictions on Access to Care Don’t Improve Medicaid Beneficiaries’ Health
Incentives for Healthy Behaviors Have Mixed Results
Some recent Medicaid demonstration projects (or “waivers”) purport to test the use of incentives — or penalties — to promote healthy behaviors that lead to better health. Penalties that restrict access to care are unlikely to be effective and can cause harm for those who don’t participate.Research shows that offering Medicaid beneficiaries immediate rewards like cash or gift cards for engaging in healthy behaviors can be successful in increasing behaviors such as attending a diabetes management class or participating in tobacco cessation programs. But penalties that restrict access to care, such as increasing cost-sharing charges for beneficiaries who don’t complete preventive care visits, are unlikely to be effective and can cause harm for those who don’t participate. States should be mindful of these negative effects. Lessons learned from complex incentive programs also have implications for states that are considering, or have implemented, waivers that will impose penalties on beneficiaries who don’t pay premiums or take away coverage for not meeting work requirements. The evidence shows that:
- Rewards are most likely to change behavior, but even they have limited long-term impact. Incentive programs that offer immediate rewards for specific healthy behaviors or activities, such as providing a gift card for participation in a weight loss class, are most likely to be effective in changing behaviors and least likely to have harmful effects. But rewards are often poorly targeted, going to people who would have engaged in the behavior regardless of the incentive, and they have largely been shown to produce small, short-term results. For example, ten states received $85 million in grants to provide rewards to Medicaid beneficiaries as incentives for healthy behaviors in the Medicaid Incentives for the Prevention of Chronic Disease program beginning in 2011 and lasting five years. The results of the program were mixed: some states reported moderate success in increasing specific behaviors, such as participation in diabetes management classes, while others reported that rewards were not associated with significant changes.
- Penalties don’t work and are often harmful. Policies that restrict access to services, such as taking services away or increasing out-of-pocket costs of people who don’t pay premiums, can in turn restrict access to needed care. Nor have these policies been shown to be effective in increasing compliance with the desired behavior. For example, in order to reduce unnecessary use of the emergency department (ED), West Virginia limited access to benefits for people who did not enroll in a plan in which they received enhanced benefits in exchange for agreeing to rely on a medical home and limit their use of the ED to emergencies. Only 12 percent enrolled in the enhanced benefit plan, and the state ended up seeing a substantial increase in unnecessary ED use by beneficiaries in the limited benefit plan.
- Beneficiaries often lack awareness of incentives, leading to confusion and limited impact. Research examining several incentive programs in Medicaid shows beneficiaries and providers had very low awareness of the programs, and among those who were aware, many reported confusion about how they worked. In Michigan, for example, beneficiaries were largely unaware of the state’s policy to reduce co-pays for those who completed a health risk assessment (HRA); less than 19 percent of beneficiaries who had been enrolled for at least six months had received credit for completing the HRA. And according to a prior survey, of those who did complete an HRA, only 0.1 percent reported that they did so in order to save money on co-pays.
- Incentive programs have high administrative costs. Rewards above and beyond the cost of services can be costly to provide, and offering them requires upfront funds and staff time. Likewise, penalties such as premiums and cost-sharing can be complex to administer, and their administrative costs generally exceed the revenue they collect from beneficiaries.
The evidence from healthy behaviors incentive programs also has implications for states taking Medicaid coverage away from beneficiaries who don’t meet work requirements or pay premiums. In approving some of these waivers, the Centers for Medicare and Medicaid Services (CMS) cited an Indiana program’s supposed successes in using premiums to encourage healthy behavior in beneficiaries. In fact, the Indiana program’s premium structure, like Michigan’s, has been found to confuse beneficiaries, suggesting that it has likely had little effect on behavior but has led to fewer people enrolling in and maintaining coverage. Likewise, in states newly seeking to impose work requirements or premiums, many enrollees will likely lose coverage simply because they do not understand the new rules or paperwork requirements. (In addition, as explained in the text box below, waivers that seek to use Medicaid coverage as an incentive to achieve other goals are inconsistent with the purpose of the program and an inappropriate use of CMS’ waiver authority.)
Instead of using penalties or incentives, states would be better served investing in improved access to appropriate health services, better care coordination, and increased integration of health and social services. Providing transportation, for example, can enable more beneficiaries to keep appointments. And programs in several states have been shown to reduce ED use by expanding access to primary care services and targeting interventions to populations that use the ED frequently. Investing longer term in improving Medicaid beneficiaries’ access to timely and appropriate care is likely to provide greater returns — and is more likely to help Medicaid beneficiaries improve their health.
Rewards Can Be Effective in Encouraging Short-Term Behaviors
Incentive programs that reward healthy behaviors have been studied in both Medicaid and the private sector. These programs offer different types of rewards for a wide range of activities:
- Rewards for one-time activities or behaviors aim to incentivize people to keep primary care appointments or call a tobacco cessation hotline, participate in cancer screenings, or meet a health benchmark such as reducing high blood pressure. Rewards can be cash, a gift certificate, a health-related reward such as a soccer ball or yoga mat, a contribution to a Health Savings Account, or access to a service such as a dental appointment.
- Rewards for long-term behaviors aim to incentivize ongoing behaviors such as participating in weight management classes or achieving health goals such as lowering blood sugar to mitigate diabetes. Rewards are usually similar to rewards for one-time activities, although there are frequently caps on the number or amount of rewards per year.
Limited evidence from private-sector worksite wellness programs suggests that small financial incentives such as cash or gift cards can affect some health behaviors. Financial incentives appear to increase participation in healthy activities such as preventive screenings, obesity and diabetes prevention programs, routine vaccinations, and tobacco cessation. However, a robust review of incentives aimed at increasing medication adherence found inconsistent effects, with limited improvement at best.
In Medicaid, financial rewards for one-time or short-term activities have shown moderate success, such as increasing the use of some preventive health services. Michigan beneficiaries in a focus group noted that the more immediate receipt of a gift card as an incentive — rather than future reduction in cost-sharing — was a greater incentive to complete a healthy behavior program requirement. Some studies have shown that offering as little as $5 or $10 can incentivize healthy behaviors. However, the evidence is mixed as to whether incentives can be effective in changing long-term behaviors that require ongoing engagement, such as sustained weight loss or smoking cessation.
Medicaid Incentives for the Prevention of Chronic Disease
Ten state Medicaid programs were part of the Medicaid Incentives for the Prevention of Chronic Disease program (MIPCD) that began in 2011 and lasted five years. States received $85 million in grants to pay for program incentives and administration. Rewards included gift cards, cash, points redeemable for health-related services or items, and reduced premiums or cost-sharing. Some rewards were awarded at the time of participation while others were redeemable later, such as reductions in cost-sharing in the following year. Participating states had to target at least one health prevention goal such as tobacco cessation or controlling diabetes; several states targeted multiple behaviors or conditions.
Focus group participants reported satisfaction with the programs, according to the MIPCD final evaluation. Three-quarters of participants strongly agreed that the incentives had encouraged lifestyle changes to improve their health. Immediate incentives such as gift cards were more effective than those provided later such as future reductions in cost-sharing.
The final evaluation found mixed effects from offering Medicaid beneficiaries rewards for healthy behaviors. The rewards were associated with an increase in the use of preventive services in most states’ programs; for example, participants receiving incentives to attend diabetes prevention program classes in Montana, New York, and Minnesota attended significantly more of the classes than the control group. While Minnesota found a small — but statistically significant — increase in weight loss between the group receiving incentives and the control group, New York and Montana found no statistically significantly difference between the groups’ weight loss. This is consistent with other interventions, which found mixed effects and limited associations between the incentives and outcomes.
Penalties Have Not Improved Health Outcomes
Penalties designed to reduce undesired behaviors, meanwhile, have been shown to reduce access to care, and likely don’t produce the desired results. For example, states are imposing premiums and cost-sharing to increase the use of preventive care or the appropriate use of services — such as increasing cost-sharing for people who visit the ED for a condition that could have been treated in a primary care setting.
Indiana: Increased Cost-Sharing, Less Preventive Care
Indiana’s Medicaid expansion program, Healthy Indiana Plan (HIP) 2.0, includes two types of coverage: HIP Plus and HIP Basic. HIP Plus beneficiaries must pay premiums, but it only charges co-pays for non-emergency use of the ED. Those who don’t pay premiums are moved to HIP Basic or lose coverage altogether, depending on their income level. HIP Basic has significantly narrower coverage — for example, it excludes dental and vision care — and charges co-pays for many services.
This structure — charging premiums for Plus and higher cost-sharing for services in Basic — amounts to a dual penalty for beneficiaries. First, many who missed Plus’ premium payments said they found them unaffordable, were confused about how to pay, or didn’t know a payment was required. Second, Basic’s cost-sharing responsibilities may be effectively discouraging beneficiaries from engaging in healthy behaviors. An interim evaluation of HIP 2.0 found that Plus enrollees used more preventive, primary, and specialty care and prescription drugs than Basic enrollees. Basic enrollees were more likely to use the emergency room — including for non-emergencies — and less likely to use primary care and preventive care. These findings suggest that Basic enrollees were more likely to lack adequate access to ordinary health care, probably due in part to the co-pays charged in Basic.
Basic members also were less likely to adhere to their prescription drug regimens for certain chronic conditions such as asthma, arthritis, and heart disease. This isn’t surprising, because Basic members must refill their prescriptions every month and make a co-payment, while Plus members can obtain a 90-day supply of maintenance medications with no co-pay. This is concerning because access to maintenance medications can affect health outcomes. It’s particularly concerning for African Americans in Indiana, who are more likely than other groups to be in the Basic plan; fully half of African Americans enrolled in HIP 2.0 are in Basic rather than Plus.
Waiver Approvals Cite Indiana Despite Its Flawed Cost-Sharing
Despite the significant harm to beneficiaries caused by increased cost-sharing, CMS cited the supposed success of HIP 2.0 when it announced its approval of Medicaid waivers in Wisconsin and Kentucky. Among other harmful provisions, both states’ waivers will take Medicaid coverage away from those who don’t meet work requirements, and both will impose premiums on Medicaid beneficiaries. (See text box.) In its approval letters, CMS claimed premiums would incentivize beneficiaries to engage in healthy behaviors, based on higher preventive and primary care use by HIP Plus enrollees.
However, there is no evidence that Plus’ premium requirements led to these results, as CMS claims. CMS’ own evaluators note that the greater use of preventive care by those in HIP Plus came in spite of their limited understanding of the cheaper cost-sharing incentives, “suggest[ing] that other factors — such as intrinsic beneficiary motivation or prompts from care providers — might have been equally important.” In other words, those who paid premiums may have used preventive care at higher rates even if they didn’t have to pay premiums as part of Indiana’s waiver. If anything, the evidence suggests that HIP 2.0’s premium structure confuses beneficiaries and has likely prompted fewer people to enroll in and maintain coverage.
West Virginia: Emergency Visits Rose After State Limited Access to Benefits
Prior to Indiana’s HIP 2.0, West Virginia began testing its Mountain Health Choices program in 2007 with the primary goal of reducing unnecessary ED use. Beneficiaries could choose an “enhanced plan” that offered more comprehensive benefits by committing to program rules such as relying on a medical home for services and keeping appointments with providers. Alternatively, they could choose or default into a “basic plan” that limited them to four prescriptions per month and restricted access to services such as behavioral health care, tobacco cessation programs, and podiatry services.
An evaluation conducted after the program’s termination in 2010 found that Mountain Health Choices actually raised unnecessary ED use. Individuals in the basic plan used the ED more, both overall and for non-emergency visits, than those in the enhanced plan and those in a county that didn’t offer Mountain Health Choices. Although ED use fell slightly for those in the enhanced plan, the decline was overwhelmed by the increase in ED use among those in the basic plan.
Among all Medicaid beneficiaries in the counties offering the program, the likelihood that an individual would go to the ED rose by an average of 6 percent, and the likelihood that they would go to the ED for a non-emergency rose by 12 percent. Limiting access to benefits for individuals in the basic plan ended up increasing their ED use, both for conditions that could have been treated in a primary care setting and for mental health and substance use.
Taking Coverage Away as an “Incentive” to Change Behavior Is Not a Proper Use of Waiver Authority
Despite the evidence showing penalties are ineffective in changing behavior and can be harmful, the Centers for Medicare and Medicaid Services (CMS) is doubling down on their use. CMS is allowing states to terminate Medicaid coverage altogether under the guise of incentivizing beneficiaries to work or engage in work-related activities, pay premiums, or comply with other requirements. This is contrary to the purpose of waivers, which are intended to test new policies and approaches to help meet Medicaid’s core objective — providing comprehensive health coverage to low-income people so they can get the health services they need. Withholding coverage from eligible beneficiaries in an attempt to promote economic mobility or other goals cannot advance that objective.
Wisconsin’s recently approved Medicaid waiver, for example, conditions some adults’ eligibility on their submission of a health risk assessment to the state, ongoing payment of monthly premiums, and working or participating in work-related activities for at least 80 hours a month (and properly reporting it). The waiver’s premium requirements are strict, locking people out of coverage for failing to pay premiums beginning at just 50 percent of the poverty level — or an income of $505 per month for a single person.a CMS claims that these policies are an “incentive for beneficiaries to take measures that promote health and independence.”b
CMS also re-approved a waiver for Kentucky (after a court initially disallowed it) that takes coverage away from state beneficiaries who don’t work or engage in work-related activities, pay premiums, or renew their coverage or report changes in a timely manner. In doing so CMS claimed that, in order to create “an effective incentive for beneficiaries to take measures that promote health and independence, it may be necessary for states to attach penalties to failure to take those measures, including with conditions designed to promote health and financial independence.”c
In these approvals, CMS is recasting Medicaid eligibility itself as a reward for behaviors that it claims will lead to better health, and taking coverage away as a penalty for those who don’t comply. Such claims ignore substantial research on the use of incentives in Medicaid, which shows that rewards have limited effects and that penalties are even less effective in increasing desired behaviors, as this paper shows. Moreover, there is little or no evidence that the behaviors CMS is trying to incentivize will lead to better health,d while taking away coverage will likely worsen health.e
But even if submitting a health risk assessment or engaging in work or work-related activities for a set number of hours would improve health outcomes for some beneficiaries, making coverage contingent on these behaviors is outside the scope of the agency’s authority. It’s mandatory to provide Medicaid to those who meet eligibility factors such as income and citizenship or immigration status; it’s not a reward for eligible people who comply with extraneous behaviors having nothing to do with the eligibility conditions enacted by Congress. Waivers that undermine Medicaid’s mandatory nature can’t promote Medicaid’s core objective of providing comprehensive health coverage to low-income people. As recent experience in Arkansas showsf they lead to widespread loss of coverage, and they impair rather than improve health.
a CMS, BadgerCare Reform.
c CMS, Kentucky HEALTH.
d Larisa Antonisse and Rachel Garfield, “The Relationship Between Work and Health: Findings from a Literature Review,” Kaiser Family Foundation, August 7, 2018, https://www.kff.org/medicaid/issue-brief/the-relationship-between-work-and-health-findings-from-a-literature-review/.
e Benjamin Sommers, Atul A. Gawande, and Katherine Baicker, “Health Insurance Coverage and Health — What the Recent Evidence Tells Us,” New England Journal of Medicine, August 10, 2017, https://www.nejm.org/doi/pdf/10.1056/NEJMsb1706645.
f Jennifer Wagner, “Another 3,815 Arkansans Lost Medicaid in November Due to Rigid Work Requirement,” Center on Budget and Policy Priorities, November 16, 2018, https://www.cbpp.org/blog/another-3815-arkansans-lost-medicaid-in-november-due-to-rigid-work-requirement.
Beneficiaries’ Confusion, Lack of Knowledge Weaken Incentive Programs
Incentives can’t be successful unless they are understood. Research examining several incentive programs in Medicaid shows beneficiaries and providers had very low awareness of the programs, and among those who were aware, many reported confusion about how they worked.
States participating in MIPCD found beneficiary engagement to be more challenging than expected. Only two of the ten states met their enrollment goals despite substantial, ongoing investment in a range of outreach activities. Overly complex programs with multiple steps also contributed to confusion and lack of engagement, CMS reported in its final report on MIPCD. According to the report, “One State said that if their design had been simpler, it might have been easier for both participants and [managed care organizations] to understand the incentive process.”
Iowa and Michigan included incentive programs aimed at increasing the use of preventive care in their Medicaid expansion waivers. In Michigan, newly enrolled adult beneficiaries have to pay co-payments for most services, but their cost-sharing is reduced if they complete a health risk assessment and agree to participate in certain activities. Focus groups in Michigan — both Medicaid beneficiaries and providers — in 2016 reported confusion about the purpose and use of health accounts and account statements. As of March 2018, less than 19 percent of beneficiaries who had been enrolled for at least six months had received credit for completing the HRA. Another 2016 survey found that of those who did complete an HRA, only 0.1 percent reported that they did so in order to save money on co-pays. 
Iowa was the first state granted permission to charge premiums to beneficiaries with incomes below the poverty line — they are charged to members between 50 and 133 percent of poverty. Enrollees don’t have to pay premiums in the first year they are enrolled and premiums are waived in later years if enrollees complete a health risk assessment and get a wellness exam. As of December 2017, more than 80 percent of beneficiaries below the poverty line and 71 percent above the poverty line had failed to complete the healthy behaviors required to have premiums waived. In a 2016 survey of individuals who had been disenrolled due to non-payment of premiums, most reported being unaware of the incentive program before they were disenrolled.
Indiana’s experience also shows that beneficiaries need to be aware of and understand incentives if these programs are to succeed. All enrollees in Indiana’s HIP 2.0 have “POWER Accounts,” a kind of Health Savings Account that contains state contributions and deposits from beneficiaries. After the first year of enrollment, enrollees in HIP Plus can use a portion of the funds remaining in their HSAs to reduce their premiums for the following year. The amount they can roll over is doubled if they get preventive care services recommended for their age and gender.
The rollover does not appear to be working effectively as an incentive for a large share of enrollees, however, because they evidently lack basic knowledge about the accounts and how they work. Only 60 percent of respondents to a survey said they had heard of the POWER Accounts, according to a 2016 evaluation of HIP 2.0. Those who said they’d heard of the accounts were asked whether they have one, and only about three-quarters of those who had heard of the accounts said they did. This means that fewer than half of all enrollees (three-quarters of the 60 percent who had heard of the accounts) even knew they had an account, when all enrollees have one. Large shares of respondents also showed a lack of understanding when answering a series of true-false questions about their POWER Accounts.
As implementation of HIP 2.0 has continued, beneficiaries’ low awareness has persisted: 39 percent reported that they had not heard of the POWER accounts, according to a survey in late 2017, while 26 percent had heard of them but were not consistently making the required payments. HIP 2.0’s lockout provision, in which non-medically frail adults with income from 100 to 138 percent of the poverty line are locked out of coverage if they fail to pay premiums within 60 days of enrollment, is also widely misunderstood: one focus group found that all enrollees, regardless of income, believed the lockout provision applied to them. One participant said, “You have to sit out for six months and pray you don’t get sick for six months,” suggesting people may be forgoing needed care.
Medicaid Programs Face Challenges in Implementing Incentive Programs
Administrative challenges are common among Medicaid agencies administering complex programs. Providers’ engagement and participation can be particularly difficult given their large patient loads and low reimbursement rates. Providers in West Virginia’s program felt that the state failed to adequately educate them as to their role in reporting whether patients completed healthy behaviors. An evaluation of Iowa’s incentive program found that although providers were compensated for completing health risk assessments with their patients, the majority of providers interviewed “either had a vague idea about the program or knew nothing about it.” Most “reported that they did not hear about this program from anyone.”
Identifying and engaging beneficiaries to participate can also be challenging due to inaccurate contact information, changes in beneficiaries’ eligibility or health status, and difficulties identifying eligible individuals.
In addition, states tend to incur significant costs in implementing and administering these programs. Rewards above and beyond the cost of services can be costly to provide, and offering them requires upfront funds and staff time. Taking the steps described above will cost states and the federal government (and in some cases counties) tens of millions of dollars for eligibility system changes, notices, and increased staff to track compliance, address questions, and handle appeals. For example, Arkansas paid over $9 million in contracts to manage their “independence” accounts and collected $426,000 in total premiums from Medicaid beneficiaries.
In the final MIPCD evaluation, the Department of Health and Human Services reported that administrative costs for the incentive programs accounted for 42 percent of overall expenditures. Penalties such as premiums and cost-sharing can also be complex to administer, and their administrative costs generally exceed the revenue they collect from beneficiaries.
Providing Access to Care Is Far More Effective Than Incentives
The evidence is clear: the threat of penalties for Medicaid beneficiaries is unlikely to substantially increase their participation in healthy behaviors. For those who are sanctioned, limiting or eliminating access to health care has a negative effect on health outcomes and can increase unnecessary use of the ED. And while reward programs may increase participation in some short-term healthy behaviors without negatively affecting care, there is little evidence that they facilitate long-term changes in behavior and better health outcomes. Many individuals who engage in the rewarded behaviors would likely have done so without the reward. This lack of targeting, combined with reward programs’ administrative costs, suggests they offer a low return on investment.
Reward programs are premised on the assumption that Medicaid beneficiaries often don’t make the healthiest choices available to them, and that this can raise Medicaid costs especially when it results in more ED visits. Studies show, however, that while Medicaid beneficiaries do use the ED more often than privately insured individuals, they also are in poorer heath than the general population and that the majority of ED visits by non-elderly Medicaid beneficiaries are appropriate. Medicaid beneficiaries have been shown to be the most likely to report that their ED visit was due to the severity of their condition, versus adults with private insurance, who were the most likely to report visiting the ED because their primary care provider’s office was closed.
Some states continue to propose steep co-pays for Medicaid beneficiaries who use the ED for non-emergency care. However, co-pays for non-emergency use of the ED didn’t change beneficiaries’ use of the ED or primary care, a recent study showed.
Rather than investing scarce financial and administrative resources in offering rewards or imposing penalties to encourage individuals to make healthy choices, states may be better served by investing in improved access to appropriate health services, better care coordination, and increased integration of health and social services. Providing transportation, for example, can enable more beneficiaries to keep appointments.
Similarly, various ED diversion and care coordination initiatives have been found to promote healthy behaviors and save states money. Successful programs in states including Georgia, Indiana, Minnesota, New Mexico, Oregon, Washington, and Wisconsin show that states can reduce ED use by expanding access to primary care services and targeting interventions to populations that use the ED frequently. For example, a recent evaluation of Oregon’s Coordinated Care Organizations — integrated, community-run organizations responsible for providing all medical, mental health, and dental care services for their members — showed that they have reduced ED utilization by 50 percent since 2011, while increasing enrollment in patient-centered primary care homes that coordinate beneficiaries’ physical and behavioral health care.
Thus, while rewards may prove to have some benefits for short-term behaviors, a longer-term investment in improving Medicaid beneficiaries’ access to timely and appropriate care is likely to provide greater returns.
 The program was enacted as part of the health reform law in section 4108 of the Affordable Care Act.
 Soeren Mattke et al., “Workplace Wellness Programs Study: Final Report,” RAND Health, 2013; Nola M. Ries, “Financial Incentives for Weight Loss and Healthy Behaviours,” Healthcare Policy, February 2012, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3298019/; Niteesh K. Choundry et al., “Assessing the Evidence for Value-Based Insurance Design,” Health Affairs, November 2010.
 R. Nieuwlaat et al., “Interventions for Enhancing Medication Adherence,” Cochrane Database Systematic Reviews, November 20, 2014, https://www.ncbi.nlm.nih.gov/pubmed/25412402; see also Kevin G. Volpp et al., “Effect of Electronic Reminders, Financial Incentives, and Social Support on Outcomes After Myocardial Infarction: The HeartStrong Randomized Clinical Trial,” JAMA Internal Medicine, August 2017, https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2633258.
 Karen J. Blumenthal et al., “Medicaid Incentive Programs to Encourage Healthy Behavior Show Mixed Results to Date and Should Be Studied and Improved,” Health Affairs, March 2013; Rob Saunders et al., “Are Carrots Good for Your Health? Current Evidence on Health Behavior Incentives in the Medicaid Program,” Duke Margolis Center for Health Policy, https://healthpolicy.duke.edu/sites/default/files/atoms/files/duke_healthybehaviorincentives_6.1.pdf; Amanda Van Vleet and Robin Rubinowitz, “An Overview of Medicaid Incentives for the Prevention of Chronic Diseases (MIPCD) Grants,” Kaiser Family Foundation, September 2014; RTI International, “Medicaid Incentives for Prevention of Chronic Diseases,” April 2017, https://downloads.cms.gov/files/cmmi/mipcd-finalevalrpt.pdf; Melinda B. Buntin, John Graves, and Nikki Viverette, “Cost Sharing, Payment Enforcement, and Healthy Behavior Programs in Medicaid: Lessons from Pioneering States,” Vanderbilt University, June 2017, https://www.vumc.org/health-policy/files/health-policy/public_files/Cost%20Sharing,%20Payment%20Enforcement,%20and%20Healthy%20Behavior%20Programs%20in%20Medicaid.pdf.
 MaryBeth Musumeci et al., “An Early Look at Medicaid Expansion Waiver Implementation in Michigan and Indiana,” Kaiser Family Foundation, January 2017, http://files.kff.org/attachment/Issue-Brief-An-Early-Look-at-Medicaid-Expansion-Waiver-Implementation-in-Michigan-and-Indiana.
 Jessica Greene, “Medicaid Efforts to Incentivize Healthy Behaviors,” Center for Health Care Strategies, July 2007, http://www.chcs.org/media/Medicaid_Efforts_to_Incentivize_Healthy_Behaviors.pdf.
 The program was enacted as part of the health reform law in section 4108 of the Affordable Care Act.
 Medicaid and CHIP Payment and Access Commission (MACPAC), “The Use of Healthy Behavior Incentives in Medicaid,” August 2016, https://www.macpac.gov/publication/the-use-of-healthy-behavior-incentives-in-medicaid.
 RTI International.
 Blumenthal et al.; Pat Redmond, Judith Solomon, and Mark Lin, “Can Incentives for Healthy Behavior Improve Health and Hold Down Medicaid Costs?” Center on Budget and Policy Priorities, June 1, 2007, https://www.cbpp.org/research/can-incentives-for-healthy-behavior-improve-health-and-hold-down-medicaid-costs.
 RTI International.
 People with incomes above the poverty line must pay premiums and enroll in HIP Plus.
 Jesse Cross-Call, “HHS Extends Indiana's Medicaid Waiver — Though It’s Hurting Many Hoosiers,” Center on Budget and Policy Priorities, February 2, 2018, https://www.cbpp.org/blog/hhs-extends-indianas-medicaid-waiver-though-its-hurting-many-hoosiers.
 The Lewin Group, “Indiana Healthy Indiana Plan 2.0: Interim Evaluation Report,” July 6, 2016, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/in/Healthy-Indiana-Plan-2/in-healthy-indiana-plan-support-20-interim-evl-rpt-07062016.pdf The terms and conditions of the demonstration project require this interim evaluation as well as a final evaluation at the end of the three-year demonstration project.
 Judith Solomon, “Indiana Medicaid Waiver Evaluation Shows Why Kentucky’s Medicaid Proposal Shouldn’t Be Approved,” Center on Budget and Policy Priorities, August 1, 2016, http://www.cbpp.org/research/health/indiana-medicaid-waiver-evaluation-shows-why-kentuckys-medicaid-proposal-shouldnt-be/.
 Centers for Medicare & Medicaid Services (CMS), Demonstration Approval, BadgerCare Reform, October 31, 2018, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/wi/wi-badgercare-reform-ca.pdf; CMS, Demonstration Approval, Kentucky HEALTH, November 20, 2018, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ky/ky-health-ca.pdf.
 Rachel Miller, Kristin Maurer, and Katharine Bradley, “Beneficiary Understanding of Incentives: Evidence from Interim Demonstration Evaluation Reports in Indiana, Iowa, and Michigan,” July 2017, https://www.medicaid.gov/medicaid/section-1115-demo/downloads/evaluation-reports/beneficiaries-understanding-incentives.pdf.
 Tami Gurley-Calvez et al., “Impacts on Emergency Department Visits from Personal Responsibility Provisions: Evidence from West Virginia’s Medicaid Redesign,” Health Services Research, August 2016, https://www.ncbi.nlm.nih.gov/pubmed/26762205.
 RTI International 2017.
 Musumeci et al..
 Michigan Department of Health and Human Services, Medical Services Administration, Bureau of Medicaid Care Management and Quality Assurance, “Healthy Michigan Plan – Health Risk Assessment Report,” March 2018, https://www.michigan.gov/documents/mdch/HMP_HRA_Report_FINAL_468616_7.pdf.
 Susan Door Goold and Jeffrey Kullgren, “Report on the 2016 Healthy Michigan Voices Enrollee Survey,” University of Michigan Institute for Healthcare Policy and Innovation, January 17, 2018, https://www.michigan.gov/documents/mdhhs/2016_Healthy_Michigan_Voices_Enrollee_Survey_-_Report__Appendices_1.17.18_final_618161_7.pdf.
 Iowa Department of Human Services, “Iowa Health and Wellness Plan Annual Report,” April 2018, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ia/Wellness-Plan/ia-wellness-plan-annl-rpt-2017.pdf.
 Natoshia Askelson et al., “Healthy Behaviors Dis-enrollment Interviews Report: In-depth interviews with Iowa Health and Wellness Plan members who are recently disenrolled due to failure to pay required premiums,” University of Iowa Public Policy Center, January 2017, http://ppc.uiowa.edu/sites/default/files/ihawp_disenrollee_report.pdf.
 Lewin Group.
 Benjamin D. Sommers et al., “New Approaches in Medicaid: Work Requirements, Health Savings Accounts, and Health Care Access,” Health Affairs, July 2018, https://www.healthaffairs.org/doi/10.1377/hlthaff.2018.0331.
 Musumeci et al.
 Michael Hendryx et al., “Evaluation of Mountain Health Choices: Implementation, Challenges, and Recommendations,” Robert Wood Johnson Foundation, August 2009, https://www.rwjf.org/en/library/research/2009/08/evaluation-of-mountain-health-choices.html.
 Natoshia M. Askelson et al., “Health Behaviors Incentive Program Evaluation Interim Report,” March 1, 2016, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ia/Wellness-Plan/ia-wellness-plan-bhvrs-int-rpt-mar-2016.pdf.
 Buntin et al.
 Jennifer Wagner and Judith Solomon, “States’ Complex Medicaid Waivers Will Create Costly Bureaucracy and Harm Eligible Beneficiaries,” Center on Budget and Policy Priorities, May 23, 2018, https://www.cbpp.org/research/health/states-complex-medicaid-waivers-will-create-costly-bureaucracy-and-harm-eligible.
 Joseph Thompson et al., “Arkansas Experience with Health Savings Accounts in a Medicaid Expansion Population,” AcademyHealth Annual Research Meeting, June 27, 2017, https://academyhealth.confex.com/academyhealth/2017arm/meetingapp.cgi/Paper/18272.
 RTI International.
 Thompson et al. See also Jessica Schubel and Judith Solomon, “States Can Improve Health Outcomes and Lower Costs in Medicaid Using Existing Flexibility,” Center on Budget and Policy Priorities, April 9, 2015, https://www.cbpp.org/research/health/states-can-improve-health-outcomes-and-lower-costs-in-medicaid-using-existing.
 Gurley-Calvez et al.; Sommers et al.; Samantha Artiga, Petry Ubri, and Julia Zur, “The Effects of Premiums and Cost Sharing on Low-Income Populations: Updated Review of Research Findings,” Kaiser Family Foundation, June 1, 2017, https://www.kff.org/medicaid/issue-brief/the-effects-of-premiums-and-cost-sharing-on-low-income-populations-updated-review-of-research-findings/; Astha Singhal et al., “Eliminating Medicaid Adult Dental Coverage in California Led to Increased Dental Emergency Visits and Associated Costs,” Health Affairs, May 2015; Roy Grant, “In Kentucky’s New Medicaid Plan Evidence Takes A Back Seat,” Health Affairs, August 2016.
 Tamyra Carrol Garcia et al., “Emergency Department Visitors and Visits: Who Used the Emergency Room in 2007?” Centers for Disease Control and Prevention, NCHS Data Brief No. 38, May 2010; Anna Sommers et al., “Dispelling Myths About Emergency Department Use: Majority of Medicaid Visits Are for Urgent or More Serious Symptoms,” Center for Studying Health System Change, July 2012.
 Renee M. Gindi et al., “Reasons for Emergency Room Use Among U.S. Adults Aged 18-64: National Health Interview Survey, 2013 and 2014,” National Health Statistics Reports, February 2016.
 Mona Siddiqui, Eric T. Roberts, Craig E. Pollack, “The Effect of Emergency Department Co-payments for Medicaid Beneficiaries Following the Deficit Reduction Act of 2005,” JAMA Internal Medicine, January 26, 2015.
 Samina Syed, Ben S. Gerber, and Lisa K. Sharp, “Traveling Towards Disease: Transportation Barriers to Health Care Access,” Journal of Community Health, October 2013, https://www.ncbi.nlm.nih.gov/pubmed/23543372.
 Oregon Health Authority, “Oregon’s Health System Transformation Quarterly Legislative Report,” September 2016, http://www.oregon.gov/oha/analytics/Documents/LegislativeReport_Q1_2016.pdf.