HHS: 'Medicaid Demonstrations & Impacts on Health Coverage – Review of Evidence'

WASHINGTON, March 14 -- The U.S. Department of Health and Human Services issued the following issue brief (No. HP-2021-03) on March 11, 2021, entitled "Medicaid Demonstrations and Impacts on Health Coverage: A Review of the Evidence":

KEY POINTS

* State Medicaid section 1115 demonstrations are often used to test innovative or new policies in the program, but some demonstrations that have placed new conditions on eligibility have led to unintended coverage losses and other adverse effects.

* Multiple studies indicate that Medicaid work requirements (also called "community engagement" requirements) can lead to significant coverage losses and worse access to care, without improvements in employment, job training, or other related activities.

* Other demonstration programs, including those using health savings account-like arrangements or healthy behavior incentives, are frequently confusing and produce administrative challenges for beneficiaries, with some evidence that these harmful effects are larger for racial and ethnic minorities.

INTRODUCTION

On January 28, 2021, President Biden released an Executive Order declaring the policy of the Administration to protect and strengthen Medicaid and the Affordable Care Act (ACA)./1

Section 3 of this Executive Order calls for the immediate review of agency actions related to Medicaid and the ACA, including demonstrations and waiver policies that may reduce coverage.

This issue brief examines policies in four major areas of state Medicaid section 1115 demonstrations: 1) work requirements (also referred to as "community engagement" requirements), 2) healthy behavior incentive programs, 3) health savings account-like arrangements, and 4) capped federal funding and other financing changes. Section 1115 of the Social Security Act gives the Secretary of Health and Human Services authority to approve experimental, pilot, or demonstration projects that, in the Secretary's judgment, are likely to promote the objectives of Medicaid and the Children's Health Insurance Program (CHIP). This issue brief reviews the evidence to date on the impact of the above-mentioned demonstration policies on Medicaid coverage and access to care.

WORK REQUIREMENTS

Many states have sought to implement community engagement demonstrations, which require beneficiaries to report work or other qualifying activities to the state in order to maintain Medicaid eligibility or receive additional benefits. These demonstrations have the stated goals of improving beneficiary health, employment, and income.

A total of 23 states have submitted section 1115 demonstration projects that include work requirement policies to the Centers for Medicare & Medicaid Services (CMS), 13 of which are in states that have adopted Medicaid expansion under the ACA./2

Of these 23 demonstration projects, 13 were approved by CMS (4 approvals have been blocked by the courts), 8 are pending CMS review, and 2 were rejected by the state after CMS approval.

* Five states with approved work requirements - Arkansas, Indiana, Kentucky, Michigan, and New Hampshire - proceeded far enough in their demonstrations to obtain early data on potential enrollment effects. Only Arkansas proceeded far enough in its implementation to disenroll any beneficiaries due to noncompliance. In many cases, this was due to beneficiaries not reporting their work or community engagement activities as required by the demonstration.

Arkansas was the first state to implement a work requirement as a condition of eligibility for its Medicaid program. In order to maintain eligibility for coverage, Medicaid beneficiaries subject to the requirement had to report completion of 80 hours per month of work or another approved community engagement activity or a qualifying exemption. Qualifying community engagement activities included job training, job search, education, and community service activities, while exemptions included pregnancy, certain caregiving activities, and experiencing certain medical conditions.

Implemented in June 2018, Arkansas' work requirement policy was in effect until March 2019, when a federal court halted it. During this period, approximately 18,000 beneficiaries were disenrolled from Medicaid due to noncompliance. A research study found a significant increase in the uninsured rate in the target population (low-income 30 to 49 year-olds), with no increase in employment or other community engagement activities./3

Overall, work requirements led to a reduction in the share of this population who had Medicaid coverage by 12 percentage points. While most coverage losses were reversed in 2019 when the policy was halted, the temporary loss of Medicaid coverage had adverse consequences. One study found that adults with chronic conditions in Arkansas were more likely to lose coverage,4 and another found that many who lost Medicaid coverage in 2018 experienced negative effects. Specific findings include:

* 50 percent reported serious problems paying off medical bills;

* 56 percent delayed seeking health care because of cost; and

* 64 percent delayed taking medications because of cost./5

Moreover, research suggests that more than 95 percent of adults in this population were already meeting the work requirements or should have qualified for an exemption. A survey of low-income individuals in Arkansas ages 30 to 49 found that, in 2017, 42 percent reported working more than 20 hours per week (thus meeting the 80 hours per month work requirement), 39 percent reported they were disabled and therefore should have qualified for an exemption, and 15 percent had another qualifying activity or exemption such as community service, job training, or caring for a family member./5

The decrease in Arkansas Medicaid enrollment is likely explained by a pervasive lack of awareness and confusion among many Medicaid beneficiaries about reporting requirements related to community engagement activities./6

One year after implementation began, a survey of individuals subject to work requirements found one-third of them had not heard anything about the policy, while 44 percent were unsure whether the policy applied to them./7

Evidence suggests awareness of state work requirement provisions was lower among beneficiaries with less education./8

These findings from Arkansas are consistent with data from New Hampshire and Michigan. New Hampshire's Medicaid work requirement policy would have removed 17,000 beneficiaries from the program (67 percent of those subject to the requirements) had it not been halted by the courts./9

In Michigan, 80,000 beneficiaries (almost one-third of those subject to the requirements) were in danger of losing coverage for failing to submit documentation when the policy was vacated by the courts./10

In both New Hampshire and Michigan, the potential for coverage losses was largely linked to limited awareness among beneficiaries and low rates of reporting activities to the state, similar to the findings in Arkansas./9,10,11

This evidence highlights that largescale difficulties with meeting reporting requirements have posed risks of coverage loss for many beneficiaries across multiple states implementing work requirements.

Other studies indicate that most individuals potentially subject to work requirements are either already working or face substantial barriers to employment. For example, in Kentucky (where implementation of a work requirement was blocked by a federal court before any beneficiary disenrollment took place), among those who were not exempt or working, nearly three-quarters faced one or more significant barriers to doing so: 59 percent lived in a household with someone with a serious health limitation (including 38 percent who had a serious health limitation themselves); 25 percent did not have internet access; 24 percent had not completed high school; and /11 percent did not have access to a vehicle./11

Poor physical and mental health also were more common among those in Kentucky not yet engaged in activities that would satisfy the work requirements./12

A survey in Arkansas found that among beneficiaries who were not employed but wanted to work, more than 80 percent indicated that job training and 70 percent reported that assistance with transportation to work would boost their ability to find a job if these services were provided by the state./13

Obstacles to finding new employment have likely grown even larger during the pandemic-related economic downturn of the past year, particularly since job and income losses have been highest among low-income and minority workers, who are disproportionately enrolled in Medicaid./14

Fifty-two percent of lower income Americans live in households where someone has lost a job or taken a pay cut due to the pandemic, compared to 32 percent in upper income households./15

Minorities also experience greater rates of unemployment: in December 2020, the unemployment rate for black workers was 9.9 percent compared to 6.0 percent for whites, and 9.2 percent for Hispanic workers compared to 5.9 percent for non-Hispanic workers./16

Households that experienced a pandemic-related job or income loss were two to three times more likely to experience economic hardship than those who did not experience a loss./17

In recognition of some of these challenges, one state - Utah - suspended its planned community engagement requirements in April 2020, citing rising unemployment and "the unique challenges" created by the pandemic./18

Furthermore, consistent with the requirements for receiving an increased Federal Medical Assistance Percentage (FMAP) in the Families First Coronavirus Recovery Act, no state is currently implementing community engagement requirements.

HEALTHY BEHAVIOR INCENTIVES

Several states have designed healthy behavior incentive programs with the goals of improving health outcomes, reducing costs, and increasing patient involvement in health care. Nine states have received approval to implement healthy behavior incentives in their Medicaid section 1115 demonstration programs, although two states' demonstrations (Kentucky and Michigan) have been blocked by the courts, and Kentucky later withdrew this demonstration authority. Five of these state demonstrations apply to non-expansion populations./2

These incentive programs typically promote participation in certain activities theorized to help improve beneficiaries' health. For example, Medicaid beneficiaries in Wisconsin, Indiana, Michigan, and Iowa were encouraged through financial incentivizes or required as a condition of continued eligibility (depending on the state) to engage in certain healthy behaviors, such as completing a wellness exam and/or a health risk assessment (HRA). The programs included incentives such as reduced premium amounts for completing specific behaviors, and in two states, completing an HRA was a condition of continued eligibility for some beneficiaries./19

Healthy behavior incentive programs have experienced low rates of completion of required activities due to limited program awareness. In Michigan, despite financial incentives encouraging beneficiaries to do so, only 27 percent of beneficiaries in the Medicaid expansion population completed an HRA. Of those who completed the assessment, however, nearly 90 percent selected a healthy behavior to change. Enrollees who agreed to address at least one healthy behavior change had lower emergency department visit rates, compared to those who did not complete an HRA, though this may reflect underlying differences between those two groups rather than an effect of the HRA itself./20

In Iowa, less than 20 percent of enrollees complied with requirements in the first two years of the program, leaving the other 80 percent of enrollees potentially subject to higher premiums and/or disenrollment for premium nonpayment./21

Research in Iowa indicates that members who were younger, male, or racial or ethnic minorities were at significantly higher risk of not completing the required activities./22

The state conducted a disenrollment survey in 2019, which found that only 39 percent of respondents subject to the healthy behavior program had heard of it. Sixty percent had not completed an HRA, and most cited being unaware they were supposed to complete an HRA as the reason. Moreover, 22 percent of respondents were unaware they had been disenrolled and, of those who were aware, 70 percent had done nothing to prepare for disenrollment. Disenrollment had a significant impact on the respondents, with 54 percent lacking health coverage at the time of survey and many reporting delays filling prescriptions or delays obtaining needed medical care./23

A multi-state federal evaluation conducted by CMS found some evidence that financial incentives can increase rates of preventive visits, but found mixed results on whether these policies improve chronic disease management or reduce emergency department visits./24

Despite these challenges, research suggests that states that expanded Medicaid while incorporating healthy behavior incentive programs still experienced substantial coverage gains and improved access to care, compared to non-expansion states./25

HEALTH SAVINGS ACCOUNT-LIKE ARRANGEMENTS

Five states have had section 1115 demonstration projects approved that incorporate health savings account-like arrangements in their Medicaid programs. Again, Kentucky's demonstration (which had features of both healthy behavior incentives and a rewards account similar to a health savings account) has been blocked by the courts and was later withdrawn by the state./2

While these provisions are intended to increase beneficiary involvement in their health care, research across several states indicates that many beneficiaries do not participate in these consumer-directed provisions, often because of confusion or lack of awareness./26

For example, Medicaid beneficiaries participating in Indiana's demonstration with incomes above 100 percent of the federal poverty level (FPL) were required to make monthly contributions to POWER health accounts and faced a coverage lockout period for non-payment. Those below 100 percent of FPL were also required to make monthly contributions, but in the event of non-payment, they stayed enrolled in Medicaid but lost vision and dental benefits. Many individuals lost or did not obtain coverage due to these provisions. A state evaluation identified 46,000 people who did not enroll in coverage between February 2015 and November 2016 due to failure to meet the initial POWER account contribution requirement./27

One survey in 2015 found 39 percent of demonstration enrollees had not heard of POWER accounts and 26 percent had heard of them but were not making the required payments. Only 36 percent of eligible enrollees were making required payments, with the remaining nearly two-thirds potentially subject to loss of benefits or coverage. Nine percent reported being locked out of coverage for failure to pay. Half of individuals cited either affordability (31 percent) or confusion about the accounts (19 percent) as the main reason for nonpayment. Latinos, men, and those with less education were significantly less likely to have heard about the accounts compared to whites, women, and those with more education./26,28

These findings are consistent with the federal evaluation of Indiana's program, which found that POWER accounts were administratively burdensome to operate, enrollee understanding was low, and former beneficiaries reported forgoing medical care after being disenrolled from Medicaid coverage for failure to make required monthly payments. However, the evaluation also concluded that coverage rates in Indiana in 2017-2018 were significantly higher than would have occurred if the state had not expanded Medicaid and were similar to what would have occurred had Indiana expanded Medicaid without a demonstration./29

By 2019, four years into the demonstration, a second state evaluation in Indiana found that a higher share of respondents - 77 percent - were making POWER account contributions, and most members interviewed had a general understanding of the POWER accounts. However, nearly 20 percent reported difficulties making payments and only 19 percent understood the consequences of non-payment. Black beneficiaries had a higher likelihood of disenrollment compared to white beneficiaries. Non-payment of POWER account contributions represented a small percentage (less than 2 percent) of reasons for disenrollment. Of all disenrolled members, there was a large share of individuals reporting disenrollment due to failure to verify information (22 percent) or submit paperwork for redetermination (23 percent). /30

Overall, the findings from Indiana are similar to those from a study of Michigan's experience. In Michigan's demonstration, beneficiaries have health savings account-like arrangements and are required to make monthly payments; unpaid monthly premiums can become a collectible debt (though enrollees maintain coverage)./31

A state evaluation found of all enrollees who owed payments, 23 percent paid in full, while 48 percent had made none of the required payments./20

The evaluation concluded that required monthly contribution amounts may increase disenrollment among beneficiaries, particularly those without chronic conditions./32

CAPPED FEDERAL FUNDING AND OTHER FINANCING CHANGES

The traditional Medicaid financing structure is open-ended, allowing the 50 states and the District of Columbia to receive federal funds based on their actual expenditures for the Medicaid program, which enables funding streams to rise and fall according to a state's needs in changing economic circumstances. Through the Federal Medical Assistance Percentage (FMAP), the federal government funds a portion of most state Medicaid service expenditures using a statutorily defined formula based on state per-capita income. The statutory formula results in minimum and maximum FMAP rates of 50 percent and 83 percent, respectively. In Fiscal Year 2021, the base FMAP percentage ranged from 50 percent in 13 states to 77.76 percent in Mississippi./,33

In some circumstances, such as the Medicaid expansion under the ACA, states are given an enhanced matching rate. The federal government initially paid 100 percent of the service costs for newly eligible beneficiaries in the Medicaid expansion population, which gradually decreased to 90 percent in 2020 and thereafter./34

In recent years, there has been interest among some federal and state policymakers to make changes to this structure to cap or limit federal spending, sometimes through what is referred to as a block grant approach.

In January 2020, CMS introduced the Healthy Adult Opportunity (HAO) initiative, which invited states to submit section 1115 demonstration applications to make major changes to Medicaid financing./35

Under an HAO demonstration, states agree to a limit on federal funding in exchange for waivers to many federal requirements related to Medicaid eligibility, benefits, delivery systems, and program oversight. Oklahoma is the only state that applied for the HAO demonstration, but the state withdrew its application in August 2020 due to passage of a state referendum on Medicaid expansion.

Another state - Tennessee - proposed an alternative financing arrangement with some similarities to the HAO model. In January 2021, CMS approved Tennessee's section 1115 demonstration request to change the state's financing structure to an aggregate cap, with new broader state authority in how funding can be used, for a 10 year period./36

Under this agreement, CMS and Tennessee will work together to establish a funding cap over the life of the demonstration, which will increase with enrollment. The cap has been established for the first five years and will be re-established, based on the first five years of actual expenditures, for the second five years of the demonstration period. However, unlike under Medicaid's traditional financing arrangement, federal Medicaid funding for Tennessee will not automatically keep up with rising per-person Medicaid expenses./37

Thus, if per-beneficiary costs rise faster than the negotiated inflation factor, the state would be at financial risk./38

Critics of this approach fear the potential for enrollment cuts and benefit reductions and have pointed to the increased importance of Medicaid and federal funds during public health emergencies and economic downturns, when states typically face declining revenues and the need for Medicaid coverage increases./39

Since Oklahoma withdrew its application for the HAO demonstration and Tennessee's has not yet been implemented, there is no evidence to date on the impact of a capped federal funding model on coverage or access to care among program beneficiaries.

OTHER ENROLLMENT-RELATED POLICIES

The demonstration projects discussed in this report are not the only policies that risk leading to coverage losses within Medicaid. States have made other changes under section 1115 and other existing authorities such as imposing premiums,/40 eliminating or reducing the statutory period of retroactive eligibility,/41 and increased verification and periodic data matching to identify potential changes in circumstances,/42 all of which can create barriers to enrollment and reduce coverage rates. Between 2017 and 2019, Medicaid enrollment declined by 2.6 percent or nearly two million people, including over one million children./43

State-by-state trends indicate that the decline in Medicaid was too large to be explained by the improving economy, /44 and a federal evaluation linked premiums as a contributor to reduced enrollment in several states./24

At the same time, national survey data show a rise in the uninsured rate between 2016 and 2019, raising additional concern about the impact of these policy changes in Medicaid./45

However, enrollment is increasing during the public health emergency, in part due to the requirement to maintain enrollment for most Medicaid beneficiaries as a condition of states' receiving a temporary increased FMAP./43

CONCLUSION

In recent years, states have sought demonstration authority to implement work requirements, healthy behavior incentives, and health savings account-like arrangements. These initiatives and other policy flexibilities, to varying degrees, have reduced enrollment and access to care. While capped federal funding demonstrations and other financing changes have not been implemented to date, concerns about the potential for enrollment cuts and benefit reductions are prevalent in the literature.

While one of the stated motivations for several of these demonstration features has been the desire to improve beneficiary health,/31 the loss of coverage evident in multiple studies of these policies suggests they carry a significant risk of having the opposite effect - harms to access to care and adverse health effects. Given the strong evidence linking health insurance coverage to positive health and economic outcomes,/46,47 policies that lead to loss of Medicaid coverage, increase rates of uninsurance, and heighten barriers to medical care can have significant negative public health consequences, particularly during emergencies such as the current pandemic.

* * *

REFERENCES

1 The White House (2021). Executive Order on Strengthening Medicaid and the Affordable Care Act. https://www.whitehouse.gov/briefing-room/presidential-actions/2021/01/28/executive-order-on-strengtheningmedicaid-and-the-affordable-care-act/

2 Kaiser Family Foundation. (2019). Medicaid Waiver Tracker: Approved and Pending Section 1115 Waivers by State.

3 Sommers, B. D., Goldman, A. L., Blendon, R. J., Orav, E. J., & Epstein, A. M. (2019). Medicaid work requirements--results from the first year in Arkansas. New England Journal of Medicine, 381(11), 1073-1082.

4 Chen, L., Sommers, B.B. (2020). Work Requirements and Medicaid Disenrollment in Arkansas, Kentucky, Louisiana, and Texas, 2018. American Journal of Public Health, 110, 1208-1210.

5 Sommers, B. D., Chen, L., Blendon, R. J., Orav, E. J., & Epstein, A. M. (2020). Medicaid Work Requirements in Arkansas: Two-Year Impacts on Coverage, Employment, and Affordability of Care. Health Affairs, 39(9), 1522-1530.

6 Greene, J. (2018). Medicaid recipients' early experience with the Arkansas Medicaid work requirement. Health Affairs Blog, 5.

7 Sommers, B. D., Goldman, A. L., Blendon, R. J., Orav, E. J., & Epstein, A. M. (2019). Medicaid work requirements--results from the first year in Arkansas. New England Journal of Medicine, 381(11), 1073-1082.

8 Greene, J. (2019). Medicaid Recipient Awareness of Work Requirements: Importance and Challenges. Health Affairs Blog, December 10, 2019. DOI: 10.1377/hblog20191205.883142

9 Hill, I., Burroughs, E., & Adams, G. (2020). New Hampshire's Experience with Medicaid Work Requirements: New Strategies, Similar Results.

10 Wagner, J., & Schubel, J. (2020). States' experiences confirming harmful effects of Medicaid work requirements. Center on Budget and Policy Priorities.

11 Gangopadhyaya, A., & Kenney, G. M. (2018). Updated: Who Could Be Affected by Kentucky's Medicaid Work Requirements, and What Do We Know about Them?. Washington, DC: Urban Institute.

12 Venkataramani, A. S., Bair, E. F., Dixon, E., Linn, K. A., Ferrell, W., Montgomery, M., & Underhill, K. (2019). Assessment of Medicaid beneficiaries included in community engagement requirements in Kentucky. JAMA network open, 2(7), e197209-e197209.

13 Sommers, B. D., Chen, L., Blendon, R. J., Orav, E. J., & Epstein, A. M. (2020). Medicaid Work Requirements in Arkansas: Two-Year Impacts on Coverage, Employment, and Affordability of Care. Health Affairs, 39(9), 1522-1530.

14 Gangopadhyaya, A., Garrett, B. (2020). Unemployment, Health Insurance, and the COVID-19 Recession. Urban Institute.

15 Parker, K., Horowitz, J.M., & Brown, A. (2020). About Half of Lower-Income Americans Report Household Job or Wage Loss Due to COVID-19. Pew Research Center.

16 Congressional Research Service (2021). Unemployment Rates During the COVID-19 Pandemic: In Brief.

17 Despard, M., Weiss-Grinstein, M., Chun, Y. & Roll, S. (2020). COVID-19 job and income loss leading to more hunger and financial hardship. Brookings Institution.

18 Meyer, H. (2020). Utah suspends Medicaid work requirement due to COVID-19. Modern Healthcare.

19 Kaiser Family Foundation (2020). The Landscape of Medicaid Demonstration Waivers Ahead of the 2020 Election.

20 Clark, S., Cohn, L., & Ayanian, J. (2018). Healthy Michigan Plan Evaluation: Report on Health Behaviors, Utilization, and Health Outcomes in the Healthy Michigan Plan.

21 Askelson, N. M., Wright, B., Brady, P. J., Jung, Y. S., Momany, E. T., McInroy, B., & Damiano, P. (2020). Implementation Matters: Lessons From Iowa Medicaid's Healthy Behaviors Program: Iowa's Medicaid Health Behaviors Program provides incentives to enrollees to receive a wellness exam and complete a health risk assessment annually to waive a monthly premium. Health Affairs, 39(5), 884-891.

22 Wright, B., Askelson, N. M., Ahrens, M., Momany, E., Bentler, S., & Damiano, P. (2018). Completion of requirements in Iowa's Medicaid expansion premium disincentive program, 2014-2015. American journal of public health, 108(2), 219223.

23 Public Policy Center University of Iowa (2020). Iowa Wellness Plan Summative Evaluation Report For the Period Ending 12/31/19.

24 Bradley, K., Niedzwiecki, M., Maurer, K., Chao, S., Natzke, B., & Samra, M. (2020). Medicaid Section 1115 Demonstrations Summative Evaluation Report: Premium Assistance, Monthly Payments, and Beneficiary Engagement. Mathematica.

25 Nelson, D. B., Sommers, B. D., Singer, P. M., Arntson, E. K., & Tipirneni, R. (2020). Changes in Coverage, Access, and Health Following Implementation of Healthy Behavior Incentive Medicaid Expansions vs. Traditional Medicaid Expansions. Journal of general internal medicine, 35(9), 2521-2528.

26 Sommers, B. D., Fry, C. E., Blendon, R. J., & Epstein, A. M. (2018). New approaches in Medicaid: work requirements, health savings accounts, and health care access. Health Affairs, 37(7), 1099-1108.

27 The Lewin Group (2017). Healthy Indiana Plan 2.0: POWER Account Contribution Assessment.

28 Buntin, M. B., Graves, J., & Viverette, N. (2017). Cost sharing, payment enforcement, and healthy behavior programs in Medicaid: lessons from pioneering states.

29 Long, S.K., Coughlin, T. A., Ramos, C., Bart, L., Samuel-Jakubos, H., Govil, M. & Peterson, G. (2020). Federal Evaluation of Indiana's Healthy Indiana Plan -- HIP 2.0. The Urban Institute Social & Scientific Systems.

30 The Lewin Group (2019). Healthy Indiana Plan Interim Evaluation Report: Final for CMS Review.

31 CMS (2017). Beneficiary Engagement Strategies In Medicaid Demonstrations.

32 Hirth, R., Cliff, E., Kullgren, J., Fendrick, A., Clark, S., Beathard, E. & Ayanian, J. (2018). Healthy Michigan Plan Evaluation: Report on the Impact of Cost Sharing in the Healthy Michigan Plan.

33 Federal Financial Participation in State Assistance Expenditures; Federal Matching Shares for Medicaid, the Children's Health Insurance Program, and Aid to Needy Aged, Blind, or Disabled Persons for October 1, 2020 Through September 30, 2021, 84 FR 66204 (December 3, 2019).

34 Kaiser Family Foundation (2015). Medicaid Financing: How Does it Work and What are the Implications?

35 CMS (2020). SMD# 20-001 RE: Healthy Adult Opportunity.

36 CMS (2021). CMS Approves Innovative Tennessee Aggregate Cap Demonstration to Prioritize Accountability for Value and Outcomes.

37 Galewitz, P. (2021). Trump Administration Approves First Medicaid Block Grant, in Tennessee. Kaiser Health News.

38 Buntin, M. B. (2019). Tennessee's opening bid for a Medicaid Block Grant. N Engl J Med, 381(18), 1701-1703.

39 Rosenbaum, S., Somodevilla, A., Handley, M., & Morris, R. (2019). Inside Tennessee's Final 1115 Medicaid Block Grant Proposal. Health Affairs Blog, December 6, 2019. DOI: 10.1377/hblog20191205.927228.

40 Dague, L. (2014). The effect of Medicaid premiums on enrollment: A regression discontinuity approach. Journal of Health Economics, 37, 1-12.

41 Medicaid and CHIP Payment and Access Commission (2019). Medicaid Retroactive Eligibility: Changes under Section 1115 Waivers.

42 Kaiser Family Foundation (2019). Recent Medicaid/CHIP Enrollment Declines and Barriers to Maintaining Coverage.

43 Kaiser Family Foundation (2019). Analysis of Recent Declines in Medicaid and CHIP Enrollment.

44 Broaddus, M. (2019). Research Note: Medicaid Enrollment Decline Among Adults and Children Too Large to Be Explained by Falling Unemployment. Center on Budget and Policy Priorities.

45 Finegold, K., Conmy, A., Chu, R.C., Bosworth, A., & Sommers, B.D. (Issue Brief No. HP-2021-02). Washington, DC: Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. February 11, 2021.

46 Sommers, B.D., Gawande, A.A., Baicker, K. (2017). Health Insurance Coverage and Health - What the Recent Evidence Tells Us. N Engl J Med, 377(6):586-593.

47 Goldin J., Lurie I.Z., McCubbin J. (2021). Health Insurance and Mortality: Experimental Evidence from Taxpayer Outreach. Q J Econ, 136(1):1-49.

* * *

View full text of the report at https://aspe.hhs.gov/system/files/pdf/265161/medicaid-waiver-evidence-review.pdf