To date, 4 states (Arkansas, Indiana, Kentucky, and New Hampshire) have federal waivers to impose work requirements as a condition of eligibility for Medicaid (although a judge recently stayed Kentucky’s waiver1), and 7 other states (Arizona, Kansas, Maine, Mississippi, Ohio, Utah, and Wisconsin) have submitted waiver applications. The governor of Kentucky2 claimed that excluding “able-bodied” adults will reduce Medicaid enrollment by 16%, ensuring the program’s “fiscal sustainability.” However, such claims may be overstated because many Medicaid enrollees already work, and waivers have specified that many others (eg, people with disabilities and caregivers of young children) will be exempted. To our knowledge, no studies have quantified the potential influence of work requirements on Medicaid spending. We estimated the number of Medicaid enrollees at risk of losing coverage if work requirements are implemented with the exemptions specified in approved waiver applications, and we calculated current Medicaid spending for those enrollees at the national level and among states with approved or pending waivers.
The Institutional Review Board at Cambridge Health Alliance deemed this study exempt from review and waived the need for informed consent of participants. We identified characteristics of Medicaid enrollees and Medicaid spending on their behalf using the 2015 Medical Expenditure Panel Survey.3 We calculated Medicaid enrollment and expenditures for categories of persons already fulfilling or exempt from work requirements specified in approved waiver applications (Kentucky [https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ky/ky-health-ca.pdf], Indiana [https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/in/Healthy-Indiana-Plan-2/in-healthy-indiana-plan-support-20-ca.pdf], Arkansas [https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ar/ar-works-ca.pdf], and New Hampshire [https://www.dhhs.nh.gov/pap-1115-waiver/documents/nh-pap-stcs-05072018.pdf]) including persons with the following characteristics: older than 64 years or younger than 19 years, employed or actively seeking work, receiving Supplemental Security Income for disability, disabled but not receiving Supplemental Security Income, and serving as caregivers for disabled household members or children younger than 6 years, as well as those who are pregnant, students, or medically frail. These exemptions are common to all approved waivers.
Using total Medicaid enrollment and total Medicaid spending from the Medicaid Budget and Expenditure System, we then calculated the proportion of total Medicaid enrollment and total Medicaid spending accounted for by the residual group of nonworking, nonexempt adults (“work-eligible adults”). We estimated enrollment figures and spending for the work-eligible adults nationally and for states requesting waivers.4,5 We repeated our national analysis using varying age limits, including 50 years and older (Arkansas waiver) and 60 years and older (Indiana waiver). All analyses account for the complex survey design of the MEPS.
Most Medicaid enrollees were already working or exempt (Table 1). If work requirements were applied nationally with all of the exemptions applied, 2.1 million persons could be at risk for disenrollment, representing 2.8% of current Medicaid enrollees, who account for 0.7% (95% CI, 0.4%-1.1%) of Medicaid spending (Table 2). Among those at risk for disenrollment, the mean annual income is $3820, 42.0% are of black or Hispanic race/ethnicity, 18.6% have children younger than 18 years, 26.6% have hypertension, 33.7% have depression/anxiety, and 31.0% report an activity limitation.
In the 6 states (Arizona, Arkansas, Indiana, Kentucky, New Hampshire, and Ohio) that have applied for waivers and have expanded Medicaid through the Patient Protection and Affordable Care Act, the affected group represents 3.4% (95% CI, 1.6%-5.2%) of the current total Medicaid enrollment and 1.1% (95% CI, 0.3%-1.9%) of total Medicaid spending (Table 2). In nonexpansion states that have applied for waivers (Kansas, Maine, Mississippi, Utah, and Wisconsin), work requirements would affect 0.8% (95% CI, 0.0%-1.6%) of current total Medicaid enrollment and 0.04% (95% CI, 0.0% to 0.3%) of total Medicaid spending.
After applying exemptions and removing working adults, 2.8% of current Medicaid enrollees would be subject to new work requirements nationally; they account for less than 1% of Medicaid spending. In states that applied for waivers but did not expand Medicaid, potential savings represent less than 0.1% of Medicaid spending.
Our study was limited by several factors. The analysis likely overestimates the population size and Medicaid spending of the work-eligible group because accurate data on several exemption or work categories were unavailable, including addiction or behavioral health treatment, homelessness, volunteer work, and history of foster care. In addition, exemptions specified in waivers vary among states, and we opted to only include exemptions that were present consistently in all states with approved waivers. Exemptions present in only a subset of waivers, such as receipt of unemployment benefits, were not included, which may also have resulted in overestimation of the population size and Medicaid spending by individuals affected by work requirements. Conversely, we may have underestimated the population size and Medicaid spending by the work-eligible group because we did not have information on the number of hours worked per week by employed individuals (adults must work 20 hours per week or 80 hours per month to meet the work requirement); therefore, we classified all individuals reporting employment as fulfilling work requirements. We also classified individuals with an exempt status lasting less than a full year, such as pregnancy, as exempt for the full year. We did so because data on Medicaid expenditures are only available as annual totals, making it impossible to calculate expenditures for less than a full year. Furthermore, there is wide variability in the frequency of redetermination of exemption: in Kentucky, redetermination is required only on an annual basis, while redetermination is monthly in Arkansas. Finally, data in the MEPS are self-reported and thus subject to inaccuracy and may not correspond with data obtained through Medicaid documentation.
Results of our analysis indicate that savings to Medicaid would be minimal if exemptions were precisely applied. However, savings might be much larger (as Kentucky’s governor2 suggested) if work requirements caused a “spillover effect” (ie, loss of coverage by exempt Medicaid enrollees unable to comply with documentation requirements).6 If spillover loss of coverage is the primary mechanism by which work requirements decrease Medicaid enrollment and spending, these savings would likely come at substantial cost in terms of human health.
Accepted for Publication: July 14, 2018.
Corresponding Author: Anna L. Goldman, MD, MPH, Harvard T.H. Chan School of Public Health, 677 Huntington Avenue, Room 430, Boston, MA 02115 (firstname.lastname@example.org).
Published Online: September 10, 2018. doi:10.1001/jamainternmed.2018.4194
Author Contributions: Dr Goldman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: Goldman, Woolhandler, Himmelstein.
Drafting of the manuscript: Goldman.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Goldman, McCormick.
Obtained funding: Bor.
Administrative, technical, or material support: McCormick.
Supervision: Woolhandler, Himmelstein, Bor, McCormick.
Conflict of Interest Disclosures: None reported.
Funding/Support: Dr Goldman’s salary was supported by National Research Service Award in Primary Care grant T32HP12706 from the National Institutes of Health. Dr McCormick’s salary was supported by internal funds from Cambridge Health Alliance.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.