Abstract
Policymakers have intensified calls to expand work requirements in Medicaid across the United States, which could have implications for low-income adults who experience a high burden of cardiometabolic risk factors and disease. In this difference-in-differences analysis, we found that the implementation of Medicaid work requirements was associated with decreased health insurance coverage, no change in employment status, and a trend towards worse access to care. Our findings suggest that the expansion of work requirements could have major implications for the cardiovascular health of working-age adults in the US.
Introduction
Medicaid work requirements, which require that able-bodied adults participate in work-related activities to remain eligible for Medicaid, have been at the forefront of national health policy debate in the United States.1 Proponents argue that these policies increase employment, reduce poverty, and improve health, while critics have raised concerns that work requirements may lead to loss of coverage and impede access to care.1 Policymakers are now intensifying calls to expand work requirements across the US, and some states have recently adopted this policy change.1,2 Low-income adults have a disproportionately high burden of cardiovascular risk factors and disease and may be more likely to suffer adverse clinical sequalae due to disruptions in insurance coverage and access to care3. Therefore, the implementation of Medicaid work requirements could have major implications for the cardiovascular health of this population. In this study, we evaluated the impact of Medicaid work requirements in Arkansas on working-aged adults with and without cardiovascular disease.
Methods
We used the Behavioral Risk Factor Surveillance System (BRFSS), a nationally representative interview administered to more than 400,000 participants annually by the Centers for Disease Control and Prevention (CDC). The BRFSS collects state-level information about health-related behaviors, health conditions, and healthcare utilization in the United States. We used the BRFSS to identify adults eligible for work requirements (30–49 years old) in Arkansas and in a control group of similar, neighboring states (Kentucky and Louisiana) that did not implement work requirements. Arkansas was selected because it was the first state to successfully implement work requirements in its Medicaid program between June 2018 and March 2019. The BRFSS survey response rate was 50.4%. The pre- and post-implementation periods were defined as January 2017 to July 2018 and August 2018 to December 2019, respectively. The post-implementation period included the first five months after work requirement de-implementation to capture possible delayed policy effects.
We fit survey-weighted linear regression models to compare changes in outcomes in Arkansas versus control states, adjusted for age, sex, and income level. Our models included interaction terms between states (Arkansas versus control) and time period (pre- versus post-policy implementation). The coefficients of the interaction terms represented the difference in outcomes between adults residing in Arkansas versus control states during the post-period as compared with the pre-period. Outcomes were self-reported and included: 1) insurance coverage, 2) employment, 3) presence of a personal healthcare provider, 4) inability to see a doctor due to cost, and 5) self-reported good health. We then repeated this analysis among the subset of adults age 30–49 years with cardiovascular risk factors and/or disease (hypertension, hyperlipidemia, diabetes, myocardial infarction, stroke).
Survey weights, which account for nonresponse and noncoverage, were applied to generate state-representative estimates. Complete case analyses were performed given low missingness rates (<1% for all outcomes). Two-sided p-value <0.05 defined statistical significance. All analyses were performed using SAS EG version 7.15 (SAS Institute Inc). IRB approval was not required due to the use of publicly available de-identified data, as per institutional policy.
Results
The unweighted study population included 12,298 adults aged 30–49 years (2,673 in Arkansas and 9,625 in the control states). Prior to policy implementation, the weighted mean age (39.0 vs 39.2 years, p 0.34), proportion female (50.8% vs 49.7%, p=0.60), and distribution of racial and ethnic groups was similar between Arkansas and control states.
The implementation of Medicaid work requirements in Arkansas was associated with a significant decrease in insurance coverage compared with control states (adjusted differential change: −4.3% [95% CI: −8.4, −0.1), but no increase in employment rates (−2.2% [95% CI: −7.0, 2.6]). There was a trend toward decreased likelihood of having a personal healthcare provider (−4.7% [95% CI: −9.9%, 0.5%]), but no change in being unable to see a doctor due to cost (3.8% [95% CI: −1.0%, 8.7%]) or self-reported good health (−0.2% [95% CI:−4.7%, 4.3%]) (Table 1). Among the subgroup of adults with cardiovascular risk factors and/or disease, we found no statistically significant associations between policy implementation and outcomes, although changes in insurance coverage and health status were directionally similar to the overall analysis (Table 1).
Table 1.
Changes in Insurance Coverage, Employment, Health Care Access, and SelfReported Health Among Adults Aged 30 to 49 years After Medicaid Work Requirements1
| Pre-period (Jan 2017-July 2018) | Post-period (Aug 2018-Dec 2019) | Adjusted Differential Change (CI)2 |
||
|---|---|---|---|---|
| Working-Aged Adults | ||||
| Covered by health insurance | Arkansas | 87.0% (84.6%, 89.5%) |
81.0% (78.3%, 83.7%) |
−4.3% (−8.4%, −0.1%) |
| Control states | 89.0% (87.8%, 90.3%) |
88.1% (86.7%, 89.4%) |
||
| Employed | Arkansas | 69.3% (66.1%, 72.5%) |
69.8% (66.8%, 72.7%) |
−2.2% (−7.0%, 2.6%) |
|
Control
states |
71.0% (69.2%, 72.9%) |
73.8% (72.0%, 75.6%) |
||
| Personal healthcare provider | Arkansas | 77.6% (74.5%, 80.7%) |
73.9% (70.8%, 76.9%) |
−4.7% (−9.9%, 0.5%) |
|
Control
states |
75.0% (73.2%, 76.7%) |
74.2% (72.4%, 76.0%) |
||
| Unable to see doctor due to cost | Arkansas | 20.1% (17.2%, 23.1%) |
21.8% (19.0%, 24.6%) |
3.8% (−1.0%, 8.7%) |
|
Control
states |
17.4% (15.9%, 18.9%) |
14.8% (13.5%, 16.2%) |
||
| Selfreported good health | Arkansas | 79.8% (76.9%, 82.6%) |
80.6% (78.1%, 83.1%) |
−0.2% (−4.7%, 4.3%) |
|
Control
states |
80.7% (79.1%, 82.2%) |
81.4% (79.7%, 83.0%) |
||
| Working-Aged Adults with Cardiovascular Risk Factors and/or Diseases | ||||
| Covered by health insurance | Arkansas | 88.7% (84.2%, 93.2%) |
83.7% (79.4%, 88.0%) |
−2.1% (−9.3%, 5.1%) |
| Control states | 90.7% (88.5%, 92.8%) |
90.1% (87.9%, 92.2%) |
||
| Employed | Arkansas | 63.7% (57.4%, 69.9%) |
62.9% (57.7%, 68.2%) |
−1.0% (−9.7%, 7.8%) |
|
Control
states |
62.8% (59.5%, 66.1%) |
64.1% (60.6%, 67.6%) |
||
| Personal healthcare provider | Arkansas | 81.3% (75.9%, 86.8%) |
81.1% (76.3%, 85.8%) |
0.6% (−8.0%, 9.2%) |
|
Control
states |
84.2% (81.8%, 86.7%) |
82.9% (80.2%, 85.7%) |
||
| Unable to see doctor due to cost | Arkansas | 25.7% (19.6%, 31.8%) |
25.3% (20.4%, 30.2%) |
−0.5% (−10.0%, 9.0%) |
|
Control
states |
19.9% (17.2%, 22.6%) |
17.4% (14.8%, 20.1%) |
||
| Selfreported good health | Arkansas | 68.4% (62.3%, 74.6%) |
67.5% (62.5%, 72.5%) |
−3.0% (−12.4%, 6.4%) |
State-representative estimates of adults aged 30 to 49 years after the application of BRFSS survey weights are shown. Control states were Kentucky and Louisiana.
Regression models adjusted for age, sex, and income level.
Discussion
We found that the implementation of Medicaid work requirements was associated with decreased insurance coverage, no change in employment, and a trend towards a lower likelihood of having a personal healthcare provider. Although some policymakers have intensified calls to expand Medicaid work requirements nationwide, our findings suggest that doing so could have major implications for the cardiovascular health of working-age adults, who have experienced a concerning rise in cardiometabolic risk factors over the past decade4. We also found directionally similar changes in coverage specifically in the subgroup of adults with established cardiovascular disease, which is important given the strong link between these socioeconomic factors and cardiovascular outcomes5. These findings underscore the need for the Centers for Medicare and Medicaid Services to assess the public health impacts of Medicaid waiver programs.
While very little previous research has been done in this area, our findings are consistent with a prior study that found insurance coverage losses following the implementation of Medicaid work requirements but was limited to a survey with low response rates.6 We extend upon this work by evaluating measures of health status and access to care, and focusing on a higher-risk subgroup with cardiovascular diseases, using a CDC-administered survey with a high response rate. Limitations of this study include the self-reported nature of the insurance coverage and employment data. In addition, we could not determine whether the disenrollment seen in this study was a result of the work requirement itself or the administrative burden of work reporting, although prior research suggests that the latter may have played an important role6.
Our findings suggest that the expansion of work requirements in safety-net programs like Medicaid, which has been a major source of ongoing policy debate, could have major ramifications for the cardiometabolic health of working-aged adults in the US.1,2
Acknowledgments
Dr. Wadhera receives research support from the National Heart, Lung, and Blood Institute, American Heart Association Established Investigator Award, and the Donaghue Foundation, and serves as a consultant for Abbott, CVS Health, and Chambercardio, outside the submitted work.
Footnotes
Disclosures:
All other authors report no disclosures.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Goldstein A Push to tie Medicaid to work is making a comeback. Georgia is at forefront. Washington Post. https://www.washingtonpost.com/health/2023/06/28/medicaid-work-requirements/. July 2, 2023. Accessed January 9, 2024. [Google Scholar]
- 2.Rep. Gaetz M [R F 1. Text - H.R.1079 – 118th Congress (2023–2024): Medicaid Work Requirements Act. February 24, 2023. Accessed March 11, 2024. https://www.congress.gov/bill/118th-congress/house-bill/1079/text [Google Scholar]
- 3.Brownell NK, Ziaeian B, Jackson NJ, Richards AK. Trends in Income Inequities in Cardiovascular Health Among US Adults, 1988–2018. Circ Cardiovasc Qual Outcomes. 2024;17(5):e010111. doi: 10.1161/CIRCOUTCOMES.123.010111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Aggarwal R, Yeh RW, Joynt Maddox KE, Wadhera RK. Cardiovascular Risk Factor Prevalence, Treatment, and Control in US Adults Aged 20 to 44 Years, 2009 to March 2020. JAMA. 2023;329(11):899–909. doi: 10.1001/jama.2023.2307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Oseran AS, Sun T, Wadhera RK. Health Care Access and Management of Cardiovascular Risk Factors Among Working-Age Adults With Low Income by State Medicaid Expansion Status. JAMA Cardiol. 2022;7(7):708–714. doi: 10.1001/jamacardio.2022.1282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sommers BD, Goldman AL, Blendon RJ, Orav EJ, Epstein AM. Medicaid Work Requirements — Results from the First Year in Arkansas. New England Journal of Medicine. 2019;381(11):1073–1082. doi: 10.1056/NEJMsr1901772 [DOI] [PubMed] [Google Scholar]