Medicaid Eligibility for Teenagers in South Carolina
Introduction
South Carolina provides free healthcare to low-income children via the Medicaid program, but childless adolescents lose coverage at age 19. Jácome (2020) uses this eligibility cutoff to study the role of Medicaid (and, in particular, mental health care) in reducing crime.
Jácome (2020) finds that men who lose access to Medicaid eligibility at age 19 are 15% more likely to be incarcerated and have lower earnings in the subsequent two years after loss of eligibility. The author employ a difference-in-difference (DD) strategy where individuals enrolled in Medicaid are compared to similar low-income individuals who were not enrolled in Medicaid before age 19. (The intuition behind this choice of control group is that these comparison individuals should be less affected by the loss of Medicaid eligibility at age 19.) The author uses these estimated impacts on earnings and incarceration along with assumptions about the value of crime reductions to compute an MVPF for the expansion of Medicaid eligibility around this age 19 cutoff.
The paper conducts two MVPF calculations: one using conservative and one using aggressive assumptions. The conservative assumptions yield an MVPF of 1.77; the aggressive assumptions yield an MVPF of 14.96. The numbers on this page primarily reflect the conservative estimates, but the inputs in the aggressive specification are also reported below.
MVPF = 1.8
Net Cost
Jácome (2020) reports three primary components that are used to calculate the long-run net cost of Medicaid. The goal is to understand the total cost associated with extending Medicaid eligibility by two additional years for young men in South Carolina.
First, there is the direct cost of insurance via Medicaid. For the conservative specification, this amounts to $14,643,793. For the aggressive specification, Jácome (2020) assumes, as in Finkelstein et al. (2019), that the majority of Medicaid expenditures (60%) are recouped by the government in the form of reductions in uncompensated care spending by the government. In that specification, the net cost to the government falls to $5,857,517.
Second, there is the lower government spending on incarceration due to reductions in crime. In the conservative MVPF calculation, this amounts to a savings of $2,069,931. In the aggressive MVPF calculation, Jácome (2020) also includes the cost savings from future recidivism, leading to a cost savings of $3,220,491.
Finally, there is the increased tax revenue resulting from the positive impact of Medicaid on labor earnings. To estimate this, Jácome (2020) uses estimates of incarceration’s negative effects on employment from Mueller-Smith (2015). Applying a 20% tax rate to these earnings effects yields between $194,616 and $257,282 in increased tax revenue, depending on the specification. (The lower number here is used for the conservative specification, while the higher number is used for the more aggressive specification.)
Combining, these estimates suggest a net cost of $12,379,246 for the conservative specification and $2,379,744 in the aggressive specification.
Willingness to Pay
Jácome (2020) considers three components of the willingness to pay associated with a Medicaid expansion.
First, individuals in society have a willingness to pay for the reduced incidence of crime. Jácome (2020)’s specifications suggest this value of crime reduction ranges from $16.4 million to $33 million, which are used in the conservative and aggressive specifications, respectively.
Second, individuals who avoid incarceration value the earnings they receive when employed, rather than incarcerated. Jácome (2020) estimates this to be $973,100 and that figure is used in the conservative estimate. For the aggressive estimate, Jácome (2020) includes earnings losses for incarcerated individuals in the five years after they are released. Those losses are $313,300 producing a total earnings effect of $1,286,410.
Finally, there is the value of the health insurance to the beneficiaries of this policy. For the conservative specification, Jácome (2020) assumes individuals themselves have no willingness to pay for this insurance, but there is a willingness to pay by uncompensated care providers of $4,461,761 for reduced costs of this care. In the aggressive specification, Jácome (2020) assumes individuals value insurance at its full cost, or $1,358,491.
Combining these estimates leads to a willingness to pay of $21,865,137 in the conservative specification and $35,602,065 in the aggressive specification.
Finally, the author notes that this calculation has not included other sources of positive WTP. For example, beneficiaries would also have direct WTP for the reduced probability of being incarcerated, as long as they value their own lack of incarceration above and beyond the extra opportunity to generate earnings. Including this benefit would imply a higher MVPF value in both specifications.
MVPF Summary
Combining the willingness to pay and the cost estimates lead to an MVPF of 1.77 in the conservative specification. The more aggressive specification leads to an MVPF of 14.96.
References
Finkelstein, Amy, Nathaniel Hendren and Mark Shepard (2019). “Subsidizing Health Insurance for Low-Income Adults: Evidence from Massachusetts,” American Economic Review, 109(4), 1530-1567. DOI: https://doi.org/10.1257/aer.20171455
Jácome, Elisa. “Mental Health and Criminal Involvement: Evidence from Losing Medicaid Eligibility.” Working Paper. (2020).
https://elisajacome.github.io/Jacome/Jacome_JMP.pdf
Mueller-Smith, Michael. “The criminal and labor market impacts of incarceration.” Working Paper (2015).
http://www.irp.wisc.edu/newsevents/workshops/2015/participants/papers/10-Mueller-Smith-IRP-draft.pdf
Policy Features
- Category
- Social Insurance
- Sub-Category
- Child Health
- Beneficiary Type(s)
- Children and Youth
- Average Age
- 19
- Country of Implementation
- United States
- Year of Implementation
- 2013
- Empirical Method
- Difference in Differences
- Research Type
- Primary
- Peer Reviewed
- No
- MVPF Publication Link
- elisajacome.github.io/Jacome/Jacome_JMP.pdf