The Employment Impact of the Provision of Public Health Insurance: A Further Examination of the Effect of the 2005 TennCare Contraction

Missouri’s Medicaid Contraction and Consumer Financial Outcomes

“…Our point estimates also show a 2.2pp increase in the uninsured rate and an 18 percent drop in Medicaid spending, though these estimates are not statistically 11 We also note that the procedure outlined by Conley and Taber (2011) (hereafter, CT) does not apply to our empirical setting because their method is only valid when the number of control clusters approaches infinity (Hagemann, 2020). CT present evidence that their procedure has good finite sample properties with the number of groups equal to as low as G=42 (their Monte Carlo simulations are done with G=50), while Ham and Ueda (2021) argue that they can use the CT procedure with G=16 when reanalyzing the TennCare disenrollment. Since we only have eight groups in our setting, it is highly unlikely that we have large enough G to for us to have asymptotically valid CT estimates.…”

“…The authors find that 97 percent of the change in employment is explained by workers working for private employers who provide ESI. In contrast, Ham and Ueda (2021) and DeLeire (2019) separately conduct updated analyses of the TennCare disenrollment and find that the Garthwaite et al (2014) results are sensitive to both the choice of data used in the analysis and the type of standard errors used to conduct statistical inference. The two papers conclude that there is a large amount of uncertainty regarding the impact of the TennCare Medicaid change and that a zero effect of the TennCare disenrollment on employment cannot be ruled out.…”

In-Kind Welfare Benefits and Reincarceration Risk: Evidence from Medicaid

“…For example, there was no evidence of employment effects in the Oregon Health Insurance Experiment population (Baicker et al, 2014) but Wisconsin adults who were able to access an earlier limited Medicaid waiver program returned to the work force more slowly than those who were waitlisted for the program, with net employment declines of 3-5 percentage points (Dague et al, 2017). One paper finds evidence of large increases in employment in Tennessee following the TennCare disenrollment event (Garthwaite et al, 2014), suggesting strong work disincentives, although other work has challenged that interpretation (DeLeire, 2019;Ham and Ueda, 2021). Recent cross-sectional studies of ACA Medicaid expansions do not typically find any declines in employment, which could be due in part to policy uncertainty, different affected income groups, labor market conditions, or to the availability of phased out Marketplace subsidies in non-expansion states (Gooptu et al, 2016;Kaestner et al, 2017;Leung and Mas, 2018).…”