Job Corps

Schochet et al. (2006) estimate the program costs to be $16,158. Hendren and Sprung-Keyser (2020) then use the earnings and tax series estimated in Schochet (2018) to form their primary source of fiscal externalities and willingness to pay. For the fiscal externalities, Hendren and Sprung-Keyser (2020) sum the tax liability impacts that are directly observed for years 6-20 in Schochet (2018), discounting at 3%. This yields $52. For years 1-5, Hendren and Sprung-Keyser (2020) do not observe tax and transfer revenue, so Hendren and Sprung-Keyser (2020) instead use an estimated tax rate from the Congressional Budget Office of 0.3% applied to the earnings decrease in the first 5 years, which decreases tax revenue by $0.18. Hendren and Sprung-Keyser (2020) sum this with the value of the produced output from the participants of $220, which when combined with the programmatic cost implies a net cost of $15,886.

If Hendren and Sprung-Keyser (2020) extrapolate to 65 years, it implies a net cost of $15,832. Hendren and Sprung-Keyser (2020) also consider an alternative specification that includes cost reductions from crime reduction, AFDC receipt, and spending on other forms of public assistance that are estimated in Schochet et al. (2006). Including these cost reductions implies a net cost of $12,624.

The net cost and WTP estimates above yield an MVPF of 0.15 (95% CI: [-0.23, 0.58]).

If one assumes the earnings impacts persist in percentage terms throughout the life cycle to age 65, it implies an MVPF of 0.18 (95% CI [-0.53, 1.06]). An alternative assumption is that Job Corps is valued at its programmatic cost. This would imply an MVPF of 1.00 (95% CI [0.96, 1.06]). And, yet another alternative assumption would include the additional cost reductions from crime, AFDC, and other public spending. This implies an MVPF of 0.180 (95% CI [-0.27, 0.81]).