Hope Tax Credit, Single Filers at Phase Start

Bulman and Hoxby (2015) estimate that a dollar decrease in income leads to $0.064 fewer tax credits claimed by the household. That $0.064 is the direct cost of the program. The second stage regression from Bulman and Hoxby (2015) finds that, at the kink point, a dollar decrease in income decreases college attendance by 5.01 \cdot 10^{-6} percentage points. That decrease in attendance results in lower educational costs for the government. Hendren and Sprung-Keyser (2020) project out the impact those attendance changes onto changes in income. Then, they use a tax and transfer rate of 19.6% to determine the impact on the government’s budget. They find that government revenue falls by $0.120. Putting these components together, the total net cost of the policy is $0.16.

The MVPF of this policy is -1.64. The 95% confidence interval is [-6.45,\infty]. For context, if the Hope credit had no effect on college attendance, we would expect the MVPF of this policy to be around 1 (the HOPE credit would simply be a tax cut that induced a muted behavioral response). As these calculations show, the observed causal effect of the HOPE credit policy is a sufficiently large MVPF estimate that deviates meaningfully from 1. That said, the confidence interval around the estimate is quite large and so we cannot rule out an infinite MVPF or an MVPF of 0. In other words, this means that one cannot statistically rule out that the HOPE credit either pays for itself or provides no benefit to the beneficiaries, and suggests the value of further work to increase the statistical precision of the impact of the HOPE credit.