,
11 tweets,
4 min read
Read on Twitter
🚨PAPER ALERT! Happy to share “What does free community college buy? Early impacts from the Oregon Promise”, conditionally accepted at JPAM: ogurantz.github.io/website/Gurant… 1/n
Oregon was an early adopter of state-level “free” community college, beginning with the 2016 cohort of high school graduates. I find that community college enrollment increased by a significant amount (~4-5 percentage points) in the 2016 and 2017 cohorts. 2/n 
In the 2016 cohort this appears driven by students shifting out of four-year colleges, whereas in 2017 Oregon began to see overall increases in postsecondary attendance, though there is some uncertainty in these estimates across specifications 3/n
Why do we see shifting out of four-year colleges, at least early on? Likely as the program is a last-dollar scholarship layered on top of federal aid and a state-level need-based grant program, which shifts larger aid packages towards higher-income individuals 4/n
Oregon’s own data confirms most money went to higher-income families in the first year, before they began instituting a maximum EFC limit for eligibility. See citations in paper. 5/n
Nonetheless, the signaling of free college appears to be a powerful motivation for students (see HAIL @dynarski @cjlibassi @michelmorek and other related work). 6/n
What now? Time will tell in Oregon, along with data coming in from other fabulous researchers (@linzcpage @ckcarruthers @TimBartik and colleagues) studying financial aid and Promise programs. 7/n
As for methods, I use College Board data and a diff-in-diff with Oregon and a number of control states which offer essentially universal 10th grade PSAT statewide 8/n
The benefits are twofold: (1) less self-selection into my sample within a year and over time; (2) students would have taken the PSAT prior to the announcement of the policy. Together, these helps avoid endogenous and unrepresentative samples. 9/n
Big thanks to Ferman and Pinto paper (mitpressjournals.org/doi/abs/10.116…) for helpful code on standard error adjustments! (I don't see them on twitter) 10/n
And big thanks to @MizzouTSPA @MizzouAandS @Mizzou @Lael_Keiser for all the support I get here at the Truman school. /fin
