The 81 Percent Effect
A couple weeks ago I pointed out that one could predict with 77% accuracy the amount that states asked for in Race to the Top (RTT) funding by looking at only two things, neither of them related to the state's RTT application. The first is state student enrollment, and the second is the size of a state's reported 2010 budget shortfall.
At the time, I expressed doubts about what this might mean regarding RTT's ability to change the way the federal government and the states do business and for President Obama's State of the Union promise that, under RTT, "Instead of rewarding failure, we only reward success. Instead of funding the status quo, we only invest in reform." I wondered whether this reflected familiar state alms-collecting and budgeting-by-formula more than a serious attempt to craft reform-centric budgets. But I also noted, as many RTT enthusiasts reminded me, that this was only the initial apps. The emphasis on plugging budget gaps and on budgeting by body count would undoubtedly be attenuated once applicants were vetted based on their reform agendas.
Well, last Thursday, round one RTT finalists were named. So, my crack research assistant Daniel Lautzenheiser and I ran the numbers once again. And, to our great surprise, it turns out that the relationship between budget shortfalls and body counts is stronger among the finalists than it was among the applicants as a whole. Together, enrollment and the state fiscal situation allow us to predict with 81% accuracy how much money a given state requested in their RTT application (for the researchers out there, that's an "r-squared" of .81 in a simple two-variable regression that includes all round one RTT finalists).
The simple correlation between student enrollment and the size of a state's RTT ask was .90 (a 1.00 would mean the numbers march in lockstep). This suggests that, whatever the ins and outs of their actual proposed reforms, state requests were largely a product of head counts. So far, it's looking like we could have saved millions in consultant fees and tens of thousands of hours in staff time if the Department of Education had just disbursed RTT funds based on an enrollment formula.
Even more peculiar, the correlation between a state's 2010 budget shortfall and its (seemingly unrelated) RTT ask was an impressive .66--and actually up a bit from .63 on the first go-round. That figure was helped along by the curious inclusion in the pool of reform-minded finalists of cash-strapped Democratic strongholds like New York and Ohio, despite their apparent lack of, well, reform-mindedness.
I'm a cynical guy by nature. I admit it. But the Department of Education is really making it awfully easy for me.