Memo: Data For Progress Achieved Excellent Targeting in the 2019 Fuck Gerrymandering Program
By Sean McElwee, Executive Director, Data for Progress
Executive Summary:
To study a targeting program, the variable of interest is not wins or losses, but closeness of margin to zero, since the goal is to direct money where it has the most chance of influencing the outcome (and accepting some loses in the process). Too high of a win rate suggests that resources are not being optimally allocated to the most competitive seats, and too low of a win rate suggests that either the program is not successful, or that resources are again not being optimally allocated.
To begin analyzing Data for Progress targeting, we examined all races decided by less than 5 points, of which there are 16. Of those closest races, FG targeted 14 of them and won 5, enough to help put the Democratic party in the majority in Virginia. Most importantly, the two races decided by less than a five point margin we did not target (HD-75 and HD-10) were both won by Democrats. To reiterate: there was no race decided by less than five points where Democrats came up short that was not targeted by the Fuck Gerrymandering program.
