'9' L153 blull U Estimate the average causal eect of gender on fundraising by regressing log.total . raised. candidate on running.variable in each of the newly created subsets and then taking the difference in predicted outcomes at the discontinuity (i.e., running.variable = 0). Express the causal effect in dollars (as opposed to log dollars). Give brief substantive comments on the viability of this estimate as a causal effect, as well as its internal and external validity