I noticed you utilized a measure of rain-fed cereal suitability to estimate local crop yield. You mentioned you ran other regression on other possible geographic indicators. While you noted that none of the other indicators were robust, I wonder if some sort of interaction model between geographic indicators maybe. Or it there is a nonlinear model that might yield better estimates?
Thank you for presenting!
I noticed you utilized a measure of rain-fed cereal suitability to estimate local crop yield. You mentioned you ran other regression on other possible geographic indicators. While you noted that none of the other indicators were robust, I wonder if some sort of interaction model between geographic indicators maybe. Or it there is a nonlinear model that might yield better estimates?