We study treatment-effect estimation using panel data. The treatment may be nonbinary, non-absorbing, and the outcome may be affected by the treatment lags. We make parallel-trends assumptions, but do
In many social science experiments, subjects often interact with each other and as a result one unit's treatment influences the outcome of another unit. Over the last decade, a significant progress ha
Accuracy is R2. Local corresponds to CEs for individual regions, global for the meta-analysis mean. For global CE accuracy and coverage the first value comes from using MR-Egger and the second from th