Mai Mahmoud presents a study on blast tolerant wheat in Bangladesh. The study examines the allocative efficiency of agricultural input subsidies. Using a two-stage experiment, Mai randomizes the price subsidy for a new seed variety in stage one. Subsidy amounts range from full to zero subsidies. In stage two, she randomizes free distribution across the self-selected sample of non-buyers from stage one. This design allows her to compare realized returns across the entire population with the returns among the sample choosing not to buy in stage one. She finds that the stage two free distribution increases adoption by an amount almost equal to the effect from stage one. Thus, higher prices screen out farmers who are willing to plant the new seed. In addition, she finds no evidence that choosing not to buy in stage one is associated with having lower returns from adoption. The results suggest that prices do not have a selective ability in the context of agricultural inputs.