I think the interesting question then moves one level down: how well do those elasticity assumptions describe the actual H-2B labor markets in the study?