My argument isn't quantitative, it's predicting how the policy will impact incentives and thus peoples' future behavior. Incentives matter in policy construction, but not everything that matters can be put on a chart or in a table of numbers.