This is an argument that the soft reform option would seldom be exercised (e.g., only when positive externalities are huge); and this, in turn, argues against wchurchill's concern that it might substantially disrupt incentives.

Words and ideas I offer here may be used freely and without attribution.