As a basis for welfare comparisons, the idea that if a policy change (such as a tariff reduction) could be Pareto improving if it were accompanied by appropriate lump-sum transfers from winners to losers, then it is viewed as beneficial even when those transfers do not occur.