Not Everything Has To Be Pareto Optimal

A reader asked whether in economics we should always strive for Pareto improvements. For those unfamiliar with the term, a Pareto improvement is any action that makes at least one person better off and makes nobody worse off. That is not the same as saying that society enjoys a net benefit, because that can happen while still seeing some people being made worse off.

A new village pub could make 90% of the villagers better off while still making 10% of them worse off. The only way to turn that into a Pareto improvement would be to compensate the 10%. Besides, Pareto optimality isn't much of a measure of a society anyway - a society can have numerous Pareto efficiencies and still be a pretty fraught place to live.

Pareto efficiencies are relatively rare in society because life is a series of trade offs, and it is difficult to do anything in society without making someone else worse off. The prospect of people feeling like they've been made worse off often leads to a stalemate.

For example, suppose your 23 year old son Jack is performing his first piano concert at the local school hall, but you and your wife get a one-off offer of a holiday to stay in a luxury apartment overlooking Central Park courtesy of your best friends Jim and Sue who've won a New York holiday for four in a competition.

The holiday clashes with the piano concert - and what usually would happen is that someone is going to be disappointed, and it must be weighed up with the question of who should be the one disappointed? Personally I think the son should be disappointed.

When there is a stalemate, the economists Nicholas Kaldor and John Hicks came up with a watered down version of the Pareto criterion, called the Kaldor–Hicks criterion, which is a measure of economic efficiency that runs alongside the Pareto-efficiency, but makes less of a demand on the outcome. The Kaldor-Hicks criterion states that a decision can be seen as efficient as long as in theory everyone can be compensated to offset any potential costs.

So in the above case, Jack's parents are made better off by the holiday, whereas Jack is made worse off because his parents won't be there for his big concert performance, so they would need to find significant financial restitution or compensatory action to make things beneficial for all concerned. As long as this can be done in theory, the Kaldor-Hicks criterion is satisfied - and this is a good basis on which to assess cost-benefit analyses.