I've been quite critical of the Make-A-Wish Foundation in the old days. That's about the opposite of what you would do if you tried to design an extremely effective charity.

Does it target the poorest people? Not really. Although it targets terminally ill children, it does not work in the poorest countries of the world.

Does it help most people in need? Again, no. He does not try to save lives. It tries to inject a little joy into the lives of sick children, sometimes – as in the famous case of Batkid in San Francisco – at considerable financial cost.

The average wish costs $ 10,130 fill. Given that Malaria Consortium can save the life of a child under 5 years old for about $ 2,000 (Getting a precise figure is difficult, of course, but it's around that), you could probably save the lives of four or five children in sub-Saharan Africa at the cost of providing a pleasurable experience to a single child in the United States. For the cost of the chaotic Batkid cascade – $ 105,000 – you could save the lives of about fifty kids.

That's why I was hard Make-A-Wish in the pastand why effective altruists like Peter Singer also criticized the group.

But now, I'm reconsidering. A new study in the journal Pediatric research, comparing 496 patients from the Nationwide Children's Hospital in Columbus, Ohio, who granted their wishes for 496 "control" patients of similar age, sex, and illness, found that patients for whom these wishes had been granted were less urgent, and were less likely to be readmitted to the hospital (apart from scheduled readmissions).

In a number of cases, this reduction in hospital admissions and emergency room visits has saved more than $ 10,130, the cost of the average wish. In other words, Make-A-Wish helped and contributed in a profitable way.

This study is not the last word. It's not random, and some measures seem odd to me. For example, instead of simply measuring the effect of fulfilling a wish on the number of hospitalizations in the following year, it measures "the probability of having fewer unplanned hospitalizations and visits to 'emergency".

Instead of directly estimating an effect on admissions to the hospital, it introduces a binary variable ("has fewer admissions" as opposed to "not less than"). admissions ") and performs an analysis based on this binary variable. I'm not a medical researcher, but I'm a big consumer of social science like this, and it seemed strange to me. Andrew Gelman, professor of statistics at Columbia, who I asked about the study, said, "The practice of discretizing variables is common in medical statistics and I think it's usually bad idea."

Similarly, the study does not directly determine whether the savings resulting from the reduction in admissions in the treatment group are greater or less than the cost of all wishes. This may be due to a lack of data, but Make-A-Wish should really order a fully randomized study to see if this finding holds.

But it seems plausible that the psychological impact of Make-A-Wish can reduce admissions and that this alone is enough to offset the costs. If this is true, it seems to me that it eliminates the need for Make-A-Wish as a charity. If this really saves on hospital bills, health insurers should have an interest in paying directly for Make-A-Wish experiences to reduce their own claims for compensation. For Make-A-Wish, this would be a way to resize the model more broadly.

Do you already have trouble figuring out where to make a donation that will have the greatest impact? Or what kind of charities to support? Over 5 days, in 5 emails, we will guide you through research and executives who help you decide how much and where to giveand other ways to do good. Sign up for the new Future Perfect pop-up newsletter.