# The Angelina Jolie Dilemma

Angelina Jolie decided to have a preventive double mastectomy performed after she was informed by her doctor that there was an 87% risk that she would, in time, develop breast cancer, since her genome had a mutation of the BRCA1 gene. The generally accepted opinion nowadays is that women carrying this gene have an 80% risk of developing breast cancer by the age of 90.

We can all sympathize with her having to take such a horrendously difficult decision based on a probabilistic estimate. She could have decided to gamble on the chance that possibly she was part of the 13% who would not develop the disease even though she had that gene mutation. But she decided to play it safe − better safe than sorry.

All gamblers know how difficult it is to take a decision on a single roll of the dice, even though the probabilities of the different outcomes are well known. It is an entirely different situation when we are dealing with a large number of repetitive events in which the individual probabilities are known.

The resulting average of many events is known to be the probabilistic expectation. This is the result of the Law of Large Numbers, and even casino operators who have not studied mathematics know it. So do people in the insurance business who base the rates they set for insurance policies on their analysis of a large mass of statistical data, from which the average to be expected from a large number of similar events can be deduced. Under these circumstances, the decisions are easy.

Can we make rational decisions based on probabilistic forecasts regarding a single event, like Angelina Jolie did? If the outcome is of no great consequence, why not? You win some, you lose some.

But if the outcome is important, crucial, irreversible − what then? The probabilities that Angelina faced in making her decision were based on solid statistical data, on the analysis of the outcome of many similar events in which women carried this gene. What if the probabilities in question were themselves questionable, not much more than guesstimates? Who knows what Angelina’s decision would have been in such a case?

Many of the important decisions taken in the life of individuals or nations are not based on objective probabilities attached to future events. That information is usually nonexistent, and we may have recourse to game theoretic considerations or intuition in making decisions. When probabilities are introduced, they generally are on pretty shaky ground.

That is the case with the demographic projections that are promoted by those urging an Israeli withdrawal from Judea and Samaria. Here we are advised by some that we should be prepared to cut out parts of the Land of Israel − Judea and Samaria, the biblical heartland − based on certain demographic prognostications indicating that in time the Jewish population would constitute a minority in the State of Israel unless this decision was taken now.

The Danish-Jewish physicist Niels Bohr famously said that prediction is very difficult, especially about the future. That is certainly true in this case. Most past demographic forecasts in Israel have turned out to be wrong.

If this was no more than an academic exercise, there would be nothing to get excited about; the forecasts could be published in an academic journal and we could revisit the data in another 10 or 20 years. But those using these demographic forecasts hold them as a Damocles sword above our heads, insisting that we take a decision now and abandon Judea and Samaria, a decision that would be irreversible.

Yet it might turn out in the years to come that their forecasts were off by 10, 20, maybe 30 percent. Hurry, they shout, there’s no time to lose − the window of opportunity is closing. They are not talking about a mastectomy, they are talking about cutting the heart out of the Land of Israel.

haaretz