Decisions Under Uncertainty

While intelligence services may provide estimates regarding the enemy’s intentions, the government must realize that these are no more than intelligent, and sometimes not so intelligent, guesses. The rest is up to the government.


By Moshe Arens

(A version of this column appeared in Haaretz on September 28, 2004.)

Every year at this time the people of Israel return to reexamine that fatal decision taken by the Israeli government on the eve of Yom Kippur 1973. The decision not to fully mobilize the IDF’s reserves, when it was known that the Egyptian and Syrian armies were mobilized and poised for an attack on Israel, was not, as is frequently argued, an intelligence failure.

It was simply a wrong decision taken by the government of Golda Meir, a decision that reflected its lack of competence in dealing with uncertainty. All the intelligence needed regarding the military capability of the enemy was available. The government’s mistake was that it attempted to guess the intentions of the enemy. As is well known, even the best intelligence services can usually not divine the enemy’s intentions – they were wrong before Chamberlain’s surrender to Hitler at Munich, and before the Japanese attack against the US fleet at Pearl Harbor, and again before the Al-Qaida attacks upon America on 9/11.

Good intelligence services will usually provide adequate information on the enemy’s capability. Taking the measure of the actions that are in the realm of the possible with these capabilities means stretching the imagination and putting yourself in your enemy’s shoes. The capabilities of the Japanese navy were known in December 1941, but that these capabilities would be used to strike at Pearl Harbor was evidently not seriously considered. After the many hijackings of commercial airliners, the ability of hijackers to fly them into buildings was obviously there, but that possibility was not seriously considered before it actually happened. So the intelligence services are faced by the challenge of “thinking about the unthinkable.”

Although intelligence services may also be called upon to provide estimates regarding the enemy’s intentions, the government must realize that these are no more than intelligent, and sometimes not so intelligent, guesses. The rest is up to the government. It is responsible. As Harry Truman said “the buck stops here.” That in a nutshell is the relationship between the intelligence community and the decision-makers at the governmental level. In putting the blame for not mobilizing the IDF reserves on the eve of Yom Kippur 1973 on the chief of staff rather than on the government ministers, the Agranat Commission committed a serious error.

During the period preceding the recent Allied military operation against Saddam Hussein, Israeli intelligence, like their American colleagues, were wrong about Iraqi capabilities, as well as about Iraqi intentions regarding Israel. The Israeli government’s decision to distribute gas masks to the population at the time was reached on the basis of “worst case” assumptions regarding Iraqi capabilities and Iraqi intentions toward Israel, as depicted by Israeli intelligence. As it turned out, they were wrong on both counts.

On occasion, politicians depict their decisions as “taking calculated risks,” whereas in fact, there is no way to calculate the risks involved when dealing with one-time events. That methodology – probability theory – is relevant to repetitive events on which a great deal of statistical data has been accumulated. Insurance companies are in that business and not governments. When dealing with uncertainty, government decision-makers are left with the difficult and unpleasant task of considering “worst case” alternatives and then making a decision, in full knowledge that the responsibility rests on their shoulders and only on their shoulders.

Menahem Begin’s decision to order the Israeli Air Force to destroy the Iraqi nuclear reactor in 1981 is a good case in point. The intelligence community provided reliable assessments of the impending Iraqi nuclear capability, and Begin – giving consideration to “worst case” scenarios of Saddam Hussein’s intentions regarding Israel – reached the conclusion that the reactor would have to be destroyed before it went critical.

Although arousing much criticism at the time, throughout the world and even in Israel, it is now generally considered to have been a decision of ultimate importance to Israel and the entire free world.

Much has changed since those days. The ability to gage the enemy’s capability has improved tremendously with the introduction of reconnaissance satellites and the proliferation of news media providing real-time world-wide information. Government ministers do not have to rely only on the information provided by the intelligence services. But nothing has substantially changed regarding the uncertainty about the enemy’s intentions. So government ministers are left with the age-old problem of reaching decisions without full knowledge of the enemy’s intentions. Under the circumstances, an introductory course in game theory might do them some good.


Translate »