Thinking About Intelligence

(Originally published May 17, 2022.)

An intelligence success you probably didn’t know about saved us from nuclear Armageddon in October 1962. An intelligence failure you certainly were aware of but may not have fully understood resulted in the death of a US President in November 1963. A lot happened in those 13 months.

What I hope to do by looking honestly at these and other intelligence successes and failures is learn some lessons. What are the keys to success that we should try to repeat? Where did we go disastrously wrong and how can we be on guard to avoid similar errors in the future? Learning these lessons is enormously important for American citizens and taxpayers.

To do this, we need a framework for our thinking. This framework needs to be simple enough that we can draw general lessons from cases that differ widely in their nature and complexity. Importantly, the framework must be true to the way that intelligence work is really done.

Collection – Analysis – Leadership

In its essence, American intelligence work is easy to understand.

Intelligence collectors use sources to determine what we think are facts -- this is happening on the battleground, this is what XYZ leader said to a confidante, this is the nature of a country’s nuclear weapons program. The collectors are trusted because they have US Government security clearances. The sources are varied, however; some are more trusted, some less. So, the intelligence report must provide some sense of how confidant we are that the facts as reported are true.

To use a Cuban Missile Crisis (CMC) example:

- The Soviets are moving a large number of troops and their equipment to Cuba – 100% certain fact.

- Some sources say that the troop movement is defensive in nature – 60% probability.

- Some sources say that the Soviets are moving troops to protect the deployment of offensive nuclear missiles – 40% probability.

Intelligence analysts then take the raw factual reporting, combine it with everything else they know, and give an assessment. My apologies to my analyst friends, but I’ll use the CMC again as an example. The analysts, after extensive debate and consideration, reach the conclusion:

- We think it unlikely the Soviets would move offensive nuclear missiles to Cuba.

The leadership of the American Intelligence Community -- the Director of National Intelligence (DNI) after 9/11 or the Director of Central Intelligence (DCI) between 1947 and 9/11 -- will typically support the judgement of the analysts in meetings of the National Security Council at the White House. Only rarely has the DNI/DCI openly dissented from the analytic judgement, as DCI McCone did in the case of the CMC. When they do, the President of the United States should pay particular attention.

The CMC vs the Kennedy Assassination

Using a simple scale of (+) to indicate they got it right, (–) to indicate they got it wrong, and (o) to indicate neutrality, I think the collectors get a (+) in the Cuban Missile Crisis, the analysts get a (–), and McCone’s leadership gets a (+). The key, determinant factor that brought about success was McCone’s leadership.

In the Kennedy Assassination, the collectors bear the full weight of responsibility for the tragic failure. The Secret Service driver Greer should have accelerated. The FBI Special Agent Hosty should have warned the Secret Service that Oswald was a threat. Most importantly, Fitzgerald and Sanchez at the CIA should have been much more alert to the danger that Rolando Cubela was a double agent who was keeping Fidel Castro informed of our plotting to kill him.

Neither Greer nor Hosty advanced in their careers after the assassination. Fitzgerald was promoted to the powerful position of CIA Deputy in charge of all collection. Sanchez advanced to Chief of the Latin America Division and, after his retirement from CIA, became Deputy Assistant Secretary of Defense for Latin America.

Of special note, the CIA plot to kill Castro was a “covert action.” This means that the analysts would have been totally cut out and unaware of what was going on. Leadership would have been aware of the plot but busy with their other responsibilities and unlikely to see potential problems. Only the CIA collectors were in a position to see the potential problem, ring alarm bells, and get something done about it.

Using our simple scale then, the collectors get a (–) in the Kennedy Assassination, the analysts certainly get a (o), and leadership probably gets a (o).

As we look into future cases, I think you will find this simple framework useful for thinking about complex intelligence successes and failures.