Security Studies Program Seminar

Reshaping Intelligence for an Age of Terror: Where Do We Stand?

Gregory Treverton (RAND)

March 16th, 2011

The aim of the presentation was to give “a scorecard” on the ways in which intelligence operations have changed since 9/11, and the extent to which those changes have been successful. There were five key issues that Dr. Treverton dealt with in order:

1. Change in the target of intelligence

During the Cold War (and before), intelligence gathering had the following characteristics: the primary targets were large, rich states – the US’s peer competitors; those states were clearly defined geographically, organized more or less hierarchically in bureaucracies; intelligence operations occurred largely “over there” – in the target state, and originated from relatively few sources. Further, there were relatively few customers for the intelligence (i.e. senior federal officials), and although intelligence was viewed as important, it was not “primary”. Deterring and containing the Soviet Union, for example, did not require a great deal of intelligence.

Today, all of these characteristics have been partially or completely reversed.   Transnational actors like terrorists are the primary target as the source of threats to the United States.  So, too, weak, small and failed states are import. There are a far greater range of sources of intelligence – including many domestic sources within the United States, and a far greater number of customers for that intelligence – at all levels of the US government. Intelligence has taken on a far greater importance in the effort to prevent terrorist attacks.

2. The Cold War legacy is mismatched to the new target

The way in which intelligence was conceived during the Cold War is no longer appropriate. For example, the distinction between intelligence and law enforcement was dichotomous during the Cold War when domestic intelligence was mostly counter-intelligence against Soviet (and other hostile nations’) spies.  Today’s threat blurs the distinction, with much intelligence coming through law enforcement efforts, and many adversaries operating (and potentially breaking laws) inside the United States.

The clean distinction during the Cold War between foreign and domestic operations has also become murkier. Whereas domestic intelligence was a relatively minor FBI mission during the Cold War (focused on countering Soviet espionage efforts), it is now central to actions in the fight against terror.

US intelligence operations still reflect this legacy to a significant degree – and it is no longer appropriate. The new model should be similar to the Combatant Commands within the military (of which the National Counterterrorism Center (NCTC) is the closest prototype to date). In this model, the existing agencies would take on the role of the military services – training and equipping operatives and executing operations. However, the agencies are not enthusiastic about this role, just as the services were unenthusiastic about the Combatant Commands.

3. The challenge for analysis is daunting

Intelligence products serve three functions:

Intelligence needs should be organized by the type of issue, the likely demand from policy officials, and the demands on the time of the official to understand the intelligence. For example, specific tactical intelligence solves a puzzle, has high demand, and requires low time to understand. “Sensemaking” helps officials’ understanding of complexity, but there is unknown demand for such intelligence, and it requires significant time from officials in order to fully process such information.

Dr. Treverton proposed the following:

4. Domestic Intelligence

Domestic intelligence can be divided into three tasks.

The refocusing of the FBI has been dramatic but remains a work in progress. The mission has changed from primarily one of law enforcement to one of intelligence-led prevention. But the challenge continues, for example in increasing the status of intelligence within the organization, and in beginning to think about cases not in terms of building evidence to secure a conviction, but rather as “collection platforms” for intelligence.

Questions of values loom large in domestic intelligence. Privacy is the critical liberty at issue – and suggests that the emphasis should be on what is done with collected information than whether or not it is collected.

5. Information sharing

Information sharing is not the right label for what it needed – there is a need to reconceptualize the process not as sharing, but more of the “co-production” of intelligence.  In particular, there is a need to build on comparative advantages between local and federal organizations.

Co-production must utilize the comparative advantages of each organization. If co-production is to truly work, it must be a two-way street – local organizations must feel like they are benefiting from the arrangement – at the moment it is not clear that that is happening.


Is there a danger in dedicating more resources to modern threats, that the US does too little re: traditional great power threats?
There are inevitably tradeoffs and opportunity costs (in the domestic realm, for example, the FBI is now spending less attention on gangs). Although this is a worry, the big threats are probably still getting a lot of attention.

Does the US need an Mi5 – a purely domestic intelligence agency?
Many countries do have a separate intelligence agency, and if the US was starting out from scratch, there probably would be one. But given that the US isn’t starting from scratch, the question is whether we would get better intelligence after such a reform, and whether it would be worth the inevitable transaction costs of doing so. It isn’t clear that the answer to either question is yes. Further, there are some advantages in having consolidation of both intelligence and law enforcement in one agency – for example in terms of the variety of people you can attract to work in the organization.

What is the status of enhanced interrogation techniques?
The US pays a very high price to achieve very shaky intelligence when they use these techniques. We have a lot to learn from Israeli training of interrogators – for example by focusing training on languages, and only using experienced agents to conduct interrogations. The US faced a problem in ramping up interrogation capabilities quickly after 9/11 – and did a poor job in many respects.


Rapporteur: Mark Bell

back to Wednesday Seminar Series, Spring 2011