Skip Navigation
Resource

Rethinking Intelligence: Interview with Erik Dahl

Dr. Erik Dahl is the author of “Intelligence and Surprise Attack: Failure and Success from Pearl Harbor to 9/11 and Beyond,” Dr. Dahl served 21 years as an intelligence officer in the U.S. Navy, and is currently an assistant professor at the Naval Postgraduate School, Department of National Security Affairs and the Center for Homeland Defense and Security.

Published: June 6, 2014

Dr. Erik Dahl is the author of “Intelligence and Surprise Attack: Failure and Success from Pearl Harbor to 9/11 and Beyond,” Dr. Dahl served 21 years as an intelligence officer in the U.S. Navy, and is currently an assistant professor at the Naval Postgraduate School, Department of National Security Affairs and the Center for Homeland Defense and Security.

Mike German, a Fellow at the Brennan Center for Justice at NYU Law School, interviewed Erik Dahl on June 2, 2014. The following is an edited transcript of that interview.


Q: Hi. My name is Mike German. I'm a fellow with the Brennan Center for Justice at NYU Law School and I'm engaged in a research project where I'm interviewing intelligence professionals and researchers who are writing about intelligence reform in a way that challenges the conventional wisdom with the hope that that this discussion can lead to more comprehensive reform of the intelligence policy and national security policy in the United States. And today I'm with Dr. Erik Dahl. He's a professor here at Naval Post-Graduate School, Department of National Security Affairs and the Center for Homeland Defense and Security. He was a Naval Intelligence Officer for 21 years, and then obtained a PhD from the Fletcher School at Tufts University. He is the author of last year's Intelligence and Surprise Attack: Failure and Success from Pearl Harbor to 9/11 and Beyond. Dr. Dahl, you served in and have studied the intelligence community for almost three decades- a little more than three decades.

Dahl: That's right.

Q: It's a $70 billion dollar a year enterprise. Five million cleared employees and staff. I tend to think of it more as an enterprise than just a community anymore, but what is the mission of this intelligence enterprise?

Dahl: Well, it's a good way to put it- intelligence enterprise. Certainly more than any single intelligence community, which is what we usually talk about. But still, the fundamental domain, the central part of the American intelligence enterprise as you put it, was established after World War II to prevent what we call strategic surprise. To really try to -bottom line- to prevent another Pearl Harbor-type surprise devastating attack on the United States. The intelligence community and intelligence agencies have always had a lot more to do during the Cold War, keeping track of current events, of world events, keeping our military prepared- but that was the main goal. And the question since 9/11 has been whether or not that still is sufficient for a single primary goal. Is it enough to try to keep America safe from a single devastating attack or are we today- and perhaps we'll get a chance to talk about this. Are we today expecting our intelligence community to keep individual Americans safe- not just America safe, but all Americans safe- and how do we do that?

Q: And despite the effort and the expense that we put into this effort, intelligence often fails and in your book you mention two spectacular failures: Pearl Harbor and 9/11. But based on your studies of surprise attacks like this and you focused on that surprise attack. What's the conventional wisdom regarding why intelligence fails?

Dahl: Well the subject of intelligence failure is probably the most studied question in the intelligence studies literature. And in fact there is, as you say, there is a conventional wisdom. There is a standard understanding about why intelligence fails and how intelligence fails.  It really was first established back in one of the most famous books after Pearl Harbor by Roberta Wohlstetter. She termed it that, we weren't able to find the important warning signals in the middle of all the noise of other things going on. We had a similar finding after the 9/11 attacks. The 9/11 commission had a different image but a similar concept. They said that the information was there, but we couldn't connect the dots. So what this really means in intelligence terminology is that we find that intelligence analysis fails. At least that's the standard understanding about intelligence failure. That usually, the warning signs were there, but we couldn't understand it. We perhaps weren't imaginative enough to understand it- at least that's the conventional wisdom.

Q: Dr. Dahl, for your study you adapted a methodology that's fundamental to social science: the use of a control group. But this actually hasn't been used in a lot of failures-or studies of intelligence failures. You compared well-known intelligence failures like Pearl Harbor and 9/11 to cases where the surprise was actually thwarted or stopped. What did you learn from studying it that way?

Dahl: Well it's fascinating that most of what we know about intelligence, and especially about how intelligence works, comes from intelligence failures when things don't go right. What I wanted to do was find out what happens when intelligence works, when intelligence succeeds in preventing a surprise attack or preventing a terrorist attack. And what I expected to find was that conventional wisdom would be borne out. In other words, I would find that when intelligence works, it's because intelligence analysts connect the dots. They put together widely scattered bits of information, they come up with a theory, a hypothesis, bring it to a decision maker and the decision maker says, “Ah you're brilliant, you're imaginative, let's go stop something bad from happening!” But I found, that's not what happens. In fact, there are many cases of brilliant analysts connecting the dots that didn't do any good. What does seem to work is when intelligence collectors, not analysts, but collectors- it's actually the harder job of finding the intelligence in the first place. But collectors can find very specific, precise information- what in the military sense, we call tactical or operational intelligence. And then they need to take it to a decision-maker, a policy-maker, a leader, who is receptive, who understands intelligence and is willing to listen. And often, in fact, that second part is the hardest part of the intelligence business.

Q: And that's what I found fascinating reading your book- that it's often the intelligence that gives sound warning and the policy maker fails to heed that accurate warning. What causes that policy maker not to heed the warning?

Dahl: You know, many intelligence experts and insiders, and also many academic scholars of intelligence have argued that the main reason for intelligence failure, the main reason why our assessments aren't used, is because policy makers don't listen. And I didn't find that's exactly the case. What I found was that policy makers think they want long-range, big-picture strategic intelligence. They want somebody to connect the dots and come up with a brilliant assessment. But that's not what convinces them. Decision-makers, whether they're military, governmental, civilian. Decision makers, I found, don't want to make a difficult, dangerous decision based on a hunch, on an assessment, on somebody just connecting the dots. They want to know specifically what the nature of the problem is and what specifically I need to do in order to solve it. That's what makes them receptive to intelligence.

Q: And of course, one of the successful surprise attacks that you studied were the 9/11 attacks, and you document really an incredible amount of advance warning going back for several years. And that's one particularly strong aspect of the book I found where it's all in a very concise way, one after another, it's really quite compelling and even more astonishing how the policy makers didn't adapt to these increasing warnings. Why do you think the conventional wisdom, that this was a failure of imagination, is wrong?

Dahl: Well it is amazing. And I tried to put together, as you say. In my book, all the different warnings that we had before 9/11. And it's amazing. On the one hand you might think, how is it possible that our nation's leaders didn't understand these warnings? But then I realized that almost all of those warnings were of this broad, general, non-specific nature, what we call strategic warning. And usually we think, our experts think we need strategic warning. But our decision makers didn't want to listen to it. In fact there's one case where the FAA, the Federal Aviation Administration- they have their own intelligence organization- and their intelligence analysts put together the dots. They assessed that there was a real significant threat to terrorists associated with al-Qaeda, perhaps using airplanes as bombs. But they didn't have any specific facts to back it up. You would still think that the FAA, that organization within our government would be the most willing to listen to warnings like that. But even the FAA leadership wasn't willing to do anything about that when all it was, was analysis. They hadn't been able to get inside the al-Qaeda plot and give specific warning. So it's not that strategic warning that's really needed. Instead, what we need is even better collection, which is tougher on a specific threat. And that's what we didn't have before 9/11.

Q: Could it also be that the amount of warning, that the repetitiveness of the warning, actually dulled its effect?

Dahl: We do see that happen fairly often. In other studies of intelligence failure, such as the bombing of the Marine barracks in Beirut in 1983, there were lots of warnings. Dozens of warnings in fact, of the possibility of car bombs against our facility there, and they did dull the warning. In fact the Marine leadership said afterward, “Hey we didn't know – [we didn’t] need anybody else telling us to watch out, we were watching out as much as we could.” What we didn't have in that case, was we didn't know who to watch for. We didn't know what kind of specific threat to have happen. And that's what also happened before 9/11. So again we talk about connecting the dots, we like to think that's what the 9/11 Commission said, we need to connect the dots better. We need to collect the dots much more specifically. We need much more fine-grained intelligence if we're going to use that intelligence to stop bad things from happening.

Q: And when you say fine grained, do you mean focusing our intelligence effort?

Dahl: I do. In fact, that gets into a topic that often people talk about today- big data, collection of  large data sets.  What I found is, it's not big data that prevents bad things from happening, it's little data. It's really fine-grained intelligence. In fact, for a terrorist plot you need someone inside that plot, and you need to actually know what precisely a decision-maker can do to prevent that.

Q: And one of the concerns I've raised about big data is that the more data the more alarms get set- maybe lower level alarms, but more alarms. And in your discussion of 9/11 attacks, with the possibility of false alarms, the fact that so many of these hair raising incidents didn't actually come to fruition. I remember in the FBI in 1999 in the run-up to the Millennium there was all sorts of concerns that you documented in the book that all sorts of groups were going to take advantage of the dawn of the new Millennium to have attacks- and then when there weren't any, you know, the next warning was maybe a little less likely to be heeded. Is that a possibility?

Dahl: Well that's certainly something that we see in intelligence all the time. In fact that's one of the important ways that intelligence for domestic purposes, Homeland Security issues- is different from international security intelligence, in the sense that to some extent, when you're focusing your intelligence assets outward toward the rest of the world, toward outward enemies, false alarms and false positives don't matter quite as much because, well, we're on alert anyway and a wrong report about a potential foreign enemy that doesn't come to fruition may not be that dangerous, that bad. But when that false report is about a domestic person, domestic organization, not only do you risk over-warning your decision-makers who then become numb to those warnings, but possibly, you've blamed or you've accused an American citizen of doing something that they weren't trying to do. So it's a double problem domestically.

Q: You also, for the book, compiled a database of successfully thwarted or prevented terrorist attacks. What did you learn from studying those incidents?

Dahl: I found really two major, major points. One is that the nature of the domestic terrorism threat in the United States today is actually more serious, more severe than many believe it is- especially when you consider that there are a number of plots that have been thwarted since 9/11 from domestic right-wing or other sorts of organizations. It's certainly not just international terrorist groups. It's certainly not just Muslim groups. That's the nature of the threat. But then the most important question I thought that we need to look at is how are these threats thwarted? How do we stop these things from happening? And here I found that for the most part, the tools and techniques that are used to stop these would-be terrorists are usually very similar, the same tools and techniques that law enforcement has been using for many years. It's tips from the public, it’s undercover officers, it's the use of informants. It's not so much when we're talking domestic homeland security threats, it's not so much the national intelligence assets such as spies overseas or satellites or even signals intelligence efforts.

Q: And so over the course of the last year, there were some intelligence leaks, and many Americans were shocked about the scope of the domestic electronic surveillance that's going on. What did you find about the role of that type of electronic surveillance in stopping terrorist attacks?

Dahl: I think that what we're finding is that the domestic plots that have been foiled --I found over 100 domestic terrorism plots that have been prevented since 9/11- at least from what I can tell, and from what we've seen from a number of the official studies into the National Security Agency programs for instance. We aren't finding that it's that sort of program that stops domestic plots and homeland security plots. Those sorts of programs may certainly be effective overseas, we don't have the information on that, but domestically it's mostly this low level, much more of a law enforcement approach that seems to be effective.

Q: You actually also wrote an essay called “Domestic Intelligence Today: More Security but Less Liberty” where you describe how we've created this new domestic intelligence enterprise that collects a lot of information at the state and local level. And you said that the creation of this has altered the balance between security and liberty, and that while it's clear we're moving toward more government control, we've seen a benefit in security. But what metrics do we have to measure whether that investment in this domestic intelligence enterprise is actually improving security?

Dahl: That's a great question. On the issue of the balance between security and liberty, I think one important point is that soon after 9/11, most of our government officials and the 9/11 Commission itself argued that there didn't have to be a trade-off between security and liberty. We could have both. I think that's wrong. I think that what we are finding and the Obama administration is acknowledging that today- we are finding that in order to have greater security on the domestic front, we may have to give up some civil liberties. We may have to have more intrusive government interference and surveillance. The important thing though is to realize that the place we put that balance should be something that we, as a society, can recalibrate. I don't think that's happened. We haven't done that since 9/11. We haven't been able to get a good handle on an agreement among our government agencies on what the nature of the threat is today and whether or not we have put that balance too much towards security and not enough towards civil liberties. But I'll say that a way that I think we should measure, a gauge to determine whether domestic intelligence efforts are proper, is that we should use the same thinking that the Food and Drug Administration uses to decide whether a drug should be able to be used by the public. And that is of course that a drug needs to be safe and effective. If we look at it that way, first of all that means that a domestic intelligence effort needs to be effective. It needs to actually be able to either stop bad things from happening, or have some other value. And here I don't think our government has done a good enough job of being able to make that case. But the other part is it needs to be safe for civil liberties, individual freedoms, at some level that we, in society agree is the appropriate level. Right now it looks as if the balance is off. We haven't been able to determine that these programs are sufficiently effective, and we do think it's possible that they may be causing problems for our individual liberties.

Q: And do you think the excessive secrecy- certainly around our national intelligence programs, but even around the domestic intelligence programs is actually inhibiting both that discussion and the democratic control over these institutions that would require that kind of analysis?

Dahl: I think we definitely need to have what we started to have, but we've only started to have it, which is a much better national discussion about intelligence. Certainly back during the Cold War and during the time when I was busy in my military intelligence career, intelligence was a topic for spy novels and occasionally for public discussion, but for the most part, not very much. But today, especially because the information that's being collected, the dots that are being collected are about you and me, are about the American people, the American people need to have more oversight, more availability to discuss these programs. Although I’ve got to admit, when I was an intelligence officer, I wouldn't have wanted greater public freedom and public information because that would seem to harm the process.

Q: But as a Naval Intelligence Officer your focus was at hostile foreign nations or hostile organizations.

Dahl: That's right. And that's one of the big shifts in intelligence since 9/11 that maybe the American people don't quite realize sufficiently, it's that we really have changed the focus for much of our American intelligence community. When I was a Naval Intelligence Officer tracking the Soviet Navy back during the Cold War for instance, one of our guiding principles was don't spy on Americans. Well, after 9/11 we realized, and I believe it's absolutely true we realized that there are threats that can involve Americans. There are threats that can be within our country. So now sometimes we need to spy on Americans. But I'm not sure whether the same intelligence agencies, the same intelligence officers who used to spy overseas or spy on the Soviet Navy, I'm not sure whether those organizations and individuals are the right ones to be doing the spying on Americans. For instance I don't think you, as an American would want to hire me, as someone who trained and grew up in the old ways- you wouldn’t want to hire me to conduct domestic surveillance. I don't have that skill set.

Q: And with the terrorism incidents that were thwarted, it was traditional law enforcement methods that mostly were most effective. Is there a way that we can turn the domestic intelligence activities, turn those resources into more of that traditional policing do you think?

Dahl: I think that it's important if we can somehow, find a way to have our domestic intelligence agencies achieve the same level of trust and understanding with the American people, that law enforcement does. Now certainly there are problems with law enforcement and different communities. In New York City for instance, the NYPD is under fire for civil liberties violations, charges, those sorts of things. But still law enforcement as a whole, throughout America, is generally respected by the American people. Our communities know that if they're not happy with their sheriff's department, chances are they can vote in a new sheriff. But we don't have that opportunity at the national level. That's why one of the things that we need to do is re-emphasize in some way the state and local, the lower level intelligence and surveillance and law enforcement assets, and bring those to bear. You can much more easily calibrate the nature of your domestic surveillance effort at a local level than you can at the national level.

Q: And with traditional law enforcement, oversight mechanisms are built in through the courts, through civil procedures. How does intelligence oversight fail? Have you looked at that at all? And did that play a role in any of the successfully thwarted attacks, that it was actually oversight mechanisms that compelled the intelligence agencies to act more effectively?

Dahl: I think that certainly at the national level, the 9/11 Commission pointed out, they're absolutely correct that one of the problems before 9/11 was that the intelligence community did not have sufficient oversight, in the sense that intelligence oversight really has two parts to it: One is to try to make sure that the intelligence agencies don't violate the law, don't do something wrong. But the second part that often doesn't get enough attention is that intelligence oversight can and should help intelligence agencies and officials focus on the things that need to be focused on. Before 9/11 really nobody was focusing on international terrorism. The intelligence overseers never grilled the Director of the CIA on why you're not spending more time worrying on those things. That way, today, we need to have that type of oversight as well. But one issue is that when you get down to the state and local level, oversight involves many different organizations, many different localities and jurisdictions, so it's a more complicated problem.

Q: So how do Americans evaluate the performance of the intelligence enterprise? You quote Robert Jervis who said that for Americans to expect the intelligence community to predict non-routine political occurrences is unrealistic and you yourself say that some sort of surprise is inevitable. So, if prevention of strategic surprise is the mission, and surprise is inevitable, how do Americans evaluate whether the $70 million investment, the risk to their civil liberties are actually worth the security product they receive in return?

Dahl: Measuring intelligence value is a tough job. In fact I remembered during my intelligence career, one of my bosses once asked me in frustration, “Can you ever tell me what a pound of intelligence is worth?” And there's no way to do that. In a military sense, you can judge how much a gallon of gasoline is worth, or how much a trained pilot is worth. But intelligence is tougher. What we do usually is revert back to that maybe most simple measurement, which is whether or not bad things happen. And that's the quote that political scientists like Robert Jervis have typically used. We want intelligence to keep us safe from bad things. And during the Cold War, that usually meant keep us safe from nuclear annihilation. And that worked pretty well. But during the Cold War lots of Americans died in lots of incidents and small wars around the world. Today the concern is, and I think it's legitimate, that our intelligence officials are saying we're trying to hold the intelligence community to a higher standard, trying to prevent the death of one individual American in a terrorist attack for instance. And that really is impossible. But what think we need to do is realize that intelligence and the American intelligence community does many different things. When we're looking domestically, the intelligence community needs to be able to better explain its performance against this very specific problem of domestic terrorism. And here, we have a problem because it hasn't been shown in even the official studies that looked at it that the National Security Agency programs have actually had a hand in stopping terrorist plots. But it's also important to keep in mind that the intelligence community is involved with many other things, many other things than just that. And I think that having a better national discussion, a national dialogue, the American people will realize that there are many other problems and issues that the intelligence community around the world and within the United States focuses on and that often provides a great deal of value.

Q: And it's nice that we're having some of that dialogue, but it's somewhat by accident because there were leaks. Do you think the intelligence community could be better at providing more information so the American people can have a better understanding of what the debate is about?

Dahl: Absolutely. The intelligence community is typically very bad at public relations- although really today it's much better than it used to be. I remember back during the Cold War we used to talk about the National Security Agency as “no such agency.” During my Navy career, I was in fact a reporter for a little while, and once I tried to call the NSA for comment and they didn't even have a press office. So we're much better than we are today, but we're still not good enough. And the intelligence community needs to realize that what I call receptivity, in other words the trust and understanding between the intelligence agencies and its consumers, the people who receive intelligence and pay the bills, that receptivity and trust is vital. And right now we definitely don't have that on a domestic level between the American people and the intelligence community. Basically the intelligence community needs to step up its act when it comes to public relations.

Q: And what reforms would you recommend?

Dahl: I think two major, major changes are needed: One on the national level and another on the state and local level. On the national level, as I mentioned earlier, we need a new type of intelligence agency, a new type of intelligence mindset, when we focus the telescope of intelligence, if you will, away from the rest of the world and focus it on the domestic population. What I would like to see is a new domestic intelligence agency similar to the British MI5, their domestic security agency. Now, it's probably doubtful that we'd ever have the national will and the budget and the money to set up an agency like that, but we might be able to do something with either continuing to reform the FBI intelligence effort or perhaps turn the DHS intelligence effort, which is usually considered a small player in the intelligence community, but beef up that effort and have an organization that uses the tools and techniques of intelligence, but at the same time it is very understanding about the need for civil liberties and privacy, the sorts of things that law enforcement officials are typically used to. But at the same time we're doing that at the national level. We need to have a more robust and even more professional state and local intelligence effort. There are a lot of efforts going on around the country establishing new intelligence fusion centers and developing law enforcement intelligence, policing intelligence efforts. I think we need to step that up, because as I mentioned before it's at the local level that you could calibrate those intelligence efforts much more appropriately. You may find you need a different type of intelligence effort in New York City than you do in Seattle or Portland for instance.

Q: And again, when you're using the term intelligence, you're not talking about bulk data surveillance, you're talking more focused.

Dahl: That's right. I really think that unless the national intelligence community can explain to the American people better, what the value has been so far of the bulk surveillance programs- I think that what I found in my book is the case, and that is that that it's not big data, it's little data that stops bad things from happening.

Q: One of the things I really appreciate about what you're writing is how much it's important for the American people to have this discussion, and your book and your articles are certainly a contribution to that discussion. What other reading materials would you point to that viewers could look at and get a better understanding of the intelligence community?

Dahl: Well, I'd like to say that in addition to my book, [which] I hope, is a good introduction to the broader questions about intelligence and intelligence failure- I think actually we have a bit of a problem, because although there is a lot of good work out there that's been done looking at the current national security agency and domestic intelligence problems, a lot of good work done by some of the official studies such as the presidential review group that issued a several hundred page report- the problem is those reports are awfully long. A lot of good information there, but I would encourage that someone interested in this take a look and skim reports like those. For instance there are also the dueling opinions by two federal judges. One that ruled that these programs are constitutional, the other ruled that they're not. A lot of good information there, but kind of hard to wade through, I've got to admit. One of the most useful single articles that I've seen recently on this was just in Foreign Affairs recently by Daniel Byman and Benjamin Wittes called “Reforming the NSA” in the Foreign Affairs May/June 2014 issue. And they struck a good middle ground. Usually the op-eds and the opinion pieces we read are either that the NSA is doing a terrible job and Edward Snowden was a hero, or on the other hand the NSA is doing a great job and Edward Snowden was a traitor. Well, the truth may be somewhere in the middle there. And I also think in fact that one of the authors of that Foreign Affairs piece, Benjamin Wittes, has a very useful blog called The Lawfare Blog, he's a fellow at Brookings and has a lot of good interesting coverage on these issues on his blog.

Q: Great. Well I thank you so much for your contribution and for speaking with me today. So thanks very much.

Dahl: Thanks very much. Thank you.