The article below was first drafted several years ago and approved for release by my SSO (Special Security Officer), but never published. I have added some minor points to update the article in light of the incident on Christmas Day, but the fundamental conclusions are unchanged. This article represents a condensation of the book I am writing on Why Intelligence Fails and What We Need to do to Fix It.
“Another Intelligence Failure.” Once again elements of the Intelligence Community failed to ‘connects the dots’ and we (the US) almost fell victim to another terrorist attack.
Strictly speaking, the events of Christmas Day, wherein a man, by all indications associated with an al Qaeda related organization in Yemen, was able to board an airplane with an explosive device of some sort, are not the purview of the intelligence community. The intelligence community collects data, analyzes data, stores data and disseminates data. It is up to others to decide what to do with it. That, at least, is the mantra of the Intelligence Community leadership.
To give the IC credit, there are any number of events that the IC has acted on over the past one, five, ten and 40 years in which the right information was collected, the right analysis took place, the right people were informed, and the right decisions were made.
Such a claim misses the mark, however. A doctor who performs 5000 successful operations is still brought to task if he botches number 5001; an airline that flies several hundred thousand flights a year for a decade, without any incidents or injuries, is still ‘dissected’ by the FAA if they have an accident. The IC cannot claim a special exemption.
Furthermore, the fact remains that in nearly every case that comes to mind (and many that have received little if any attention), more then enough information had come to light before the incident and failure was not the result of some analyst to recognizing the impending event, but the failure of the system to act when notification began to move ‘up stream.’
In fact, a look at the failures of the past 10, 20 or 30 years and there are many, reveal few that were terribly nuanced once the facts were presented. In nearly every case, the information had been collected early enough, was fairly obvious, was correctly assessed, and then the system failed. Why? The answer requires a little background on the origins of the current US Intelligence Community.
A Brief History
The IC was created by the National Security Act of 1947 and a host of follow-on legislation. But, while each succeeding piece of legislation has changed elements of the IC, the fundamentals remain the same. And those fundamentals are a direct result of the Japanese attack on Pearl Harbor and the series of investigations that followed.
The essential lesson learned from the Pearl Harbor investigations was that there needed to be independent reporting directly to senior decision-makers in Washington, reporting that went around analytic debates and parochial – military service - interests. Senior personnel in Washington: the President, the Secretaries of Defense and State, the Chiefs of the services, needed instant reporting, without the delays that would result from internal analytic debate – from a communal sharing, as well as the delays that might be caused by service animosities.
The result was to separate from the Army and Navy large segments of their intelligence and cryptologic capabilities and create two new organizations: the National Security Agency and, with the remnants of the WWII Office of Strategic Services, the Central Intelligence Agency. These two offices, with their own collection and analytic and reporting systems, would report directly and independently to the President and other senior decision-makers.
But this conclusion – and the resulting organization and intellectual infrastructure - was based on a massive, unspoken assumption: that the problem being watched was clear and well defined. And in the case both of Japan and the situation facing the US in 1947 (when the legislation was drafted and passed) this was true.
In the case of Japan, the US identified the Japanese as a ‘competitor’ as early as 1900 and the capabilities of the Japanese navy were key drivers of US Navy construction and training for several decades. The US understood Japanese grand strategic designs (the East Asia Co-Prosperity Sphere) and the intentions of forcing the US and European powers from East Asia; the US understood Japanese political and economic strategies in China, South-East Asia and Micronesia; the US understood Japanese military strategy – both naval and ground (army), the intended conquest of Manchuria and coastal China and South East Asia, the likely move into Indonesia (the Dutch East Indies) and the Philippines by the end of the 1920s, before any invasion took place; the US understood Japanese operational issues – how the navy and army would move, their strengths and weaknesses; and we understood their tactics and their weapons – how they fought.
The failure at Pearl Harbor was specifically tactical and technical intelligence: the US lost contact on the Japanese Combined Fleet in November when the fleet changed its encryption, and the US failed to recognize that torpedoes could be adapted for use in the relatively shallow waters of Pearl Harbor.
It is worth noting that the first ‘attack’ of Pearl Harbor by aircraft from aircraft carriers was in a fleet exercise several years before World War II when US Navy aircraft ‘attacked’ the harbor during a fleet exercise.
In 1946 the US faced a similar situation: the Soviet Union. And the Soviet Union obliged the IC over the next 45 years by presenting itself as a symmetric intelligence target that always gave the IC something to report on.
As with the Japanese, with the Soviet Union the US had a clear grand strategic picture: the spread of global communism; there were clear Soviet political and economic strategies that were tracked by the IC; there was a clear and massive military strategy that was dissected in great detail. Over the course of the next 4+ decades the IC followed Soviet military plans, identified the major theaters of operations of the Soviets, and developed detailed analyses of Soviet operational, tactical and technical capabilities. The IC tracked Soviet ships and aircraft, monitored the readiness and training status of Soviet regiments across the Warsaw Pact and kept detailed matrices that equated to warning of this or that Soviet operation.
In both cases, the US faced a clearly identified problem, that could be reduced to specific tactical and technical issues, and a great deal of time and effort was spent to track those tactical and technical issues.
With the arrival of nuclear weapons the issue became even more focused, with Soviet nuclear weapons becoming the keystone of the IC collection, analysis and warning problem. Everything else took a second seat to this one element.
And so, the US was surprised by nearly every single strategic development over the past 60 years. The US designed, developed, built and reinforced an intelligence community that tracked in excruciating detail a clearly defined problem; anything that fit within the intellectual construct was symmetric; anything outside was asymmetric. Anything that did not fit into that clearly defined problem was, to some extent, ignored. The focus had to remain on the one central problem. But strategic surprise is, almost by definition, asymmetry.
The list of failures is long and unpleasant. An abbreviated list includes:
Soviet development of atomic weapons
Korea 1950
Hungary 1956
Czechoslovakia 1968
Soviet Invasion of Afghanistan 1979
Fall of the Shah 1979
Iran-Iraq War 1980
Collapse of Warsaw Pact 1989
Iraq Invasion of Kuwait 1990
Collapse of Soviet Union 1991
Arab-Israeli Conflict – nearly every incident
Terrorist Incidents for three decades
In literally every single one of these incidents, and in scores of other failures that have fallen into the shadows of history, there was more than enough information available to provide some sort of ‘heads-up’ to our leadership. In virtually every case there were one or two analysts (often more) who had made the case for the warning; they had sorted through the intelligence, understood the associated environment and made the correct assumptions, and then had constructed a logical argument that showed their conclusion was valid. It must also be noted that in most of these cases there were case officers and their agents who had repeatedly risked their lives to acquire critical pieces of intelligence for the analysts to dissect and then integrate with other intelligence.
But in each case the senior leadership, the President and his closest advisors, were ‘surprised,’ presented with a situation as it happened rather then made aware before it happened so that they might preempt the event, in many cases presented with situations for which there was not even a planned response.
The Contradiction
Anyone who has ever had any dealing with the Intelligence Community will tell you the same few things: they have never met so many bright, hard working people. Second, the technology that the IC uses to collect information and otherwise assist these people in doing their jobs is simply remarkable. Third, the level of dedication and patriotism among the members of the IC is unmatched by anything they have seen anywhere. Tens of thousands of people, often laboring in true anonymity, have committed their lives to defending this country. It is no exaggeration to say that a walk through the halls of the CIA or NSA or any of the other agencies, or better still, working with these people in the field, reveals a professionalism and patriotism that is truly remarkable and humbling. And this is as true today as it has been for six decades.
Then why does it fail? Why does the IC fail to alert leadership? Tens of thousands of very bright, diligent and committed people and yet they continue to fail? How can this be?
Four Problems
The answer lies at the intersection of four tightly interconnected issues: organization, leadership, authorities, and investments.
Organization
First, the IC is organized and structured to do one thing: provide ‘alertment’ to the leadership by tracking a clearly defined problem. As noted earlier, the construction of the IC took place in the long shadow of the attack on Pearl Harbor and the many investigations that followed. The focus was, and is, to provide a flow of information ‘up’ the organization to the key decision-makers: the President, the Secretaries of State and Defense, the Joint Chiefs. The organization is not designed to share information; it is designed to push it ‘up,’ once it passes certain thresholds.
What does it push ‘up?’ Information that it is clear that the decision-makers want. Once again, if the decision-makers, and the support mechanisms around them, haven’t identified something, whatever it is, then the information, no matter what it is, won’t get moved ‘up.’
But both Congress and the President (actually, at least the last five Presidents) have called for better sharing of intelligence. Why hasn’t intelligence sharing improved? To be fare, in some specific, narrowly defined areas, there has been some improvement. But, for the most part, the IC grinds on as it always has.
The reason for this is fairly simple: the organizations remain, at their most basic levels, unchanged; this despite decades of Presidential and Congressional efforts. At senior levels new offices have been created, new lines of authorities established, and new requirements stated. But at the production level the situation, despite all these ‘fixes,’ has only gotten ‘set in stone.’
The situation might be compared to a sports team: a football team is organized specifically to play football. The team members are all excellent athletes, but without a complete reorganization (as well as changes in training plans and support systems), they would fail miserably as a baseball team. By analogy, the IC was organized as a football team but is being asked to not only continue playing football, but also to play baseball and hockey as well.
The IC was built as a vertically aligned set of organizations that would track a clearly defined problem. But, in a very real sense the US Intelligence Community (IC), which, with apologies to Voltaire, is neither intelligent nor a community, nor was it intended to be either; in the sense that a ‘community’ is a set of interdependent organisms or organizations. And emerging threats demand an intelligence community with a higher degree of interdependence then our current construct allows.
In any case, the IC is – rightly – being asked to identify new, emerging problems while it is expected to continue to track the clearly defined problem. The same organization cannot do both, and it has proven this repeatedly over the last 60 years.
To fix the organizational conundrum requires separating the community into two major pieces: one that is a predominantly vertically aligned community, focused on defined problems with clear, deductive processes to address defined problems; and a second, horizontally aligned community, with a much more rapidly iterative, inductive process to address the problem of identifying emerging problems.
The two IC ‘Halves’ would overlap at two points: the management of the collection process and the management of the information collected. Both must be placed in the same hands; the DNI – the Director of National Intelligence – must be given sole and absolute control over both collection and that which is collected. No one, not DOD, not NSA, not CIA, not State, etc., would own any information. Once information on any foreign activity or event, or person or thing is collected, the DNI owns it and he alone (or the DNI’s boss – the President) set the rules for use.
To reiterate, the US needs both capabilities; but one organization cannot do both well. This will not be easy or quick, but it must be done if we expect the IC to provide real strategic warning.
Leadership
The leadership of the IC has failed the people they lead, the people they work for and the nation they serve. This isn’t because these are evil or incompetent people. But it is because the organizations that make up the IC routinely select senior leaders who will function well within the bureaucratic environment of Washington: they are the people who are expert at crafting procurement programs, defending budgets and protecting organizational turf. Rarely are they either gifted leaders or gifted intelligence professionals. In the few cases where there have been gifted intelligence officers in leadership positions they have failed to perform as competent leaders who can manage to force through real change. The gifted leaders and the superior intelligence professionals normally do not promote well within the IC, and the IC suffers as a result.
The IC needs leadership that is willing to change organizations, even when that means one organization ‘gains’ and another ‘loses.’ IC leadership has traditionally resisted real change, choosing to ‘rearrange the deck chairs’ rather than change course, fighting each other to protect ‘budget share,’ then developing complex programs that ensure each is protected (a ‘you scratch my back, I’ll scratch yours’ solution), but failing in the end to address new concerns, help the actual ‘workers’ within the community or improve the intelligence product.
In the Iraq Study Group Report (issued in the summer of 2008) it was noted that the “embassy of 1,000 [people] has 33 Arabic speakers, just six of whom are at the level of fluency” (pg 60). That 5 years after moving into Iraq we still had less then 1% of the embassy staff fluent in Arabic is a disgrace. And it represents a clear failure of leadership. Later, the report notes that: “We were told that there are fewer than 10 analysts on the job at the Defense Intelligence Agency who have more than two years’ experience in analyzing the insurgency. Capable analysts are rotated to new assignments, and on-the-job training begins anew.” (Pg 62)
It is not new. At the time of the seizing of the US embassy in Tehran in 1979 the author recalls being told that the DOD had only 2 fluent Farsi speakers in uniform. While that number may have improved a bit, it is enlightening to note that the US Navy (8 years after the US pushed the Taliban out of Afghanistan) started a program in 2009 to develop more Afghani and Pakistani expertise. Sadly, they began it by moving a large number of senior officers into the program, officers who will work at large, flag level operations centers, and who will be departing the service within several years – rather than bringing in junior personnel who can be committed to developing real expertise for the future, while working in the tactical centers and forward bases where the real understanding can be generated.
Authorities
Instituting the changes needed – the massive reorganization – splitting some offices, joining others – will be impossible if we expect it to happen within the current construct. The IC cannot fix itself. The system is too large and the level of bureaucratic inertia too great. A new Director needs to be appointed, with a new team, and a new plan. And Congress needs to give him the authorities to execute the plan. The new Director needs to brief Congress, but then Congress needs to bless the plan and get out of the way.
The new director needs to be given the authority to fix the collection architecture, expand personnel hiring, and change a wide range of personnel policies that have no place in the world of intelligence, and realign tasks and internal authorities. The new director will need true authority over the budget, and true ‘hire and fire’ authority. And the new director should be given a clean slate: all current admirals, generals and SES in any intelligence or intelligence related position should tender their resignation, and the new Director be allowed to ‘cull the herd.’
True excellence in analysis is the result of years of focused effort. Yet the bureaucratic and personnel policies of the government as a whole work against developing this excellence. Does it develop anyway? To some extent, yes it does. But it does so despite the IC, not because of it. And here again the leadership has failed to make the case for excellence over bureaucratic ‘correctness.’ Personnel training policies and standards, as well as assignment practices, need to be revamped to allow the creation of a large number of truly expert, all source analysts.
In addition to this very necessary and massive reorganization, the IC needs bureaucratic streamlining. The current structure – while well intentioned – has resulted in a complex layering of bureaucracy. The large and top-heavy superstructure that now exists, with a host of directors and deputy directors, each with their own staffs and staff processes, has done nothing to either improve the product or streamline the process. This bureaucracy needs to be reduced. There are any number of ways to do this, but it needs to be done.
Investment
Finally, we need to invest in our IC, and it is ‘our IC,’ the entire nation’s intelligence community. As with the DOD, the IC has found its tasking increasing over the past seven years. Unlike the DOD, the IC’s tasking has been increasing regularly for much of the last 20 years. And this increased tasking has taken place even as the national investment in intelligence – as a percentage of GDP – has decreased.
As the nation reduced the size of military force, the argument that was used by one and all was that it was the quality of our intelligence that would allow us to safely reduce the size of the DOD. Intelligence was the ‘ace in the hole.’ Yet, since we began the massive DOD reductions in 1986, and with the exception of recent increases in the Army, the Marine Corps and the various Special Forces units, that drawdown continues to this day, we have also seen a dramatic decrease in the effective level of funding for the IC. While the numbers are difficult to find, it is reasonable to estimate that 40 years ago the nation was spending considerably more than 1% of GDP on intelligence; today that number is less then ½ of 1%. When we recognize that relative pay has increased (which is a good thing), we can see that a larger percentage of that budget goes to pay for a smaller number of over-tasked people, and an aging intelligence infrastructure is facing obsolescence.
The nation can, if it chooses, continue to reduce the size of the DOD, but to reduce both the size of the military and the organization that is supposed to provide even more of a warning then ever before, while we reduce our ability to defend or respond militarily seems to be the height of folly.
Conclusion
The nation needs an IC that can both reliably track and provide warning on major threats – the nuclear forces around the world, and an IC that can provide strategic warning, that is, notify the President and other senior decision-makers with adequate time for them to act preemptively and thereby defend the United States, its people and its interests. To do the latter is to develop a capability to identify and provide notification of emerging problems. But this is a fundamentally different problem then tracking an identified problem, for one is deductive and requires fine-grained analysis, the second is inductive and require developing understanding.
But the IC as it is organized can only provide warning on defined problems. No amount of’ rewiring’ at the top, or orders to improve sharing will change this. It is time we face the hard task of building the IC the nation needs and deserves.
