The recent revelation of Edward Snowden's successful duplicity in procuring a gullible colleague's password brings yet more evidence of the intelligence community's internal failures in the Snowden fiasco. Amidst multiple ongoing investigations and damage assessments, an unforgiving spotlight continues to shine on the intelligence community (IC) while it struggles to address its own deficiencies and culpabilities. Yet neglected in much of the ongoing debate about the American intelligence community's surveillance and collection methods are the systemic pressures placed on intelligence by policymakers.
Last Friday I participated in a conference on security and privacy co-sponsored by the Chicago Council on Global Affairs and Notre Dame's International Security Program, and organized by Professor Mike Desch. As decidedly unromantic as the conference was (it took place on Valentine's Day, after all), it also marked the NSA's recent and much-welcome decision to partially lift the veil and participate more in public debates about its role in protecting a free society, evidenced by Gen. Keith Alexander's keynote address.
This post draws on my remarks at the conference. I tried to emphasize a point that has been neglected in the Snowden tumult of the past year: intelligence is an instrument serving policy, and intelligence activities are carried out at the behest and under the ultimate authority of the presidency. For all of the ongoing controversy over the NSA's methods, at the end of the day the NSA and other three-letter agencies are merely serving the mandate of senior policymakers.
While policymakers have many justifiable frustrations with the IC -- whether for occasional intelligence errors and failures, lapses in tradecraft, or surreptitious efforts to shape policy -- the intelligence community also faces a considerable demand-side challenge. In my experience, policymakers are also prone to placing contradictory, sometimes impossible demands on the intelligence community. Here I speak as a former policymaker and consumer of intelligence -- and thus admittedly one of the erstwhile perpetrators of these pressures, which are persistent and transcend administrations. Following are seven of the impossible demands that policymakers place on the intelligence community:
1) Do whatever it takes to get me the intelligence I need -- but don't break any laws, offend any allied nations, prompt any Congressional investigations, make any mistakes, or let the media find out in the process. This is the overriding challenge facing the IC: the constant pressure from policymakers to provide accurate warnings, infallible predictions, and profound assessments, all while avoiding any of the manifest pitfalls that can make the policymakers' job even harder. I suspect this one might especially resonate these days at Fort Meade.
2) Give me accurate and precise forecasts about the future, but don't make any mistakes. Policymakers don't like hedged and vague predictions such as "there is a good chance the sun may or may not rise in the East sometime in the next year." Policymakers also don't like specific intelligence predictions that don't pan out exactly as such. Consider an intelligence forecast that "In Moscow on April 1, 2015, Vladimir Putin will rise at 6:47 am, eat a poached egg for breakfast, and order yet another missile test in violation of the INF treaty." Pity the poor analyst if it turns out that Putin does all of the above -- except eats oatmeal that day instead.
3) Produce intelligence on what I want to know about, but also tell me what I don't know about. Policymakers naturally demand intelligence on issues they are interested in, but also expect the IC to brief them on topics the policymaker is unaware of, yet might need to know about because of a possible looming crisis. This may sound like a reasonable enough balance, but for the IC briefer it can often result in being subjected to harangues such as "I want more on Iran, but you keep sending me analyses on Mali." While a place like Mali may not have mattered to policymakers in 2011, it suddenly did in 2012.
4) Don't just tell me about what I already read in the New York Times, but don't deviate from what I read in the New York Times. One of the dirty little secrets of intelligence analysis is how many products are overclassified and contain information about global events that is just as readily available in open source journalism. Yet while policymakers get frustrated if the intelligence they read only mirrors what they read in the morning newspaper, they also can get suspicious if intelligence products contain assessments that significantly differ from open source reporting, either in topic or viewpoint.
5) Warn me about what can go wrong, but encourage me about what can go right. The IC is largely structured to provide warnings about potential threats to American security and interests. But policymakers also want to find opportunities -- a hard ask of an IC that is congenitally wired to be pessimistic. Several years ago an NSC staff colleague and I met with a senior team of IC analysts for a discussion of reporting on global trends, and at the end of the meeting we asked them to work on an assessment for us of specific ways things might be going right somewhere in the world, and thus offer possible opportunities for US policy. Our query met with blank stares and puzzlement, and we quickly realized this was a community firmly conditioned by decades of policymaker demands to only watch out for what can go wrong. Perhaps understandably so, given that the costs of missing a warning can be much higher than missing an opportunity.
6) Don't just produce lowest common denominator consensus assessments, but don't disagree with each other too much either. Policymakers often complain that too many IC conclusions are hedged and watered-down to gloss over internal disagreements among analysts. This can sometimes lead to assessments afflicted with vague generalities. On the NSC staff we used to craft parodies of intelligence analyses that began "We judge with strong confidence that China is large, important nation in East Asia..." But if different elements of the IC instead produce multiple reports on the same topic that reach dramatically different assessments, policymakers will also complain "why can't the IC get its act together and come to agreement?"
7) Give me raw intelligence so I can make my own assessments, but also give me polished analytic products so I can understand what the raw intelligence means. Sometimes policymakers like to serve as their own analysts and read the raw take of signals intercepts and human intelligence reports. Other times policymakers want to read synthesized assessments that summarize and analyze reams of fragmentary data. To the vexation of the IC, too often those conflicting demands come from the same policymaker.
Some of these paradoxical demands admittedly just reflect the necessary balancing that must take place in the craft of intelligence. And within the U.S. Government, the policy and intelligence communities are engaged in a constant dialogue that at its best produces useful intelligence and effective policy. My worry is that in the wake of Snowden's treachery this dialogue and balancing have been disrupted. While the world now sees too much of what the intelligence community has done, it hears too little of the policy demands that made the IC do it.
MANDEL NGAN/AFP/Getty Images