Was the 2016 Election an Intelligence Failure?
Editor’s Note: This is the sixth installment of “The Brush Pass,” a column by Joshua Rovner (@joshrovner1) on intelligence, strategy, and statecraft.
Russia tried to help Donald Trump win the 2016 presidential election. Its method was simple: steal emails from prominent Democrats and leak them to the public. According to multiple government and private sector reports, Russian intelligence organizations deceived email users into undermining their own communications security. It then used cutouts like Wikileaks and DCLeaks to reveal their communications. Russia also engaged in a massive disinformation campaign to sow confusion and doubt among the public. Spreading “fake news” supported Trump’s claims that the electoral system was rigged, that mainstream media outlets were untrustworthy, and that existing institutions were designed to favor elites at the expense of the people. All of this created an environment of distrust and disillusionment.
This was not the first time that Moscow employed so-called active measures to interfere with U.S. politics, but Cold War efforts were laughable flops. This case was different. The rise of social media created new opportunities to spread misinformation rapidly, and the razor thin electoral margin in key states meant that even minor shifts in public opinion could change the outcome.
Some observers claim Russian success was also an American intelligence failure. Writing in The New Yorker, Dana Priest charges the intelligence community with failing to predict the Russian disinformation campaign, and then failing to warn the public and Congress when it occurred. “Only after the fact,” she concludes, “when a Russian disinformation campaign had already tainted the 2016 Presidential election, did the Office of the Director of National Intelligence, another vast post-9/11 creation, disclose the Kremlin’s interference.”
Others have made similar arguments, including longtime intelligence officials. Former CIA Acting Director Michael Morrell suggested that the election was a multi-layered failure. Part of the problem, said Morrell, was a “lack of imagination.” While analysts had been worried about Russian cyber operations for years, they had failed to imagine how Moscow could use social media platforms as political tools.
The failure to anticipate Russian actions is a particularly serious charge, given that the intelligence community had made anticipatory intelligence one of the key tenets of its reform effort. According to the 2014 National Intelligence Strategy, the goal of anticipatory intelligence is to “detect, identify, and warn of emerging issues and discontinuities.” The concept is more flexible than traditional strategic warning against states and non-state groups. Instead, it encourages creative approaches to predicting new kinds of threats and opportunities, especially including novel quantitative methods. In addition, it focuses on “trends, events, and changing conditions” instead of specific adversaries. The strategy describes anticipatory intelligence as one of the community’s “foundational intelligence missions,” and the Intelligence Advanced Research and Projects Activity is currently sponsoring a variety of research efforts to develop new technologies, methodologies, and approaches to measuring accuracy.
Organizational changes have complemented this new approach to warning. In 2011 Director of National Intelligence James Clapper eliminated the position of national intelligence officer for warning, perhaps trying make good on the old adage that every analyst should also be a warning analyst. The idea was that subject matter experts with deep knowledge of specific regions or issues would be best suited to spot indicators of change. Better to get them in the habit of providing warning, rather than leaving the task to an office of generalists.
The community’s effort to grapple with Russian election meddling is arguably the first real test of America’s new approach to anticipatory intelligence. While we do not have access to classified assessments, what we know suggests it missed the mark. Intelligence officials frequently warned of Russian activities, noting that it was beefing up its ability to use cyberspace for espionage, military operations, and possibly sabotage. But they largely missed the key ways in which Russia interfered in the election. While anticipatory intelligence is meant to flag discontinuities, assessments of Russian activity took an evolutionary approach.
The unclassified annual threat briefings given by the director of national intelligence to Congress are instructive. In 2014, Clapper mentioned Russia’s efforts to integrate cyber-operations with conventional military operations and warned that it would continue to exploit cyberspace for espionage. In 2015, he mentioned how Russia was using the Internet for propaganda purposes, but focused more attention on the threat of cyber-attacks against industrial control systems.
The picture became a bit clearer in the January 2016 briefing. Clapper used familiar language about Russian propaganda but also included an important warning: “Russian cyber actors, who post disinformation on commercial websites, might seek to alter online media as a means to influence public discourse and create confusion.” This was a fair summary, though it wasn’t very precise. The tone of the briefing, especially when compared to previous years’ assessments, suggested that the community was more concerned about Western solidarity than with the effects of fake news on U.S. politics. And the briefing was silent about doxxing – the practice of stealing individuals’ private information and posting it online.
At a glance this looks like a clear a failure of anticipatory intelligence, though it will be many years before we have a full picture of what analysts actually concluded in the years before the election. Their classified reports may have been more prescient. We should also be cautious about accepting claims that the election was failure of imagination. The 9/11 Commission used the same phrase to describe intelligence before al-Qaeda’s attack, but it was wrong. Calling for more imagination is dubious advice in any case, because we are capable of imagining just about anything.
In this case, however, the critics may be right. It is certainly possible that the community failed to anticipate the scope and importance of Russia’s activities, despite all the investment in anticipatory intelligence. There are several reasons why assessments may have gone wrong. First, while analysts were aware of past Russian efforts to meddle in U.S. elections, those efforts never amounted to much. There were clearly other reasons to worry about Russian cyber-operations, as the annual threat briefings attest, but this wasn’t high on the list. Second, analysts may have been befuddled by the bizarre confluence of events in the campaign season. Russian influence operations may have been important, but only because they occurred during the strangest election in modern U.S. history. Third, the intelligence community may have missed the mark because anticipatory intelligence is still an immature concept. Despite its prominent placement in the National Intelligence Strategy, the community had only recently begun funding efforts to put the concept into practice.
Finally, intelligence leaders may have been too eager to make every analyst a warning analyst. While this idea wasn’t new, warning intelligence had previously been treated as a distinct discipline requiring its own training regimen. Asking line analysts who specialize in other areas to add this to their portfolio is a recipe for disaster, especially given that the community was simultaneously redefining what it meant by warning. And according to a recent survey, there was no complementary effort to increase the amount and quality of warning education. In short, analysts were being asked to perform a separate and still ambiguous mission on top of their existing duties and without sufficient preparation.
What about the claim that intelligence officials should have gone public sooner? The intelligence community, after all, was reportedly tracking Russian activities by at least spring 2016. The FBI privately confirmed to reporters that Russia was behind the DNC hack shortly thereafter, and over the summer of 2016 the CIA reported that Vladimir Putin was personally involved. Shouldn’t they have done more to protect the integrity of the election by sounding the alarm about Russian-sponsored doxing and fake news?
This criticism is less convincing. Intelligence agencies are responsible to the president, not the public. The decision to release intelligence is not theirs to make. It’s harder to judge Priest’s claim that intelligence officials kept congressional intelligence committees in the dark, as she provides no evidence that this occurred. Instead she cites a State Department official who was unimpressed with the level of detail in intelligence reporting, and infers that members of Congress were similarly uninformed.
In any case, while the community may have failed to anticipate the nature or consequences of Russian meddling, it deserves credit for attributing Russian actions, giving the Obama administration months to respond. Whether it chose the right response is another question.
Injecting intelligence into domestic controversies is inherently dangerous because it invites politicization. If the public expects intelligence agencies to weigh in, policymakers will be tempted to pressure them to reach certain conclusions. This has terrible consequences for the quality of intelligence and for intelligence-policy relations. Over time a certain cynicism is likely to emerge. Critics will accuse agencies of partisanship and the community as a whole will lose any vestiges of independence and objectivity; witness the claims that intelligence professionals are card-carrying members of the “deep state.” Intelligence agencies have struggled for decades to maintain political neutrality while still remaining responsive to political leaders. Asking them to join the fray will make that much more difficult. As U.S. officials work to prevent foreign states from influencing future elections, they should keep this danger in mind.
Joshua Rovner is Associate Professor in the School of International Service at American University. He is the author of Fixing the Facts: National Security and the Politics of Intelligence (Cornell, 2011), and writes widely about intelligence and strategy.