A bipartisan Senate report, “Examining the U.S. Capitol Attack,” holds nothing back in blaming both the U.S. Capitol Police leadership and the intelligence community for security lapses in the Jan. 6 assault on the Capitol by President Donald Trump’s supporters.
The report strongly faults intelligence analysis from key U.S. agencies as contributing greatly to the police’s lack of preparedness. “A key contributing factor to the tragic events of January 6,” states the report, “was the failure of the Intelligence Community to properly analyze, assess, and disseminate information to law enforcement regarding the potential for violence and the known threats to the Capitol and the Members present that day.”
In his testimony before two Senate committees, former Capitol Police Chief Steven Sund said, “We had planned for the possibility of violence, the possibility of some people being armed, not the possibility of a coordinated military-style attack involving thousands against the Capitol.” He added that the Capitol attack was the result of a “lack of accurate and complete intelligence across several federal agencies.”
But review the details of the report and a different story emerges. The domestic intelligence analysts, in fact, did warn the right people in a timely manner. The report acknowledges that far from failing to warn, the Department of Homeland Security and the Capitol Police’s own intelligence analysts both had offered timely strategic and tactical warnings on Trump’s Jan. 6 rally. Even so, the Senate report decided to deem this “an intelligence failure.”
To understand why the Senate report gets this wrong, it is worth putting in perspective what intelligence analysis is, and what is meant by “intelligence failure.” The job of intelligence is to provide clarity and to assist in making coherent decisions in an uncertain environment. Intelligence analysis is often successful, but sometimes fails to warn. No one knows what the right “batting average” is for intelligence analysis. To a certain extent, some intelligence surprises are inevitable.
Waiting for Perfect Intelligence Is a Mistake
Intelligence analysis often is “estimative.” In the absence of absolute certainty, analysts must make informed guesses. But lack of certitude is no excuse for the police holding back on planning. It is important to be aware of intelligence estimates’ limitations and not have important decisions hanging on the next bit of intelligence reporting. Thinking that intelligence always can offer a complete and accurate picture in an ambiguous environment is delusional.
Likewise, analysts cannot justify bold predictions without ample supporting evidence. They must pay attention to the “base rate”—the likelihood of a security incident occurring. Pro-Trump political rallies in Washington, D.C., last November and December had occurred without major incident. For an analyst to have written a report explicitly saying Trump supporters would attack the Capitol to stop a routine election certification would have required substantial collaborative intelligence. Picking out lone reports after the fact from dubious social media bloggers inciting violence, as the Senate report did, doesn’t necessarily invalidate the original analysis. Intelligence analysts must weigh this information against reams of other information giving different and even contradictory messages.
Aside from these challenges, domestic intelligence analysis also was operating under two major constraints. One is that the objects of their analysis were American citizens exercising their constitutional rights to free speech and assembly. As officials noted in the report, most of the social media intelligence did not cross the threshold into actual threats. In the report, a Homeland Security intelligence official cautioned that in social media it is difficult to distinguish between constitutionally protected rhetoric and real threats. Under Homeland Security’s criteria, much of the social media reporting wasn’t specific enough to justify a warning. The analysts may have erred in being too cautious about a direct attack on the Capitol, but they said the social media reporting failed to support a stronger judgment.
Threat estimates can also be exaggerated. One analyst made headlines by predicting that based on his social media analysis, the Trump followers would be armed, but that threat was overestimated. Additionally, recent well-publicized predictions that domestic violent extremism would rise haven’t been borne out by recent reporting, which now shows a much lower rate of incidents than last year.
Another major constraint is that intelligence analysts often are restricted from analyzing the political leadership’s policies. In this case, analysts had to report on a potential security event arising from a political rally that the president had called for and would lead himself. Their ultimate customer for intelligence analysis, the president, was contributing to the security risk. How much freedom did Homeland Security analysts or others have to report that Trump himself might encourage, if not incite, a march of his angry followers on the Capitol?
A Series of Warnings Was Issued
Yet despite these obstacles, in late 2020 Homeland Security produced numerous strategic warning assessments on violent extremism and potential violence toward government officials, buildings and even events related to the 2020 presidential election and transition period.
As for tactical warnings, the analysts of the Capitol Police’s Intelligence and Interagency Coordination Division appear to have performed quite well. They reported a “dangerous situation for law enforcement” and “that Congress itself is the target” on Jan. 6. The division leaned forward as early as Dec. 21, highlighting blog posts that included references to the tunnels under the Capitol and indicating the participation of extremist groups such as the Proud Boys and the Oath Keepers.
In a Jan. 3 special assessment, the division warned of a “significantly dangerous situation for law enforcement and the general public alike.” As usual in these cases, better clarity on the demonstrators’ intentions had been emerging a few days before the event. Chief Sund and the Capitol Police should have adjusted their operational planning accordingly.
The Senate report criticized the inconsistent messaging in the division’s intelligence reporting. This may be valid given the fast pace of reporting, but facts on the ground can change quickly. What looks like good analysis today might be wrong tomorrow. The analysts appear to have been diligently updating the story as more evidence came in.
Often what we call “intelligence failures” are in fact policymakers failing to heed warnings. As Columbia University Professor Richard Betts wrote years ago, “Intelligence failure is political and psychological more often than organizational.” In this case, the Capitol Police leadership and other senior security officials probably could have used the reporting to improve security, but they chose not to.
That said, some perspective on security is necessary. The Capitol Police’s task on Jan. 6 would have been daunting even with better planning. Defending the Capitol—a building covering four acres with numerous entrances and windows—from an angry mob would have been difficult even with superb preparation. Any security system or plan is designed to deal with likely events, and no security system or plan can prepare for every contingency.
The essence of a good security system is how well it copes with the unexpected. An assault of this type has no precedent in our nation’s history. Yet the Capitol Police accomplished their prime mission: keeping the lawmakers safe. They achieved this by resorting to lethal force in only one incident, the fatal shooting of unarmed rioter Ashli Babbitt, which is remarkable under the circumstances.
Perhaps a larger police presence or a National Guard deployment might have deterred the attackers. But perhaps not: As Portland, Ore., witnessed last year, a heavy federal police presence in fact redoubled the determination of the rioters. For this and other reasons, senior officials in Washington had already decided that the D.C. National Guard would not be used to guard the Capitol. Some disasters simply cannot be easily prevented or mitigated.
Of course, intelligence officials should always be examining events such as Jan. 6 for ways to improve their tradecraft and processes. But blaming allegedly faulty intelligence is the first line of defense for policy failure. Before Jan. 6, senior security officials in Washington had timely and accurate warning of the attack, as the Senate report itself makes clear, even if that is not its intended message. The “intelligence failure” storyline is aimed at deflecting blame from security officials and elected leaders who failed in their responsibility for the public’s safety.