Published by The Lawfare Institute
in Cooperation With
It’s been four years since the FBI began its national security investigation of the Trump campaign, and Americans remain deeply divided over the probe. Democrats think the investigation was more than warranted by the number of suspicious contacts between Team Trump and the Russian government. Republicans think the investigation was a partisan hit job on an anti-establishment candidate.
They’re both right.
It would have been national security malpractice not to investigate possible Russian influence over the Trump campaign. Hostile foreign governments will always be tempted to use the openness of American presidential contests to boost their favored candidates or sabotage others. More such investigations will be needed in the future. After spending four years advertising the success of Russia’s interference campaign, the U.S. should not be surprised if other countries get the message and launch their own. Given the risks, national security agencies can’t be gun-shy about probing foreign government efforts to infiltrate the U.S. political system.
At the same time, there is a lot more evidence than many people realize that the 2016 investigation was pervasively tainted by hostility to Donald Trump. In part, that comes with the territory. Any time government officials order national security surveillance of people who want to kick them out of office, they will be suspected of partisan motives. Put charitably, the Obama administration bungled this dimension; it failed to recognize just how partisan its investigation of a political rival would look, and it did far too little to avoid the appearance of partisanship. Less charitably, there is reason to believe that the Obama administration milked the investigation for partisan advantage.
That less charitable view deserves respect. First because it’s backed by considerable evidence. And second because it’s unpersuasive to tell half the country that their suspicions are mere conspiracy theories that they should just get over. The U.S. needs a national security system that the whole country has confidence in.
Especially now. The United States has spent nearly 50 years guarding against one kind of intelligence abuse—the government turning its intelligence machinery against individual rights and unpopular minorities. It hasn’t had to worry much about a different kind of abuse—employing national security surveillance to achieve partisan political ends.
It’s not that it can’t happen here, as anyone would know who studied J. Edgar Hoover’s collection of dirt on politicians—or his willingness to share that dirt with presidents when they felt the need. The United States has been lucky in recent decades. Divided government and a narrow range of political differences discouraged incumbents from using intelligence capabilities against the opposition.
Now, not so much. If it sees members of the other party not just as wrong but as borderline treasonous, why wouldn’t the party in power use national security authorities against them? As that temptation grows, institutional reforms are needed to keep officials from yielding to it and, just as important, to show skeptics that the reforms actually worked.
The Obama administration clearly flunked the second requirement. They quite possibly flunked the first one too. Here are the most salient facts in support of that view—a much more detailed accounting of which is available, complete with footnotes, in my forthcoming testimony to the Privacy and Civil Liberties Oversight Board.
The DNC and the Steele “Dossier”
A major part of the Crossfire Hurricane investigation and the public disclosures it produced was the “dossier” created by Christopher Steele. We all now know that it was a salacious and unverifiable hit job assembled not by a network of intelligence sources but by a mix of Steele’s friends, their drinking buddies, and probably a few disinformation specialists from GRU (Russia’s military intelligence agency). Worse, Steele assembled that hit piece as a subcontractor to the Democratic National Committee, and judging by his conduct, he thought his role was to lobby the FBI to use its formidable national security powers against the Republican campaign—and to leak both the investigation and the now “FBI-validated” dossier in hopes of ruining Trump’s candidacy.
There are reasons to suspect that, despite its denials, the DNC intended that outcome: It hid its ties to Steele behind multiple cutouts and a dubious claim of attorney-client privilege, then falsely denied its connection to Steele for months after the story broke. In the end, Steele’s work didn’t pay off for Democrats until after the election. But during the transition it stoked the Russia collusion narrative that put a cloud of illegitimacy over the first two years of the Trump administration. That is a remarkable, if unseemly, achievement for a partisan hit job. Other political actors will learn the lesson and can be expected to use cutouts in the future to lobby the national security agencies against their domestic enemies.
Partisan Bias and the Carter Page FISA Application
The one really detailed examination of how the Crossfire Hurricane investigators treated the evidence against the Trump campaign is the inspector general’s dissection of the Carter Page wiretap application. That story does not exactly rebut the suspicion that partisanship tainted the probe. The application was full of errors and omissions, and all of them cut against Page and the Trump administration. Almost no one in the Justice Department or FBI stopped to ask if it was wise to pursue a surveillance order against a prominent member of the opposing party without taking a hard look at the evidence. As a result, the investigators left out—or even lied about—a raft of information that would have raised doubts about whether Page was a legitimate surveillance target.
For a while, it was possible to put these errors down to a different cause—not partisanship but a complete collapse in the Foreign Intelligence Surveillance Act (FISA) fact-gathering process. That comforting line of thinking rested on two findings by Inspector General Michael Horowitz—first that he found no evidence of bias and second that he found pervasive errors in 29 unrelated FISA applications. On closer examination, neither of those findings offers much support to the “FISA is broken” hypothesis.
First, on partisan motivation in Crossfire Hurricane, what the inspector general actually found was that no one at the FBI was foolish enough to say in writing or in testimony that they or others at the FBI were operating with a partisan bias. As the inspector general acknowledged in his Senate testimony, the absence of bias evidence didn’t prove an absence of bias. In fact, the inspector general did find written evidence of bias—in the texts of Peter Strzok, which are full of animus toward Trump. Strzok had great influence over the Crossfire Hurricane investigation, but the inspector general decided that Strzok’s bias didn’t count because Strzok never acted completely alone in the investigation. Really, that’s it. If I’m ever accused of a crime, I want Michael Horowitz on my jury.
Second, the errors he found in 29 other FISA applications evaporated on a closer look. They were, it turns out, almost all failures to properly footnote the FBI’s sources. When the FISA court ordered a review of all 29, the Justice Department found only two material errors, and neither of them cast doubt on the issuance of the wiretap order. That contrasts starkly with the Carter Page application, where the department has admitted that the errors were so serious that at least two and perhaps all four FISA orders should never have been issued.
In short, the only FISA application that targeted a partisan opponent of the administration was corrupted by numerous material omissions and errors and at least one false statement, one of the most influential investigators was a voluble Trump hater, and others may have harbored a bias against Trump that they were too prudent to articulate. Since the FISA process in general now seems to be careful and accurate, if not perfect, the deviation from norm in the case of Carter Page strongly supports the view that anti-Trump bias was at work.
A Conveyor Belt from Press Reports to Surveillance
Actually, there’s more. The inspector general passed over in silence the remarkable reliance of the Page application on media reporting. Fully a third of the core FISA case against Page consists of summaries of news stories. By itself, relying on media reports was a likely source of bias against anyone associated with Trump. (If you want to argue about that, all I can say is that I want you on my jury too.) But we don’t have to argue about media bias in the abstract. It can be found in the Page application itself, which relies on a Washington Post opinion piece, without disclosing to the court either the source or the fact that it isn’t, strictly speaking, a news report at all. Almost as bad, the opinion piece claims that the Trump campaign diluted the GOP platform on Ukraine in ways that favored Russia. (In fact, the campaign accepted a mildly diluted version of an amendment offered by a Ted Cruz delegate, which is a lot more accommodation than delegates for defeated candidates usually get at conventions.) The claim has been investigated extensively, including by Robert Mueller and the Senate Intelligence Committee, without finding any wrongdoing. The nicest thing you can say about the article in retrospect is that it was slanted to take the worst view of the Trump operation. An equally fair summary would be that the story became part of an FBI conveyor belt for turning media bias into a wiretap order. If that doesn’t worry you, imagine today’s Justice Department obtaining a FISA order against Biden campaign advisers by relying on an article from Breitbart, and simply telling the court, as the Page application does, that the information comes from “an identified news source.”
Targeting Michael Flynn
That’s not the worst of it. Viewed from the standpoint of partisan abuse, the Michael Flynn story is especially troubling. He had been investigated and cleared by the FBI on Jan. 3, 2017. But two days later, on Jan. 5, the White House obtained a wiretap of Flynn talking to Russian Ambassador Sergey Kislyak about Russia’s response to the Obama administration’s recent sanctions. The wiretap of Flynn’s remarks was legal, because the “target” of the tap was Kislyak not Flynn. But the legality of the collection does not fully resolve what you might call an analytical reverse-targeting after the fact. That’s because the White House was only really interested in Flynn’s side of the call.
After an Oval Office meeting about Flynn’s remarks, Obama administration officials began a concerted campaign to use those remarks against him. Within three weeks, he’d face leaks accusing him of violating the criminal Logan Act, he’d be reinvestigated under an implausible counterintelligence theory, and he’d find himself ambushed by the FBI in a perjury-trap interview. He’d also become the first American to have a FISA-tapped conversation leaked to the press by political rivals. Within four weeks, he’d be gone from government, disgraced and facing criminal prosecution.
By any measure, this was a political use of a FISA wiretap that targeted an American. It may have been a lawful political use of a FISA tap, but that’s not something people should be comfortable with. The Obama administration, however, had gotten comfortable with it a few years earlier. When Israel was fighting Obama’s Iran nuclear deal in Washington, it worked closely with Hill Republicans. The U.S. apparently tapped the Israelis, again legally, since they were foreign government officials. And the taps may have offered some national security insights; any time a government, however friendly, lobbies Congress against the American president, we ought to know what it’s up to. But the foreign intelligence value of understanding what the Israelis were saying paled next to the political value of getting real-time intelligence on the GOP’s Hill strategy for stopping the Iran deal. The unfortunate lesson the Obama administration learned in that battle was that the president can use FISA taps against his political enemies as long as he checks the right legal boxes. If it worked against the congressional Republicans, why wouldn’t it work against Team Trump?
But turning FISA into just another partisan weapon means it’s going to be used like one. If it hurts the other side, it’s going to be leaked. Which is what happened with Flynn’s conversation. The leak was unprecedented in national security circles, but in Washington politics, it was just another Thursday. More than 40 years had elapsed before the first FISA tap of an American was leaked to the press. I doubt it’ll be that long before the second.
The Need for Reform
To be clear, apart from the Flynn leak, none of this was plainly illegal, and no one should want the government to ignore indications that a prominent political figure is working with a hostile government. But the Republicans who were on the receiving end of these intelligence operations have every reason to doubt the good faith of the administration that carried them out. And that in itself will prove fatal to the bipartisan support the intelligence community needs as it responds to foreign influence operations. What’s needed are reforms that will prevent future administrations from using the intelligence community against the opposition in this way.
Unfortunately, most of the reform proposals are warmed-up leftovers beloved of individual rights advocates—more paperwork and audits and amici curiae for all FISA applications, not just the ones that pose partisan risk. Others could make things worse, such as the measures to require that the attorney general be briefed on FISA taps with partisan risk. Is there anyone on the GOP side who would be relieved to hear that the Flynn matter was overseen by Sally Yates, who chose partisanship over Justice Department tradition in refusing to defend the new administration’s immigration policy in court? Is there anyone on Team Biden who’ll be comforted to hear that William Barr will decide whether to investigate the former vice president for ties to Ukraine or China? It’s fine for the case to get high-level review; top officials often have better instincts than those in the ranks. But it’s not enough. We need to create a career position for a nonpartisan FBI agent or lawyer to challenge the FISA application and every other stage of the investigation. (The attorney general’s supplemental reforms memo of Aug. 31, 2020, takes a good step in this direction by requiring that politically sensitive surveillance and search applications be reviewed by a special agent from a field office not involved in the investigation.) The career official should also take the lead in reporting on the investigation to majority and minority congressional leadership, not after the fact but as it proceeds.
And when an operation has both political and national security value, the intelligence it produces needs special and far more limited handling, especially when it goes to political appointees. Every one of them should be required to sign a receipt explaining why he or she needs to read it, and the intelligence community should routinely include tags on some reports that will disclose which one was leaked.
Other measures are simple. The FBI should offer media reports to the FISA court only rarely, and it should disclose their source and any credible claims of bias that have been leveled against the news outlet. Anyone who pays a third party—directly or indirectly—to try to influence the FBI or other national security agency should disclose that fact, just as lobbyists trying to influence Congress or political appointees must.
There’s plenty of room to argue about which safeguards will best limit the partisan misuse of the United States’s security machinery. I hope that this piece—and my longer testimony to the Privacy and Civil Liberties Oversight Board—are at least sufficient to establish that, without new safeguards, the United States will slowly lose its ability to respond as it must to foreign influence operations.