Let’s leave aside for now the obvious partisan nature of the Nunes memorandum—the six-page document released Friday by the chair of the House Intelligence Committee, California Rep. Devin Nunes. Let’s leave aside the fact that everyone who has read the underlying classified FISA application for a warrant to surveil former Trump campaign adviser Carter Page says the memo is incomplete and out of context. Let’s leave aside as well the equally obvious partisan nature of the Republican response, as House members and the president try to delegitimize special counsel Robert Mueller’s investigation and demonize the deputy attorney general, Rod Rosenstein. All of that, by itself, is good and sufficient reason to think that the Nunes memo is not a serious effort at oversight.

But let’s put that aside. Instead, let’s take the Nunes memorandum on its merits and assume that it is what it purports to be—an accurate summary of a purported problem with the FISA application process. What then should we make of it?

Under the most fair reading of the memorandum, the argument it makes is as follows: The Steele dossier—a collection of reports filed by former British intelligence officer Christopher Steele—is a biased and flawed document produced by someone out to “get” the president; FBI and Justice Department officials knew of the bias and did not disclose it to the judges of the FISA court who approved the FISA warrant; as a result, the rights of an American citizen (Carter Page) were violated; and (more importantly, from the perspective of the Republican Party) the FBI’s reliance on the Steele dossier corrupted the FBI investigation of the Trump campaign—all of which is, in effect, the “fruit of a poisonous tree” that should no longer be credited.

Given that story line, one can only conclude that the Nunes memo fails to make its case—and fails quite badly at that.

Consider, first, the obvious timing problem. The Nunes memo begins with a FISA application dated October 21, 2016. That date is significant for a number of reasons. As an initial matter, coming less than 20 days before the election, it seems a particularly poor way of trying to influence the outcome of the election. A FISA application just a few days before November 9 would not actually have produced any evidence until well after the election—making Nunes’ implicit charge of a corrupted investigation chronologically implausible. In addition, the focus on this date has to deal with the uncomfortable fact that the surveillance of Page it authorized started roughly a month after Page officially left the Trump campaign—so, again, it is a poor way of effectuating a bias against Trump to collect evidence relating to the actions of a former campaign volunteer.

But the most important reason to focus on the 2016 date is that it ignores another, earlier date. We know from public reports that the FBI opened its inquiry into Page’s Russian connections as early as 2013, at which time the bureau already had probable cause to think Russian intelligence operatives might be trying to recruit him. (The Russian spies, by the way, thought that Page was “an idiot,” according to court documents.) Any story of the investigation of Page that starts in the middle is incomplete at best—and since we don’t know what was in the earlier Page application or in the rest of this October application, we can’t really know how incomplete it is. But that careful incompleteness is, by itself, grounds to doubt the memo’s conclusions.

The other timing problem arises from the effort to tie this allegedly flawed FISA application to Deputy Attorney General Rosenstein (for whom the president apparently has great disdain). Of course, before the election Rosenstein wasn’t the deputy attorney general—he was a U.S. attorney for the district of Maryland. To link him to the earlier “Steele-based” October application, the Nunes memo has to tie that original application to the application for a renewal of the FISA surveillance order that Rosenstein authorized in 2017, after he was appointed by President Trump and confirmed by the Senate.

But that timeline weakens the Nunes case for bias rather than strengthening it. Between the original application and the Rosenstein renewal, the Page surveillance was renewed two other times, for a total of four approvals altogether (and public reports say that there were four separate judges who did the reviewing—suggesting that four independent reviews validated the FBI’s investigation). But these renewals mean that it is utterly implausible (if not borderline impossible) for the renewal that Rosenstein approved to be reliant on the Steele memorandum.

Here’s why: When a FISA order is obtained to conduct surveillance on an American, the FBI must get a reauthorization from the FISA court every 90 days. In seeking renewal they cannot simply recycle the original application—they must demonstrate that the surveillance has been fruitful. In other words, they need to show the judge that the surveillance has developed foreign intelligence that reaffirms the original probable cause determination and shows that their suspicions had merit and the target is acting on behalf of a foreign power. If the FBI cannot show new evidence like this, the surveillance is likely to be terminated. In other words, the fact that the FISA order was renewed means that the original “poison” of the Steele memorandum did not taint the subsequent renewals—it means that there actually is a “there there”—at least in the eyes of the renewing judges.

Next, the Nunes memo reduces its credibility by including language that is intended to create a misimpression. For example, in order to buttress the idea that the Steele dossier was central to the original application (and thus a critical tainting factor) the memo says that former FBI Deputy Director Andrew McCabe had testified that “no surveillance warrant would have been sought” without Steele.

This is potentially explosive. It is meant to leave the impression that Steele was the central, critical basis for the probable cause submission to the FISA court. But that isn’t what the memo says. What it actually says—a much better reading (one likely to be borne out when the transcripts are eventually published)—is that McCabe acknowledged that the Steele dossier was part of the impetus for seeking the warrant—which is not the same thing as saying it was the probable cause basis for obtaining it.

Perhaps even more significant is how the memo tries to bury the admission deep in the document that the entire FBI counterintelligence investigation of Russian influence was not triggered by the Steele dossier. Instead, as the memo admits only in its final paragraph, it was information about another Trump campaign adviser, George Papadopoulos, and his meetings that began the inquiry.

Given that the investigation began from a different source, it is almost certain that the motivations behind the Steele dossier were irrelevant to the FISA court—judges routinely grant warrants based upon information provided by sources who have an ax to grind. The question is whether or not the information is corroborated. (As one wag put it, it only matters that someone told the FBI Page was talking to the Russians—it could have been Stormy Daniels for all they cared, so long as they could corroborate the information.) Here, we will not know unless and until the full FISA application is unsealed, but it is notable that the Nunes memo nowhere says that the Steele dossier was the exclusive basis for the FISA application and acknowledges the existence of other information. So, there are reasons to think that Steele’s alleged bias played no significant role in the FISA process. All of which doesn’t mean that everything in the Steele dossier is true. But his alleged bias doesn’t necessarily ruin the credibility of his research.

Much of this is speculation of course—and we may never know the full truth since the complete FISA application from October 2016 is likely never to see the light of day. But even on the limited data we have in the Nunes memo—and even disregarding any concern that the memo was motivated by partisan purposes—it simply doesn’t prove misconduct, much less game-changing misconduct by the FBI or DOJ.

Image credit: KongNoi