President Trump meets with Brazilian President Jair Bolsonaro in the Oval Office of the White House, Tuesday, March 19, 2019, in Washington. (Evan Vucci)

Donald Trump Jr. has been on a mission recently, writing articles for RealClearPolitics and The Hill arguing that conservatives are the target of censorship by technology companies. To make his case each time, he pointed to a barrage of articles from Breitbart.com, a site that’s both championed and benefited from the idea that conservative voices are being squelched on Twitter on Facebook. Breitbart itself hasn’t been squelched, mind you, given that it is consistently among the top publishers on Facebook in any given month.

Rep. Devin Nunes’s (R-Calif.) lawsuit against Twitter, unveiled on Monday, picked up similar themes.

“Well, how is it that every day there’s conservatives that are being banned?” he asked Fox News’ Sean Hannity in an interview Monday evening. “How is it possible,” he added later, “that I can be attacked relentlessly hundreds of times a day by fake accounts that they claim in their terms of service should not be there?"

Nunes claimed that his lawsuit was part of his ongoing effort to investigate anything possibly related to Russia save President Trump.

“This is part of the continuing Russia investigation,” he told Hannity. “We’re not going to just let all these fake news stories that were written about this investigation, about this hoax that were lies, we’re going to challenge every single one of them in court.”

President Trump was asked by a reporter from the conservative Daily Caller about the role of tech companies in spreading misinformation during a joint press conference on Tuesday. He didn’t directly answer the question.

“Well, we have to do something. I tell you, I have many, many millions of followers on Twitter and it’s different than it used to be,” Trump said. “Things are happening, names are taken off, people aren’t getting through. You’ve heard the same complaints. And it seems to be, if they’re conservative, if they’re Republicans, if they’re in a certain group, there’s discrimination and big discrimination. I see it absolutely on Twitter and Facebook, which I have also, and others I see. But I really focus more on the one platform.”

He went on to suggest that tech companies were colluding with one another and that it was “very, very fair to say that we have to do something about it.” As with Breitbart’s railing against conservative voices being quieted while it dominates Facebook, Trump went on to note that, even with this purported censorship, Republicans still won elections.

To bolster his case for rampant censorship, Trump Jr.'s essays point to various anecdotes, including that he himself had an image removed by Instagram (which is owned by Facebook). In an interview with Fox News’ Tucker Carlson, Trump Jr. said he heard from a number of people after complaining about that removal.

“Hundreds of people sending me [direct messages]. ‘Don, my account was shut down because I tried liking one of your tweets,’ or ‘I tried liking one of your father’s Instagram posts that was him and his grandchildren.’ Nothing even controversial,” Trump Jr. said. “And time and time again, ‘I had to follow you, Don, three times in the last week because they keep unfollowing you.’”

There is, however, no evidence of a systemic effort on the part of Facebook, Twitter or Google (which owns YouTube) to specifically shut down conservative users or content. Trump Jr. points to a report from James O’Keefe’s Project Veritas in which a former Facebook employee identifies a flag applied to content to limit their reach, a flag which O’Keefe argues was targeted to conservatives. But the documents shared show that the real focus of Facebook’s action was trolling or other negative behavior. (In a statement, Facebook indicated that the employee was fired in part for working with Project Veritas.)

We’ve seen this debate before. Last summer, a number of conservatives claimed they’d been “shadow banned” by Twitter -- a phrase that was used to describe their no longer appearing in Twitter search results. (Nunes claims this happened to him, too.) The most likely culprit, though, was that the affected individuals had triggered Twitter’s own anti-trolling measures, which had, at that point, only recently been implemented. It wasn’t only conservatives affected by the change in Twitter policy, but it was conservatives who seized upon the idea that they were being targeted.

Why were social media networks like Facebook and Twitter implementing systems to try and change the user experience over the past several years? Certainly in part because of the criticism they faced after the 2016 election, during which Twitter was a hotbed of fascist and racist behavior and Facebook was a key vector for the spread of misinformation.

This behavior may not be balanced evenly on the political spectrum. Conservatives, a study after the election found, were 30 times as likely to share content created by Russian trolls during the 2016 election. That despite the fact that much of the content being shared was targeted to those on the left. Another study found that the people most susceptible to sharing fake news on social media were older Americans -- and conservatives. (There’s obviously substantial overlap between those groups.)

When BuzzFeed spoke with a group of Macedonian teens who were making money hand-over-fist during the 2016 election, they were explicit: Content focused on appealing to Trump supporters did much, much better.

There was a similar effect when Facebook was revealed to have been using human moderators to shape its news feed. It came under fire for limiting conservative sites though the content being muted often overlapped with known fake-news purveyors. When human moderators were revealed, fake news quickly appeared on the now-removed news feed.

None of this is to suggest that Trump Jr. or Nunes were sharing racist content on their Twitter accounts (though Trump Jr. has certainly shared factually questionable stories in the past). But that he and other conservatives are looking for specific actions means that they may isolate examples as reinforcement of their points. Confirmation bias.

That’s a dangerous tendency in a space like social media, where small groups can quickly seem to be outsized. Nunes complaining about hundreds of accounts hassling him and insisting that they must be fake accounts simply because of the scale misses the point of social media. He’s a nationally visible figure engaged in sharply partisan politics. It’s not at all surprising that, especially during a campaign, he’d be targeted by a few hundred people a day who had unkind things to say to him.

Likewise with Trump Jr.'s anecdotes: Get dozens of people who’ve had posts deleted for unclear reasons or whose privileges were temporarily suspended (as Trump’s social-media director was on Tuesday) for some reason, layer on top random people not clicking the button they thought they had and you’ve got proof of a conspiracy.

No member of the Trump family has ever offered an explanation for the why. Why would Facebook, through Instagram, try and keep people from clicking “like” on a photo of Trump Jr.'s kids? What’s the benefit? There’s clearly a reason to crack down on abusive behavior, but what’s the benefit to the companies in cracking down on conservatives as conservatives? The only answer given is some version of “they (implicitly or explicitly) hate conservatives," a claim that’s often bolstered by pointing to the same perceived “censorship.”

Having Trump Jr. complain about getting an Instagram post removed is one thing. Having the president declare, based on this same sketchy evidence, that he must “do something” about the perceived censorship is something else again.

correction: This article has been updated to correct Facebook's description of its former employee.