Perhaps the most surprising part of Chris Matthews’s departure from MSNBC was people referring to him as a journalist.
Chris Matthews is not and never has been a journalist. Given this, I think it’s valuable to consider what Matthews and our other culturally prominent political talking (and writing) heads are.
I believe that we can categorize these folks into the following groups: analysts, pundits, hacks, barkers and propagandists.
Limbaugh, Hannity, Ingraham, et al., are propagandists, people willing to suspend reason and evidence in the service of propping up their favored politician or cause. Propagandists will often put an analytical gloss on their opinions, but being unbound by facts makes them primarily a vector for misinformation.
Matthews was a barker. Barkers are like their carnival namesake, whose job is to point at and pump up the spectacle in order inflame the emotional temperature of the audience. The ne plus ultra of barkers now that Matthews is gone is CNN’s Chris Cuomo. Cuomo prefers guests like Michael Avenatti, Anthony Scaramucci and Kellyanne Conway because they are fun to “spar” with. The fight is the point. A propagandist leaves you misinformed. A barker leaves you uninformed because information is irrelevant as compared to the spectacle itself.
Pundits, hacks and analysts have some surface similarities in that they all claim to be marshaling evidence in the service of argument and opinion, but there are some important distinctions.
The Meet the Press/Morning Joe roundtables are pure punditry, even when those panels include people who on occasion practice journalism. Pundits tend to start with their a priori conclusions and then go looking for evidence to support their claims. They are big on predictions.
Despite “What’s going to happen?” not being a question for journalists and journalism, huge swaths of our institutions ostensibly dedicated to journalism spend lots of time on this question.
A hallmark of the pundit is being frequently wrong. This is because the pundit's proclamation is not meant to be predictive, but instead to shape opinion with writing that carries the sheen of analysis. The New York Times's stable of opinion writers is lousy with pundits -- David Brooks, Maureen Dowd, Tom Friedman -- writers who use their exalted position to make consistently incorrect pronouncements when attempting to explain the world.
Again, this is because they are not trying to explain the world, but to alter the world into something closer to their own preferences. Thomas Friedman’s work prior to and during the Iraq War is a good example. Friedman’s recent column theorizing that Democrats could insure a landslide victory over Donald Trump if they form some sort of political Voltron, which would not only include every 2020 Democratic candidate in a cabinet position, but also Mitt Romney as commerce secretary, is classic punditry.
Never mind the dubious political calculus, in what world does Mitt Romney step down as a United States senator to become a relatively lowly commerce secretary for a Democratic president? As argument, Friedman is spouting nonsense. His goal, however, is merely to move sentiment and promote an undefined (and undefinable) notion of “unity” as a higher good. He will bemoan our polarization even as he promotes a remedy that is past unlikely all the way to impossible.
A hack is mostly a lousy pundit or a pundit gone sour with time. The speed with which they are proven wrong or the depth of their wrongness is primarily what distinguishes the hack from the pundit. Jonathan Chait of New York magazine is a hack, as shown in his classic prediction of Donald Trump having no chance of winning Michigan the day before Donald Trump won Michigan and the presidency.
Bret Stephens of the Times is also a hack. He was so eager to argue that the Democrats misplayed the 2018 midterms he wrote his column “warning” of the lackluster results before the results were in. A 27-seat margin at the time of filing swelled to a 41-seat landslide. The Times kept updating the number without changing the analysis. Only a true hack would stand by that work.
On the other hand, Jamelle Bouie of The New York Times is an analyst. A recent column, “Where Trumpism Might Take Us?” is analysis via historical analogy attempting to illuminate the political forces that animate Trumpism and where and when we’ve seen them before. It is reasoned speculation without falling into facile prediction.
When it comes to television, my favorite political analyst is Anand Giridharadas, author of Winners Take All: The Elite Charade of Changing the World and frequent contributor to MSNBC, including the occasional stint on Morning Joe, where he acts as the turd in the punchbowl insisting on marshaling evidence in the face of all that punditry.
A clip where Giridharadas engages with MSNBC host Stephanie Ruhle over whether or not taxing the “ultrawealthy” punishes success shows the benefits of being able to engage with an analyst if the goal is to better inform the audience. For over 10 minutes, an eternity on cable television, they engage in a back-and-forth argument that touches on various aspects of our capitalist system and political discourse. You can’t help but be better informed at the end of the exchange.
Please note that analysts are no less opinionated than pundits (or hacks), but the way one engages in argument with an analyst is fundamentally different. Giridharadas is trying his best to be as convincing as possible, but even if you do not agree with him, the broader issue has been illuminated through the process of argument.
A pundit seeks agreement, not argument. Consider the difference between David Brooks (pundit) recently declaring in a column that Bernie Sanders is “the end of liberalism,” and the analysis of the column by Lawrence Glickman, a historian at Cornell who reveals the ahistorical and internally inconsistent claims of Brooks’s piece.
Because his goal is to illuminate, Glickman shows his work. Because Brooks is trying to shape public opinion, he will shirk that responsibility while mangling historical understandings of “liberalism” in the process.
The question of whether or not Sanders is electable is now consuming Democratic political discourse with the punditry on the question rocketing across the airwaves and internet. Of course this is a question which can be analyzed, as Rachel Bitecofer, a professor of political science and up-and-coming election forecaster who nailed the 2018 midterms does in her piece at The New Republic on the impact of “negative partisanship” on voting choices.
Why does CNN rely on Chris Cillizza (barker/hack hybrid) to opine on electability and not people like Bitecofer?
In the end, it’s because the vast majority of our political discourse has devolved into spectacle, and the duration of Chris Matthews’s career demonstrates that this has not happened overnight. Donald Trump’s particular talent is his eagerness to embrace spectacle as the MO for his administration, the first president to do so.
Trump’s disinterest in governing makes him perfectly suited to the atmosphere barkers like Matthews created. There is an endless stream of nonsubstantive, yet impossible to ignore stimulus coming out of this White House, often as cover for actions that should be more thoroughly understood, like the recent non-peace accord peace accord with the Taliban or whatever the hell is going on with the coronavirus.
As a teacher of writing, I try to urge students towards the mind-set and behaviors of analysts, but in a word where the pundits, hacks and barkers are rewarded so handsomely, it is tough to convince them of the value of being careful with your work.
I bear no personal ill will toward Chris Mathews, or any other barker, pundit or hack, for that matter.
Don’t hate the player, hate the game, as they say.
Man, do I hate this game.
 Propagandists on the left exist, but don’t have such prominent positions because there is no liberal equivalent to Fox News. You’ll see them on Twitter, though, and I’m not going to name any because I am afraid of them.
 I believe this to be true of so-called “data journalism” as embodied by the 538 website. Recent research published in the Journal of Politics suggests that 538’s polling horse-race coverage may decrease voter participation.
 A recent report from Ashley Feinberg at The Huffington Post about leaked comments from New York Times opinion page editor James Bennet to his staff suggests where some of the problem may lie. Bennet described columnists as “people who are paid to have very, very strong convictions, and to believe that they’re right.” Notice how Bennet doesn’t say “be right,” but “believe they’re right.”
 Kathleen Parker and Peggy Noonan are interesting examples who arguably started as analysts but quickly morphed into pundits and now have entered their hack years.
Ross Douthat of the Times is also an analyst.
It is worth noting that the pundit/hack class is exclusively reserved for white people, with the vast preponderance of them men. To speak authoritatively without bothering with underlying evidence is much easier for groups that carry the pre-existing cultural imprimatur of authority.
Previous administrations found managing the press a hindrance to their work of governing. Trump’s conducting of the daily press narrative is his work.