Is Americans’ declining trust in media driven by political messaging?

1 hour ago 4

Why has trust in American news media dropped so much in the past 50 years — much more than it’s declined in other rich western countries?

To some people, I’m sure, the answer will seem obvious: Journalism’s just gotten worse. The news used to be produced by upstanding proud Americans who only cared about the truth, and now it’s the product of a cabal of latte-sipping leftist elites willing to lie about anything in order to promote their dark, anti-Christian agenda and destroy all that is good in the world. Let’s call this the Fox News hypothesis.

That argument, as common as it is, leaves something to be desired. People who lose trust in “the media” aren’t doing so solely in response to their own observations and experiences. They live in a social environment filled with pre-baked messaging on who’s got your back and who’s out to get you, and “the media” makes for a pretty convenient target. How much of the decline in media trust is actually about the media’s actions, versus the narratives that get spun about them?

That eternal question was one reason I was excited to come across a brand new dissertation written by a journalist-turned-academic named David Beavers. Beavers spent five years working as an editor and producer at Politico before moving to a PhD program at Harvard. Six years later, he’s a freshly minted scholar headed to Wake Forest, where he’s been hired as an assistant professor in the Department of Politics & International Affairs.

Beavers’ dissertation is titled “Undermining the Fourth Estate: Elite Criticism and Journalism’s Fragile Foundation,” and it aims to untangle the question of media distrust causality through a creative mix of journalism history, natural language processing, and field experiments. (It just won Harvard’s prestigious Charles Sumner Prize.) Beavers advances two major data-driven claims. First, the American system of journalism has historically relied more on rhetorical appeals than structural forces to proclaim and defend its legitimacy. The First Amendment lets anyone be a journalist, which is great — but its lack of professional gatekeeping has also left the field uniquely vulnerable to attack from those who seek to weaken it. And second, the decline in public trust in the media was both preceded by and driven by attacks on the press by political elites, especially at two key pivot points — white southern Democrats defending Jim Crow in the 1960s and 1970s and the Lee Atwater/Newt Gingrich generation of Republicans in the early 1990s. As a result, criticism of the press — which used to be relatively equal on both ends of the political spectrum — has become a decidedly conservative behavior.

Here’s the full abstract:

American democracy is predicated on having a public that is sufficiently informed to hold elected officials accountable and to make sensible decisions at the voting booth. But a decades-long decline in Americans’ trust in the press threatens to undermine the fundamental role the media plays in cultivating an informed populace. This dissertation asks: What explains Americans’ declining trust in the press? I address this question in three parts. In Part I, I examine the historical development of journalism as a profession to offer a theory for why American journalism’s institutional forms make it uniquely vulnerable to political efforts to undermine its legitimacy. In Part II, I quantify congressional rhetoric about the press between 1939 and 2023 and relate it to long-standing survey evidence to show that elite press criticism has driven down Americans’ trust in the media over the last half-century. Finally, in Part III, I turn from causes of Americans’ declining trust in the media to consequences and possible solutions. I examine the effects of media trust and elite press criticism on news consumption using web-browsing and newspaper circulation data, and I conduct a trio of original survey experiments to assess interventions aimed at restoring trust in reputable journalism. I conclude with reflections on the state of American media and American democracy — and how those two are fundamentally intertwined.

I gave Beavers a call to talk through his findings; here’s a lightly edited version of our conversation.

Joshua Benton: Let’s start with your backstory. What leads someone working at Politico to think, “Hey, I know what I want to do, I want to go get a PhD”?

David Beavers: My undergrad is actually in civil engineering, so I’ve lived a few lives. Working at Politico, two things that pointed me in that direction. One was I kept wanting to answer causal questions. I wanted to not just be able to say, “Okay, A and B are occurring at the same time,” but “Is A really causing B?” And I felt like academia was the better route to be able to do that — the type of tools that social scientists have as opposed to journalists.

And the second was my journalism career, brief as it was, pretty much perfectly coincided with the first Trump term. The word “unprecedented” was thrown around pretty much daily at that time. And I kept finding myself wanting to be able to dig more into the history, to actually understand what was genuinely unprecedented versus what maybe only felt unprecedented in the context of a 24-hour news cycle.

Benton: Let’s start out with the first part of your dissertation, where you argue that the American journalism system is “uniquely vulnerable to political efforts to undermine its legitimacy.” Why’s that?

Beavers: I look back on history, particularly at the Progressive Era and the early New Deal — so the very end of the 1800s and the first few decades of the 1900s — because I think this was the time where a lot of the form of modern journalism really crystallized. This is when like the norm of objectivity was really articulated. It’s when journalism trade publications started circulating, when journalism schools were first founded. I was interested in looking back at the actual historical process that went into developing those forms.

Particularly in the early New Deal, some conservative elements within journalism — especially newspaper publishers and the American Newspaper Publishers Association — really used First Amendment press freedom as a bit of a political cudgel to try to keep the federal government away from doing any regulations of the press. Business regulations, just as an employer — things like maximum permissible work hours or minimum permissible wages. They basically said, “You can’t do that, because that’s akin to government censorship.”

And I think what that had the effect of doing was leaving journalism with fewer of what sociologists call boundary-setting mechanisms. Journalists in the U.S. tend to participate less in mass membership associations. Journalism obviously doesn’t have degree requirements — I’m proof of that. I worked as a journalist for five years, and I have no journalism degree. There’s no licensing, there’s no testing.

There certainly are good reasons for some of that, but that does make journalism in the United States an outlier compared to other professions. In the U.S., think of medicine or law — like, the American Bar Association practically has the authority of a government unto itself in regulating the profession. And, to a somewhat lesser degree, journalism in the U.S. is a little bit different than journalism in other countries. I think of the Italian case, for instance, where the order of journalists, the Ordine dei Giornalisti, basically controls access to the profession in the same way that American lawyers and American accountants do.

Without those type of things, I think American journalism rests heavily on rhetorical means of policing its borders — talking about things like its role as the fourth estate and its importance in undergirding democracy. All of which is true! But when your societal legitimacy is really wrapped up in that type of rhetoric, I think it makes American journalism vulnerable to rhetorical counterclaims, which is really what we’ve seen over the past 50-plus years, with predominantly conservative politicians criticizing the media in what I argue is an attempt to fundamentally undermine its legitimacy.

Benton: Over the years, I’ve talked with quite a few young Italian journalists who are so frustrated by what they see as the order making it very, very difficult for a young person to enter the profession. They see it as a sort of incumbent protection — the older journalists who are already in the system at a time when jobs are disappearing. They say they’ve done such good boundary setting that no one can get a job.

Beavers: Yeah, and you can see parallels in other professions in the U.S. too. We arguably have a shortage of doctors in large part because the American Medical Association wants to keep the number of doctors a little bit down, so they can keep wages up. There is that economic advantage-seeking that every profession is going to do.

It’s not that I would advocate for, like, the U.S. adopting the exact type of system the Italians have. There are certainly trade-offs there. But yeah, I do think the degree to which American journalism really rests on kind of a rhetorical foundation is pretty unique.

Benton: You get at this by looking at Hallin and Mancini’s typology of different countries’ journalism systems. Could you explain what separates liberal systems versus polarized pluralist systems versus democratic corporatist systems?

Beavers: Hallin and Mancini created these groupings that look at different countries’ media systems.

You have the polarized pluralist countries, which is for example Southern Europe — the Italians would fall under them. These are places where journalism tends to be somewhat of an elite profession that speaks to other elites. So newspaper circulation has historically been lower, and journalists often come from the same type of elite schools that also feed, like, the bureaucracy. So it’s a little bit less of a “mass” media, less viewed as a sort of public service in some way.

The countries of Northern Europe and Germany make up a different group, which is democratic corporatists. Newspaper circulation there tends to be very, very high. You have a very rigorously professionalized journalism, embodied in things like Sweden’s uber-powerful press council.

And then there’s the liberal model — liberal not in an ideological way. That would be the U.S. and the rest of the Anglo world, like the U.K. and Canada. Here, you combine very strong public broadcasting — elsewhere, not in the U.S. so much — with often a two-tiered press, the elite newspapers like The New York Times and then a big tabloid press. It’s much more capitalist. It’s professionalized as well, but with sort of a market-driven business structure to it.

When you look at those different groups, media trust has definitely declined by the most in the U.S. and in the liberal systems. Trust has declined much more than in the countries of Northern Europe and really Europe in general. The United States, I find, has experienced the second sharpest decline in trust in the press from 1981 to roughly today — second only to Hungary, which is really not the company you want to keep.

Benton: So is there a sort of alternate future that was an option 100, 150 years ago where American journalism would have ended up with a more systematized, more institutionalized structure?

Beavers: In the very early 1900s, there was a push by some within journalism to actually turn to the state as a kind of benevolent legitimizing force. This was very much steeped in the logic of the time of the Progressive Era, which saw the government that way.

There were efforts in some states — Pennsylvania, Illinois, Connecticut — to request that state government basically issue licenses for journalists. These were all pretty short-lived and all unsuccessful. There was a bill, for instance, in Pennsylvania around 1913, but it never got a vote and died in committee. Then things pretty quickly shifted towards professionalizing within the walls of journalism, centered especially in state press associations and state journalism schools. In Illinois, the head of the University of Illinois’ department and then later school of journalism, Lawrence Murphy, had this multi-year effort to issue professional certifications for practicing and aspiring journalists, based on a combination of a journalism degree, testing, and actual professional work for a newspaper.

That also died out after a handful of years. You had, I think, the rise of European totalitarianism and press censorship in Germany and Italy and elsewhere, as well as the recollection of wartime propaganda in World War I, that made journalists a little bit less interested in anything that even resembled state control over the profession. I do a little case study on the American Newspaper Guild, which was founded in 1933, largely as an attempt to gain a seat at the table with newspaper publishers’ negotiations with FDR’s National Recovery Administration, in an effort to boost pay and reduce the employment precarity plaguing journalists during the Depression. Within the guild itself, there was really a split about how much to make it a professionally minded association versus a more narrowly minded labor organization. And what I find is that some of the publishers’ rhetoric did ultimately push it towards the latter, towards being a more narrowly focused kind of trade union.

Benton: One thing that struck me reading your work was that, while you make the case that there wasn’t as much professionalization in American journalism compared to in some other countries, you did see higher levels of institutionalization in some senses. U.S. cities were significantly more likely to develop really strong, really profitable monopoly newspapers than you saw in Europe. The one big daily in a city could be extremely powerful in that city’s politics or civic life, and it could attain a huge amount of institutional force. It wasn’t a situation where “you can’t mess with the local journalists guild,” but “you can’t mess with the Kansas City Star, or the Cincinnati Enquirer, or the Baltimore Sun,” or whatever the big powerful daily was in your town. The institutional power sort of accumulated at a different level of the system.

Beavers: Yeah, that makes me think of the political economy of media kind of literature, and Bob McChesney’s work. I know he’s argued that a lot of the professionalization was sort of a smokescreen for publishers to kind of legitimize their own, like you were saying, monopoly holdings. The era of one-newspaper towns extends pretty far back in time. And these publishers were interested in trying to legitimize, in the public’s eyes, how they could have such a strong control over their readers’ information diets, and professionalism is one way to kind of do that. I definitely think there’s some truth to that. These newspapers had a lot of power of public perceptions of politicians, so it’s easy to see why politicians might be motivated, in that sort of an environment, to delegitimize them.

Benton: So your big data project here is looking at the entire Congressional Record from 1939 to 2023 and analyzing how these politicians talk about the media and journalism. Why start at 1939 specifically?

Beavers: I mostly started there because I wanted to capture this sort of post-professionalization period. If you go back to, like, the 1870s, the relationship between elite rhetoric and journalism would be really quite different.

So I gathered about 10.7 million congressional floor speeches. That’s every congressional floor speech between 1939 and 2023. I broke them up into short text segments and identified text segments that mentioned the media, using a dictionary of media-related terms that I developed for this project. That resulted in about 559,000 media-related text segments over that 85-year period. I then used a natural language processing tool to label each of those text segments as either positive, neutral, or negative toward the media. So, effectively, +1, 0, or -1. I then use that to track how different groups of members of Congress talk about the press over time.

I find that in the very beginning of the 1970s, both Democrats and Republicans start to get more negative in how they talk about the press — up until about the mid-1990s, when they start to diverge. Then Democrats start speaking more positively about the press and Republicans start speaking ever more harshly about the press — to the point where today, Republicans in Congress are on average more negative about the press than they’ve ever been, at least since 1939. And Democrats today are approaching their high watermark of positivity about the press, a level that they were last at around the 1950s or early 1960s.

When you look at this sort of chart and compare it to polling on Americans’ trust in the press, the two almost look identical. Whereas if you look at Gallup polling, for instance, you’ll see that both Democrats and Republicans in the mass public came to trust the press less and less starting around 1972 or 1973, when they first asked this question, and then starting in the mid-1990s, they really diverge. And I find that this isn’t really a coincidence.

There are two big pivot points that I’d reflect on. The first is the late mid to late 1960s — call it the Civil Rights Movement era. That’s where you start to see a shift from everyone in Congress in trending from being more positive about the media to becoming more negative. And the tip of the rhetorical spear there is Southern Democrats, which is consistent with some historical work.

David Greenberg has done some really excellent work on this, locating the modern origins of the liberal media bias critique with white Southern Democrats — folks like George Wallace, who talked about a liberal northern elite media coming down to the Jim Crow South and kind of telling them how wrong their way of life was. So you do see white Southern conservative Democrats, I think, lead the first sort of shift towards media negativity. Then that gets followed quickly thereafter by Republicans, in large part because those Southern Democrats switch parties. Today, for instance, you don’t see a meaningful gap between Southern Democrats and non-Southern Democrats, largely because the conservative Southern Democrats have left the party and the ones who remain are much closer to the rest of the party.

The second big pivot point in the data is the mid-1990s — really the late 1980s to the mid-1990s. And this is very much centered around the Rush Limbaugh talk radio era and Newt Gingrich’s rise to House speaker. If you want to understand politics in the 21st century, I think you really can’t do it without understanding Newt Gingrich’s rise to power and especially the kind of conservative media ecosystem that helped to sort of fuel that. And this is where we start to see a strong divergence with Democrats speaking more positively about the media and Republicans speaking much more negatively about it.

I think this is the period at which press criticism, for Republicans, really became kind of a central organizing feature of their rhetoric. And certainly we know that Gingrich was very instrumental in centralizing control and influence over how the party spoke and campaigned.

Benton: There’s a new paper I saw out from Jesper Strömbäck and some others titled “Exploring possible explanations for the modest relationship between news media trust and use.” It’s trying to understand why lower levels of media trust don’t necessarily correlate with much lower levels of media consumption. There’s some effect there, but less than you might expect. From your research, how does low media trust change the media that people consume, both in terms of quality and in quantity?

Beavers: Yeah — there is, in existing work, a pretty modest relationship between media trust and media consumption — which is a little bit surprising. I mean, you’d certainly think theoretically that there would be a very strong linkage there.

Some of that is due to methodological challenges. We often ask people on a survey, “hey, do you trust the media? Yes or no?” There’s going to be some nuance there that isn’t getting captured. And when you ask people what media they actually consume, people often don’t remember their media diet accurately, or there’s a social desirability bias for people to report something different than reality.

What I’ve tried to do in my research is rely on some actual behavioral measures of news consumption. So in my dissertation, I draw on — it’s a pretty small sample, but 400 Americans’ web browsing data for four weeks. In that time they viewed about 2.8 million unique site visits. And what I do is ask these people, “hey, do you trust the media? Tell me how much or how little.” And then I actually look retrospectively at what did their actual media diet look like online in the month leading up to that study.

And I don’t really find that people who say they trust the press less consume much less news. It’s more that they consume, if I may be a little normative, worse news. They’re consuming more hyperpartisan stuff, and more sites that have occasionally or frequently been known to peddle outright false information or conspiracy theories. With the caveat that this is a small sample, this is much more true among Republicans who say they don’t trust the media than among Democrats who don’t trust the media.

Benton: Yeah. It’s a longtime complaint of mine that asking people how much they trust “the media” requires a shared definition of what “the media” is, and we don’t have that anymore. I mean, someone who watches Fox News all day would likely say, “I don’t trust the media” — but they clearly trust Fox News, right? Just asking about “the media” is asking people to think of it as a sort of nebulous force in the universe, because “the media” is so wildly varied. Versus if you’re asking about “the media” in the 1970s, people were more likely to have a common idea that you’re talking about your middle-of-the-road local daily newspaper, the friendly faces on your local TV newscast, and Walter Cronkite. Like, yeah, you should trust those more than you trust anything on the internet that could plausibly be called “media.”

Beavers: Exactly — it’s not apples to apples. I’ve asked people open-ended questions on surveys — after I ask the media trust question, I’ll say, “hey, what were you actually thinking about when you just answered that question about whether you trust the media? Were you thinking of specific organizations? Were you thinking of specific mediums? Like, what would you say the media is?”

And largely I do find that people name the sort of organizations that they would have had in the 1970s. They say the nightly news, they say The Washington Post, The New York Times, The Wall Street Journal. But then the huge caveat is that they really say Fox News, CNN, and MSNBC. And that’s a huge difference.

Benton: So in the final part of your dissertation, you evaluate the impact of several proposed or executed journalistic reforms, if we can call them that — various attempts to try and increase media trust. What did you find?

Beavers: Yeah, I can unfortunately answer that pretty quickly: They don’t really work.

I test basically three different types of interventions. The first was around increased journalistic transparency. The second one looked at the efficacy of correcting people’s misperceptions about the press. There’s a slightly naive logic in this one — “Oh, if only they knew more about the press, they would trust it more!” That one doesn’t really work either.

The third is with a co-author, Kevin DeLuca, who’s at Yale University, and we tested the efficacy of media literacy tip sheets, aimed at helping people differentiate between kind of high-quality and low-quality news sites. More specifically, looking at algorithmic news sites that involve little or no human intervention. And we look at whether that was differentially effective based on people’s pre-existing media trust, and again, we kind of find nothing.

Benton: I gotta say I roll my eyes a little every time I get another press release from some project aimed at a news org “laying out our journalistic principles” or “being transparent about our methods” and expect that trust will somehow follow.

Beavers: Yeah, you wish it could be that simple. I mean, there is, I think, an understandable desperation in American journalism, and you can understand how people are in a kind of throw-spaghetti-at-the-wall mode. But our findings were not optimistic about any of the interventions we tested.

Benton: So let’s imagine that, through some terrible governmental shift, for some reason you have just been appointed the king of all American journalism in America, and you are tasked, based on everything you’ve learned in the process of writing your dissertation, with increasing media trust. What would you do?

Beavers: Well, let me start by poo-pooing some of the work that I’ve actually already done. The type of work that I’ve been able to do in my dissertation is really limited by the type of methodologies that are most available in the social sciences — testing things via survey experiments. There’s a really apt analogy called the potato chip problem. We might think that potato chips are fattening — that if you eat a lot of potato chips, you’ll gain weight. But if you then run a survey experiment and ask a treatment group to eat one potato chip and a control group to eat none, you’re not going to find a difference. That doesn’t mean that potato chips don’t do anything. It just means that your method wasn’t really good at finding it. In my own work, I’ve really had to test the type of things that I can do in a five-minute online survey experiment, where I know that people are really only half paying attention to the task that I’m giving them anyway.

So what I really want to do, as an academic, as a scholar, is actually get out and work in partnership with news organizations in more of what we’d call a field experiment route. And the idea that I’ve been ruminating on, and this is going to be a little bit sort of theoretical or speculative, is to move more towards a relational level. We know that people aren’t forming their levels of trust in response to individual news articles or newspapers or journalists or whatever, based on, like, really cognitively effortful processing of the information that a newspaper publishes.

It’s more emotional. It’s more intuitive. It’s more, “Does this information make me feel good or not?” And it happens more at a level of identity. It’s “Well, I’m a Republican, and my president tells me that I shouldn’t trust it, so I don’t trust it,” right? I’m being a little bit glib, but when you really boil it down, that plays a big role. What I think is increasingly important is for there to be more relational connection between journalists and turned-off audiences. Because one of the big issues is that there have been such enormous staffing cuts in journalism in the U.S. over the last 25-plus years. Fewer people actually simply know a journalist. Fewer people are in a PTA meeting with a journalist. Fewer people are seeing a journalist at church, fewer people are in a beer softball league with a journalist.

And I think that if more people actually had a little bit of face-to-face exposure with a journalist — not necessarily to understand what journalists do, but actually just to see that there’s literally a human behind what’s being written — that could actually bolster trust. Maybe that feels maybe a little naive, maybe that feels a little bit idealistic, but I think we need to think more about solutions like that than just simply, you know, communicating information about journalism’s ethical procedures and things like that.

Benton: Yeah, I mean, they should learn that there’s a human being behind that story, at least for however much time we have left before it’s all AI-generated and there aren’t any human beings behind the curtain.

Last question. I searched your dissertation for the word “internet,” and it only appears five times — and most of those are when you’re describing your methodology. You’re looking all the way up to 2023, but you don’t really talk about the internet much — which is refreshing, as someone who talks about the internet in journalism all the time. But how intentional was that omission, to not make the internet as central a focus? Do you think of the internet as amplifying the larger trends that you’re seeing outside of it, or would you expect its impact to differ from what you’re describing?

Beavers: That’s funny — I would have guessed it appeared more often than that. But yeah, it was largely intentional. I’m somewhat of a paradox — I’m a political communication scholar who really doesn’t do very much work on social media. It’s certainly not to say that the internet hasn’t upended business models and norms and everything within journalism to a massive degree. It certainly has, and I would not claim otherwise.

I think for me though, what actually struck me was how consistent the effect of elite rhetoric was on public trust in the press. I kind of thought there might be a big jump in one direction or the other during the digital age — now politicians have many more direct avenues to actually communicate with constituents. And I really didn’t find that. So I think what my research shows is that there are broader forces at play that have, to some degree, even transcended the type of changes that the internet brought.

Obviously, for any kind of potential solution to journalism’s crisis, you’re gonna have to think about it being done in the context of a media environment where people are predominantly getting their information through the internet, particularly young people. And an internet that has fundamentally upended business models and allowed increasing media concentration, particularly in the hands of private equity and hedge funds who now have huge stakes in all the largest newspaper chains in the U.S. — I mean, you can’t ignore these features when thinking about kind of the implementation or the implementability of solutions.

Benton: I’d suspect — and I have no hard data on this — that one of the biggest impacts of the internet on what you’re researching is that the internet has radically reduced the cost of not trusting the media. Like, if it’s 1960 and you decide you don’t trust your local newspaper, you’re in a big bind. You probably don’t have another newspaper to switch to. And you’re using the newspaper to meet a ton of different needs. If you want to buy a used car, you need to check the classified ads. If you want to go see a movie, you need to see what time it’s playing. If you want to know if the local high school won the big game Friday night, you’ve got to check the sports section. So deciding you don’t trust it and won’t consume it meant making a global decision for a whole set of, like, 50 different information needs.

Whereas the internet makes it very easy for someone to, for example, not trust The New York Times on their political coverage but be okay with their reviews on Wirecutter. Or to get your sports news from a different place than you get your city hall news. Or to say “I’m going to not trust the media about Trump” and then to have this whole world of Fox News or The Daily Wire or Gateway Pundit to turn to. Saying you don’t want anything to do with the local daily used to have a big cost, and the internet’s removed almost all of that cost.

Beavers: Yeah. I mean, there was also social pressure decades ago. If everyone around you was consuming your local newspaper and you weren’t, you looked kind of out of it, right? I mean, you kind of had to consume it, even if you weren’t sure you actually trusted it. And the fragmentation of the media industry has very much been an internet-driven phenomenon.

Photo of an antiwar protest in San Francisco by Peter Lewis used under a Creative Commons license.

Read Entire Article