Is Social Media Destroying Democracy—Or Giving It To Us Good And Hard?
It’s easier to blame the algorithm than the bewildered herd.
One of our era’s most influential narratives is that social media is destroying democracy and perhaps civilisation itself. For the liberal establishment, this story helps to explain the surging success of right-wing populism, as well as collapsing institutional trust, growing polarisation, and an apparent explosion of misinformation and deranged conspiracy theories.
The standard formulation of this narrative treats social media as a dysfunctional technology. Because algorithms and other platform features are designed to capture people’s attention and keep them scrolling, they amplify content that is sensationalist, bias-confirming, and divisive. This viral content then infects public opinion and political debate, driving large numbers of people to adopt misinformed and hateful ideas hostile to liberal democracy.
I’ve criticised this narrative. Although social media platforms undoubtedly reward low-quality discourse, narratives that place significant weight on this fact to explain recent political developments are misguided. They rest on implausibly rosy pictures of legacy media and pre-social media history. They’re not well-supported by scientific studies. They overstate the public’s manipulability and underestimate organic demand for low-quality content. And they conveniently overlook more consequential causes of anti-establishment backlash, including the objective gap between the cultural preferences of elites and those of many voters.
Nevertheless, it would be a mistake to see no connection between social media and the rise of populism. To make sense of this connection, however, we should focus less on social media as a dysfunctional technology and more on its status as a democratising technology.
The End of the Gatekeepers
Democracy entails equality among citizens in their power to influence collective decisions. “One person, one vote” embodies this principle, but such formal equality can obviously co-exist with extreme inequalities in practice. It’s easy to get bogged down in tedious philosophical debates about how much political equality is required for societies to be “truly” democratic. Still, we can safely say that societies become more democratic as they increasingly equalise citizens’ power to shape the political process.
By that measure, mass media and the public sphere have been extremely undemocratic throughout most of history. As Brian Klaas observes, although previous communication revolutions (e.g., the printing press, radio, and television) expanded the audience for information, information production remained primarily in the hands of wealthy, well-connected elites.
The internet and then social media changed this. By removing barriers to entry and the influence of elite gatekeepers, they radically democratised the public sphere. As Klaas puts it, “With an affordable device and an internet connection, anybody could hope to influence how people halfway around the world thought about global events.”
However, this producer-centric analysis undersells social media’s democratising impact. In addition to excluding the public from voicing their views, elite gatekeeping makes the media unresponsive to widespread audience preferences. If people want to hear perspectives that elites reject, they are unlikely to find them within elite-controlled media. Social media changes that, too.
On Scapegoating the Algorithm
Of course, these two stories—social media as dysfunctional and democratising technology—are not mutually exclusive. Both factors likely interact in complex ways. Nevertheless, they involve distinct forces. And these days, most of the alarmist discourse about social media’s dangers revolves around the former story.
If only we could adjust the algorithm and tweak platform design, the thought goes, we could address social media’s poisonous impacts.
Even when the democratising story is hinted at, the core issue is typically expressed in technocratic euphemisms. The problem is not that social media has broadened access to the public sphere. The problem is that platforms facilitate the spread of “misinformation”, “malinformation”, and “disinformation”. The problem is not that the voiceless have been given a voice. The problem is that content-moderation policies don’t do enough to protect the public from algorithmic manipulation.
It’s easy to see why the liberal establishment finds social media’s democratising character awkward. Democracy is supposed to be a good thing, whereas social media is supposed to be bad—so bad, in fact, that it poses an existential threat to democracy.
Moreover, one distinguishing feature of liberal elites is their symbolic egalitarianism. It’s partly by performing anti-elitism—by displaying their commitment to equality and social justice—that they differentiate themselves from bigoted hoi polloi. So, although you might catch them dismissing large segments of the public as hopeless deplorables in their private moments, this kind of attitude is restricted to those with disfavoured political and demographic identities (e.g., white working-class Trump voters). And even then, there is a recognition that the attitude should not be expressed openly.
For these reasons, the liberal establishment typically frames the dangerous impacts of social media through the lens of the dysfunctional technology story.
From Hope and Change to Despair and Panic
It wasn’t always like this. When social media was associated with developments that liberals liked, such as the election of Obama and the early stages of the Arab Spring, its democratising character was highlighted and celebrated. However, in our era of surging right-wing populism, this celebration has been replaced by an elite panic about algorithms and platform design.
To be clear, some people do tell versions of the democratising story. Perhaps most influentially, Martin Gurri’s “The Revolt of the Public” argues that by shattering elites’ information monopoly, the digital age has exposed the public to elite failures and hypocrisies, fuelling an anti-establishment backlash. But most of those who endorse such stories are broadly sympathetic to social media and the populist politics it seems to support. (Gurri himself is now a Trump supporter.)
In other words, those who dislike populism typically embrace some version of the “social media as dysfunctional technology” story, whereas those who like it typically embrace some version of the “social media as democratising technology” story”.
A Less Comforting View
I will outline a different and more uncomfortable view: Social media’s democratising nature is the most critical factor in understanding its political effects, including its negative ones. It is precisely because social media has democratised the public sphere that it has contributed to trends liberals (including myself) are so worried about.
In some ways, this analysis aligns with Brian Klaas’s excellent article, “The Democratization of Information Production is Killing Democracy.” However, Klaas’s argument focuses on how media fragmentation and engagement-maximising algorithms increase the production and consumption of “bad” information, which dupes voters into supporting Trump and other populists.
The story1 that I find plausible is different. For the most part, social media doesn’t manipulate “good” people into accepting “bad” information. It simply reveals popular perspectives on reality that elites previously excluded from mainstream discourse, often for good reason. It is this public revelation and normalisation of popular ideas that explain social media’s most dramatic and dangerous impacts, including its connection to right-wing populism.
The Case for Democratic Pessimism
“Democracy,” said H.L. Mencken, “is the theory that the common people know what they want, and deserve to get it good and hard.”
The quote expresses a pessimistic view about democracy that is politically incorrect these days, but it was the norm among elites and intellectuals throughout most of history.
Although part of this hostility to democracy was undoubtedly self-serving, it was also driven by the recognition that democracy is absurd on its face. A political system that gives everyone an equal say conflicts with the reality that not everyone has equally valuable things to say. People differ in their wisdom and virtue. A large number of people possess neither of these traits.
Such differences are especially salient in politics, where it’s been known for a long time that most voters are shockingly ignorant and misinformed, with many approaching politics more in the manner of sports hooligans and religious fanatics than the rational deliberators of liberal fantasies.
One reason for this is that politics brings out the worst in us. People are typically fairly rational when it comes to issues of immediate practical importance. But modern politics involves distant, abstract, and complex issues. It is challenging to form accurate opinions in most cases, and there are few incentives to do so. Given this, many people treat politics symbolically, embracing facile slogan-based worldviews that resonate with pre-scientific intuitions and help them signal their tribal allegiances and demonise people they dislike.
As Joseph Schumpeter observed,
“The typical citizen drops down to a lower level of mental performance as soon as he enters the political field. He argues and analyzes in a way which he would readily recognize as infantile within the sphere of his real interests. He becomes a primitive again.”
Lippmann’s “Mental Barbarians”
For these reasons and more, the prophetic journalist Walter Lippmann published two highly influential critiques of democracy in the early twentieth century, first in Public Opinion (1922) and then in The Phantom Public (1925). Although Lippmann was a liberal and progressive, he argued that for liberal societies and progressive policies to succeed, “the public must be put in its place [...] so that each of us may live free of the trampling and the roar of a bewildered herd.”
Part of Lippmann’s argument involved a sophisticated analysis of why even intelligent and well-meaning citizens will inevitably form distorted political opinions. But he was also sensitive to the large number of citizens who, for want of a better term, are complete write-offs.
In a passage that is outrageous to modern sensibilities, he observed,
“The mass of absolutely illiterate, of feeble-minded, grossly neurotic, undernourished and frustrated individuals, is very considerable… Thus a wide popular appeal is circulated among persons who are mentally children or barbarians, people whose lives are a morass of entanglements, people whose vitality is exhausted, shut-in people, and people whose experience has comprehended no factor in the problem under discussion.”
You don’t have to go quite this far or use this language to acknowledge that there are more than a few grains of truth here. And you can’t understand many pathologies of social media without confronting the fact that its ruthlessly democratising character has given all such people a voice.
The Problems and Positives of Elite Gatekeeping
Elite gatekeeping sounds bad. In many ways, it is bad. The central problem with elitism, including elitist critiques of democracy, is that elites are also human. They might dress up their self-interest, prejudice, and unreason in fancy language, but even the highest-quality punditry and legacy media are biased by propaganda, groupthink, and worse.
These days, right-wing populists draw attention to how much of establishment discourse is subtly and often not-so-subtly biased by progressive (“woke”) values, and they have a point. But this is just one bias among many economic, social, cultural, and political forces that corrupt elite opinion and news media, both today and throughout history.
At the same time, a fair-minded analysis must also acknowledge that elite gatekeeping has many benefits. Outlets like the BBC and the New York Times might subtly select, omit, frame, contextualise, and package reality in misleading ways. But they also typically impose basic standards of professional journalism and exclude many people who have nothing of value to contribute.
Social media welcomes such voices into the conversation. More precisely, platforms that have relaxed “content-moderation” (i.e., elite-gatekeeping) policies do. The result? Deranged conspiracy theories about Jews, medieval discourse about demons and occult forces, the most hyperbolic forms of bigotry imaginable, and countless other popular ideas that were previously excluded from mainstream discourse. More concretely: some of the biggest stars of the social media age, including Tucker Carlson, Andrew Tate, Candace Owens, and Tommy Robinson.
When highly educated, liberal professionals encounter such content on social media platforms, they often assume that there must be something dysfunctional about the platforms. As Francis Fukuyama puts it,
“There is an internal dynamic to online posting that explains the rise of extremist views and materials. Influencers are driven by their audiences to go for sensational content. The currency of the internet is attention, and you don’t get attention by being sober, reflective, informative, or judicious.”
There is a grain of truth here, but also wishful thinking. The more fundamental reason social media features a vast amount of unsober, unreflective, uninformative, and injudicious content is that it gives a platform to large numbers of shockingly unsober, unreflective, uninformed, and injudicious people who were previously excluded from mainstream discourse.
Popping the Bubble
Consider conspiracy theories. Although we’re often told that social media has ushered in an era of conspiracy theories, research by Joe Uscinski and others suggests that such beliefs are no more prevalent now than in the past. This is unsurprising. Conspiracy theories resonate with deep-seated intuitions and provide attractive tools for demonising minorities, elites, and rival tribes. They are pervasive throughout human history.
When the kinds of people who write think pieces about how we’re living through an era of conspiracy theories encounter such ideas on social media, they assume that the platforms created them. But in reality, they’re simply being exposed to popular ideas that their bubble and preferred media outlets previously shielded them from.
The point is not that engagement-maximising algorithms play no role here. But to place the majority of blame on such algorithms is to miss the forest for the trees. If the world consisted of Francis Fukuyamas, social media would look radically different, engagement-maximising algorithms be damned. The reason why bad content maximises engagement is that it resonates with many people’s bad beliefs and revealed preferences.
I’m not making a narrowly partisan point. There are plenty of “mental barbarians” across the political spectrum, including on the left. One can find many of them on platforms like Bluesky, in the inexplicably vast audiences of people like Hasan Piker, and among the uncomfortably large number of unreflective partisan fanatics who celebrate the murder of CEOs and conservative pundits.
I’m also not claiming that social media’s destruction of elite gatekeeping has no upsides. As I have argued repeatedly elsewhere, legacy media and academia produce content that is often highly biased and flawed. By democratising the public sphere, social media provides a platform for brilliant, heterodox thinkers who would never have achieved prominence through more traditional routes. I’m confident that my information diet is much better—much higher quality and more varied—than it would be in the absence of social media for this very reason.
However, it’s also true that there is a vast amount of terrible content on social media and that the primary reason for this is that its democratising character amplifies ideas that elites previously excluded from mainstream discourse.
As Joseph Bernstein notes, establishment hysteria about algorithms and platform architecture misrepresents this situation, transforming
“a huge question about the nature of democracy in the digital age—what if the people believe crazy things, and now everyone knows it?—into a technocratic negotiation between tech companies, media companies, think tanks, and universities.”
Why Democratic Media Benefits Populism
The standard narrative about how social media drives support for right-wing populism is that flawed algorithms and platform design amplify the spread of bad information, which dupes people into holding bad beliefs, which leads them to support bad politics.
The story I find more plausible is that social media caters to ideas and prejudices that are already popular. It is less a broken machine manufacturing mass persuasion than the first competitive marketplace of ideas in human history, producing content that ruthlessly satisfies audience demand. Bad features of algorithms and platforms matter—they make popular content worse than it would otherwise be—but it’s pure cope to place most of the blame on such features.
This story offers an alternative and more plausible explanation of why social media helps right-wing populism: Right-wing populism is popular, and social media provides a platform for popular ideas.
More precisely, right-wing populism is popular among non-elites but largely rejected by elites, which is exactly the kind of perspective that social media’s democratising character amplifies.
Social Media “Normalises” Popular Ideas
The fact that right-wing populist views are far more popular among non-elites than elites is difficult to deny.
As Laurenz Guenther has documented, there is a huge gap in many Western countries between the views of political elites and those of voters on cultural issues connected to topics like immigration, multiculturalism, gender relations, and crime. Elites are far more culturally liberal. This gap is so big that across Europe, “the difference between the average voter and parliamentarian is as large as the difference between the average conservative and socialist parliamentarian. ”
This general pattern is also borne out by research on educational polarisation in the US, which shows that highly educated elites who dominate prestigious institutions have much more progressive cultural values than non-elites. The average voter’s preferences on topics like immigration and crime are, therefore, far closer to the right-wing populist Republican Party than the Democratic Party.
This representation gap suggests that elite-run media didn’t just exclude the most egregious forms of misinformation and conspiracy theorising. It also marginalised culturally conservative values and viewpoints associated with the populist right.
The democratising character of social media has destroyed such gatekeeping. Consequently, previously marginalised ideas have become mainstream and “normalised”. As John Burn-Murdoch puts it,
“This proliferation of views and narratives formerly considered beyond the pale, spread via individuals and platforms outside the control of erstwhile political and media powers, has shattered norms that previously kept radicals on the fringe.”
In other words, as the public sphere became democratised, audiences, content creators, and entrepreneurial politicians have learned that right-wing populist ideas were far more popular than they realised. The “inexorable rise of the populists”, writes Burn-Murdoch, is therefore “not so much an upending of the natural democratic order as an unveiling of the electorate’s unfiltered inclinations.”
Why Not Both?
This story is obviously awkward for liberals who oppose right-wing populism in the name of defending democracy. As a result, deep fears about social media’s democratising impacts have been sublimated into a superficial discourse about algorithms. But algorithms are not the problem. The problem is that right-wing populism is popular. The problem is democracy itself.
Or is it? Someone might argue as follows: Yes, social media normalises some popular bad ideas, but it also offers an opportunity for people to expose the bad ideas of elites, and it gives ordinary people who merely have different, more conservative values than the powerful a greater say in democratic decision-making. Even if one doesn’t share those values, it’s unfair for the unrepresentative progressive values of elites to exert disproportionate influence on democratic societies, especially under the hypocritical banner of “defending democracy”.
I see things differently. I think something extremely dark and worrying is happening on the populist right at the moment. This is especially true in the USA, where the MAGA coalition has evolved into a highly authoritarian, extremely corrupt political movement that attempts to overthrow democratic elections and is associated with increasingly extreme conspiratorial, anti-scientific, and crypto-fascist ideas. However, similar trends can be observed in many other right-wing populist movements worldwide.
Maybe this is all just another liberal moral panic, but I’m personally extremely worried about where all this is going.
In this sense, this essay’s title presents a false dichotomy: Perhaps social media is a genuine threat to democracy precisely because it gives us democracy good and hard. Perhaps democracy can’t survive too much democracy.
Further Reading
John Burn-Murdoch has two excellent articles that have influenced how I think about this topic: The End of the Gatekeepers and Did the Political Establishment Pave the Way for Trump and Farage? He cites Vicente Valentim’s 2024 book The Normalization of the Radical Right, which I haven’t read yet but which seems highly relevant. There is also a very interesting conversation between Burn-Murdoch and Alice Evans about social media and politics here.
Brian Klaas: “The Democratization of Information Production is Killing Democracy
Joseph Bernstein, Bad News
Matt Yglesias, The Misinformation Cope and A Boring Theory of the Populist Right
Martin Gurri, The Revolt of the Public
Sacha Altay and Hugo Mercier, Misinformation is a Symptom
My ideas here partly develop (and in some cases revise) arguments I’ve made elsewhere, including in The Focus on Misinformation Leads to a Profound Misunderstanding of Why People Believe and Act on Bad Information, The Fake News About Fake News, The Marketplace of Rationalisations, Scapegoating the Algorithm, The Case Against Social Media is Weaker Than You Think, and Status, Class, and the Crisis of Expertise
Shortly before I published this piece, Richard Hanania published Bring Back the Internet Gatekeepers, which covers some similar ground.
Two smart people who think I’m very wrong about this general topic and present alternative analyses are Renée DiResta and Nathan Witkin.
This story about the impact of social media on right-wing populism synthesises ideas I’ve come across in the writings of Joseph Bernstein, John Burn-Murdoch, and others, as well as ideas I’ve developed elsewhere (see Further Reading).
I believe this takes a very wrong turn about here: "By removing barriers to entry and the influence of elite gatekeepers, they radically democratised the public sphere."
Excuse me for a moment: ELON MUSK IS THE RICHEST PERSON IN THE WORLD! He is *AN ELITE* - of elites - by any but the most tortured definitions. And that's just one example.
The public sphere has not been "democratised". Instead, civic institutions have been destroyed, in favor of strengthening the influence of propagandists (sigh - Internet necessity - what didn't I say? I didn't say "Propagandists never existed before the Internet, or in previous media". The word I used was *strengthening*, as in existed before but now have greater power). This isn't a technological determinism claim for social media. Rather, that's one part of a very big story, and shouldn't be considered in isolation from very deliberate decisions to deform the public sphere.
A key problem is an implicit restriction of "elites" to apparently mean something like "some concern about facts", rather than more at political power. Thus, public health officials will then count as "elites", but right-wing anti-vax lunatics won't count as "elites", even if they're Secretary Of Health and come from a family political dynasty.
I was around for the first wave of Social Media evangelism, and got slammed for repeatedly pointing out that the "No Gatekeepers" slogan was just different and far worse gatekeepers. Which I suppose is sort of self-proving the uselessness of making that point.
I think you’re right that social media didn’t corrupt democracy so much as expose it for what it was. But it doesn’t stop there. Once reflection became measurable, it became reflexive: the crowd trains the algorithm, and the algorithm trains the crowd. The mirror turns into a feedback loop. What began as democratization evolves into conditioning—the system teaching us what to want next.