Saving Democracy From Democracy

The journalists who came to save us from ourselves

Audio Version:

Something very strange has happened to journalism. I do not mean that reporters have political opinions. They always have. Newspapers were often viciously partisan throughout American history, and anyone imagining Walter Cronkite descending from Mount Sinai carrying two stone tablets labeled Objectivity probably needs to read more history. What changed is more fundamental.

A large part of the mainstream press no longer seems to believe that its primary responsibility is to report what happened, present competing arguments fairly, investigate the powerful and allow readers to reach their own conclusions. Increasingly, journalists have assigned themselves a different job. They aren’t merely reporting on society. They are helping society move toward what they believe it ought to become.

“Protecting democracy” is the polite phrase usually attached to this mission, particularly since Donald Trump's election. But it goes considerably deeper than protecting elections or constitutional government. Modern advocacy journalism has increasingly attached itself to a collection of moral and political causes: climate activism, transgender ideology, racial equity, DEI, expansive immigration, abortion rights, hostility toward traditional religious morality, suspicion of nationalism, and enthusiasm for international institutions and technocratic government. These positions differ from newsroom to newsroom, and certainly not every journalist subscribes to all of them. But taken together, they form a remarkably consistent worldview.

The problem isn’t that journalists hold these beliefs. They are entitled to them. The problem is that many journalists have ceased to recognize them as political beliefs at all. They have been elevated into moral truths. Climate policy isn’t a political dispute; one side is “denying science.” Gender ideology isn’t a political or scientific dispute; one side is attacking “trans rights.” Mass immigration isn’t a legitimate disagreement over national interest, labor, culture and sovereignty; opposition becomes “anti-immigrant.” Traditional Christianity isn’t simply another moral framework; increasingly it becomes “Christian nationalism.” National sovereignty becomes “nationalism.” Populism becomes a “threat to democracy.”

Once you have classified your own political preferences as morally correct rather than politically contestable, objectivity becomes almost impossible. Why would you give equal consideration to people who are wrong? Worse, why give a platform to people whose ideas you believe cause harm?

There is an arrogance buried inside advocacy journalism that rarely gets discussed. It assumes that the journalist is not merely better informed than the reader, but morally better equipped to determine which political outcomes are acceptable. The reader becomes less a citizen to be informed than a subject to be guided.

That is not journalism.

It is social engineering with a press credential.

And that is precisely where the intellectual justification for advocacy journalism enters. The old journalistic aspiration toward objectivity has increasingly been attacked as “bothsidesism,” false balance or the “view from nowhere.” Prominent journalists and journalism scholars have argued instead for “moral clarity,” an explicitly “moral voice,” and journalism committed to defending an egalitarian conception of democracy and confronting structural inequities.

The transformation of journalism into something closer to organized advocacy did not occur in a financial vacuum. The Gates Foundation openly operates a “Global Media Partnerships” program, funding news organizations specifically to expand coverage of health, development, climate and gender issues. Its own grant database documents millions flowing to The Guardian, including $3.6 million in 2023 to continue its Global Development operation and its “positive influence on priority audiences on key foundation strategic topics,” following earlier multimillion-dollar grants; Gates has also funded NPR's Goats and Soda, Africa Check, Bhekisisa and the Solutions Journalism Network, including grants explicitly intended to promote the “solutions journalism approach.”

George Soros's Open Society Foundations has pursued a parallel model, explicitly funding “independent journalism,” journalists and media organizations while simultaneously defining its institutional mission around advancing rights, equity, justice and democratic practice. Open Society has funded journalism programs addressing racial justice, criminal justice, and inequality, including the Ida B. Wells Society for Investigative Reporting, Soros Justice Media Fellowships, and 31 investigative media grants specifically intended to stimulate discussion about racism and inequality after Hurricane Katrina.

This documents something much larger than philanthropic institutions simply supporting journalism. These NGOs have an ulterior motive. They are helping shape journalism before reporters ever enter the newsroom, funding programs and curricula that teach students to regard advocacy journalism not merely as acceptable, but as a moral obligation.

These progressive colleges and universities are teaching progressive political and economic frameworks, including expansive government intervention, central planning, redistribution, socialism, and related ideological priorities.

At the same time, these foundations finance journalism around subjects they have already identified as social priorities, helping determine which issues get covered by reporters, dedicated desks, fellowships, travel budgets, institutional prestige, awards, and years of sustained attention. The result is not simply funding journalism. It is funding the formation of journalists, the framework through which they are taught to understand their profession, and the subjects they are subsequently paid to cover.

Notice what has happened. The journalist is no longer merely an observer of the political argument. The journalist has become a participant in it, while retaining the institutional authority of someone who claims merely to be telling you the news.

That is a very powerful position.

And it creates an equally powerful temptation. If MAGA threatens the society you believe ought to exist, opposing MAGA isn’t partisan journalism. It is “protecting democracy.” If Germany’s AfD party threatens the European project, opposing AfD isn’t political advocacy. It is combating “extremism.” If parents object to gender ideology in schools, taking the other side isn’t activism. It is protecting a marginalized community.

Advocacy disappears by being reclassified as morality.

That may be the central trick.

And advocacy journalism, stripped of the polite terminology taught in journalism school, has another very old name: propaganda.

We wrote about this extensively in PsyWar. Edward Bernays, of all people, once rested part of his defense of propaganda on the assumption that an independent press would serve as a check on it. Newspapers, editors and reporters were supposed to act as gatekeepers precisely because the propagandist could not completely control them. That assumption now seems almost quaint.

By the twenty-first century, journalism schools were openly discussing and teaching advocacy journalism. Objectivity was increasingly portrayed not as an ideal that human beings imperfectly attempt to achieve, but as an impediment. Reporters were encouraged to understand that neutrality might itself perpetuate injustice. Giving both sides a hearing became derided as “bothsidesism.”

There are circumstances in which that criticism is perfectly reasonable. A reporter covering whether the Earth is spherical does not need to find a Flat Earth Society spokesman for balance. But politics is not geometry.

Whether immigration levels are too high, whether biological sex matters in women’s athletics, whether national sovereignty should supersede decisions made by unelected transnational institutions, whether energy policy should favor nuclear power, whether parents should have greater authority over schools, whether marriage and family deserve special protection, or whether a nation’s government owes a greater duty to its own citizens than to people who arrived yesterday are political questions. They are supposed to be debated.

Instead, an extraordinary number of these questions have migrated from the category of political disagreement into the category of extremism. That migration may be one of the most consequential stories American journalism has failed to cover, perhaps because journalists themselves helped make it happen.

Consider Germany. Pick up almost any Reuters or Associated Press story about Alternative für Deutschland and you do not merely encounter the party’s name. You encounter “the far-right Alternative for Germany.” Again and again. Reuters used the formulation this month while reporting that AfD is polling around 41 percent in Saxony-Anhalt and 36 percent in Mecklenburg-Vorpommern. The label arrives before readers learn what the party actually proposes. (reuters.com)

Think about what that language tells an American reader before he or she learns anything else about the party. Germany. Far right. Rising rapidly. You do not need a degree in neurolinguistic programming to understand the association being planted. AfD becomes associated first with extremism, and only afterward with policy.

When Reuters calls a political party “far-right” in virtually every article, most readers are not reaching for Cas Mudde’s academic definition of the European populist radical right. They hear something much simpler: extremist.

And then something remarkable happens. Policies that would have been recognizably conservative, or even mainstream, within living memory now serve as evidence for the label. Now, policies such as reducing mass immigration, deporting people who are in the country illegally, protecting the traditional family, restoring nuclear power, resisting further transfer of national sovereignty to the European Union, preserving German culture, and putting German national interests first all become “far-right.”

These positions may be right or wrong. They may be practical or impractical. They are open to criticism. But exactly when did believing that the German government should put the interests of Germans first become presumptive evidence of extremism? Apparently somewhere around the same time that believing the American government should put Americans first did.

And this isn't history. We can watch the same model operating around the second Trump administration today. Coverage of Kennedy's restructuring of ACIP frequently began not with what the new members had actually done professionally, but with ideological labels telling readers how they should be understood: vaccine critics, vaccine skeptics, Kennedy allies, misinformation figures. When newly released Fauci texts raised legitimate questions about what officials privately discussed concerning vaccination during pregnancy, FactCheck.org immediately framed the story as Republicans reviving a “false miscarriage statistic” and “twisting” Fauci's texts. And when Kennedy challenged CNN's vaccine coverage last week, Dana Bash finally dispensed with the pretense altogether: “I'm not debating nonsense.” There, in four words, is advocacy journalism's central conceit. The journalist is no longer merely responsible for asking difficult questions, testing competing claims and giving the public enough information to decide. She has become the boundary keeper, deciding which questions are legitimate enough for the public to hear in the first place.

The same linguistic trick has been performed on Donald Trump and MAGA. “Far-right.” “Extreme right.” “Christian nationalist.” “Authoritarian.” “Threat to democracy.” The vocabulary has become so familiar that people barely notice it anymore.

Yet many of the supposedly radical positions associated with MAGA would have been utterly unremarkable in American politics twenty or thirty years ago. Bill Clinton campaigned against illegal immigration. Barack Obama deported enormous numbers of illegal immigrants. Democratic politicians opposed same-sex marriage. The existence of two biological sexes was not a controversial political proposition. Flying an American flag did not require a political explanation. Wanting American industry located in America was not fascism. Skepticism toward China was not xenophobia. And the proposition that the United States government has a special obligation to United States citizens was essentially the reason for having a United States government.

The people who still believe these things did not necessarily move. The journalistic reference point moved. The political center of American politics shifted - “we are all far-right now.”

That distinction is almost entirely missing from mainstream coverage. If the political and cultural establishment moves substantially left over twenty years while you stand still, you eventually find yourself on “the far right” without having gone anywhere. Convenient trick, that.

For some people, the hardest part is admitting that the political ground beneath them has moved. They still cling to the identity of being Democrats while embracing positions that are now routinely labeled far right. The cognitive dissonance is simply too great a distance to cross. The political left has been remarkably effective at using Trump as the ultimate moral barrier: question the prevailing narrative and you are told that you are siding with a “rapist,” a “dictator,” a “pedophile,” or whatever accusation happens to be circulating that week. It creates a powerful social and psychological trap. Changing your mind about an issue is no longer treated as changing your mind. It becomes an admission that you have joined the enemy. For many people, that makes acknowledging that they no longer believe the mainstream narrative almost impossible.

Journalists themselves tell us that their profession has changed. Pew Research Center surveyed nearly 12,000 American journalists in 2022 and asked whether journalists should always strive to give every side equal coverage. Only 44 percent said yes. A majority, 55 percent, said every side does not always deserve equal coverage. The American public saw the issue very differently. Roughly three-quarters of Americans continued to say journalists should strive to give all sides a fair hearing. (pewresearch.org)

Again, proportionality matters. Journalism is not obligated to pretend that every fringe theory has equal evidentiary support. But who decides when a political view no longer deserves equal coverage? The reporter? The editor? The owner? The journalism professor? The NGO that has labeled an organization extremist? The government agency supplying the reporter with information?

This becomes particularly dangerous when journalists overwhelmingly inhabit the same political, educational and cultural ecosystem. They cease to recognize their worldview as a worldview. It simply becomes reality. Everyone else has an ideology. They have facts.

The year 2016 broke something.

Two events happened that the transatlantic political and media establishment clearly did not expect: Brexit and Donald Trump. Millions of ordinary people rejected policies endorsed by nearly every prestigious institution available to instruct them. The newspapers told them. The economists told them. The universities told them. The celebrities told them. The NGOs told them. The intelligence establishment told them. The European establishment told them.

And the voters said no.

That should have produced a period of intense institutional introspection. Why don’t these people believe us? Why don’t they want what we are selling? What are we missing?

Instead, much of the establishment settled on another explanation. The voters had been manipulated. Russian disinformation. Facebook. Cambridge Analytica. Fake news. Misinformation. Disinformation. Malinformation. Algorithms. Populism. Extremism.

Suddenly, the problem was not that citizens had looked at establishment policies and rejected them. The problem was that citizens had been exposed to the wrong information.

This distinction matters enormously. Once the political problem is defined as bad information causing bad democratic choices, information control begins to look like defending democracy. Journalists become natural partners in the project.

In PsyWar, we argued that this was the period in which techniques developed for combating foreign propaganda and influence increasingly migrated inward, toward Western populist movements. Brexit, Trump and AfD all came to be treated not merely as political challenges, but increasingly as information-security problems. The book describes how post-2016 institutions began speaking openly about protecting democracy from populist movements and how media, technology companies, NGOs and government actors increasingly operated inside the same information ecosystem.

The change in language is impossible to miss. Protecting democracy increasingly came to mean protecting institutions from the choices voters might make. There is a rather large philosophical problem hiding in there.

Trump accelerated another change. During his first administration, journalists began talking openly about whether traditional objectivity was adequate for covering someone they considered uniquely dangerous. The argument went something like this: Trump lies, Trump violates democratic norms, Trump attacks the press, therefore ordinary journalistic neutrality is not sufficient. Treating Trump like a normal politician “normalizes” him.

And we now have enough distance from recent political events to see just how consequential this transformation became. Russia really did attempt to interfere in the 2016 election, including hacking Democratic targets, conducting influence operations on social media and probing election infrastructure.

What never happened, however, was the much larger story subsequently embedded in American political consciousness: that Trump and his campaign had conspired with Russia to steal the presidency. And the government officials who supplied oxygen to that story were not anonymous bureaucrats. CIA Director John Brennan, FBI Director James Comey, FBI Deputy Director Andrew McCabe, Director of National Intelligence James Clapper, FBI counterintelligence official Peter Strzok, FBI attorney Lisa Page and others occupied key positions as the Trump-Russia investigation and narrative developed.

Brennan briefed President Barack Obama, Vice President Joe Biden, Comey and other senior officials in August 2016 on intelligence alleging that the Clinton campaign itself intended to vilify Trump by tying him to Russian election interference.

The FBI nevertheless proceeded with Crossfire Hurricane, with Strzok opening the investigation under McCabe's direction, and later used reporting from former British intelligence officer Christopher Steele, whose work had been commissioned through Fusion GPS for the Hillary Clinton campaign, to support surveillance of Trump adviser Carter Page. Durham subsequently found that investigators could not corroborate the substantive allegations in Steele's reporting.

The FBI relied heavily on Steele's reporting even though it had not corroborated its substantive allegations, knew the same information was being circulated to journalists, and later learned from Steele's own primary sub-source information that substantially undermined the reliability of the dossier. Despite that, Steele's allegations continued to support surveillance applications, while reporting derived from the same information stream helped give the allegations the appearance of independent public corroboration.

Journalists then amplified a torrent of allegations, leaks and intelligence claims, often with remarkably little skepticism. This was the feedback loop: political opposition research entered the intelligence and law-enforcement system, government investigations gave the allegations institutional credibility, leaks and official statements carried them into the press, and press coverage then made the underlying allegations appear independently corroborated. That is a very different history from the false story that courageous journalists uncovered a Russian conspiracy inside the Trump campaign.

The Steele dossier became part of the evidentiary basis for surveillance of a Trump campaign adviser despite serious reliability problems later documented by the Justice Department inspector general. The resulting Trump-Russia narrative did not prevent Trump from winning in 2016, but it helped define his presidency and taught government, intelligence agencies, technology companies and journalists how effectively the language of foreign interference could be used to shape domestic political information.

Four years later, the machinery worked differently. When the New York Post published material from Hunter Biden’s laptop weeks before the 2020 election, former intelligence officials publicly suggested that it bore the hallmarks of Russian information operations, social-media companies restricted its circulation, and much of the institutional press treated the story as suspect. The laptop itself was authentic.

This was far more consequential than the failed effort to shape perceptions in 2016 because this time the information intervention occurred immediately before voting and concerned authentic material damaging to one of the candidates.

In my worldview, the Biden administration stole this election, although that depends on what one means by stealing. There is no evidence that ballots or vote totals were altered, and we cannot know with certainty how people would have voted in a counterfactual election with unrestricted coverage. But voters were deliberately deprived of politically consequential, authentic information immediately before they chose a president, in substantial part because the information-security establishment had conditioned journalists and technology companies to treat inconvenient information as a potential foreign operation. That is election interference of a different kind, and this time it came from inside the system supposedly protecting the election.

The truth is that if the public had known that Hunter Biden's laptop was authentic and had been free to examine what was actually in it, many voters might well have viewed Joe Biden very differently. Probably enough to swing the election in Trump’s favor.

The material raised legitimate questions about Hunter's lucrative relationships with foreign interests in Ukraine and China, his repeated invocation of the Biden family name and access, and communications surrounding proposed Chinese business arrangements, including the now-famous reference to “10 held by H for the big guy.”

Subsequent congressional testimony added important context: Hunter's former business partner, Devon Archer, described Joe Biden as “the brand” Hunter brought to these relationships and testified that Hunter put his father on speakerphone in the presence of business associates roughly 20 times. If the press had done their job, these facts would have come out before the election.

But that was precisely the point of journalism: voters were entitled to see the authentic evidence, hear the competing explanations, and decide for themselves whether it mattered. Instead, immediately before an election, much of the press and social-media establishment helped turn the question from “What does this evidence show?” into “Is this Russian disinformation?” and were told by the deep-state, 51 former intelligence officers, that it was fake. The voters made their decision without the scrutiny the story deserved.

There was a legitimate issue buried inside this. Journalists should call a demonstrably false statement false. They should not pretend facts are evenly divided merely because two political parties disagree about them. But the principle metastasized.

Soon it was not merely false factual statements that required special handling. It was Trump’s politics. Then his supporters. Then MAGA Republicans. Then Christian nationalists. Then parents protesting school boards. Then people questioning COVID policies. Then people questioning vaccine mandates. Then people questioning immigration.

The distinction between reporting on a political movement and opposing that political movement began disappearing. The press became part of “the resistance.” And somehow the resistance continued calling itself journalism.

This is where I refuse to use the euphemism. If you enter a story having already determined which political outcome is socially desirable, and you select facts, experts, adjectives and context to move the reader toward that outcome, you are not practicing objective journalism. You are practicing advocacy.

If your objective is to change the reader’s political attitude or behavior, you have crossed another line. That is propaganda.

Propaganda does not have to consist of lies. This is one of its most misunderstood characteristics. Some of the best propaganda is factually accurate. It simply selects certain facts, excludes others, repeatedly associates particular words and images, invokes trusted authorities and guides the audience toward the desired conclusion.

Call AfD “far-right” in the first sentence. Call its immigration policy “anti-immigrant,” rather than immigration restriction. Call sovereignty nationalism. Call traditional views of sex anti-LGBTQ. Call opposition to supranational government Euroscepticism. Call populist distrust of institutions anti-establishment. Call skepticism toward the media misinformation.

Eventually, an entire political worldview is communicated through adjectives. The reporter barely has to make an argument. The nouns arrive pre-contaminated.

Germany makes this particularly potent because it has the nuclear weapon of political analogies sitting permanently in the room: Hitler, Nazism and the Holocaust. No sane German wants those horrors repeated. Precisely because of that history, allegations of right-wing extremism carry extraordinary power in German society. The domestic intelligence service can monitor political organizations suspected of threatening the constitutional order. Parties can, under extraordinary circumstances, actually be banned.

That historical structure deserves serious treatment on its own. But American media often take the German government’s classifications and political language and transmit them to American readers with remarkably little skepticism.

The German state says AfD is extremist. Therefore AfD is extremist. Germany’s established parties build a Brandmauer, a firewall refusing cooperation with AfD. Therefore cooperation with AfD would normalize the far right. AfD gains support. Therefore Germany has a growing far-right problem. AfD reaches 41 percent. Therefore 41 percent of voters are supporting the far right.

There is another possibility. Perhaps a large percentage of Germans genuinely oppose the policies being imposed upon them.

Reuters’ own current reporting shows AfD at roughly 41 percent in Saxony-Anhalt and 36 percent in Mecklenburg-Vorpommern. (reuters.com)

At what point does a journalist have an obligation to reconsider the calibration of the label? If five percent of voters occupy the “far right,” perhaps they are a fringe. If forty percent do, we have at least two possibilities. Forty percent of the electorate suddenly became extremists, or the political establishment’s definition of the acceptable center no longer corresponds particularly well to the population it supposedly represents.

Only one of those explanations receives much attention.

This is why language matters. The modern censorship and propaganda apparatus does not always need to delete your words. It can simply classify them. Misinformation. Extremism. Far-right. Conspiracy theory. Anti-vaccine. Election denier. Climate denier. Christian nationalist.

Put a sufficiently powerful label around a person or an idea and millions of people will stop examining the underlying evidence.

We described this in PsyWar as a battle over the information environment itself. Control the framing, and you do not necessarily have to censor the argument. People learn to censor it themselves.

That is much more efficient. And much harder to see.

There is one little problem with all of this. The public noticed.

Gallup began measuring American trust in mass media in the 1970s. In 1976, roughly seven in ten Americans said they trusted the mass media a great deal or a fair amount to report the news fully, accurately and fairly. By 2025 that number had fallen to 28 percent. Among Republicans, it was 8 percent. (news.gallup.com)

Eight percent.

Journalists can blame Donald Trump for that. They can blame Fox News, social media, Elon Musk, podcasters, misinformation, Russian bots, declining education or the alignment of Saturn and Jupiter. At some point, they might want to consider another possibility.

Maybe people stopped trusting journalists because journalists stopped behaving like people who deserved their trust.

Pew reported in 2026 that a majority of Americans have little or no confidence that journalists act in the public’s best interests. Among Republicans and Republican-leaning Americans, distrust is dramatically higher. (pewresearch.org)

That is not a healthy information system. It is certainly not one that journalists should celebrate as evidence that the public needs more guidance.

The purpose of a free press is not to ensure that citizens reach the correct political conclusion. It is to give citizens the information they require to reach their own conclusions.

If you believe people are entitled to vote only after receiving information curated to produce an acceptable result, then what you actually believe in is something quite different. You believe in managed democracy.

The institutions decide the acceptable boundaries. The media enforce them. The fact-checkers police them. The algorithms amplify them. The public gets to choose from what remains. Then everyone congratulates themselves for having saved democracy.

This is exactly backward.

A free society requires a press willing to offend its own political tribe, investigate the institutions it instinctively trusts and accurately explain the arguments of people its reporters may personally despise. That includes MAGA. It includes AfD. It includes populists. It includes nationalists. It includes progressives, socialists and communists. And yes, it includes genuine extremists.

That used to be the job.

JGM/RWM

If you value journalism that still believes you are capable of making up your own mind, please consider becoming a paid subscriber.

Researching pieces like this takes time. It means going back to the original documents, reading the reports everyone else cites but apparently few people actually read, checking what was said against what was reported, and following the money and institutional relationships wherever they lead. Increasingly, that kind of work is being replaced by journalism that tells you not only what happened, but what you are permitted to think about it.

We don’t have a Gates Foundation grant, a Soros-funded fellowship, a government contract or an NGO paying us to advance its preferred narrative. We have readers.

And frankly, we prefer it that way.

Paid subscriptions are what make our brand of independent journalism possible. If you find this work useful, please consider subscribing. It allows us to keep asking uncomfortable questions without first having to ask whether the people writing the checks will approve of the answers.