We Built an Accountability Machine. It Runs on Clicks, Not Consequences.

There is a photograph I keep returning to. It was taken in a Washington newsroom in 2018, the day the Pentagon Papers archive went fully digital. Rows of filing cabinets, emptied. Fifty years of investigation reduced to searchable PDFs. What struck me was not the technology—that was inevitable—but the expression on the face of the archivist in the corner. She looked relieved. As if the weight of all that paper, all those secrets laboriously extracted and preserved, could finally be set down.

I thought of that photograph last month when the Reuters Institute published its annual Digital News Report. Trust in news media has fallen to 40 percent globally, down from 51 percent in 2015. In the United States it is 32 percent. Among adults under thirty-five it is 26 percent. These are not marginal declines. These are structural collapses.

What happened was not a single catastrophic failure but a series of small surrenders, each one justified by the economics of the moment. We built an accountability machine and then, slowly, systematically, we optimized it for something else entirely.

The Arrangement

The bargain used to be straightforward. Journalists would spend months, sometimes years, documenting abuses of power. Publishers would absorb the costs—the lawyers, the travel, the risk of libel—because the business model could sustain it. Advertisers wanted reach. Readers paid subscriptions. The apparatus was expensive but it functioned.

Then the internet arrived and the arrangement dissolved. Not immediately—these things never happen immediately—but gradually, in ways that seemed at first like expansion. More readers, more access, more democracy. What we did not see, or chose not to see, was that the economic foundation had been severed. By 2008, classified advertising revenue had fallen 70 percent. By 2020, newsroom employment in the United States had been cut in half, from 71,000 journalists in 2008 to 35,000.

The question became: what do you do when the model that sustained accountability reporting no longer exists? The answer, it turned out, was that you change what accountability means.

What Replaced It

I know what replaced it because I participated in it. We called it engagement. We measured it in clicks, shares, time-on-page. We ran A/B tests on headlines. We discovered that outrage performed better than explanation, that confrontation drove more traffic than context, that the rhetorical question ('Is Democracy Dying?') generated more engagement than the declarative statement ('Here Is How Power Actually Works').

Accountability reporting became something performed rather than practiced. We published the leaked document, we live-tweeted the congressional hearing, we wrote the searing editorial. What we did not do—what the economic model could no longer sustain—was the unglamorous work that comes after. The follow-up six months later. The investigation into whether the policy actually changed. The documentation of who was hired to implement the reform and whether they did.

There is a reason why we can all recall a dozen political scandals from the past five years but struggle to name a single structural reform that resulted. The scandals were reported exhaustively. The reforms—or their absence—were not. One generates clicks. The other requires the kind of sustained, expensive monitoring that no longer fits the business model.

The Incentive Structure

What emerged was not a conspiracy but a convergence of incentives. Journalists needed traffic to justify their positions. Editors needed engagement metrics to show advertisers and investors. Platforms needed content that kept users scrolling. The result was a media ecosystem optimized for virality, not verification. For speed, not depth. For reaction, not reckoning.

Consider what happened to investigative journalism. In 2003, there were approximately 5,400 journalists in the United States working primarily on investigative projects. By 2022, that number had fallen to fewer than 2,000. The work did not disappear—it migrated to nonprofits sustained by philanthropic funding, which introduced its own distortions. Foundations prefer projects with clear outcomes and measurable impact. They are less enthusiastic about the kind of open-ended inquiry that might produce nothing, or might take five years to produce something.

The word 'investigation' became a marketing category. Publications labeled stories 'investigations' that consisted of aggregating public records requests filed by someone else, or repackaging academic research, or—in the most cynical cases—simply adding the word 'exposed' to a headline. Real investigation, the kind that requires cultivating sources over years and risking legal action and producing evidence that changes policy, became economically unsustainable for most publishers.

What It Means to Lose

I think often of a conversation I had in 2019 with a city council member in Youngstown, Ohio. The local newspaper had been sold to a hedge fund and gutted. The reporters who used to attend council meetings were gone. The paper still published—if you can call aggregating press releases and crime reports publishing—but the apparatus of accountability had been dismantled.

'We used to worry about what would be in the paper,' the council member told me. 'Now we know nothing will be in the paper. You'd be surprised how much that changes behavior.'

This is what we actually lost. Not the journalists—though we lost many of them—but the mechanism by which their presence altered the behavior of people in power. The expectation of scrutiny. The knowledge that decisions would be documented, explained, potentially exposed.

The research is unambiguous. A 2020 study published in the Journal of Financial Economics found that when local newspapers close, municipal borrowing costs increase, corruption prosecutions rise, and voter turnout declines. A separate study by researchers at the University of Notre Dame found that after a local newspaper closes, fewer candidates run for local office and incumbents face less competitive elections. Democracy, it turns out, requires witnesses.

The National Stage

At the national level the collapse manifested differently but with similar consequences. Outlets that survived did so by narrowing their focus to the most commercially viable subjects: national politics, cultural conflict, technology, crime. What dropped out was everything that required specialist knowledge and patient explanation. Agricultural policy. Regulatory capture. The mechanics of procurement. The implementation gap between legislation and enforcement.

We covered the passage of major legislation extensively. We covered the political theater around it even more extensively. What we rarely covered was what happened in the eighteen months after passage, when agencies wrote the actual rules and industries deployed lobbyists to shape them and the original legislative intent got quietly renegotiated in conference rooms no journalist attended.

The result was a journalism that excelled at documenting spectacle but failed at documenting process. We knew who said what on Twitter. We did not know which regulatory agencies had been captured by the industries they were supposed to oversee, or how, or when. That kind of reporting requires beat reporters who spend years cultivating sources and learning complex systems. It cannot be done by generalists working on three-day deadlines.

The Reckoning

I am not sure what I expected when I started thinking through this problem, but it was not the clarity I found in the data. Trust collapsed not because journalism became more biased—polling shows most people cannot accurately identify the political lean of major news outlets—but because it became less useful. Readers could not see how the work connected to consequences. They saw exposés that led nowhere, scandals that were forgotten within a week, promises of accountability that dissolved into the next news cycle.

What accountability reporting must become, if it is to survive, is something structurally different from what it has been. It cannot be episodic—the occasional investigation that generates outrage and then evaporates. It must be iterative. It must return to the same institutions, the same failures, the same promises, year after year, documenting with precision what changed and what did not.

This means abandoning the story forms that have become standard—the 1,200-word takedown, the viral thread, the explanatory piece that explains everything except whether anything will change. It means building newsroom structures that reward patient monitoring over rapid reaction. It means measuring impact not by traffic but by whether the documented abuse stopped, whether the promised reform was implemented, whether the people harmed received remedy.

There are models. ProPublica has pioneered the long-term accountability beat, returning to the same industries and agencies across years and documenting patterns rather than isolated failures. The Texas Tribune funds statehouse reporting through a membership model that prioritizes readers over advertisers. The Bureau of Investigative Journalism in London operates as a nonprofit, insulated from quarterly earnings pressures. These are not perfect solutions, but they are attempts to rebuild the economic and editorial structures that make sustained accountability possible.

What We Owe

The photograph I mentioned at the beginning—the archivist in the emptied newsroom—stays with me because of what it represents. The archive went digital and became searchable, which is progress. But searchability is not the same as use. The Pentagon Papers mattered not because they were preserved but because journalists spent months analyzing them, contextualizing them, explaining what they meant and why it mattered. The revelation was inseparable from the interpretation.

We now have more information available than at any point in history. Government databases, court filings, corporate disclosures, leaked documents—all searchable, all accessible. What we lack is the institutional capacity to transform information into accountability. To connect the documented abuse to the structural failure to the policy change that might prevent recurrence.

This is not a technological problem. It is a problem of will, resources, and editorial priorities. It requires publishers to fund work that may not generate traffic. It requires journalists to commit to beats that take years to develop. It requires readers to support journalism that does not provide the immediate satisfaction of outrage but the slower, more demanding satisfaction of actual change.

The alternative is what we have now: a media ecosystem that documents everything and holds nothing to account. That produces the appearance of scrutiny without the substance of consequence. That allows the powerful to weather scandals by waiting for the next news cycle, secure in the knowledge that no journalist will be there six months later to document whether promises were kept.

We tell ourselves stories in order to live. The question is whether the stories we are telling—the viral exposé, the shared outrage, the performative demand for accountability—are the stories that actually produce the change we claim to seek. I am not sure they are. I am certain we owe it to ourselves to find out.