Before the Algorithm, There Was the Penny Press: How 1840s America Built the Attention Economy

The standard Silicon Valley origin story positions the attention economy as a product of the digital age — an emergent property of social networks, behavioral data, and the optimization engines that learned to serve human psychological vulnerabilities with inhuman precision. It is a compelling narrative. It is also, historically speaking, about 180 years too late.

The attention economy was not invented in a Menlo Park conference room. It was invented in a cramped New York print shop in the 1830s, by a Scottish immigrant named James Gordon Bennett who had figured out something that every platform engineer since has independently rediscovered: that outrage, scandal, and the arousal of strong emotion reliably outperform accuracy as drivers of audience engagement. Bennett didn't have an A/B testing framework. He had a nose for what sold papers, and he was ruthless about following where that nose led.

The results were, depending on your perspective, either a triumph of entrepreneurial innovation or a civilizational warning that nobody heeded.

The Technological Convergence That Changed Everything

To understand what happened in the 1830s and 1840s, it helps to appreciate that the penny press didn't emerge from a vacuum. It was the product of a specific technological convergence — the simultaneous arrival of high-speed cylinder printing presses, cheap wood-pulp paper, and rising urban literacy rates — that suddenly made mass-circulation newspapers economically viable for the first time in American history.

Before this moment, newspapers were expensive, subscription-based, and targeted almost exclusively at the merchant class. A copy of a major New York paper in the 1820s cost six cents — roughly equivalent to a working man's hourly wage — and its contents reflected its audience: shipping news, commercial intelligence, political correspondence, and the kind of formal prose that assumed a reader with both leisure and education.

The penny press blew this model apart. At one cent per copy, sold by street vendors rather than delivered by subscription, papers like Benjamin Day's New York Sun and Bennett's Herald could reach an entirely new audience: the urban working class, recent immigrants, the newly literate. The potential circulation numbers were staggering. The advertising revenue that those circulation numbers could command was more staggering still.

But reaching that audience required feeding it content it would actually pay a penny to read. And what that audience wanted — what, it turned out, almost every audience wants when given the choice — was not shipping news.

Bennett's Algorithm

James Gordon Bennett was, by most accounts, a genuinely unpleasant man with a genius for understanding human appetite. He pioneered the coverage of crime and scandal as front-page news, personally reporting on murder investigations with a lurid attention to physical detail that scandalized polite society and sold papers by the tens of thousands. He invented the financial interview, the society column, and the practice of sending reporters to actively generate stories rather than passively receive them from official sources.

Most consequentially, he understood that controversy was a renewable resource. Bennett regularly published content designed to provoke responses from rivals, readers, and public figures — responses that he would then cover as news, generating a self-sustaining cycle of attention that required no external stimulus to maintain. He was, in the vocabulary of a later era, deliberately manufacturing engagement.

When the moral reformer and newspaper editor William Cullen Bryant organized a public "Moral War" against the Herald in 1840, gathering competing editors and civic leaders to denounce Bennett's methods and urge advertisers to withdraw their support, Bennett covered the campaign against him with evident delight. The controversy boosted his circulation. He had grasped, intuitively, what platform designers would later encode into recommendation engines: that negative engagement is still engagement, and that the appearance of being attacked by respectable opinion can function as a powerful endorsement to an audience that distrusts respectable opinion.

The Telegraph and the Birth of Real-Time Outrage

If the penny press created the demand architecture of the attention economy, the telegraph provided the supply chain. When Samuel Morse's commercial telegraph network began expanding in the mid-1840s, it did something to journalism that is almost impossible to fully appreciate from the contemporary vantage point of ubiquitous instant communication: it made news from distant places available in hours rather than days or weeks.

This sounds like an unambiguous improvement in the quality of public information. In some respects it was. But it also introduced a structural pressure that permanently altered what journalism was for.

Before the telegraph, the time lag between an event and its coverage provided a natural — if accidental — editorial filter. A reporter covering a distant event had days to gather context, seek multiple sources, and construct a narrative that reflected some degree of complexity. The telegraph collapsed that interval to near-zero and, simultaneously, made speed itself a competitive advantage. The paper that got the story first sold more copies. The paper that got the story first and made it more emotionally vivid sold even more.

Accuracy, which requires time and verification, became a competitive liability in a market that rewarded velocity and sensation. The newspapers that thrived were the ones willing to sacrifice the former in pursuit of the latter. This is not a Silicon Valley innovation. It is a mid-nineteenth-century American business discovery that has been rediscovered, with each new communication technology, ever since.

The Line from Bennett to the Algorithmic Feed

The direct lineage from the penny press to the contemporary algorithmic feed runs through several intermediate stages — yellow journalism in the 1890s, tabloid culture in the 1920s, cable news in the 1980s and 1990s — each of which represented a new technological capability applied to the same underlying business logic. Sensation sells. Outrage retains. Complexity loses audience to simplicity, and nuance loses to narrative.

What changed in the social media era was not the logic but the precision of its application. Where Bennett relied on his own intuitions about what would provoke a reaction, modern recommendation systems can test millions of variations in real time and optimize toward measurable engagement signals with a consistency and speed that no human editor could match. The result is the same instinct — serve the audience the content most likely to produce an emotional response — executed at a fidelity and scale that Bennett could not have imagined.

This reframing carries significant implications for how we think about media reform. The popular critique of social media platforms tends to treat their addictive, polarizing qualities as the product of specific design choices made by specific engineers at specific companies — choices that could, in principle, be reversed or regulated away. This critique is not wrong, but it is incomplete.

The design choices were made within a competitive market that has been selecting for emotional engagement over informational quality since at least 1835. Every major attempt to interrupt that selection pressure — the Moral War against Bennett, the yellow journalism backlash of the early 1900s, the broadcast journalism standards of the mid-twentieth century — has produced temporary corrections followed by eventual reversion, because the underlying economic incentive was never eliminated, only temporarily suppressed.

What the 1840s Knew That We Keep Forgetting

Human psychology has not changed in the interval between James Gordon Bennett and the present day. People were drawn to scandal, aroused by outrage, and more likely to share a vivid story than a carefully qualified one in 1840, and they are drawn to, aroused by, and likely to share the same things in 2025. The technology that delivers the content has changed beyond recognition. The content that performs best has not.

History suggests that treating this as a problem introduced by any specific technology is a category error. The penny press didn't corrupt a previously virtuous information ecosystem. It revealed what the information ecosystem had always been capable of becoming when the economic incentives aligned correctly.

Silicon Valley didn't break journalism. It gave a very old business model a steroid injection and an audience of three billion people.

The question worth asking — the one the 1840s were already implicitly posing — is whether any institutional arrangement can sustainably reward the production of accurate, complex information in a free market where sensation is always cheaper to produce and easier to consume. American journalism has been trying to answer that question for nearly two centuries.

It has not yet arrived at a satisfying response.