Hearst, Pulitzer, and the Circulation War That Invented Yellow Journalism

A million copies, bought openly

Joseph Pulitzer bought the New York World in 1883 and pointed it at a strategy that seems obvious now and was audacious then: crusade against corruption, print heavy illustration, and chase circulation as the paper's central measure, not a byproduct of good reporting. It worked. Between 1883 and 1911 the World became the first American daily newspaper to sustain a circulation of one million copies, a figure that draws on outside circulation counts of the period rather than a publisher's own claim (Wikipedia, "New York World"). Pulitzer had found something structural about the newspaper business before anyone named it: a paper's real product was not the news on the page. It was the readership the page assembled, because that readership was what a newspaper actually sold, to advertisers buying space against a guaranteed number of readers.

William Randolph Hearst arrived at the same conclusion from the opposite coast. His father, Senator George Hearst, had entrusted him with the San Francisco Examiner in 1887, and the younger Hearst built it into a sensationalist success using the same formula Pulitzer was running in New York (Wikipedia, "William Randolph Hearst"). By 1895, with family money behind him, Hearst bought the New York Journal and moved directly into Pulitzer's own market, not to compete on quality but to take Pulitzer's readers away from him.

The two papers were, by design, nearly identical products. Both ran sensational crime coverage, heavy illustration, and civic crusades, because both had found the same commercial truth: a reader who felt something, outrage, pity, excitement, was a reader who came back tomorrow, and a returning reader was inventory a paper could resell to an advertiser every single day. Circulation was not a report card on the journalism. It was the asset the journalism existed to produce.

None of this began with Pulitzer or Hearst. American newspapers had been moving toward mass, advertising-supported circulation since Benjamin Day launched the New York Sun in 1833 at a price low enough for a working household to buy daily, breaking with an older model built on costlier subscriptions sold mainly to merchants and the professional class. The cluster of decades this series calls the mass press and the advertising economy runs roughly from that 1830s break to the early twentieth century, and the World and the Journal sat near its peak: the technology to print cheaply and the advertiser demand to fund it had both matured, leaving circulation itself, not the subscription fee, as the number a paper's owners answered to.

The staff raid and the price war

Hearst's method for taking Pulitzer's readers was not to out-report him. It was to buy the machinery that made the World work and move it a few blocks to the Journal's offices. He hired away much of the World's Sunday staff, the department Pulitzer had built into the paper's biggest draw, and among the hires was Richard Felton Outcault, the cartoonist behind The Yellow Kid, the comic strip that had become the World's signature attraction (Wikipedia, "Yellow journalism"). Outcault kept drawing the strip for the Journal. Pulitzer, unwilling to lose the feature entirely, put a replacement artist on a rival Yellow Kid strip in the World. For a stretch of 1896 and 1897, two versions of the same yellow-clad comic character ran in two competing papers, and the spectacle gave the whole style of journalism its name.

The staff raid was not a side story to the circulation war. It was the circulation war, conducted in its purest form: rather than build an audience from scratch, Hearst bought one, treating a rival newsroom's talent as inventory that could be acquired directly. He backed the raid with a pricing strategy that made the same point from the other direction. The Journal sold below what it cost to produce, a loss the paper could absorb only because its real customer was never the reader. It was the advertiser paying for space against whatever circulation number the paper could put in front of them. Every additional reader cost Hearst money on the copy itself and made him money on the ad inventory sold against a bigger number, multiplied across the entire press run, every day the paper printed. Circulation, in other words, was not the reward for running a good business. It was the business.

The arithmetic worked. By 1897, Hearst's two New York papers, the morning Journal and its evening edition, had a combined circulation of roughly 1.5 million (EBSCO Research Starters, "William Randolph Hearst"). By the end of 1898, in the thick of the Spanish-American War, Hearst claimed the Journal alone was moving about 1.25 million copies a day. Whether every claimed copy reflected an actual paid sale is a separate question, one publishers of the period had every incentive to inflate, but what is not in dispute is that Hearst had converted a staff raid and a loss-leading price into the largest newspaper audience in New York.

Pulitzer had not stood still either. He answered the raid by defending his own Sunday circulation with new writers and features rather than matching Hearst's price cuts outright, and the two papers spent the mid-1890s locked in a spending contest that had almost nothing to do with which one covered the day's events more accurately. A dollar spent luring a star writer or cartoonist away from a rival, or shaved off the cover price, only made commercial sense if it returned more in advertising revenue than it cost, which meant every editorial decision at both papers was, in practice, also a circulation decision.

Manufacturing the story: the Cisneros jailbreak

The same commercial logic explains an episode that reads less like reporting than like stagecraft. In late summer 1897, Hearst sent Journal reporter Karl Decker to Havana with instructions to engineer the escape of Evangelina Cisneros, a young Cuban woman jailed by Spanish colonial authorities on charges connected to the independence insurgency (W. Joseph Campbell, Getting It Wrong: Debunking the Greatest Myths in American Journalism, 2010; Wolfsonian-FIU Library, 2025). Decker did what he was sent to do. He helped Cisneros escape a Havana prison, and Hearst then staged a public welcome for her in New York, followed by a meeting with President McKinley, a sequence engineered from the first cable to the last handshake to keep her name in the Journal's pages for weeks.

Nothing about the Cisneros jailbreak was investigative reporting in any ordinary sense. It was a newspaper creating the news event it then covered, at real personal risk to Decker and to Cisneros herself, in pursuit of a story built to sell. Read against the staff raid and the price war, the jailbreak looks less like an isolated stunt by an overzealous editor and more like the same commercial pattern applied to foreign coverage: a business whose revenue depended on readers feeling something would eventually manufacture the feeling directly, once manufacturing it became more reliable than waiting for it to occur on its own.

The World had already shown Hearst how this worked. A decade earlier, in 1887, Pulitzer's paper sent the young reporter Nellie Bly to feign insanity and get committed to the asylum on Blackwell's Island, producing an expose of conditions there that drove real reform and drove circulation at the same time. Bly went on to a stunt that circled the globe in under eighty days, a feat the World tracked and promoted for weeks. Pulitzer's crusading and Hearst's jailbreak both used the reporter as an actor inside the story rather than an observer of it, because a paper whose revenue rose with a gripping personal narrative had every commercial reason to go looking for one, or to build one, rather than wait for the news to supply it.

The quote that never happened

The single most repeated line from this period is a telegram Hearst supposedly sent to the illustrator Frederic Remington, who was in Cuba on assignment and is said to have cabled home that he found no war worth covering: everything is quiet, there will be no war, I wish to return. Hearst's reply, as the story goes: please remain, you furnish the pictures and I'll furnish the war. No copy of that telegram has ever surfaced in any archive, Hearst denied under oath that he sent it, and journalism historian W. Joseph Campbell, after tracing the quote's origins, treats it as almost certainly apocryphal (W. Joseph Campbell, Yellow Journalism: Puncturing the Myths, Defining the Legacies, 2001).

The line survives because it is a tidy story: one villain, one sentence, one confession of intent. The documented history is less tidy and, in its own way, more damning. Hearst did not need to confess an intent to manufacture a war in a single telegram, because the business model he and Pulitzer were both running already rewarded exactly that behavior, story by story, without any single order having to be given. The Cisneros jailbreak needed no Remington telegram to happen. It happened because engineering a dramatic story was, commercially, a better bet than waiting for one.

From circulation numbers to a war

The readership Hearst and Pulitzer were fighting over was large, urban, working class, and disproportionately Democratic. The Republican officials making foreign policy in Washington, McKinley among them, mostly did not read the yellow press themselves (Wikipedia, "Yellow journalism," citing historian W. Joseph Campbell's scholarship). That fact complicates any version of the story in which two publishers personally talked a reluctant president into a war he did not otherwise want. McKinley was not reading the Journal over his morning coffee and being swayed by its front page.

What the yellow press did instead was shape the public mood a president has to govern within, whether or not he reads the paper producing that mood himself. Coverage of the Cuban insurgency in the World and the Journal, sympathetic to the rebels and unsparing in its depiction of Spanish colonial rule, reached a mass urban readership at a scale no earlier American medium had achieved, during the exact years the United States was deciding what to do about a war on an island 90 miles from Florida. The most careful reading among historians, and the one summarized as the current consensus (Wikipedia, "Yellow journalism"), is that yellow journalism amplified an existing appetite for intervention rather than manufacturing that appetite from nothing. Amplification is a real effect, not a rhetorical one: a mood a president has to answer to is itself a political fact, regardless of how it was assembled.

The geopolitical result is structural, not personal. Two private newspaper proprietors, competing for a purely domestic measure, circulation in New York City, ended up shaping how a nation understood a foreign conflict it stood on the edge of entering. Neither man held office. Neither had any formal role in foreign policy. What they held was the largest audience in the country's largest city, assembled through a staff raid, a price war, and a manufactured jailbreak, and that audience's mood became, whether McKinley read the Journal or not, one more input into a decision that was formally his alone to make. War with Spain was declared in April 1898, after two years in which the Journal and the World had built the emotional case for it, and the war, once declared, drove circulation higher still, a small proof of how tightly the economics and the politics had wound around each other.

The same mass medium cuts both ways, and the fair account has to hold both cuts at once. The World's and the Journal's coverage of Spain's brutal reconcentration policy in Cuba, which herded rural civilians into disease-ridden camps, informed a mass American readership about a real humanitarian crisis that a costlier, subscription-only press of an earlier era would have reached only in outline. That is the liberating side of a cheap, wide-reaching medium: more people knew more about a distant war than they otherwise would have. The concentrating side sat right next to it. The specific facts an American reader received about Cuba, which incidents were dramatized, which were ignored, which were staged outright, were set by two men answering to advertisers and a circulation count, with no reader ever voting on the selection. A medium that reaches everyone is not, for that reason, controlled by everyone; in this case it was controlled by whoever could win a staff raid and survive a price war.

What the pattern does and does not explain today

The specific mechanism does not repeat. No AI answer engine sells advertising against a headcount of outraged readers the way Hearst and Pulitzer did, and nothing about a large language model resembles a Sunday cartoon staff worth poaching. What does repeat, in a different technical form, is the underlying rule the 1890s newspaper war exposed without ever stating it directly: whoever controls the medium that carries the news controls what gets amplified, and that control follows the economics of the medium, not the personal ethics of whoever runs it.

Search engines rewarded whichever page matched a keyword. Social platforms rewarded whichever post held attention longest. The pattern now taking shape in AI answer engines is that they name whichever source their systems judge legible and trustworthy enough to cite, and a business absent from that judgment does not get named, however good its underlying offer is. That is a different mechanism from a price war fought over pennies, built on retrieval and citation rather than a printed page sold at a loss, and it is early enough that its long-run incentives are not yet as settled as Hearst's and Pulitzer's turned out to be.

None of this is a forecast of how AI answer engines will behave once their business models fully settle. It is one plausible scenario among several: engines could converge on rewarding accuracy and disclosure the way a mature, regulated press eventually did, or they could reward whatever holds a user's attention or completes a transaction, closer to the yellow press at its rawest. Which scenario plays out will depend on choices still being made about how these systems are paid for and audited, choices no more settled today than newspaper economics were in 1895. What survives the move from newsprint to answer engine is the question the 1890s war first made visible: who controls the medium that decides who gets named, and what does that medium's business model reward them for naming.