Kaplan test-prep grew to half the Washington Post Company's revenue — then collapsed and took the paper with it

In November 1984, the company that owned the newspaper that toppled a president agreed to buy a family-run tutoring outfit that coached teenagers through the SAT.² The price was small — about $45 million for a business then earning roughly $8 million a year.¹ It looked like a diversification footnote, the kind of side bet a media company makes and forgets. It was nothing of the sort. That tutoring outfit would go on to pay The Washington Post's bills for the better part of two decades, and the day it stopped being able to was the day the Graham family stopped being able to own the paper.

The official story of the Post's long independence is a story about journalism: a family committed to the mission, holding the line while other papers folded. The truer story is an accounting one. The newsroom did not pay for itself for most of that era — a for-profit education business did, and when that business broke, so did the arrangement that kept the paper in family hands.

A tutoring business quietly became the company: the newspaper's name stayed on the masthead while its money increasingly came from somewhere else

The subsidy did not arrive on schedule. The popular version has Kaplan bankrolling the paper from the moment of purchase, but Kaplan itself lost money through much of the late 1980s and early 1990s and only turned into a reliable profit engine after new leadership took hold in the mid-1990s.¹ What followed, though, was extraordinary. Kaplan expanded from test coaching into higher education — actual degree-granting online and campus colleges — and the growth compounded. By late 2007 the numbers had inverted the entire identity of the enterprise: Kaplan accounted for fifty cents of every dollar of the Washington Post Company's revenue, its single largest generator by a wide margin.³ The company that carried the paper's name on its stationery was, in cash terms, mostly an education company that happened to publish a newspaper.

This is the anatomy of a cross-subsidy. A cross-subsidy works when one arm throws off cash faster than it needs it, and another arm loses money doing something the owner values for reasons that aren't purely financial. The tuition and test fees Kaplan collected were high-margin and growing; the newspaper was a prestige asset with worsening economics. For years the education profits simply absorbed the newsroom's losses inside a single set of consolidated books, and from the outside it looked like a healthy media company. It wasn't. It was a healthy education company wearing a media company's clothes.

The losses the tuition money was hiding: the paper's own economics had been failing for years — the subsidy just kept the failure off the front page

Strip away Kaplan and the newspaper's standalone picture was grim. Across 2008 and 2009 the newspaper division's losses topped $150 million combined, and Kaplan's profits kept the parent company solvent anyway.⁴ The rot was structural, not cyclical: by the third quarter of 2011, print advertising revenue at the newspaper and online division had fallen 20% year-over-year to $57.6 million, the kind of decline no cost cut catches up to.⁵ The company also carried other bleeding assets — it had sold its money-losing Newsweek magazine the year before⁵ — but the core problem was plain: the newspaper could not fund itself, and everyone reading the consolidated statements knew what was doing the funding.

When Washington came for Kaplan, the whole structure buckled: the arm holding up the paper turned out to depend on federal rules that changed

Here is the twist the tidy version leaves out. The subsidy did not fade because the newspaper improved; it broke because the subsidizer itself came under attack. Beginning in 2010, federal scrutiny of for-profit colleges landed hard on Kaplan's higher-education division. Enrollment fell from a peak of more than 100,000 students at the end of 2009 to roughly 65,000 by the end of 2012 — a decline of over 37% in three years.⁴ The engine that had covered the newsroom's losses was now a shrinking, embattled unit fighting for its own survival. By late 2011 both arms were contracting at once, and the company posted a $6.2 million quarterly loss.⁵ Kaplan didn't so much fund the paper up to the moment of sale as fail alongside it.

The final filings make the imbalance almost surreal. In the second quarter of 2013, Kaplan brought in $548 million of the company's roughly $1 billion in total revenue, while the newspaper division managed $138 million and ran a $49 million operating loss over the first half of the year.⁶ A company that in 2012 already drew 55% of its revenue from education was about to become one that drew nearly two-thirds from it.⁸ The newspaper had shrunk to a rounding error inside its own parent — a mission the family could no longer afford to keep on the books.

On August 5, 2013, the company announced it had agreed to sell its newspaper publishing businesses — including The Washington Post itself — to Jeff Bezos, in his personal capacity rather than through Amazon, for $250 million in cash.⁷ And note precisely what was and wasn't sold. Kaplan was not part of the deal; nor were the TV stations. The company kept the education business and renamed itself Graham Holdings Company.⁷ The move was the mirror image of the myth that the Post divested Kaplan to save the paper. It divested the paper and kept Kaplan — because Kaplan, even wounded, was still the actual company. The Grahams didn't sell a business unit. They sold the reason they'd needed one.

Wasn't this just the internet killing another newspaper?: print was dying everywhere — but that's not why this particular family had to let go when it did

The fair objection is that this is over-clever: every newspaper was collapsing in these years, the internet gutted print advertising universally, and the Post was simply one more casualty of a technological wave that no cross-subsidy could have outrun forever. That's partly right — the 20% single-quarter drop in print ads was an industry story, not a Post-specific failure.⁵ But the timing is the tell. The Post did not sell in 2008 or 2009, when its newspaper losses were at their worst and exceeded $150 million.⁴ It held on. What changed between then and 2013 was not the newspaper's economics — those were bad throughout — but Kaplan's. The subsidizer's enrollment fell 37%,⁴ the cushion thinned, and only then did an independent Post become unaffordable. The internet explains why the newspaper lost money. It does not explain why this owner sold in this year. For that, you have to look at the tutoring business.

The lesson of the Post is not that great journalism can't pay its way, though it often can't. It's that for twenty years the most closely watched newspaper in America was, in the cold arithmetic of its own filings, a line item funded by test scores and tuition checks. The family kept the paper independent not by mastering the economics of news but by owning a better business next to it. When Washington's regulators reached for the for-profit college industry, they were — without meaning to — reaching for the thing that kept The Washington Post in the Graham family's hands. The paper didn't die of its own losses. It changed owners the moment its quiet benefactor could no longer afford the gift.