Patrick Radden Keefe is a staff writer at The New Yorker and the author of six nonfiction books, known for a method he calls “write-around reporting” — telling stories about powerful, uncooperative subjects by interviewing everyone around them rather than the subjects themselves. He has spent two decades refining a form of narrative nonfiction that reads like a novel but adheres to strict factual standards, and he thinks deeply about structure, scene construction, and how to keep readers engaged in an era of short attention spans.
Write-around reporting: how to write about someone who refuses to talk
Traditional journalism often depends on access — the subject agrees to an interview, and PR people and lawyers negotiate terms behind the scenes. When a subject says no, most journalists walk away.
Keefe does the opposite: when someone refuses to cooperate, he writes around them. He interviews former spouses, college roommates, business associates, administrative assistants, yoga instructors, doormen — anyone who had a vantage point on the person.
Famous examples of this tradition include Gay Talese’s Esquire profile “Frank Sinatra Has a Cold,” in which Sinatra refused to participate.
Keefe wrote a major profile of art dealer Larry Gagosian expecting a write-around, but Gagosian unexpectedly agreed — Keefe thinks Gagosian was savvy enough to realize that if he didn’t participate, the real estate in the piece would go to other voices, and he’d rather be his own advocate.
For his book on the Sackler family, none of three generations would talk to him, and they actively threatened to sue him throughout. He ended up with what he believes is a fuller, truer picture than if he’d sat down with cagey, scripted interviews.
He always makes the effort to talk to the subject. He tells them “the train is leaving the station” — they can get on or not, but the piece is happening either way.
Getting to truth without access
Keefe’s goal is that people who actually know the subject read the piece and say, “You really captured their essence. You got it right.” He is comfortable with the subject themselves disagreeing.
He tells subjects upfront: when he’s done, it won’t look like a photograph of them — it will look like a painting, filtered through his sensibility. It won’t be exactly the way they see themselves.
He aims for a picture that is multifaceted, not a caricature — drawing on a huge number of interviews and deep research — resonant to people who don’t know the person and recognizable to those who do.
Using place to reveal character
Keefe uses specific, granular details of place rather than generic ones. A “green sofa” tells the reader nothing; a vintage Victrola or a TWA poster of London is evocative and revealing.
In his book London Falling, about a 19-year-old who died after falling from a luxury London balcony while living a secret double life pretending to be the son of a Russian oligarch, place is central. He traces the family histories of the three people in that apartment — families that came to London from India, Uganda, and Eastern Europe — to understand the environment and milieu that shaped them.
Choosing which stories to tell
Keefe grew up reading mysteries — Hardy Boys, Sherlock Holmes, Agatha Christie, Dorothy L. Sayers — and his father read mass-market crime fiction. This seeped into him and he is drawn to mysteries, sometimes literal whodunits (his book Say Nothing is about a 1972 murder he solved at the very end) and sometimes simpler mysteries where he withholds a piece of information to create a question in the reader’s mind.
He thinks readers are jaded. He opened his Chapo Guzmán piece not in the Mexican desert but in Amsterdam’s Schiphol Airport, where a cartel assassin who loved European travel was arrested — subverting the reader’s expectation of where a drug cartel story should start and creating an immediate question: how did we get here?
He is explicit that he is not in the business of delivering information bullet points. He wants to seduce the reader with narrative. He doesn’t care if you’re interested in the opioid crisis or Northern Ireland — if you read the first few paragraphs and aren’t intrigued enough to keep reading, he’s failed.
What makes a good introduction
Keefe thinks of introductions as making “down payments” — giving the reader unexpected, specific details that reward their attention and make them trust you’ll take them somewhere interesting.
His Bourdain piece opened with Obama’s armored limousine (the Beast) arriving in Hanoi — a sideways entry that worked on multiple levels: it contrasted Obama’s experience of Vietnam through bulletproof glass with Bourdain’s “intravenous” experience of culture, and it nodded to the fact that for many viewers, watching Bourdain on TV was a proxy for going yourself.
He believes in “little mysteries” — withholding small pieces of information that create questions — and sprinkling in “gee whiz” factoids (the Beast carried emergency supplies of blood) to keep the reader engaged.
He rejects the old New Yorker model of two pages of landscape description before anything happens. Attention spans have changed. He needs to earn the reader’s time immediately.
Structure: dealing out the deck of cards
Keefe thinks about structure more than most writers he knows. Because he can’t invent anything in nonfiction, the artistry is in how he deals out the cards he has — when to introduce characters, when to drop revelations, when to reverse the reader’s expectations.
He used to report exhaustively and only think about structure at the end. Now he thinks about it earlier and earlier. While still reporting, he’ll sketch eight or nine beats on the back of an envelope — the beginning, the end, the big narrative moments — and then populate research beneath those beats.
This early structural thinking makes him ruthless about characters. He might encounter a fascinating person but say “nope, doesn’t fit” — avoiding the amateur mistake of shoehorning interesting people into a story where they don’t belong.
He values non-linear structure. He likes to “hop all over the place” and keep the reader guessing, partly because that’s how he likes to read.
He thinks cinematically about scenes: set something up early (the flood of scooters in Hanoi) and pay it off later (Bourdain taking him on a Vespa ride, joining the current of traffic — a moment of freedom that becomes emotionally important near the end of the piece). He realized afterward that Bourdain, a writer himself, had intuited that Keefe needed a scene and deliberately created one for him.
Articles vs. books: different scales, different effects
Keefe thinks of himself as an architect of the reader’s experience, especially in long-form articles (8,000–14,000 words), where he curates a single sitting’s emotional journey from start to finish. Books are dipped in and out of over weeks.
A London newspaper editor friend pointed out a key difference: in Keefe’s New Yorker article, a horrifying moment (a teenage boy throttling his mother) rang like a bell you can’t unring — it vibrated through the entire piece. In the book version of the same story, the moment is still arresting but not emotionally defining, because the reader has more canvas to metabolize it, knows the characters more deeply, and can contextualize it.
In books, you can place a shocking detail 200 pages in and trust the reader to understand it in context. In an article, that same detail would color everything and overwhelm the piece.
On conclusions: Keefe is wary of over-promising. He thinks about the reader’s investment (45 minutes for an article, 15 hours for a book) and doesn’t want them to feel buyer’s remorse. He has gotten better at embracing ambiguity — not wrapping things up with a neat bow just for narrative satisfaction, but letting the reader experience the ambiguity without feeling let down.
Statistics and texture
Keefe uses statistics sparingly and selectively. He thinks of his writing almost like mixed-media art — varying the texture between exposition, quotes, scenes, and statistics so the reader doesn’t hit quicksand.
He rarely goes far without including a quote, because quotes “air out” a story. But he also avoids pages and pages of dialogue. He wants variety.
When he does use statistics, he reaches for frames people can see in their minds. “Two football fields” works better than “200 yards.” The detail that opioid overdoses killed more Americans than all wars since World War II is more powerful than a raw number — though he acknowledges this kind of analogy can sometimes feel crude.
Screenwriting’s influence
Keefe has done extensive screenwriting (unproduced), and it taught him about distillation and concentration. In screenwriting, a page is roughly a minute of screen time, so a four-page scene is long. The discipline is to enter a scene at the last possible moment and exit at the first possible minute.
In nonfiction, he applies this by boiling down, say, a 300-page court transcript to five pages — capturing the haiku clarity of what matters.
Screenwriting also taught him about juxtaposition: cutting away from a scene at the point where the reader wants to know what happens next, delivering necessary exposition while the reader is still hooked, and then picking up that story thread later.
Arthur Sackler’s marketing playbook
Arthur Sackler, the patriarch of the Sackler family, was a Brooklyn-born doctor and advertising copywriter who became what Keefe calls the “Don Draper of the pharmaceutical industry.”
His key insight after World War II, during the birth of Big Pharma and branded drugs, was that you don’t sell to the consumer — you sell to the doctor, who writes the prescription. He designed elegant (and sometimes crooked) campaigns to persuade doctors that new drugs were wonder drugs, often lying about upsides and downsides.
He made his fortune marketing Valium and Librium in the 1960s and 70s — the big push for tranquilizers. He died before OxyContin was released, and his widow and children insist he had nothing to do with it, but his brothers used the playbook he invented when they sold OxyContin in the 1990s.
Keefe identifies a broader pattern in the American economy: the ability to produce a dangerous product and offload legal liability onto individual consumer responsibility — “guns don’t kill people, people kill people” — a logic now being applied to social media’s effects on children.
What he learned from his editors
Daniel Zalewski, features director at The New Yorker, has been Keefe’s editor for over 20 years. Keefe has internalized Zalewski’s thinking so deeply that by the time he turns in a draft, he’s already edited it in his mind anticipating Zalewski’s reactions.
From his first piece (about Sister Ping, a Chinese snakehead smuggler), Zalewski taught him: “I want to hear her voice more.” Even though she wouldn’t talk to him, Keefe found testimony she’d given and asked people what she said, so he could put her words in quotes and let the reader hear her — gruff, utilitarian, barking orders. When he contacted her in jail, her response was, “What’s in it for me?”
Zalewski also taught him to “cull the herd” — if you have six FBI agents, readers can’t keep track. Introduce fewer characters and do it better. Keefe calls this reducing “context load” — the reader has limited RAM, and every new name, place, or scene requires work to load. Once context is established, it works in your favor like gravity.
Lawrence Wright, a colleague, talks about finding a “donkey” — a character who pulls the reader through complicated material. You need at least one person the reader can capture their imagination around and follow. Lee Child calls this “propulsion.”
Keefe’s writing routine
He lives in the suburbs outside New York City and leads a quiet, monastic life with his family. Two days a week he goes into the city for meetings, lunches, and events — containing all his outward energy into those days.
The rest of the week he preserves silence and unbroken stillness, which he needs to hold an article or book in his head and do the kind of deep thinking his work requires. He used to be out four nights a week and found it destructive to his ability to work.
He writes fast at the end of a long research process — 90% research, then a concentrated period of writing in a “fugue state” where he dreams about the work and wakes up to write himself notes.
How to learn to write better
Keefe’s core advice: be in constant conversation between your reader brain and your writer brain. Most aspiring writers treat these as separate activities.
His example: someone sends him a 5,000-word pitch for a New Yorker story. He asks them, “Do you get email? When you open an email that looks like this, do you run?” The editor gets a hundred emails a day. You need three tight paragraphs that leave them wanting more.
The same principle applies to articles and books — writers forget their own lived experience of being a reader. His teaching would involve intensively studying pieces of writing that truly work, taking them apart like a magic trick or a Swiss watch, and then stealing the techniques.
