The Methodology Was Never Neutral

We met in the back booth of a diner on K Street that had survived three recessions by not updating anything since 1987. The laminate tables were cracked in patterns that looked geological. Egan had suggested it. She said she’d been doing research for something else and had become fascinated by a place where the prices on the menu hadn’t changed in two years even though the portions had gotten smaller three times. “The numbers stay the same,” she said when I arrived. “The thing the numbers describe does not.”

Ma was already there, reading something on her phone with the particular stillness of someone who has been waiting long enough to stop being annoyed about it. She had a cup of coffee that looked like it had been poured a while ago. When Egan slid in across from her, Ma put the phone face-down on the table with a gesture that I recognized from office life — the studied casualness of someone who wants you to know they’re giving you their full attention without appearing to have been doing anything important.

I told them about the premise. A government statistician who works on consumer sentiment. She discovers that the Bureau has been making a series of methodological adjustments — none of them individually dishonest, all of them pointed in the same direction. She starts keeping a parallel index on her personal laptop. The official numbers and her shadow numbers diverge.

“How far apart?” Egan asked.

“Far enough that one of them has to be wrong.”

“But they’re both right,” Ma said. “That’s the thing. Methodologically, both indexes are defensible. You can walk someone through either set of assumptions and they’ll nod along. The difference isn’t in the math. The difference is in what you decide to count.”

Egan was tearing the corner of a sugar packet, not to open it, just working the paper between her fingers. “I spent three weeks once reading about how the Bureau of Labor Statistics handles substitution effects in the Consumer Price Index. This was for something I never wrote. But the basic idea — this stayed with me — the basic idea is that if the price of beef goes up twenty percent and people start buying chicken instead, the CPI doesn’t register that as inflation. It registers it as consumer choice. You chose chicken. You weren’t forced into chicken by the price of beef. The methodology assumes agency at the exact point where agency has been removed.”

“That’s the story,” I said, too quickly.

“That’s one sentence of the story,” Egan corrected. “The story is the woman who understands this and has to go to work the next day and produce the number.”

Ma was nodding, but in a way that seemed more like she was tracking something internally than agreeing. “I want to talk about the laptop,” she said. “The parallel index. Because right now it sounds like a resistance gesture. Brave statistician maintains the true numbers. Keeps the faith. That’s a movie. That’s not what this should be.”

“What should it be?” I asked.

“A compulsion. She doesn’t start the shadow index because she’s courageous. She starts it because she can’t stop. The way some people can’t stop picking at a scab. She comes home and opens the laptop and runs her numbers against the official ones and the gap has gotten a little wider, and she feels — what? Not vindication. Something worse. Something like nausea. The numbers confirm what she already suspects but can’t say out loud, and now she has proof, which means she’s responsible for the proof, which means every morning when she gets on the Metro and badges into the building and sits down at her desk and produces the official number, she’s doing it with full knowledge. The laptop turns her from a technician into a liar.”

The waitress refilled Ma’s coffee without asking. Ma didn’t acknowledge it.

“I agree with all of that,” Egan said, “and I want to push on the form. Because this is a story about measurement, and measurement has a shape on the page. What does this woman’s workday actually look like? She sits in front of datasets. She adjusts seasonal factors. She writes methodology notes — those little paragraphs at the bottom of a data release that nobody reads, that say things like ‘beginning with the January 2024 release, the weighting procedure for housing costs has been updated to reflect current market conditions.’ Those notes are the most important sentences in the economy and they’re written in a voice designed to make you stop reading.”

“The story-as-document question,” I said. “The story has a risk card: it should be structured as a document of some kind. Or multiple documents.”

Egan leaned forward. “Not memos. Not corporate memos. That’s been done, and it’s too easy — the reader sees the corporate voice and feels superior to it. I’m thinking about the documents a statistician actually produces. Methodology notes. Data revision tables. Seasonal adjustment reports. These are genuinely boring documents, and the challenge — the formal challenge that actually excites me — is to make them hold a story. Not by making them secretly interesting. By letting them be boring in a way that becomes unbearable.”

“Because the boredom is the weapon,” Ma said. “That’s how the adjustments stay hidden. Not behind a wall. Behind boredom. Behind the sheer density of technical language that makes anyone who tries to scrutinize the methodology feel stupid and give up after three paragraphs.”

“Yes. The form protects the content. And the protagonist knows this because she writes the form.”

I asked whether the whole story should be documents, or whether we needed a human voice running through them.

They both answered at once, which was clarifying.

Egan: “Documents with cracks.”

Ma: “A person trying not to be visible in her own paperwork.”

They looked at each other.

“Those might be the same thing,” Egan said.

“They might not,” Ma said. “A crack in a document is structural — it’s a place where the form fails to contain the content. A person trying not to be visible is psychological — she’s suppressing herself, writing in the passive voice, using ‘the analyst’ instead of ‘I.’ Those produce different effects on the page.”

“Both,” I said. “Can we do both?”

“You keep asking that,” Ma said, not unkindly. “You want to do everything. That’s a writer’s instinct and it’s usually wrong. The story needs to choose. Either the documents crack open and the human spills through, or the documents remain intact and the human is visible only in what she’s suppressing. I think suppression is harder and better.”

Egan was quiet for a moment. Then: “I think Ling is right about suppression being the default. But I think the story earns one moment where the voice breaks through. Not a confession. Not a revelation. Just a sentence where the passive voice fails. Where instead of ‘the adjustment was applied’ she writes ‘I applied the adjustment.’ And then catches it. And changes it back.”

“She changes it back,” Ma repeated. “That’s essential. She corrects herself. The compliance is the ending of the gesture.”

The diner was filling up with the lunch crowd — people from the office buildings on either side, government workers mostly, their badges dangling from lanyards or clipped to belts. I watched one woman sit down alone and immediately open a spreadsheet on her phone. Ma noticed me noticing.

“There’s something I want to say about complicity that I don’t think I’ve said well enough in anything I’ve written,” Ma said. “Everyone talks about complicity as if it’s a moral failure. As if there’s a door marked EXIT and the protagonist chooses not to walk through it. But complicity in these systems isn’t a choice. It’s an environment. You’re complicit the way a fish is wet. The statistician doesn’t choose to produce false numbers. She chooses — every day, every small choice — to remain in a life where producing the numbers is the natural thing to do. She has a lease. She has student loans. She has a retirement account denominated in the same currency the numbers describe. Her complicity is economic before it’s moral.”

“And that’s the Ehrenreich connection,” I said. “The gap between the official economy and the lived one. Ehrenreich worked minimum-wage jobs and found that the math didn’t work — the numbers said the economy was growing and her back said she couldn’t afford a deposit on an apartment. Our protagonist sees the same gap, but from the other side. She’s producing the numbers that obscure the gap.”

“She IS the gap,” Egan said. “She’s standing in it. One foot in the official version, one foot in her laptop version. And the gap is getting wider, and eventually she either has to jump or split.”

“She doesn’t jump,” Ma said flatly. “People don’t jump. People stand in the gap until it’s so wide they can’t feel their legs and then they go numb and then the numbness becomes normal and they call the numbness ‘professionalism.’”

I asked about the DeLillo — the White Noise influence. The toxic event that’s real but mediated.

“White Noise is about proximity to catastrophe,” Egan said. “The characters can see the black cloud. They can smell it. But they can’t experience it directly because every experience of it is filtered through information — through news reports, through chemical names they don’t understand, through official reassurances that contradict what their bodies are telling them. The data becomes a kind of weather. You live inside it the way you live inside humidity.”

“For our story, the toxic event is the economy itself,” Ma said. “People can feel it. They know things cost more. They know their lives are harder. But every time they try to articulate that, someone shows them a chart that says things are fine. GDP is up. Unemployment is down. Consumer sentiment is — well, that’s the number our protagonist produces. The number that tells people how they feel about how things are going.”

“And the number is wrong.”

“The number isn’t wrong. I keep saying this and I need you to hear it.” Ma set her coffee cup down with a firmness that felt like punctuation. “The number is produced by a defensible methodology. Every adjustment has a technical justification. The number is the output of a process, and the process is legitimate, and the result is a fiction. All three things are true at once. That’s what makes it a dystopia. Not lies. Not fraud. Just a system of measurement that has quietly evolved to measure the wrong thing, and nobody with the authority to change it has any incentive to do so.”

“Because the number is good,” Egan said.

“Because the number is good. Exactly. A bad number would get scrutinized. A bad number would trigger a review. A good number moves through the system like water through a pipe. Nobody stops it. Nobody asks why it’s good. Good numbers are invisible. They’re the silence between alarms.”

Egan was doing something with the sugar packet again, folding it into smaller and smaller squares. “There’s a structural question I keep coming back to. If the story is built from documents — methodology notes, data tables, revision reports — then the reader is in the same position as anyone who tries to understand the economy by reading government releases. The reader has to do the work. They have to notice that the methodology changed. They have to compare the numbers across pages and realize the gap exists. The story doesn’t tell you the numbers are wrong. The story gives you the numbers and trusts you to feel the wrongness.”

“That’s a lot of trust,” I said.

“It’s the only kind of fiction worth writing,” Egan said. “Fiction that does the work for the reader is journalism. Fiction that trusts the reader to do the work is art. And in this case the form is the argument — the reader’s experience of trying to parse these documents IS the experience of trying to understand the economy. The boredom, the frustration, the moment where a number doesn’t look right but you can’t quite say why, the temptation to stop reading and just accept the summary — that’s not an obstacle to the story. That’s the story.”

Ma drank her coffee. “I want the protagonist to have a body. This is a story about numbers, and numbers are bodiless, and I don’t want the story to be bodiless. She commutes. She eats lunch, probably at her desk — the kind of lunch that comes in a container and costs eleven dollars and would have cost seven dollars three years ago except the methodology now counts it as a different product category so the increase doesn’t register. She has a physical life that the numbers don’t describe, and the physical life is where the truth lives.”

“The sandwich that’s been reclassified,” Egan said, almost smiling. “That’s very good.”

“It’s not a joke.”

“I know it’s not. That’s why it’s good. The comedy of economic measurement is that it’s never funny. It’s too precise to be funny. It just — lands. With this weight. You read that gambling revenue is counted as economic output and you don’t laugh, you just sit there.”

“Costa Rica,” I said. Both of them looked at me. “GDP per capita one-fifth of the United States. Life expectancy three years longer. The protagonist finds this data point and can’t stop thinking about it. Not because it’s damning — because it’s a number. Her language. Her training. And it says something her own Bureau’s numbers are structured to never say.”

Ma nodded slowly. “That’s the detail she puts in the shadow index. Not because it belongs there methodologically. Because she needs it somewhere. She needs one number that means what it says.”

Egan was looking out the window. The lunch crowd had turned the sidewalk into a current. “I want to say something about endings that might cause a fight. I think the story should not end with her doing anything with the shadow index. She doesn’t publish it. She doesn’t leak it. She doesn’t quit. She doesn’t delete it either. The story ends on a workday. A regular workday where the new numbers come out and they’re good — adjusted, defensible, good — and she badges in and sits down and the gap between the official version and her version is the widest it’s ever been, and the day is ordinary, and nothing happens, and that’s the horror.”

“That’s not an ending,” I said. “That’s a —”

“That’s the point,” Ma said, cutting me off. “The absence of an ending is the ending. The system doesn’t produce endings. It produces continuity. The numbers come out on the first Tuesday of every month. They will come out next month and the month after and the month after that. The protagonist’s tragedy isn’t that she’s trapped. It’s that nothing changes, and the nothing-changing is the thing she measures.”

I wanted to argue. I wanted a moment of rupture, something the reader could hold onto — the document where the voice breaks through, the paragraph where the passive voice cracks. But I could feel both of them pulling toward something quieter and more terrible than anything I would have chosen on my own.

“The document format,” Egan said, returning to it. “I’m thinking about what she can’t put in the methodology notes. The notes are technical. They describe procedures. They do not describe the fact that her hands were shaking when she ran the parallel numbers, or that she ate lunch at her desk again because the cafeteria reminds her of a conference room where someone explained the new weighting procedure with a PowerPoint and a tone of voice that assumed everyone was fine with it. The gap between what the documents contain and what the documents exclude — that’s the negative space of the story. That’s where the person lives.”

“In the footnotes,” Ma said.

“What?”

“Put her in the footnotes. The methodology notes have footnotes. Technical clarifications. Caveats. Things the main text doesn’t have room for. If there’s a place where the human voice leaks through, it’s in a footnote — some clarification that goes on slightly too long, that provides slightly too much context, that a reader would skip but that she wrote at eleven at night because she couldn’t sleep.”

“A footnote nobody reads,” Egan said. “Containing the only honest sentence in the document. Yes.”

The waitress brought the check. Nobody had ordered food. We’d taken up a booth through the lunch rush on three cups of coffee and I felt the particular guilt of someone who understands that this, too, is a small economy the numbers don’t capture — the waitress’s lost revenue, the opportunity cost of the table, all of it invisible to every index ever devised.

Ma was putting on her coat. “One more thing. The title. ‘Unadjusted.’ It should mean at least two things. The numbers she keeps on her laptop — unadjusted for seasonal variation, unadjusted for substitution effects, unadjusted for all the methodological decisions that make the official number palatable. And the protagonist herself. Unadjusted. The person who has not been — processed. Calibrated. Made to fit.”

“Or the person who hasn’t adjusted,” Egan said. “Who can’t adjust to the thing she knows. Everyone around her has adjusted. The adjustment is the normal state. She’s the anomaly. The outlier the methodology was designed to smooth.”

Ma paused at the edge of the booth. “Those are different. Mine is a compliment. Yours is a diagnosis.”

“I know,” Egan said.

They looked at each other across the cracked laminate, and neither of them blinked, and neither of them conceded, and I understood that the story would have to hold both meanings without choosing between them, which is the kind of problem that only fiction can solve and that fiction solves by not solving it.

Ma left. Egan stayed another minute, reading the menu — the one with the prices that hadn’t changed.

“The portions are smaller,” she said, to no one in particular. “But the prices are the same. Somewhere, an analyst would call that deflation.”

She put the menu down and walked out into the K Street lunch current, and I sat in the booth with three cold coffees and a check for eleven dollars and forty cents that did not, I noticed, include tax.