---
title: "Parity Looks Like Disappearance From the Top"
slug: "literary-men-parity-data"
language: "en"
date: "2026-10-28T16:34:56.789Z"
updated: "2026-10-28T16:34:56.000Z"
description: "289,000 short stories across 211 magazines: women passed men in the 2020s for the first time in 150 years. The bigger change is that magazines stopped being gendered."
canonical: "https://fix.slopit.io/literary-men-parity-data/"
tags: ["culture", "literature", "data", "gender"]
---

Every few months someone publishes a piece asking where all the literary men went. The New York Times ran one; Compact ran one; Jacob Savage claimed that "not a single white American man born after 1984 has published a work of literary fiction in The New Yorker," which is not true — the magazine published Adam Ehrlich Sachs, and later Nathan Blum and Andrew Martin.

Dan Sinykin decided to check the numbers. With Claude Code, he assembled data on 211 literary magazines from 1841 to 2026: more than 289,000 stories by 63,000 authors, partly from the Fiction Magazine Index, partly from his own extraction.

The result is the cleanest refutation of the "disappearing" frame I've seen, and it's more interesting than a rebuttal.

## The long arc

Women outpublished men in the mid-nineteenth century, then fell sharply in the 1880s through the 1900s, hovered around 20% from the 1910s to the 1970s, and climbed decade by decade after that. In the 2020s they passed men for the first time in 150 years, at 54.6%.

That's a group that went from near-total exclusion to bare majority. Calling that "men disappearing" requires a very specific starting point: the decades when men held 80% of the slots.

## The finding that matters more

The headline is the parity number. The more interesting result is segregation — where people published, not just how many.

In the 1930s, a woman writing short fiction typically published in a magazine that was nearly half women. A man, in the same decade, typically published in one that was one-eighth women. By the 2020s, a woman's typical venue was 57.2% women and a man's was 50.3%.

Duncan's index of dissimilarity falls from 54.5% in the 1930s to 17.6% in the 2010s; Theil's entropy index from 0.341 to 0.039. The magazine ecosystem went from a set of gender-typed venues to a mixed one.

The extremes are vivid. Adventure published 6,364 stories between 1910 and 1953; 1.5% were by women. Dime Detective: 0.4%. All-Story Love: 90% women. Playboy ran 2% women in the 1960s, 5% in the 1970s, 4% in the 1980s and 7% in the 1990s — and about a third of those were by Joyce Carol Oates.

That's what a segmented market looks like: genres with different readers, different venues, and different authors. The collapse of that segmentation is a bigger cultural change than the shift from 20% to 50%, and it's invisible in a single parity number.

## The cross-checks

Sinykin doesn't rely on one dataset. Underwood, Bammon and Lee's study of 104,000 books in HathiTrust shows a similar shape (with a different nadir). NYT hardcover fiction bestsellers through 2020 show women close to but not at parity. Major prizes from 2000–2025 show women just under 50%, and 52.7% for 2016–2025. NEA Creative Writing Fellowships go further: 56% women in the 2010s, 60% in the 2020s, with nonbinary writers at 9.1% and men at 30.9%.

Four datasets, different sources, same direction. That's what a real trend looks like, and it's why the "disappearing" claims feel data-backed: they're picking the one series that shows a decline and ignoring that it started from dominance.

## The method is the other story

The gender assignment was deliberately conservative: Wikidata, self-identification, peer-reviewed data, or pronoun inference from at least three reliable sources. No inference from names, though Underwood et al used it. He tested what name inference would do: raise women's share by as much as 5% in the sparse 1860s and 0.1% in the 2020s. The conclusion doesn't depend on the method.

The limitations are stated: missing magazines (Epoch, Prairie Schooner, TriQuarterly, Fence, Grand Street, Tin House), underrepresentation of romance (partly corrected by overrepresenting the Index's romance holdings), no complete pulp bibliography, errors likely.

That's the right way to publish this: the claim is robust to method, and the exact numbers aren't.

## My take

Two things.

**The framing war is about the baseline.** A group going from 80% to 50% has lost 37.5% of its relative share; a group going from 20% to 50% has gained 150%. Both descriptions are true, and the choice of denominator decides the story. "Disappearing" is what parity looks like from the top.

**The interesting change is mixing, not counting.** The segregation collapse means the categories themselves changed: there is no longer a "women's magazine" and a "men's magazine" in the way there was in 1930. A parity ratio can be reached by two segregated populations of equal size; that's a different world from the one the data describes.

One methodological note worth flagging: the corpus was assembled with Claude Code. Large-scale bibliographic extraction used to be a research assistant's year. That's a genuine acceleration for humanities data work, and it also means the error model is now "whatever the extractor missed," which is why the missing-magazines list matters.

The literary men aren't disappearing. They're being joined.

*Source: [Where Have All the Literary Men Gone? — Dan Sinykin](https://sinykin.substack.com/p/where-have-all-the-literary-men-gone).*