textCrunch, No. 1: Fit Check

on diversity in publishing, from the payroll to the page

Welcome to the first official issue of textCrunch!

I don’t work in publishing; I never have. I’m a researcher who studies publishing from the outside, using data and interviews and observation and archives. My goal with this newsletter is to get high-quality research into the hands of people who can do something with it. I had planned on using this first issue to talk about some exciting new research on Amazon algorithms—and I will share that with you, eventually!—but that seemed insufficient for our present moment.

When confronted with misinformation or information overload, I find myself comforted by data. Inhale statistical significance, exhale anxiety. Of course, data is not objective, but when the world feels like a dumpster fire, I turn to numbers to understand what’s going on. Then I turn to books to make sense of it all.

One of Trump’s current targets is “DEI.” This is just the most recent iteration of his racism. Language of diversity, equity, and inclusion is being scrubbed from government agencies—including important research agencies—and offices are being shuttered. I don’t know how publishing will respond. Already, the wave of industry diversification that crested after George Floyd’s murder has receded, as countless industry veterans told us it would, as it has every ten-or-so years since 1968. And while the first Trump presidency sparked a wave of inclusive publishing, the recent comments from NewsCorp CEO Robert Thompson haven’t inspired much encouragement about the years ahead.

So, I thought I’d share some data about why diversity in hiring matters for publishing. Because someone might need these receipts, sooner rather than later.

We’ve all seen the hiring studies. Two versions of the same resume, featuring the same information, are submitted to the same hiring manager, with one small difference: one resume belongs to “Chad” and one resume belongs to “Letitia.” What happens? Despite their nearly identical experience, Chad gets the job, every time. Chad is impressive. Chad inspires confidence. What is it about that Chad?

How does this situation play out in book publishing? This is precisely what sociologists Clayton Childress and Jean-François Nault set out to find in their 2019 study, “Encultured Biases: The Role of Products in the Pathways to Inequality,” published in the American Sociological Review. The authors essentially run a version of this classic hiring study in the context of trade publishing, focusing not on Chad or Letitia, but on their books, and not on hiring managers but on the people evaluating them: agents and editors. Do publishing intermediaries (agents and editors¹) prefer to work on books that reflect their own cultural background?² And, if so, why?

This study isn’t especially new, whether by academic standards (2019, with research conducted well before) or by publishing standards (2019 was, like, three diversity waves ago). But I wanted to return to it now because it shows why diversity and inclusion in publishing is so important—not because it is ethically and morally correct (though it is), but because it directly shapes the books that are published.

The authors analyze data on 1,196 literary agents who work on fiction.³ This is a huge sample—there’s no hard data on the number of literary agents working in the US, but best estimates suggest around 1,500. Data comes from the AAR member roles,⁴ and includes information about the sorts of books that members self-reported that they want to represent.⁵

The first part of the question: are publishing intermediaries (in this case, agents) choosing to work with books that match their cultural background? The authors look at a wide swath of demographic characteristics, including gender, race/ethnicity, and age.

The answer to this question is straightforward: yes.

Let’s start with gender. Women are five and a half times more likely than men to represent books about women protagonists. The authors find highly significant statistical evidence that women-identified intermediaries are more likely to work on:

• romance

• chick-lit

And that male-identified intermediaries are more likely to work on:

• action/adventure

• crime/detective/police

• espionage/military

• science fiction

• Westerns

How about age? The older the intermediary, the less likely it is that they work on children’s or young adult fiction. This can be quantified even further (and mind-blowingly): for every year an intermediary ages, the likelihood of representing YA decreases by 4.5%.⁶

Finally, let’s look at the stats on race. The “odds of non-white intermediaries selecting to work on ethnic or multicultural fiction, compared to white intermediaries, are 5.8 to 1.” Let’s reverse that: white intermediaries are almost 6 times less likely to work on ethnic or multicultural fiction than those who identify as POC.

The takeaway: the authors find strong statistical evidence that intermediaries choose to work on books that align with their own cultural backgrounds and identities, particularly in regard to age, gender, and race.

The publishing industry thrives on received wisdom—ideas that have circulated for decades, both formally and informally, that have rarely been tested with data. One of these ideas is the concept of “fit”—that an author should look for a good fit with prospective agent or editor, that an agent or an editor is looking for a good fit in a prospective client or book. Good fits, the thinking goes, make good books. This advice is everywhere, in handbooks for writers and for editors. To be clear, this is good and well-meaning advice! I have given this advice, and I have taken it in my own writing life. The intermediaries interviewed in this study (82 agents, editors, and authors) agree.

When intermediaries talk to the authors about fit, they mean cultural or personal familiarity. If they have an affinity with the book, they will be able to sell or edit it better, as their work will come from a place of personal connection. The authors describe one telling exchange:

Interviewer: You signed a roller derby book, right?

Respondent: Yes.

I: So, let’s say you didn’t know anything about roller derby, but you thought . . .

R: [cutting off question] If I didn’t know anything about roller derby I probably wouldn’t have signed that book.

This is an example of one of the many things meant by fit. And, in this case, it’s completely harmless! It means “knowing about roller derby.” (I, for one, would not be a good fit for this book!) But, like most received wisdom, fit has unintended consequences. The language of fit, the authors found, was frequently invoked to justify exclusion on the basis of race.

Tailoring the Chad/Letitia resume study to context of publishing, the authors presented their interviewees a description of a hypothetical manuscript. These manuscripts matched with each intermediary’s background and professional context, with one key difference: race.

In interviews, they pitched (say) a literary novel about a young Black woman growing up in Chicago to a white woman editor who had also grown up in Chicago. Gender: match, and was working at an imprint known for literary fiction. Age: match. Regional background: match. Literary/generic suitability: match. Race: no match. Invariably, the editor would find ways to explain why she wasn’t the right fit for the manuscript. She wouldn’t say “race,” but it would be clear from her answers that she had latched on to this difference as a reason why she was wrong for the hypothetical manuscript, no matter how much she loved it, how much she believed in “alternative voices,” or how strongly she felt it should be published.

The authors repeated this hypothetical over the course of their interviews, and found the response was typical of intermediaries (white and non-white). The authors got even more explicit: “When asked straightforwardly about whether they work on books with non-white main characters, white intermediaries would indirectly explain that they were not the right ‘fit’ for them.” The typical answer: I believe the book should be in the world, but I am not the right person to work on it. It’s not the right fit. Just as no hiring manager could name what it was, exactly, that made them choose Chad over Letitia, so too did “fit” become a rhetorical scapegoat, a way to justify negative judgment.

Presumably, if the industry were more diverse, this might not be a problem. The authors cite an interview with editor Chris Jackson, describing his first meeting with Ta-Nehisi Coates: “[W]e were set up on a lunch date . . . because . . . [w]e were both youngish black men, from somewhat similar backgrounds and with some of the same tastes and interests.” A good fit is helpful—just look at Jackson and Coates. But because the industry is predominantly white, demographic biases in matching make it all the more challenging for non-white authors to make it into the bottleneck, let alone successfully through it.⁷ One author, a Black man, described being on submission: “Chris Jackson said ‘no,’ so I’m fucked.”

What if white intermediaries simply chose to take a more activist, more inclusive approach toward their work? What if the real problem is what the data showed us up top—that white intermediaries are almost 6 times less likely to work with books with non-white protagonists? Presumably, many have made that choice since this article was written and published—this was one way that the industry responded to George Floyd’s murder and the national attention on the Movement for Black Lives. The authors give examples, even in 2019, of intermediaries who did take that approach, especially when they found a book “too good not to publish,” and knew that a writer of color faced long odds.

But the authors don’t challenge the idea that a good fit between editor/agent and author is advantageous for everyone involved—and lead to a better book. In fact, they recount a number of stories of writers of color being misunderstood or taken advantage of or discriminated against by an intermediary who didn’t understand their identity or cultural background. Just because an intermediary might take an activist approach to their work doesn’t mean that it will end well for the writer or her book. An individual’s activist approach won’t solve the problem, which is that there are so few agents and editors with whom a writer could fit in the first place, and demographic similarity (especially in regard to race) is a primary lens through which the others evaluate manuscripts.

I doubt that this last part is ever said out loud; in fact, I’m sure it’s almost entirely unconscious. But that’s precisely why data is so effective at uncovering these hidden (and encultured, institutionalized) biases. Large-scale data like this can show us how individual decisions, made on a single day, about a unique manuscript, can add up to something huge and systemic and institutionalized.

I wish that this study had been conducted more recently. So much has changed— in the world, in trade publishing— since it was published, all of which would certainly some bearing on the results. But there are, nonetheless, some really useful takeaways. Mainly: this study powerfully demonstrates the strong connection between the payroll and the page— that the demographics of the editorial board or agency payroll shapes which authors get a foot in the door. Hiring practices matter.

That said, using this data to reassert legalistic, identity-matched work is the wrong takeaway. This study isn’t prescriptive; it’s not arguing that non-white intermediaries should be the only ones working with non-white writers, or that white intermediarie should stay in their lane, while we watch lists become whiter by the season. They should be able to work on whatever they want! It shouldn’t be up to them to fix the problem of structural racism. But this essay is making a powerful case that, all things being equal, intermediaries lean on identity matching to determine fit. A reversion to the norm, in this case, is a reversion to whiteness.

Hiring practices are important, but they won’t fix everything. This is a systemic problem: without investment in sufficient support structures, POC in publishing (as well as their books and authors) are left in a precarious position. This is, of course, beyond the scope of Childress and Nault’s study. (And besides, you don’t need to take my word for it.)

But these data give us an important baseline— a starting point. And we’ll need to be very clear, in the weeks and months and years to come, about where we stand.

1. “Intermediaries” is a bit like “Anticipated Top Selling Books.” Does anyone in the industry use this term? No. Is it nonetheless rhetorically useful? Yes. I’m sticking with it.

2. This is a technical point: the authors do not match intermediaries to writers or books they already work on. This isn’t an evaluation of current lists. It’s an analysis of the books that intermediaries say they want to work on. When evaluating a match between an intermediary and a book, the authors rely on other research about genres— namely, the characters that appear in those books. Women’s Fiction, one of the genres AAR provides, features women as main characters.

3. Huge caveat alert! The world of nonfiction is different, and the following results may not apply in quite the same way.

4. The AAR changed its name to the AALA in 2020, after this study was published

5. AAR members could select their preferences from a list of 43 pre-identified genres

6. Speculative sidebar, per moi: this may be one reason why we see so many debut novels, largely repped by younger agents, featuring stories of young people. For more on this, allow me to humbly direct your attention to a fascinating forthcoming book about the history of literary agents.

7. In addition to relying on the Pub Weekly Salary Survey (the 2014 edition) to demonstrate the industry’s whiteness, the authors also rely on some research that badly needs to be updated, but is nevertheless the best we’ve got, to argue that most publishing staff come from upper-middle-class backgrounds (1982) and attended elite universities (2007). Education and social class matter a great deal, impacting who can get a foot in the door, and who can afford to take the job or unpaid internship. It’s a shame we don’t have better data on this. Call me, Pub Weekly.