Where Is the Diversity in Publishing? The 2019 Diversity Baseline Survey Results

The Diversity Baseline Survey (DBS 2.0) was created by Lee & Low Books with co-authors Laura M. Jiménez, PhD, Boston University Wheelock College of Education & Human Development and Betsy Beckert, graduate student in the Language and Literacy Department of Wheelock College of Education & Human Development

Lee & Low Books released the first Diversity Baseline Survey (DBS 1.0) in 2015. Before the DBS, people suspected publishing had a diversity problem, but without hard numbers, the extent of that problem was anyone’s guess. Our goal was to survey publishing houses and review journals regarding the racial, gender, sexual orientation, and ability makeup of their employees; establish concrete statistics about the diversity of the publishing workforce; and then build on this information by reissuing the survey every four years. Through these long-term efforts, we would be able to track what progress our industry shows over time in improving representation and inclusion.

Why does diversity in publishing matter? The book industry has the power to shape culture in big and small ways. The people behind the books serve as gatekeepers, who can make a huge difference in determining which stories are amplified and which are shut out. If the people who work in publishing are not a diverse group, how can diverse voices truly be represented in its books?

The results of DBS 1.0 were stark. 79 percent of respondents identified as White. 78 percent were women. 88 percent were straight. 92 percent were non-disabled. At a time when readers of all backgrounds were demanding to see themselves in books, the publishing industry came nowhere near to reflecting the rich diversity of the United States.

The Last Four Years

The numbers provided by DBS 1.0 contributed to a sense of urgency that has resulted in more diverse books being published in the marketplace today—at least on the children’s book side.  It is now 2020, and many powerful cultural events, changes, and movements have taken place in the four years since the first survey.
● Donald Trump was elected President.
● The #MeToo movement spread virally after sexual abuse allegations against Harvey Weinstein.
Black Panther and Crazy Rich Asians, both featuring POC leads and majority-POC casts, broke major box office records.

Within our industry, too, important conversations have changed the landscape:
Corinne Duyvis coined the term #ownvoices to highlight books where the creator and main character share some form of marginalization.
● Beth Phelan created #DVpit to help diverse creators pitch their work to agents and editors.
● Four educators of color founded #DisruptTexts to question the books commonly taught in high schools.
● Problematic books were canceled, re-illustrated, or rewritten, while The Hate U Give and other diverse titles became New York Times bestsellers.
Authors of color became more vocal about their place in the industry and their exhaustion with the diversity conversation
Drag Queen Story Hour started in San Francisco.
● The Romance Writers Association was forced to reckon with its own biases.

After all this . . . has the publishing workforce actually become more inclusive?

Diversity Baseline Survey 2019 Results

DBS 2019-final5


Methodology and Response Rate


We expanded the DBS 2.0 to include a larger sample set than the original survey, including members of the Association of University Presses (AUP) as well as literary agents. In 2015, there were 3,706 responses to the survey. In 2019, we received 7,893 responses, showing a 112 percent increase in responses from DBS 1.0 to DBS 2.0.

Like the first survey, DBS 2.0 took a year to complete, beginning in January 2019. Reaching out to companies and trying to connect with decision-makers took the most time. Some companies took a lot of convincing; others agreed to participate and then dropped out for myriad reasons. Ultimately, 153 companies participated, including all of the Big Five publishers, eight review journals, forty-seven trade publishers, thirty-five university presses, and sixty-three literary agencies of all sizes from across North America. You can see the full list of participants here.

The Diversity Baseline Survey (DBS 2.0) was sent to 2,609 reviewer employees, 17,100 trade publishing employees, 1,528 university press employees, and 516 literary agents for a total of 21,753 surveys deployed.  DBS 2.0 had a response rate of 36.2 percent. The survey was sent to both children’s and adult divisions of each company.

The DBS was sent out directly from each participating company’s internal contact person. To ensure the integrity of the data, we worked with a professional survey company called Toluna that administrated and deployed the survey for us. The surveys were completely anonymous, and companies did not have direct access to the results.

All data was analyzed and aggregated by a small team at Boston University consisting of Laura M. Jiménez, PhD (Lead) and Betsy Beckert, a graduate student in the Language and Literacy Department of Wheelock College of Education & Human Development. The team at Boston University ensured the anonymity of individual respondents, and they were the only ones with access to the raw data. Excerpts from Laura Jiménez about the statistics have been included below and are indicated by ” “.

Note: DBS 1.0 was deployed in 2015 and released in January 2016. DBS 2.0 followed a similar pattern and was deployed in 2019 and released to the public in January 2020.

Notes and Analysis: What the Numbers Tell Us

The Industry Overall

Race: 
According to the survey, 76 percent of publishing staff, review journal staff, and literary agents are White. The rest are comprised of people who self-report as Asian/Native Hawaiian/Pacific Islander (7 percent); Hispanic/Latino/Mexican (6 percent); Black/African American (5 percent); and biracial/multiracial (3 percent). Native Americans and Middle Easterners each comprise less than 1 percent of publishing staff.

The 2015 survey reported that overall, 79 percent of people who work in publishing self-report as White. Given the sample size difference, this 3 percent change in White employees does not meet the bar for statistically significant change. There is no discernible change to any of the other racial categories. In other words, the field is just as White today as it was four years ago.

Creating the list of ethnicities for a survey such as this is challenging. We received feedback from DBS 1.0 that some of the race categories were too broad or that subcategories should be separated out, e.g. Asian Pacific Islanders within the Asian American category. But we wanted to be able to compare data as well as to avoid any chance of compromising any individual’s personal identifiable information (PII), so we stuck with the racial breakdowns we offered in the first survey, which was based on the US census, with a few adjustments.

Gender:
 The survey reveals that publishing is about 74 percent cis women and 23 percent cis men. The 2015 survey reported that overall, 78 percent of people who work in publishing self-report as cis women. The current survey has 74 percent of the respondents self-reporting as cis women. Given the sample size difference, this 4 percent change in cis women does meet the bar for statistically significant change. These numbers may help explain why some feel that book publishing caters more to female readers. Among executive and board member positions this disparity evened out somewhat, with approximately 38 percent of executives and board members identifying as cis men. This reflects the reality that males still ascend to positions of power more easily, even in female-dominated industries.

Concerning our gender question, we had significant input from various experts (listed in credits on the last slide) who helped us vet the choices for this question. Even so, the gender question reveals that about 97 percent of publishing staff identify as cis men or women, meaning that they identify with the genders they were assigned at birth. How does this compare with the general population? We don’t really know. That being said, the small number of genderfluid, non-binary, genderqueer, trans man, trans woman, and intersex people in publishing points to the need for publishers to make sure that books on these topics are examined for cultural and scientific accuracy by experts before they are published. We have a separate article on the need for targeted expert readers that relates to this issue. (Note: After receiving feedback regarding the wording of “woman/cis woman” and “man/cis man” in the survey, the terminology has been corrected to “cis woman” and “cis man” respectively in the infographic.)

Sexual Orientation:
 According to the survey, about 81 percent of publishing staff identify as straight or heterosexual. This category saw the most change over the course of the study. In the initial survey, 88 percent of the respondents self-reported as straight. This survey saw a statistically significant negative change in the number of respondents that self-reported straight at 81 percent. The change can be largely attributed to the number that identified as bi and pansexual, 10 percent. Not surprisingly, most of these are White women, but that again is most likely attributed to the fact that the field is overwhelmingly White women. There was a slight drop in gay representation to 4 percent, and lesbian to 2 percent, with asexual holding at 1 percent.

During the planning of DBS 1.0, we debated internally whether to include a survey question on sexual orientation, since it is a question that human resources departments cannot ask their employees. We ultimately decided that if we did not acknowledge this aspect of diversity, this segment of the workforce would remain uncounted and invisible. Four years later, as the current administration calls for the removal of LGBTQ data collection in the next census, including LGBTQ representation in our survey seems more critical than ever.

Disability:
 We received good input from disability advocate Ace Ratcliff on the wording choices for this question. This input may have led more people to self-identify as having a disability in DBS 2.0. There is a statistically significant change from 4 percent self-reporting a disability in 2015 to 11 percent reporting a disability in 2019. The clear majority of disabilities are mental illness (45 percent), physical disability (22 percent), and chronic illness (20 percent). There is a high co-morbidity within this data of mental illness and chronic illness co-occurring. We did some initial investigation into this phenomenon, and these findings are consistent with mental health and medical findings. Although this is not the focus of this survey, we took this to signify that this data is falling into line with other, more established research strands.

One hypothesis for this growth is that it is not growth at all, but a change in the way the field, and our society in general, views mental-health issues. In other words, the field of survey respondents may have experienced an uptick in mental health-issues, but this data is also consistent with national surveys that find more people are seeking out and identifying as experiencing mental-health issues. In a subsequent blog post, we will share more detailed data related to the disability question.

Departmental Observations

The DBS 2.0 results offer the opportunity to filter responses by department, giving us a better picture of how inclusive (or homogenous) various publishing departments are in relation to each other. Space constraints prevent us from discussing every department, but we wanted to highlight the data for some of the departments that play a significant role in shaping which books are published and how they get into the world: executive positions, editorial, sales, marketing and publicity, reviewers, and literary agents. We also included interns, to provide a glimpse at what our industry’s future workforce may look like.

Data for these departments can be found in the slideshow above, and data from the previous study (for comparison) can be found here. A few interesting observations:
More diversity at the top: The Executive Level saw some fairly significant changes in its numbers, with the percentage of people identifying as White dropping from 86 percent in 2015 to 78 percent today. In addition, the percentage of people identifying as straight dropped from 89 percent to 82 percent and the percentage of people with a disability increased from 4 percent to 10 percent. This is good news, since true change in company culture almost always requires buy-in from the very top.
Editorial is even more White than before: The percentage of people in Editorial who self-identified as White increased from 82 percent to 85 percent. So, even though more diverse books are being published now, it’s fair to assume that the large majority of them are still being acquired and edited by White people.
Marketing & Publicity is more racially diverse than other departments: Of the departments we address in the slideshow above, Marketing & Publicity is the area with the lowest percentage of White employees (74 percent), although it is less diverse than other departments in the other areas we surveyed.
More people identify as disabled in book reviewing than in other areas of the publishing industry: Nineteen percent of book reviewers identify as having a disability, while in most other areas of the publishing industry, the number is closer to 10 percent. This trend is in line with data from DBS 1.0. Perhaps this is because reviewers are able to work freelance or remotely, without the constrictions of a 9-5 desk job in New York, which is notoriously inaccessible for those with disabilities.
● Literary agent demographics are in line with the rest of publishing: We did not include literary agents in the first DBS, but their inclusion has not really impacted the numbers because their demographics match the rest of the industry: mostly White, straight cis women who are not disabled.
Interns are significantly more diverse than the industry as a whole: Of the interns surveyed in 2019, 49 percent identify as BIPOC; 49 percent are on the LGBTQIA spectrum; and 22 percent identify as having a disability. These numbers are a dramatic departure from the overall industry numbers and signal a new, more representative generation of entry-level publishing staff. The question is whether many of these interns will be retained and promoted, or whether they will burn out or leave publishing for other reasons before their presence can truly change the industry.
● Comparing apples to apples, there’s still no change: For those who wonder how much the inclusion of literary agents and university presses may have skewed the data, our last data slide removes these segments for a true apples-to-apples comparison to DBS 1.0. The largest change is in the sexual orientation category, which dropped from 88 percent straight/heterosexual to 82 percent. No other category changed by more than 4 percent.

What’s next?


Although our baseline numbers have not shown a compelling change, in all cases there was change. As in other industries, increasing diversity in the publishing workforce is a challenge, and change has been slow in coming. But DBS 2.0 did have some positive takeaways. Only three of the Big Five publishers took part in DBS 1.0, so it is a sign of solidarity that all five participated this time. The addition of literary agents and university presses contributed to a bigger pool of participants and gave us a clearer picture of who makes up the industry.

The efforts to make the book community a more inclusive one is an ongoing—oftentimes herculean—struggle, conducted predominately by people of color. The encouraging numbers in the intern section indicate that publishing is trying to reach out to diverse populations. But keeping diverse employees engaged and believing they have a home in this industry is another matter. Without a clear career path and the promise of opportunities for a bright future, retention will continue to be a serious problem, and the needle will not move.

The world has changed a great deal from just four years ago. With so many diverse causes that run parallel to one another that sometimes it can be hard to keep in mind common goals. But until we all start to care about equity, we will not make progress, and any gains the industry makes will continue to be not statistically significant. So, the same questions that we asked four years ago bear repeating: How can company cultures be more welcoming for diverse staff? Do diverse staff members feel comfortable voicing their opinions? Are systems in place to make sure all staff are trained and well versed in diversity issues? And some newer questions to ponder: Have recent conversations on bias and privilege changed your perspective on the systemic problems that exist in society today? Has your empathy grown or receded toward diverse causes in the last four years?

The world is a diverse one. Publishing needs to accurately reflect the world as it is, from the books we publish to the people working in every facet of this business. A lot has happened in four years, and not all of it for the better. Four years from now, what will the next baseline survey show us? And what will those numbers tell us about ourselves?

For permission to reprint or press inquiries, please contact Hannah Ehrlich at dbs [AT] leeandlow [DOT] com

140 thoughts on “Where Is the Diversity in Publishing? The 2019 Diversity Baseline Survey Results”

  1. Two things:

    I don’t see how any of the discrepancies are evidence of current discrimination (as opposed to, say, the aftereffects of past institutional discrimination).

    2. nobody gon talk about the dearth of cis men?

  2. Are there more specific data about the 76% white part? How does that divide en detail in different groups of people who self-identify as white? Does that include white Hispanics?

    1. Hi Eva,
      We don’t have more detailed data about the White/Caucasian category–the survey asks how people self-identify so that’s what is reflected in the numbers.

  3. IMO, these numbers should be considered in light of the U.S. population as a whole and don’t seem as dire to me as the topic assumes. The US Census Bureau, for example, reports the following 2019 stats by race: 76.5% white (pretty even w/publishing, which means the “non-white” identifying group is also even). 8.6% identify as disabled (compared to 11% in this publishing census). The US Census does not track sexual orientation but the Gallup poll does, putting the US at 4.5% LGBTQ, fully half the rate in publishing. The 76% female (as opposed to 50.8% of the general population) stat doesn’t bother me. First, various statistics show women read more books than men, and – there are so many male-dominated fields, I’m happy to see this one field, that is IMO a societally gender-neutral profession, lean female. Also, let’s not forget, as Hannah E. alluded to, statistics are hard to trust and compare, there are so many variables in data collection. All in all, are we crying wolf here?

    1. I was going to say the same thing. I want to see more gender/racial diversity particularly in children’s book CHARACTERS, but on the face of it, these stats seem to reflect favorably on the publishing industry as an equal opportunity employer.

    2. No. We are not crying wolf. I’ll point you back to the question and answer in the opening of the article: “Why does diversity in publishing matter?”

      The issue is not with the number of physical people as implied by your comparison to the U.S. Census. The issue concerns BIPOC and LGBT people not having an equal voice in an industry that shapes education and culture. Gatekeeping is real. Essentially, the survey results show that white cis women continue to have the loudest voices in the publishing industry and continue to decide which books should be read by the masses.
      If a BIPOC or LGBT writer’s work is not palatable to (or most likely, not even understood by) the white cis women making decisions, then the work may never see the light of day, regardless of how essential or transformative the work may be for a minority group.

      On the flip side, the publishing industry will continue to run into “American Dirt”-like blowback because… How many Latinx voices were heard in the conference room before greenlighting that book? Based on the results from this survey and the current controversy, clearly not enough.

      1. I’m almost through the book – chapter 28. I’m listening to you, cognizant and troubled by my lack of awareness of my own prejudicial blindness, an affliction I believe we all suffer from. I’ve certainly experienced my own exclusion by gatekeepers. The problem is, publishers, all of us, are dying out. We have to make a profit. We don’t have margins that allow us to spend our resources publishing works that will never pay for themselves. Even with that truth, I do. It’s part of my mission to do this within my niche. And, I’m inches away from going out of business. All that to say, the readers are making the decisions, and data, and plenty of guessing by fallible humans. It’s terrible that profits play such a role, but true. I will put in the effort to understand this from new perspectives and appreciate your thoughtful reply to my comment.

        1. What I hear you saying is that you publish books about white cis characters because that’s what sells. Is that correct?

          I disagree. There are plenty of profits to be made in diverse books. At the children’s book store for which I work, our largest event audiences were for BIPOC, LGBTQIA+, and books about the disability community. Schools and organizations that serve children need to do diversity audits of their bookshelves, modernize their collections, and their required reading lists. These tend to be outdated, stale, white, and male compared to the award-winning literature that is currently published. Public funds should be appropriated for this. When that happens, books will be purchased and profits will be made. Publishers must go beyond the single “fill-in-the-minority” title (and narrative) for their lists. The demand is there.

          1. Anna, First, I 100% agree that schools’ bookshelf audits are needed, and that there are profits to be made in diverse titles. Let me clarify, I did not mean to imply I publish only books written for White cis women. The publishing business is so complicated, and discoverability plays a big role in whose book is seen by readers and bookstore buyers. Traditional publishers look at sales trends, big data, who’s in the public eye, and gut instincts to determine what they think will sell, then measure the author’s ability to effectively promote and be received well. They don’t take risks. This group controls the bulk of what hits the market, and has the resources to market the book effectively, to make it “discoverable,” so the cycle continues. I am not a traditional publisher. I am a very small press, hybrid, niche publisher. My authors are all unknown, so every one is a huge risk for me. They almost all have vertical markets within which they speak or work, so their books manage to pay for themselves only after a couple of years. No sky-high profits. The printers, distributors and booksellers all make far more than our authors on each book, and we earn a very small fraction of what our authors earn on book sales, a typical scenario for any non-traditional publisher. Our books are not practically discoverable by bookstores such as yours without a considerable marketing budget and an exceptional book. All costs are shared, and the author’s limited marketing funds are usually more effectively spent on a very narrow audience. I don’t want to get off topic here, but my point was that all publishers have to mitigate risk and make decisions that will pay the bills and, in the case of large publishers, satisfy the investors or board. Money drives the decision. Getting in to the school market is a whole separate issue, equally challenging. So, the demand may be there, but we who are not among the Big 5 have our own wall to cross in order for our books to be discovered. Perhaps with a little brainstorming this obstacle, still extremely over-simplified here, could become a window of opportunity. Thanks, the conversation continues.

      2. I had the same question as Leslie, and I’m not sure this response addresses it. If “the issue is not with the number of physical people,” then why produce a report which is focused entirely on the number of physical people? If the number of people of different demographics is relevant – and it must be because that’s all the report tells us – then what is the desirable mix for the publishing industry? And if the desirable mix isn’t one that reflects society, as it currently seems to be, then what is it?

        1. Matthew Anderson is making an ideological argument. What he’s saying is not just that diversity is important, but it has to be the right kind of diversity. The diversity needs to penetrate into the highest levels of power, and also reflect a certain vision of writing, art, thrust, politics, etc. In other words, a publishing company could have a totally non-white boardroom, but if those people thought like Prof. John McWhorter, Prof. Thomas Sowell, and Sandra Cisneros, it would still be a big problem. And that, friends, is a problem.

          1. Yes, I would agree regarding the makeup of the board room. Staffers at the Big 5 report the younger staff is very diverse. Are the best of the bunch being groomed for those leadership positions? Would they even be considered? Do we have to wait another 20 years for that to happen? Not ok. This same issue is endemic throughout society, in business, the non-profit sector, our churches, our local governments.

      3. Since the masses of book readers are, in fact, white women, why is it a problem that they have the loudest voices in what is getting published? Do women need gay men to tell them what they are allowed to read?

        1. Well, I’d have to say, as a White cis woman, I have gained the most by reading books I never would have read had they not been required or recommended to me by my non-White friends, professors, or trusted book commentaries, simply because I never would have known about them. I would love to read a book recommended by a gay man, or any other person different than me, who thought it might expand my understanding. I would also have to say that we women are being spoon-fed a very limited span of book subjects and tones, as if our tiny little female minds could not handle deep or controversial subjects, or subjects that feed something other than our own vanities. Here again, lies a possible opportunity to turn our rudder.

    3. Your data is not complete. In fact, because we provided participants to choose Lantinx, as well as multiracial you need to compare the with the data what aggregates the category as “white alone” which stands around 60%. Most census watchers expect that number to drop below 50% in 2020 – if they are able to get an accurate count.

    4. That includes Hispanics. Once you remove Hispanics from the equation, 60.7% of the population is White and the remaining 40% is all other racial/ethnic consumer groups. As you can see, the publishing industry has a serious problem.

  4. Looking just at outcome data on this issue and claiming inequity is like looking in the desert for the keys you lost in the woods, and claiming you’ve done a thorough search. It’s like looking at the National Basketball Association, with its essentially unchanging 75% African-American to other player ratio, and claiming there’s ongoing inequity, when there are so many more highly qualified African-American players. It’s a misguided analytic. The only analytic that counts is opportunity data. Is there evidence that publishing companies e.g. are favoring for interviews/hiring the incoming resumes of white applicants over equivalently qualified black applicants, or is it that they’re just not getting many applicants from the black, Native nations, Latino, or other minority communities? Are they rejecting a higher percentage of equivalently qualified Latino folks compared to white applicants? If that’s the case, there’s an issue. If not, there’s no issue. Talk to human resources. Maybe Lee and Low can share its own data on this to start. You must have it, at least for people who come in to interview. What’s the percentage of white interviewees who are hired, versus e.g. African-American or Asian-American?

  5. Is there any research into the number of people in the populations you claim are under-represented who have applied to positions in the publication industry? without these numbers and proof that there is discrimination in the system you can’t claim that the system is discriminatory based upon it’s current make up.

    Based upon the logos used in your article, I could make an argument if I that if I were to walk into a restaurant and notice that there aren’t any disabled people working there that the restaurant is discriminatory. Not necessarily true as there are many other factors involved.

  6. In view of the RWA problem and the large market share, it would be interesting if an analysis was conducted that excluded the romance genre. Would the results be different?

  7. Thank you for continuing to track these numbers and advocate for improvement.

    I did have a question about the inclusion of South East Indian in the Asian category. I’ve never encountered the term before outside of references to Native Americans from the southeastern region of the United States. Could you please clarify? Thank you.

    1. Hi Greg,

      The breakdown of races for that question was originally based on US census categories and updated based on feedback/write ins we got after the last DBS. The survey company we worked with ran the choices by their survey content team, but it’s still possible we got it wrong. If so, we can definitely make a note to correct in the next version.

  8. Hannah E. Is there a slide that breaks out the children’s publishing numbers. Does that change things at all? I thought we were making progress. I believe low entry salaries and NYC housing costs play a huge role here too.

      1. I am concerned about the lack of men. We want boys to grow up as competent readers, yet women are selecting the books. I used to think my selections for review were a product of both the editor (me) and most of the reviewers being women. I made a real attempt to find books with male characters, etc. But if the editors, publicists, etc. are women, maybe they are not there in the first place to find.

Leave a Reply to Anonymous Cancel reply

Your email address will not be published. Required fields are marked *