A diversity audit takes inventory and determines what's in a collection and what areas need to be better developed. It yields concrete data. This type of audit helps put the science in library science. In doing a diversity audit of my collection, my goal was to learn what percentage reflected something other than the experiences of straight, white, and non-disabled males, and to fill the gaps. I wanted to verify that I was doing a good job of transcending my worldview and internalized biases. I was not. 

As a 45-year-old white woman, I am the dominant demographic in librarianship, though definitely not in the world at large. For a long time, I took for granted that I was building inclusive collections that would best serve my teens. Conducting a diversity audit made me look honestly at my collection development and forced me to step outside my comfort zone to accomplish the work I always claimed to be doing.
 

Initial research

Establishing a procedure for a diversity audit was challenging. There were not a lot of research or examples online. I found information for conducting a classroom diversity audit, and I discovered a resource about a Montessori classroom doing a toy diversity audit, which helped me establish basic practices.

Before beginning an audit, you need baseline data. I examined  sources such as the U.S. Census to get a better idea of the world's population, and I used  We Need Diverse Books, Lee and Low’s Diversity Baseline Survey, and the Cooperative Children’s Book Center data on children’s publishing to understand the publishing landscape. The truth is, the state of diversity in publishing is not good, so those tasked with collection development need to work extra hard to ensure inclusive collections.
 

The process

With my research, I developed a system. My collection contained 2,000 items at the time. I printed a shelf list and imported the list into an Excel spreadsheet. I researched each author and made a notation as to whether the title involved characters who were something other than male, white, nondisabled, and heteronormative. 

The initial research took months. I scoured the Internet for information about authors and titles. I examined a variety of published diversity lists (which I turned into a “Diverse Reads” readers’ advisory notebook).

After annotating each title, I determined percentages. To do that, take the total counted for a part of collection and divide it by the total number of books in the collection. For example, if there are 200 titles with LGBTQIA+ representation, divide 200 by the total number of books in the collection (let’s say that’s 2,000 books). In that scenario, 10 percent of the collection includes  LGBTQIA+ representation. That number can be compares with U.S. Census data to decide if its satisfactory. Research from a 2016 survey by the J. Walter Thompson Innovation Group found that 52 percent of Generation Z (people between the ages of 13 and 20) identified as something other than straight, meaning that more than half identify in some way as LGBTQIA+.

When there are areas with low percentages, use that data to target ordering to diversify your collection. For example, I served a very large and vocal LGBTQ group of teens, so I always assumed I was doing a good job of purchasing LGBTQ titles. What I found, however, was that less than three percent of my collection was LGBTQ-friendly. With that information, I sought lists of award-winning titles and yearly “best of” for the last couple of years.
 

When a complete audit is impractical

With a large collection, auditing the entire collection may be near impossible. Instead, create random samples from each section, and audit those. If a random sample is done correctly, it provides valuable information. Remember that one of the goals is to help those involved in collection development determine their own internal bias and pinpoint areas they may need to be more consciously aware of when selecting materials.

Another type of diversity audit can be done with subject headings. This works for some areas better than others, however. For example, most LGBTQIA+ literature has a subject heading indicating that it falls into this category. However, not all books will have a subject heading that indicates, for instance, a character is black or Latinx. We can work to expand and add subject headings, but that means asking our catalogers to add those subject headings or notes. Kirkus now includes a mention in their reviews about the ethnicity of the characters, and these notes could easily be ordered into our catalogs, but this will take some intentionality and additional coordination.

There is also what I call a “reverse diversity audit.” Take a list of titles on a traditionally marginalized people or group and determine how many of those titles on the list are in your collection. For example, look at how many titles from the yearly Stonewall and Rainbow lists are in your collection. This method isn’t as concrete as a true diversity audit, but it still offers useful data.

If you’re doing a reverse diversity audit, I recommend creating an annual cycle to keep you on track, like the weeding calendars that many librarians maintain. In January, you may check for Native American representation; in February, look at Asian representation; in March, focus on LGBTQIA+ representation; and so on. That way, each month you take a targeted area and make sure it is well represented in your collection. This doesn’t mean you aren’t ordering titles throughout the year for each area of representation; it just means that you are going in once a year and holding yourself accountable.
 

Collection maintenance—and a few caveats

After that initial diversity audit in 2016, I began doing mini-audits on each book order to continue to develop an inclusive collection. I print each book order and make notes in the margins. I then use the same basic formula to determine the breakdown of my book order before submitting. I also input that data into my spreadsheet to keep an up-to-date look at the diversity of the overall collection.

Because we know representation matters, what we look for is always expanding. In addition to examining ethnicity, religion, sexuality, and disability, I expanded my criteria to include topics such as body acceptance, socioeconomic status, and mental health. Don't be overwhelmed. There are a lot of other librarians, authors, publishers, and readers out there doing this work as well, so use them as resources. We Need Diverse Books is a go-to resource, but a Google search can also yield recommended booklists on a wide variety of topics as well.

Remember—not all representation is good representation. A book may have a black main character but might be a problematic slavery narrative that reinforces stereotypes. Or it may be a book with Jewish representation but is another Holocaust book in a collection with many. There is nothing wrong with books about slavery or the Holocaust, but all titles cannot center on the same scenarios. Jewish children and teens need representation that doesn’t remind them that they were, and are, so reviled by some that they were literally marked for genocide. And black children and teens need to see themselves in books that aren’t focused on slavery and don’t include stereotypes. We’re looking for more than representation—we need healthy representation that affirms our teens’ experiences and helps people outside that group to question and challenge harmful stereotypes.

For more resources and information on conducting a diversity audit on your library collection, check out this series of posts on the “Teen Librarian Toolbox” blog.