Do Book Bans Work?

Censorship, the Streisand Effect, Statistics

On February 24, Illinois Representative Mary E Miller introduced House Resolution 7661, calling it the “Stop the Sexualization of Children Act.” This is the most far-reaching attempt by Republican lawmakers to ban books that contain “sexually oriented material,” moving beyond an ad hoc, local approach to call for federal legislation. As with all book banning efforts, this is very bad. It targets books about LGBTQ+ kids— trans kids in particular— and books about people of color, under the smokescreen of “sexually oriented content.” Publishers Weekly has called this “same ingredients, different recipe.” There’s been excellent analysis from Ron Charles on his substack, and Kelly Jensen at BookRiot. The Bill has been strongly denounced by the American Library Association. Freedom to Read advocates have pointed out that the language of HR 7661 is intentionally vague, in order that the resolution can be interpreted and applied widely, effectively assuring the potential for full and total censorship of children’s and young adult literature in public libraries.

This post is not about whether book bans are blatantly authoritarian tactics (they are). It is not about whether book bans infringe on children’s rights (they do). It is not about whether these books meet legal thresholds for obscenity (they do not) or whether they are appropriate for children (I won’t fight about that). It is not about whether Rep Miller is acting in good faith (she is not). I want to ask a more fundamental question about book bans. Are book bans effective? Do they actually achieve their stated purpose? Stated (dubious) goals of book bans aside, they have come to occupy such a significant position in our current cultural landscape. What sort of cultural work do book bans do?

A few years ago, I was awarded a grant from the Mellon Foundation to study book bans with a group of students at Temple. While discussing our research plan, one of my Deans asked an important question: don’t book bans actually help book sales? Doesn’t a ban or challenge raise public outcry, which, in turn, results in a sales bump as readers rush to show their support? She was referring to the then-recent challenge and subsequent resurgence of Art Spiegelman’s Maus, which (by my reading) could easily be banned or challenged under HR 7661. (The irony of banning a book about the Holocaust under the auspices of a Bill introduced by a woman who identified “what Hitler got right” on the steps of the Capitol does not escape me.)

This counter-narrative is known as the “Streisand effect.” Though often well-intentioned, it’s used to diminish the significance of book banning. It draws from the historic precedent of censorship challenges to adult titles, in which an obscenity trial anoints a text with the thrill of danger: Lady Chatterley’s Lover, Ulysses, Tropic of Cancer.

I explained that while a sales bump could happen in the short-term for high-profile titles, like Maus, I hadn’t seen any evidence to suggest that the Streisand effect was either widespread or lasting. And while being “banned in Boston” may have once been a badge of honor, the widespread and systematic nature of these bans presents a fundamentally different problem than the high-profile obscenity trials of the early 20th century.

That was a few years ago. Now, I have seen some evidence.

---

In 2024, a group of researchers (Marcel Goncalves, Isabelle Langrock, Jack LaViolette, and Katie Spoon) published a paper in PNAS Nexus that examined the efficacy of book bans. Do they actually work? Goncalves et al began with the PEN America dataset of banned and challenged titles in the 2020-2021 school year, and massively expanded it to include county-level population data, author demographics, descriptive information about book characters, and— the Holy Grail, for research purposes— sales data. (As a sidebar, I do not understand why this paper has not received more attention— it’s excellent.)

The authors begin by investigating topical and generic trends in the books that are most commonly challenged: book bans target books that emphasize race and ethnicity, on the one hand, and gender issues, on the other. In the 2020-2021 school year, the first year of the most recent surge in bans and challenges:

• 37% of books banned were identified as children’s books with diverse characters, including both LGBTQ+ characters and characters of color

• 22% of books banned were nonfiction about social movements and historical figures

• 10% young adult queer romance novels.

• 10% sci-fi and fantasy

• 10% women-centered fiction

• 7% mature, non-romance themes (i.e., drug-use)

• 4% unclassifiable

What’s more, regardless of the topic, banned books are much more likely to be written by people of color or LGBTQ+ authors.

• Gender: 64% of the authors of banned books are women, 3% are nonbinary, 29% are men

• Sexuality: 19% of the authors in the sample of banned books self-identify as LGBTQ

• Race and Ethnicity: 39% self-identify as people of color

We know, however, that publishing itself is discriminatory. There are more books published each year by straight authors and by white authors than there are books by LGBTQ+ or POC authors. Perhaps book banning is equal opportunity? Maybe it really is about the sexual content, and not about authorial identity? In fact, no. Relative to publishing as a whole, LGBTQ+ authors are slightly over-represented in banned books— meaning, LGBTQ+ authors are more likely to see their books banned. But POC authors are strongly over-reprsented, disproportionately more likely to see their books banned.

In fact, an author of color is 4.5 times more likely to be banned than a white author. But matters get worse when they consider book popularity— that is, just those books that are the most widely-held by public libraries nationwide. Amongst high-profile books, the odds that an author of color is banned is 12 times higher than a white author. This phenomenon, the authors argue, is driven primarily by authors who are women of color. Women authors of color are dramatically over-represented amongst banned books.“By banning children’s books,” the authors write, “women authors of color are effectively banned as well.”

This research accords with the narratives put forth both by book challengers (who protest the alleged incursion of DEI into the classroom) and Freedom to Read advocates (who argue that book bans discriminate against people of color— authors, characters, readers). This has never seemed like an especially strong argument to me against book banning— at least, not one that’s likely to convince someone who nominally supports banning books. No disrespect intended to the Freedom to Read advocates, who are doing good and hard and important work, but these statistics seem to suggest to me that book challengers are succeeding in what they set out to do: targeting people of color (though they would call it “DEI” or “CRT”).

Note that nothing about DEI or CRT appears in HR 7661. Perhaps “sexualization” is just a very wide umbrella, giving greater leeway to what can be banned and challenged. Or, perhaps, the conversation about objectionable content has moved on since the 2020-2021 school year, now targeting trans-rights, specifically. It would be interesting to re-run these data now.

---

But let’s get to the root of my Dean’s question: is banning a book an effective approach to censorship?

Imagine the best-case scenario for a book challenger. The censors carry the day: they succeed in banning a book from public schools or libraries. But their actions have far-reaching consequences, in this hypothetical scenario: they are so effective in their efforts to suppress a title that it simply fades from public consciousness, never to be heard from again. Its sales drop. The author retires into obscurity.

Now imagine the opposite. The Streisand effect works like this: any attempt to censor information backfires by driving more attention or interest to the censored material. Because many people want to show their support to a beloved book or author who was unjustly banned, they buy the book in droves. Following the banning of Art Spiegelman’s Maus in 2022, for instance, it was catapulted back to the top of the Bestseller list and its sales increased by 753%. Nothing about this situation is typical, of course. But is it possible that something like this occurs on a smaller scale with book bans? Do either of these scenarios occur with any sort of regularity?

According to the data: nope.

Goncalves et al use two indicators of reader “interest” in banned books: internet searches and book sales figures, testing both before and after a challenge or ban is issued. It is important to note, from the outset, that there is overall low interest in all titles, both before and after. (You may have heard: there’s a reading crisis on.)

They find a very small (1%) increase in attention after a ban in terms of Google searches— but not sales. They conclude that “book bans produce little change in the number of people who engage (or do not) with a book.” Book bans have almost zero effect on the public’s interest in a book.

They speculate that, perhaps, this is because challengers target books that are relatively low interest to the public to begin with. Cases like Maus or To Kill a Mockingbird are abnormal because they raise public attention about a beloved book— but also because it is less common for these books to be banned. Most banned books are unlikely to have crossed the radar of the average adult book reader.

One case in the data, I think, is worth reflecting on: Gender Queer: A Memoir by Maia Kobabe, one of the most frequently banned titles and hotly contested. Kobabe has received significant media attention as a result of eir book being banned. Gender Queer was a quietly published title with a relatively small print run, until it became the most banned book in the United States. The researchers note that there was a small increase in attention directed toward Gender Queer after it was banned. And while it’s impossible to tell what’s driving the attention— is it the NYT profile, or is it Moms for Liberty?— this is one case where the researchers did note a very small effect.

Yet, this shows us, further, what isn’t accounted for in this conversation around book banning (which, in fairness, the authors acknowledge). Regardless of why e intended to write eir memoir, or what e hoped for eir career, Kobabe has been forced by circumstance to become a Freedom to Read advocate. These methods don’t consider cancelled school assemblies with authors, or the career implications for authors (who are, again, statistically more likely to be women of color). These methods don’t show the extent of “shadow banning”— that is, when a librarian, fearing reprisal (or for her job), simply doesn’t order a book in the first place. These methods don’t consider other economic impacts, such as the loss of bulk purchases or books that simply aren’t acquired in an embattled climate. And this doesn’t include any anecdotal evidence about the harm done to kids, whose identity is now off-limits (at best), stigmatized, or vilified (at worst). Just because book bans aren’t effective at large-scale censorship doesn’t meant that they’re not effective in other ways.

But, in terms of challengers’ stated goals, the evidence shows that book bans do not work. Interest in banned books does not decrease after a challenge or outright ban is issued— not even for the most popular books. But, likewise, there is no “Streisand effect”: banned books do not benefit from the increased attention or experience any measurable sales bump. So, despite well-meaning efforts to find the silver lining, there is no upside here for books.

So what is going on here? If book bans do not work as a form of censorship, why engage in these tactics? (Aside from the obvious racism and transphobia.) The data tell us another interesting story.

Another way of asking “who is banning books?” or “what is motivating book banners?” is to look at contextual and situational data. Let’s begin: where are book bans occurring? We know, from PEN America, that most bans have occurred in Texas, Florida, Tennessee, and Pennsylvania. But what, really, does state-level data tell us, really? This isn’t very descriptive. Take Pennsylvania. I live and work in blue Philadelphia (home of Antifa Gritty, for goodness sake!) and it seems odd that a county not far from where I work, where many of my students grew up, should challenge books with such frequency. I’m sure the residents of Austin, TX, or Miami, FL, also experience this sort of cognitive dissonance.

To understand what’s really going on with book bans, we need to drill down to more local, county-level data. Goncalves et al examine a number of factors that could impact book banning: immigration patterns average income and education levels, rates of religious observance, racial demographics, and political participation. The most noteworthy difference between the counties that banned books and those that did not was the change in voting share won by Republican candidates. Examining the shift in votes between 2016 and 2020, counties controlled by Republicans, but with a decreasing Republican voting share (i.e., some people who voted Republican in 2016 then voted Democrat in 2020) were more likely to ban books than counties that didn’t. These are embattled counties, where the cultural and political landscape is changing— and crucially, where Republican strength and popularity is weakening. “In other words,” the researchers write, “Republican strongholds were not likely to ban books while counties with increasingly precarious conservative majorities were.” They are a form of political theater, a grievance escape valve, designed to rile up a shrinking voter base in order to maintain a hold on power.

In this context, the national reach of HR 7661 is especially interesting. It’s goal is national, not county- or school-board level, but it’s out of the same playbook. And I think it’s a signal of weakness. A rapist is in the White House, intent on distracting the electorate from evidence of his crimes, waging a wildly unpopular (and unconstitutional) war in Iran after staging a coup (also unconstitutional) in Venezuela, while unleashing ICE (defund now) on American citizens. This national book ban is designed to get the Republicans back on-side— to appeal to all of the moms who were horrified to watch Renee Good murdered while her kids were waiting for her, imagining the stuffies and snacks in their own glove compartments. Who imagined their five-year-olds in bunny hats, kidnapped and used as bait. To mobilize Republicans in support of the party that allegedly wants to protect their children, all while the actual people intent on sexualizing children occupy the White House.

I don’t mean to diminish the very real harm done to trans-kids, queer kids, and kids of color when books about them are banned. Or the harm done to all kids, and freedom of speech generally, when access to information is restricted. And I don’t mean to suggest that banning books doesn’t matter, or that its ineffectuality is a reason why we should just let it go.

It might not be about the books— but that’s why it matters more than ever.

---

Goncalves, Marcelo SO, et al. “Book bans in political context: Evidence from US schools.” PNAS nexus 3.6 (2024): pgae197.

Here’s what I’m reading, watching, thinking about, and teaching this week:

Reading: I’m reading Scavengers by Kathleen Boland, a delightful debut novel about a mother and daughter on a treasure hunt in Utah. And I just finished Brawler, Lauren Groff’s most recent short story collection. Groff is one of my favorite writers and I so admire her commitment to the short story form. I love her work for its smoldering rage, and the stories in Brawler (many of which I’d read first in The New Yorker) did not disappoint.

Watching: I love Slow Horses, Apple TV’s adaptation of Mick Herron’s Slough House novels, but it took me a little while to find my way to Down Cemetery Road, another Herron adaption starring Emma Thompson and Ruth Wilson. This show is taking me a long time to watch because I must routinely pause to gush about Emma Thompson’s presence and how Ruth Wilson refuses to get any work done on her perfect weird beautiful face and omg did I mention Emma Thompson. (My poor husband.)

Thinking: I’ve been thinking a lot about bestsellers this week and the stories that we tell about commercial fiction, along with the news of literary agent (and author) Al Zuckerman’s death.

Teaching: My Young Adult Literature class is reading Feed by M.T. Anderson this week! Feed was written before the age of generative A.I., which makes its critique all the more prescient. I’ll be sharing with them this research, about the effects of prolonged use of ChatGPT on your brain.

Enjoying textCrunch for its blend of data analysis and cultural criticism? You can pre-order my book, Middlemen: Literary Agents and the Making of American Fiction, coming out on April 28! You can snag a pre-order from my favorite local indie, Inkwood Books, from Bookshop.org, from Amazon, or directly from Princeton UP.