What does banning a book even mean, when is it appropriate, and how should we decide which books to “ban”?

Gavin Newsom, while stumping for Biden in Idaho, recently visited a “banned bookstore,” presumably a purveyor of all those books that dastardly DeSantis and his conservative cronies have been removing from school libraries. I, for one, am glad that there exists a bookstore from which one can easily purchase banned books, as a fair bit of my reading list runs afoul of the current regime, though somehow I doubt that I can acquire any actually dangerous books from them.

I suppose they can furnish me with a copy of The Anarchist’s Cookbook, whose detailed descriptions of the manufacturing processes of homemade weapons may prove invaluable, or perhaps they have a back issue of Inspire with classic articles such as “How to Build a Bomb in the Kitchen of Your Mom.” If it’s good enough for the Boston Marathon Bomber, it’s good enough for me. I’m sure the bookstore contained such classics as the Little Red Book, and perhaps some of Ho Chi Minh’s writings as well. I may also want to pick up The Camp of The Saints — most critics consider it xenophobic dribble, but what do they know? And speaking of racially insensitive books, what about the straightforwardly titled Race War In High School? In memory of the recently departed, I’ll check if they have a print copy of the Unabomber’s Manifesto. Finally, one wonders if they have that most archetypical of banned books, the autobiography of a certain Bavarian artist with a funny mustache who I do not dare to name. 

I should probably stop googling these books before I end up on any watchlists. 

In all seriousness, I doubt this bookstore had these books in mind when they discussed “banned books,” and that illustrates my point. When people refer to “banned books” they rarely are discussing, or evening thinking of, books like the above, which are actually difficult to obtain because of their controversial, one might even say dangerous, ideas. Rather, they are referring to books which are still simple to obtain, but have merely, for one reason or another, been removed from one or more public libraries. This does not mean that the removal was right, just that we need to be aware of what exactly we mean when we discuss “banning books”. 

What is Book Banning?

It can be a lot of fun to posture about being opposed to “banned books,” but this has a way of obscuring the key question which we actually need to discuss: What do we actually mean by “banned books?” What heuristic should we use to decide which books to ban, and by what heuristic do we decide what level of banning is appropriate for a particular book?

First, “none” is a perfectly acceptable answer to the above question. You could believe that no book, however dangerous, should be in any way legally or socially censored, but I doubt even self-described free speech absolutists would go that far. In order to understand why, we need to discuss what we mean by “banning” books.

I have elsewhere made a distinction between legal and social censure for speech, between what the law prohibits and what with which most people will generally refuse to associate. Similarly, we can distinguish between legal and social book banning. At a first glance, legal banning of books consists of laws restricting their publication or widespread distribution, interpreted so as to exclude reasonable enforcement of copyrights (where the copyright holder faces no government pressure to restrict the distribution of their books). There are plenty of books which would be banned for non-speech related reasons, such as a hypothetical book containing the leaked CocaCola formula or the nuclear codes (for If I wrote a book with the nuclear codes in it, it would be illegal to sell, so such a book would be banned) but I think it is still useful to analyze them as banned. As I said, even free speech absolutists will want to ban some books.

(One question which follows from the above is, if copyright law prevents us from printing an unabridged copy of Harry Potter and the Sorcerer’s Stone on thetexasorator.com, why is that not a violation of freedom of the press? After all, it is content based regulation of newspapers. Articulating the answer is left as an exercise for the reader.) 

Social banning is, in addition to shaming people for reading banned books and refusing to associate with them, refusing to carry the book in private bookstores, refusing to include the book in private school curricula or literary projects where it would otherwise make sense to place it, and refusing it access to publishing infrastructure, or boycotting any books published by publishing firms which print the banned book. There is, admittedly, somewhat of a spectrum here, with a particularly fuzzy line between “we will not sell your book because it is poorly written or will not do well” and “we will not sell your book because society has a moral or social objection to people reading it.” A good rule of thumb is if a the book increased tenfold in popularity, and the would be censor changes their mind, then they are merely reacting to its poor quality or unpopularity, whereas if they would hold their ground no matter how large the market was for it, then they are engaged in social censorship. 

But here there is an important gray area between social and legal censorship — public schools and public libraries. The choice whether or not to stock a book in a private bookstore is a social one, but what about stocking a book in a place where books are paid for by the government? For the government, or its representatives, to refuse to include a book in a public library or school is not entirely social censorship, for it is a decision made by the state, not private and voluntary action, yet it is clearly a different thing from the government outlawing, or even discouraging, the private consumption of books. 

We can call “banning” books, where banning means removing them from a government funded institution in a way that does not threaten their private consumption, institutional banning. I understand the objection that this is not banning books in the denotative sense of the word, but in common parlance removing a book from a school library has been described as “banning the book” often enough that I will use the terminology for the sake of convenience. So how does banning a book from a public library work?

When is Excluding a Book from a Library Banning It?

For simplicity, we will start by dividing books into three categories: books which are included in a given school library, books which are not in the school library not for reasons of banning, and books which were banned from the library. The average school library fails to contain the vast majority of books in print, and yet it would be untrue to say these books are being banned. I doubt many school libraries contain Houston Cooks: Recipes from the City’s Favorite Restaurants and Chefs (the first result on Google for “Top Cookbook 2022”) yet I do not believe anyone would argue that the book has been banned. 

Rather, any given book which is not included in a given library is excluded due to some combination of the following reasons: 

1. The librarians (or whoever chooses to stock this particular library) have never heard of it;
2. The librarians do not believe there would be enough demand for or use from the book to justify stocking it;
3. The librarians have a limited amount of money and shelf space which preclude stocking the book; 
4. The librarians cannot obtain the book easily enough to justify attempting to obtain it at all;
5. The rules or their bosses prevent the librarians from stocking the book; 
6. The librarians have some moral or social objection to the library’s patrons reading the book, or believe that it would be harmful for them to do so.

If a book is excluded for any reason under the latter reason, or some reasons under the penultimate one, then I would consider it to be banned in regards to that particular library. 

Simple, right? Well, not so fast — the third reason (the library has limited resources and the book is not worthwhile compared to the next best option) and the sixth reason (the book is dangerous or morally suspect) overlap. Someone who believes a book has a positive moral message may believe the marginal utility of stocking the book is quite high, whereas someone who believes the book has no moral value, but would nonetheless not deny anyone the ability to read the book on that account, will nonetheless rate the marginal utility of the book as low and would prefer the library use limited shelf space to stock a book with a more widely agreed upon mediocre marginal utility. 

The reason for this, if you have not spotted it yet, is that being a public librarian is an inherently political action. 

What Makes a Book Valuable? 

Consider, for the sake of simplicity, the case of a self funded librarian, perhaps one who runs a charitable library or personally stocks a local Little Free Library. Surely the librarian would seek to stock the library with the books she believed would be best, but how would she determine this? I would say that the three criteria she would use to determine which books had the highest marginal utility would be enjoyment, usefulness, and citizenship. She might choose books based on how much she thought readers would enjoy them, or on books which she thought it would be helpful for them to read — perhaps because they would expose readers to new ideas, or they would be psychologically helpful. Or she might choose books designed to promote social or moral change — books which advocate, openly or subtly, Veganism or Christianity or LGBT Acceptance or Objectivism or some other idea, or are designed to make people braver, compassionate, or otherwise a better citizen. Usually, the marginal utility calculation includes all three factors. 

With the exception of the matter of enjoyment, which is a merely factual question about how much readers will like a given book, all of these questions are potentially political questions. The idea of self improvement implies a standard to which someone ought to reach, so that they can “improve” by increasing in closeness to it. A book which encourages readers to get a job and move out of their parents basement, either by exhortation or example, is valuable for self improvement if you believe it is good to be independent and gainfully employed. A book which shows the reader they ought to be kind to people of all backgrounds will, if effective, make the reader a better citizen — but only if you believe people being kind to others is good for society. In these cases a supermajority of 21st-century America would theoretically agree with the moral elements of the book, so much that it would not usually parse as political, but within any book that seeks to change the reader lies the seed of political conflict. Once an idea becomes such that the future different positions are philosophically, socially, or economically salient to the future of society, and different positions become broadly supported, the issue can be considered political.  

The seed only germinates when the change is one about which there is a widespread difference of opinion in society. If a book promotes a change in the reader, either to improve them as a person or a citizen, and some people think the change good, and other people think the change bad, and both sides agree the change is important enough to care about, then the book can become a source of political controversy. The reason why the controversy can be considered political is that it is a question about what values society should encourage, what ought statements America ought to endorse, the utility function we ought to optimize, the idea of the good the people ought to pursue, or whatever other phrasing you prefer.

In terms of the private sector, the classical liberal answer — which America has largely settled on — is that people can read, or not read, whatever they please with a few caveats, so long as they do not try to push it on others. You may believe that everyone should read, or not read, the Bible, or How to Be An Antiracist, or The Fourth Political Theory, and you are free to read it and encourage others to do so, but they have the freedom to ignore you. You can buy the book from anyone who will choose to sell it to you, but no bookstore shall be required to furnish you with it. Although doing so wantonly is contrary to the spirit of a free society, you are also free to not associate with anyone you please based on their reading tastes. So much for the private sector.

Who Really Chooses Government Funded Books?

The difficulty with applying this to the public sphere is the question of who makes decisions. In private life, we each make decisions for ourselves and our families. And private firms have a clear decision criterion of “sell whatever books makes the most profit.” Nonprofits engaged in literary activities are usually controlled by people with a common vision of the good, at least in regards to the relevant books, and any divisions which cannot be talked through can be solved by some members leaving to start their own group. But in government, and in the libraries and schools it controls, who has the right to make the decision about which books have a marginal utility sufficient for their inclusion? 

“Government,” it is said, “is the name for all the things we do together.” There is an element of truth to this, for it is also said that the government “deriv[es its] just powers from the consent of the governed.” Whatever is done by the government is done in the name of, and by the power of, the citizens of the nation. While our republican form of government is designed to allow the government to sometimes make decisions which are in the national interest, against the temporary opinion of the majority of voters, everyone has the right to have an opinion on these decisions, and use the legitimate levers of government to push for it.

Public schools, as agencies of the state government, are ultimately answerable to the citizens of the state in which they are located. How exactly power is delegated varies by state, but generally the voters, represented in the legislature, create general policies about what schools should do and delegate implementation of the plan to high level bureaucrats in the executive branch. Meanwhile, local school districts, controlled by boards elected by the local community, set local priorities for schools, which are delegated to administrators, who in turn delegate to teachers. While each level in the hierarchy must exercise its discretion in achieving the system’s goals, it ultimately does so at the pleasure of the levels above, to whom it answers and from whom it derives its just powers. 

(Note that our education system is influenced at the local, state, and federal level in different ways, meaning the relevant voters who must be considered in any given situation will be different. Decisions made by local school boards should be made in accordance with the expressed utility functions of local voters, those by state legislatures, state voters, and those made by the federal government of all American voters. For the sake of clarity we will ignore this complication for now, and just round this off to the local level, where most library related decisions are related, and the state level, where the broader library policies are set.)

On Stocking Books, What’s Our Algorithm? 

We have all the pieces laid out now, so let’s put them together. To relate and condense the problem: the librarian must, with limited resources, choose the highest utility books to stock in the library, where the marginal utility of a book can be approximated as: 

Number of students who will read the book * (enjoyment they will get from the book + personal growth they will gain from the book + social externalities related to them reading the book,) – costs of stocking the book

Or: Utility = readership * (enjoyment per reader + usefulness per reader + citizenship per reader) – stocking costs

(Note that this is different from the marginal utility of the book, which would be given by the sum of the utilities of all the books in a particular library including the book, minus the sum of the utilities of all the books in a particular library excluding the book. This is to take into account synergies (it makes more sense to have the The Two Towers if you already have the Fellowship of the Ring) and cannibalization (if you already have 10 books on zoology and 0 on geology, the 11th zoology book may have less marginal utility than the 1st geology book, even if the former would likely have a higher utility) between books. For the sake of simplicity I will ignore these, and pretend that utility = marginal utility) 

This is somewhat of a technical problem, in which the librarians presumably have superior expertise (and if they aren’t experts on which books which kids will enjoy and profit from, why are they librarians?) However, the question of “what makes someone a better person” and “what counts as a positive cultural externality” are political, because different political values will lead to different answers. So how should the librarian handle a situation where society has no consensus on a book, where some would say the book has high enough utility to breach the bar to be included, and others that it is almost null, or actively harmful? 

I would maintain that, in a society with a republican form of government in which schools are paid for by the state, the librarian has a duty to make stocking decisions using the theoretical aggregate utility function of voters. 

What this is not saying: librarians should just rate books by what the majority of voters want, because some do not care, as much as others, which the aggregate utility function accounts for, nor on their aggregated preferences, because most voters will not be well read enough to know the utility they would give any book if they had the time to understand it. Thus, the need to appoint librarians to figure this out for us.

Of course, no one can actually calculate the utility of stocking a given book, and even if they could it would be far more involved than using the above equation. In practice we rely on approximates like “reading 1984 is an enjoyable experience for many people, can help the reader think more critically about language and understand more cultural references, and, by decreasing the risk of embracing for certain sorts of totalitarianism, has positive social externalities, so it is probably a good book to have;” or “Harry Potter is a fun book to read for most people, gives exposure to cultural motifs it is helpful to know, and does not have many social externalities one way or the other, so it is probably worth stocking.” All of these are rough utility calculations, and they are the sort librarians make with every order of new books. Thinking of things in terms of utility, while overly simplistic, is a helpful tool to analyze what is actually happening.

How Do We Actually Calculate Utility?

Now that we have a way to actually decide which books to stock in a library, we can circle back to banned books. The controversy has centered around Florida, but has spread to several other red states. The issue is that citizens and politicians have been pushing libraries to remove certain books which are deemed low marginal utility by those doing the pushing, but high marginal utility by many school staff, as well as other voters, parents, and politicians. Coincidentally, these two groups line up almost perfectly with conservatives and progressives respectively — who would have thought?

Note that the next paragraphs get a bit more in the weeds, so readers who are fine just understanding the general idea can skip down to the next heading.

Consider a toy model: suppose a school ran for the sake of two families, who agree on the marginal utility of adding most books to a library, but disagree on one particular book — one thinks its value is near zero, and the other that it well surpasses the threshold for inclusion — how should they resolve this? They could:

1. Take the average of their respective, honestly reported, utilities, and see if it makes the cut, 
2. Take the weighted average of their utilities, weighed by how much each cares, 
3. Take the median of their utilities (in this case the result is identical to A, but in broader society it would have the effect of reducing the effect of extreme values and outliers, for better and for worse, 
4. Weighted median, 
5. Minimum threshold, where it has to meet a certain threshold (say, be in the top quartile of marginal utility) for both of them to be included (when expanding this idea to broader society it will probably have to be a supermajority). 
6. Minimum inclusion threshold, where a book being in the top quartile of utility for one person is sufficient for its inclusion.  

For brevity, I will not elaborate on the arguments for each, but suffice it to say any intentional weighing misses the point of representing the preferences of everyone, and any level of real apathy, as opposed to not understanding the issue, is already baked into the utilities (though in practice the people who care more will have more influence, leading to a weighing effect anyway.) A is the best in theory but is completely unworkable on a large scale in practice, C is less ideal than A but avoids some of the difficulties, and E is partially workable at the cost of giving any faction which wants to be a jerk veto power over the entire system. F Is also partially workable, but is also too easy to game. 

Expanding these systems to a larger scale might look like this: You can estimate what society on average would like to do, as measured by what the politicians they elect would like — though this would require them to lay out a syllabus of preferred books more granular than they might wish. You can ask a randomly selected citizen what they would want done, try to figure that out, and just do that, and in another society either of those could work. In fact, in our society such a plan does work for the majority of books, which is why most school libraries will have War and Peace and not Goodnight Moon, but this breaks down when applied to politically controversial books, as both sides have an incentive to distort the issue to make it look like they represent the true embodiment of the will of the people. You could get around that by using a system like “each faction gets veto power over a certain number of books” or “we add the books everyone agrees are good, and each faction gets a certain number of slots to add whatever books they want.” Which sounds appealing, but would not only make the political element in libraries more explicit (which is bad, because then people wouldn’t want to read the “conservative” or “progressive” books, even if they would otherwise be perfectly fine with that). Another problem would be internal disagreements within each “faction” about which books it wanted to prioritize, and who would do the prioritizing. 

The Solution

But the biggest problem is that this system blows the problem entirely out of proportion. We started this analysis by noting that the “banned book controversy” was really a “disagreement over marginal utility of including a handful of books in a library controversy.” Any system which solves this problem by explicitly labeling books “progressive” or “conservative,” leaves us worse off than the current muddle — not to mention labeling books with such hamfisted poltical labels is an insult not only to good books everywhere, but to the very spirit of literature itself.

Fortunately, there is a solution which lets us adjudicate such cases, without either breaking the school library system or allowing one side to stock whatever books it desires with no input from anyone else. I will elaborate on that in part two of this series. 

For now, though, we have seen that the question of banned books is merely a sub question of how to select books for public support in the first place, and seen the theoretical solution to this question in terms of utility functions. The rest of this series will elaborate on how this works in practice, and the implications for the culture wars.