Bryan Caplan  

A Waste of Paper?

Don't apply AD to long run gro... Free the yuan...
Last week I walked the stacks of GMU's Fenwick Library with my elder sons.  Their presence was clarifying.  As we three perused the books, my main emotion was embarrassment.  I'm an academic.  A university library is supposed to be a warehouse of great thoughts.  But the vast majority of the books seemed literally indefensible.  Lame topics, vague theses, and godawful writing abounded.  Titles withheld to protect the guilty.

The indefensibility was clearest for the humanities, where many books seemed doubly pointless: A History of Literary Criticism of a Minor Genre.  But most of the social science was in the same ballpark.  90% of the books screamed, "If writing stuff like this wasn't a ticket to tenure, no one would write it."

You could dismiss me as a philistine, but that seems unfair.  On any conventional test of cultural literacy, I'd at least make the 95th percentile.  I have very broad interests.  I've heavily explored economics, philosophy, psychology, political science, history, sociology, education, genetics, evolution, and the history of music and film.  The fact that most of the books failed to minimally pique even my interest reflects poorly on them.

Could the problem be my lack of expertise?  Perhaps if I were an expert on Emily Dickinson, I'd see the value in most of the volumes written about her.  But I doubt it.  The obvious test: Do I see greater value in the median work in the areas where I have attained expertise?  Only to a slight degree.  And that small effect is readily explained by selection bias: You should expect me to be more favorable toward academic literatures I chose to carefully explore.

My question for anyone who's wandered the stacks of a university library: Do you really think that most of the books were worth writing?  What fraction of the collection literally isn't worth the paper it's printed on? 

Bonus question: What lessons do you draw about the social value of academia?

COMMENTS (31 to date)
James D. Miller writes:

You could argue that the process of writing an academic book makes you a better professor, or at least sends a strong signal that you are qualified to teach a topic related to the book.

Eh–the university library was one of the things I miss most about college/grad school. (I dropped out of philosophy grad school and am now a software engineer.) I definitely get the impression that a lot of academic publishing is resume-padding, but do university libraries actually buy much of that stuff? I have fond memories of learning the title of 1 recommended book on a subject I was interested in, getting the call number, going to that section of the stacks and then browsing every title in the general vicinity. It seemed like a good way to learn things at the time.

yarbel writes:

One metric is citation counts. While a more sophisticated metric would be needed for a more satisfactory answer (self-citing, citing superficial facts not ideas, etc.), a low cited book is a good indication of it not-being-worth-writing. In my field (law), my sense is that books generally fare badly, but that may be the effect of papers being easier to obtain online than books and some other quirks of the field.

This brings up another really important question. Are books cited more than papers, weighting the fact that books "cost",say, three times more to produce? Should we expect three times the citation count? For that matter, make it socially acceptable to cite to blog posts, and see whether papers are even worth writing.

James writes:

"Do you really think that most of the books were worth writing?"

Yes, I think. I don't think there is a possible world in which the good books get written and the bad books don't, so the production of books that are not worthwhile in their own is "worth it" because this is just a cost of producing the better books.

It would be easy to come up with ways to produce less junk by avoiding the production of anything that deviates very much from existing material known to be of high quality. However, this would also prevent the production of many good books which deviate from earlier work. The optimal amount of filtering would be where the benefits from the good books are just slightly greater than the costs of producing all of the books. Since the benefits of good books are far in excess of the cost of producing books, it is far better that a wide range of books get written, even those on topics with little promise.

A. M. writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Peter H writes:

I think James has the right argument here.

Bryan, you sound like a communist looking at the 1000 types of suit jackets on the shelf at a department store, and thinking what a waste it is that all these ugly and ill-fitting jackets are made, thinking wouldn't it be better if there were less waste.

And yes, a lot of them are ugly and ill-fitting to you, and some are objectively terrible. But it is only by having a dynamic market where people can produce terrible work that the good stuff gets produced.

phil writes:

I'm not really an expert on the economics of academic book publishing

but... I suspect the value from the worthwhile books more than covers the costs of the worthless ones

though I buy that's its plenty possible that some future digital alternative will prove better

(I find it suspicious that books tend to be of relatively standard length [usually at least 150 pages or so], despite the fact that they don't all have standard depths of arguments to flesh out, I've opened a lot of books that's I'm pretty sure could have taught me exactly as much in about 40 pages of text)

Josiah writes:

I believe Sturgeon's Law is applicable here.

Paul Geddes writes:

It's a market failure, so tax books a lot. Wait, wrong website. It's a government failure, so de-regulate academics.

Richard writes:

I know a bit about political science and law. In general, Bryan is pretty much correct.

Law lies on the more useless end of the spectrum. It would talk all day to explain in enough detail why.

A lot of political science research has a point, but it's written in too many words. For example, every paper on a topic has a literature review. This is simply wasteful. It would be just as easy to have one standard literature review everyone refers to. Sometimes a scholar will turn an article into a book. You make your work 10x longer, and at most add 30% new information. Similarly, an abstract often tells you 80% of what's important in a paper, unless you want to replicate the results.

International Relations spends its time coming up with more game theoretic models that treat states as unitary, rational actors, without a shred of evidence that they are.

My own opinion about academics in the humanities and social sciences:

1) About 10% go into it because they're intellectually curious and want to add some kind of knowledge to the world. Luckily these people have a disproportionate amount of success.
2) About 30% have some kind of agenda they want to pursue. See just about every women's studies and African American studies professor. They gravitate towards those fields that require the least empirical rigor. Unfortunately, these people also tend to have a lot of motivation, so they succeed too.
3) About 60% just go to graduate school because they like the idea of becoming a professor or because they want to stay in school forever because they're afraid of coping in the real world. I think they lack the motivation to come up with projects on their own and follow through with them, so they're least likely to succeed. However, there are a lot of these people, so when they do succeed they produce most of the useless research Bryan complains of.

Butler Reynolds writes:

This is not at all like the marketplace in that these things are subsidized heavily.

If you subsidize something (even garbage), you'll get more of it, right?

khodge writes:

Bryan, have you wandered other universities' libraries? I loved wandering the stacks in my college and graduate days and do not recall experiencing a similar frustration. I never went to a "research university" and it seemed to me that what I found in the stacks was generally useful and readable.

Nathan Smith writes:

You're exactly right, Bryan. The vast majority of academic publishing isn't worthwhile. Its production is motivated by rent-seeking. People write it not because they care about truth but because they want tenure.

It's a stable equilibrium because universities have no good way to assess teaching (student evals are worse than useless, see here: so they rely on scholarship as an indicator of professor quality.

In effect, what you're seeing on the library shelves are reams and reams of job applications.

The best defense of the system is that maybe there really isn't a better way to assess professor quality. But I suspect standardized testing of both professors and students could provide a superior signal of quality at much lower cost. And then the excessive academic publishing would disappear very quickly.

Among the beneficiaries would be science. The presumption that everyone should know the literature enforces a regime of stultifying over-specialization, since editors like to reject because "author doesn't know the literature," and the literature is so excessively large that it's impossible for anyone to know it unless they're narrowly specialized. And so wheels keep getting rediscovered (ineptly) in each little silo, since there's too much noise for people to listen to each other.

Adam Smith could never have published *The Wealth of Nations* in today's academia.

But I do disagree with one point in the post. The title. The cost of the paper is pretty negligible. Most of the mediocre over-specialized scribbling that clutters academic libraries probably *is* worth the paper it's printed on. It's just not worth the much more valuable time of readers, still less of the authors who wrote it.

Nathan Smith writes:

By the way, this is an interesting example of what I think is a broader phenomenon, namely, some things are only likely to be done well for the sake of *intrinsic rewards,* and they get spoiled when the *extrinsic* rewards become too appealing.

For example, consider a judge. One hopes that the judge seeks to discern the right and wrong of the cases that come before him because he or she cares about justice, and I think judges often really do care about justice. If the judge doesn't care about justice, but only about his own interest, he or she might still seek to discern justice for the sake of self-interest, if, for example, bad rulings will be appealed and overturned, causing the judge humiliation and reduced chance of pay raises or promotion. But bad rulings will be appealed and overturned only if the court of appeals cares about justice. Who guards the guards? It's surely impossible to design a judicial system where where judges will rule justly even if they don't care about justice at all, simply because their incentives are so well-designed.

I think the same principle applies in academia. Worthwhile scholarship occurs when you have communities of people who care, above all, about truth. Even when there are no monetary incentives are involved, other corrupting motives are at work: the desire for fame, for vengeance after losing an argument, for cheap popular appeal; the desire to flatter the powerful, or to play the prophet. Introducing monetary incentives for scholarship can heighten the incentive to do good work, but it also tends to corrupt the systems that evaluate the work that's produced. It's a delicate problem, and institutional arrangements to solve it fail far more often than they succeed, which is why intellectual history seems to consist of flashes of genius (5th-century Athens, 13th-century Paris, 18th-century Scotland) punctuating long ages of forgettable mediocrity.

Contemporary academia certainly does *not* represent one of those rare moments when the incentives are personalities are aligned to induce a flowering of valuable intellectual production.

John Samples writes:

I once ran an academic press. Let's ask whether the benefits of these books exceed their costs. We can do some rough calculations about per-unit cost. An average 250 page academic monograph would have a print run of 400 copies; the per unit publishing cost of each of those 400 would be about $140. (The retail price may or may not reflect that unit cost; check out the cost of an Elsevier monograph to get the actual price plus profit on a monograph). To this you might add the cost of writing the book. Let's say - this cost will be variable - that a professor making $90,000 annually devotes half time for 2 years to writing the book. So that's another $112.50 of cost to add to each unit. So the question might be: are all the books Bryan saw worth $252.50 each?

The libraries bought them at retail, perhaps as low as $70 or $80. So their purchasing tells us little about books costing $252.50. (Library purchasers also buy the books for someone else, and in any case, data from libraries suggest many of the books are rarely used even though they may be read at a minimal cost, i.e. going to the library). Finally, I learned from my book publishing days that individuals, even those who were the target audience for a monograph, were almost always unwilling to buy it even at the subsidized retail price mentioned above.

Students, donors, or taxpayers are subsidizing these books. Few if any of these people actually use the book though it is argued that they benefit from the outlays. How often does the benefit to "society" equal $252.50 per book?

Phil writes:

"An average 250 page academic monograph would have a print run of 400 copies; the per unit publishing cost of each of those 400 would be about $140. "

huh? walk me through why it costs $140 to print 250 pages

I get that they're not operating with the scale of say John Grisham,

but that seems several orders of magnitude off

Philo writes:

"What lessons do you draw about the social value of academia?" None. Beside the average worthless stuff there are the few outstandingly valuable productions--probably so valuable as to outweigh the costs of maintaining the overall system. The question, then, is: To get these valuable productions do we need to maintain the entire expensive system? I think the answer is negative, but that doesn't follow merely from the fact that most academic books and articles are worthless.

James Hanley writes:

I would say not. My grad mentor, who has published extensively and in multiple disciplines, has never written a book, and his explanation was that he'd never had a book length idea. In my view, most academics haven't, not through personal shortcomings but because most ideas are not book-length ideas. They can be expressed clearly with much greater brevity, and the rest is just further extensions of the idea, yet another example, etc.

Of course a lot of people don't have particularly novel or interesting ideas, and are just applying a standard formula because it's easier than thinking clearly or inventively. But success in their occupation is measured by output quantity more than quality, so there's little choice but to play the game.

Urstoff writes:

Worth writing compared to doing what? When I was in grad school, I could see the value of 90% percent of the books in my field.

John Samples writes:

"walk me through why it costs $140 to print 250 pages"

I should have been more precise. By "publishing cost" I meant all the costs associated with publishing a book. The costs of paper, printing and ink plus other production costs (editing etc.) were and probably are about 15 percent of total costs to publish including discounts and distribution and so on. A $20 cost for paper, printing and ink and production for a hardback with a run of 400 copies could easily be $20 and probably more.

guthrie writes:

Bryan, you may have many interests, but perhaps also very specific preferences regarding your tastes in writing/subject matter.

This may make you less of a 'Philistine' and more of a snob. If your question was even somewhat honest, I'll presume to venture one possible answer... perhaps instead of gaining more expertise (which is costly) you might consider 'relaxing your standards', or at least examining the possibility of being more charitable to the writers/subjects in question.

Brian writes:


Here's one way to approach the question. How many of those books in the library would YOU find worthwhile? I can't speak for you, but I know I always found loads of fascinating books when I wandered through my graduate library. Surely there were hundreds, if not thousands. I'm sure you would be similar.

Granted, hundreds is a small fraction of all the books there. But now consider how many people--faculty, students, staff, outsiders--have used that library and will continue to do so. Now multiply that number of people by several hundred, if not thousands, and you'll get the number of worthwhile "reads" in the library. I'm sure it far exceeds the number of books in the library. Of course, many people will like the same book, which is why the number of "reads" can exceed the number of books. But given this reasoning, isn't is possible, even probable, that almost all books would be considered worthwhile to some people? Consequently, almost all books could be worthwhile, even if very few seem worthwhile to any one of us.

So no, you're not a philistine. You're just failing to understand the power of human diversity.

Brian writes:

By the way, what's with your concern about paper? We can always grow more. It literally grows on trees. :)

Floccina writes:

Reminded me of the following quip:

Third, much research consists of obscure articles published in even more obscure journals on topics of trivial importance. Mark Bauerlein, a professor of English at Emory University, once estimated that 21,000 articles have been written on Shakespeare since 1980. Wouldn't 5,000 have been enough?
From here

Long ago in college I chose to write a paper on Sigmund Freud compared to Karen Horney. This was for a course in psychology. My major interest is Mathematics.

Freud and Horney were both psychologists and German contemporaries. I read translations. Freud's work was turgid and heavily burdened by clouds of unusual terms, largely undefined.

For example, Freud used the term "cathexis" everywhere. Wikipedia: In psychoanalysis, cathexis is the process of investment of mental or emotional energy in a person, object, or idea. The math part of me asked, "What is emotional energy?" Note that this is a "process", so what is really going on? Freud's patients didn't just think about something, they formed a cathexis.

Karen Horney wrote simply. Her view of psychology was that people were attracted to pleasant situations and avoided unpleasant ones. If society disagreed with their preferences, they made up reasons to justify themselves. She was skeptical of a vast unconscious directing our actions. People usually had good insights into why they did or did not do things.

Freud and Horney both wrote in German, so the clarity of the translations into English weren't just difficulty in translation. To me, Freud wrote in an appalling abstraction and confusion, and Horney wrote clearly.

The result is that Freud produced an industry devoted to analyzing his work, usually increasing the confusion. Horney was easy to understand and generally ignored.

Now to a vast generalization. When people understand what they read, they think "This can't be great research, considering that even I understand it." On the other side, they assume that unreadability is a sign of a very difficult problem analyzed as best as can be done. Of course a mere lay person cannot understand it.

So, if you want to write a book or thesis to impress people, make it unreadable. This will also reduce the amount of criticism. Worthy critics will hesitate to put in the time to unravel what you wrote. Then, you can deflect criticism by claiming correctly that the critic did not understand your work.

Lewis writes:

You should do research on how people's research agendas---their topics of study---change before and after tenure. My guess would be that afterward they public at longer intervals, longer papers.

Alex writes:

Most of the books of the library of my university are not read at all by anyone.

Tom West writes:

huh? walk me through why it costs $140 to print 250 pages

At least 20 years ago, there was a large fixed cost to printing a single copy. A 250 page hardcover might easily be $10K + $5/book.

We considered a print run of 2K to be the minimum without losing our shirts, (we sold our books for $40) but academic presses have much smaller print runs. Add in the editorial costs, proof reading costs, etc. and you could end up with a much,much higher fixed cost.

Plus warehousing. 400 academic books might take you 5 years to sell.

When I left the market, short run printing threatened to make the economics much more attractive, but then book piracy also became a thing and took it all away.

Grant writes:

...libraries still exist? ;) I was in uni ten years ago and practically never visited one. Granted I was in for computer engineering, so most all material was available on the Internet.

Most of the books I checked out related to mechanical engineering hobbies, and not school work. I found a few of them extremely informative.

I suspect book quality improves the closer one gets to an experimental (hard) science.

Tim writes:

It's a trial-and-error process, and authors (and their publishers) don't know quite they're capable of, or quite how their work will be received, until they conduct an experiment. The best work is acclaimed or reviled, and the worst work is probably discarded completely.

A lot of mediocre work sells (look at any best seller list), so if there is some doubt about the quality of a work, why not put it out on the market and see what happens?

david condon writes:

Where's the like button? Some really good comments too

Comments for this entry have been closed
Return to top