> How different is my Calculus 1 textbook Ed. 17 from this year's Ed. 18?

There's almost certainly no significant mathematical difference as far as the calculus it teaches goes.

If the target audience for the book includes students who aren't really interested in calculus, such as students who are just taking it to fulfill a requirement, and don't expect to use calculus much after they finish the class, then I'd expect the new edition to update exercises and examples to try to make them interesting and relevant to today's youth.

If the target audience is students who actually want to learn calculus, either because it is interesting to them per se, or because they know it is useful for things that do interest them, then there is probably little or no need for frequent new editions.

For example, a few top schools use Apostol's two volume text, "Calculus", either as their main calculus text, or for the more advanced track if they have multiple calculus tracks. Apostol volume 1 is currently all the way up to 2nd edition, which came out in 1967. Volume 2 is also on its 2nd edition, which came out in 1969.

Another example is Spivak's "Calculus", also used at several top schools, which is on its 4th edition, which came out in 2008. According to the preface, "Although small changes have been made to some material, especially in Chapter 5 and 20, this edition differs mainly in the introduction of additional problems, a complete update of the Suggested Reading, and the correction of numerous errors". The preface to the 3rd edition says that the biggest change was the addition of a chapter on planetary motion. It also rearranged quite a bit of material, and added problems. It looks like 2nd edition was a pretty substantial upgrade over 1st edition.

Spivak 1st edition was 1967, 2nd edition 1980, 3rd 1994, and 4th 2008.

Sometimes it's the professor writing the book so a new edition means more royalties for them. Other times it's like was mentioned in a neighboring comment, the publisher offers some perk like an online tool to let students submit their assignments and grades them automatically. The publisher wouldn't have any interest in letting one book be used multiple times for that system so if you want to use it you need the new edition.

The reason doctors will often prescribe the "next generation" drug, even if the old one works fine and is generic, and 100x cheaper, is that they believe it's better. They aren't stupid, and know its only marginally better, but they want the best for their patients. Professors also want the best for their students, and so the latest edition is prescribed. In a way, it's a point of pride, but its also a little bit lazy, because it's the simplest way they can signal "hey, I'm staying current, doing my job and not just phoning it in."

So, yeah, I don't think it's bribery, it's just a heady mix of (ignorance of) diminishing returns, moral hazard, and status signaling. Plus its a petty injustice that only hurts people in aggregate, and it only hurts people who already can afford college in rich places, so no-one is really going to care about this to change it.

> Professors also want the best for their students

Yes, but students need to buy textbooks. Is it really for "the best" if students have to worry about paying hundreds of dollars for the latest version of a textbook rather than getting 99% of the benefit from the PDF for edition n-1 they can find online?

> Plus its a petty injustice that only hurts people in aggregate, and it only hurts people who already can afford college in rich places, so no-one is really going to care about this to change it.

No, no, and no :(

> Professors also want the best for their students, and so the latest edition is prescribed.

No, that's not it at all. Calculus 17th edition is not in any way better than Calculus 16th edition. If anything, the newer one has more typos, weighs more, and is more expensive. A 600-page single-variable calculus text does not help students learn the material, since the small number of wonderful ideas that make up calculus are diluted in a fetid sea of repetitive examples, mindless busywork, pointless "real-world applications" that have no relevance to the real world, and a general tendency to give recipes that guarantee you will get the right answer even if you have no idea what you are doing. That's not math. That's the opposite of math.

The new intervention being “marginally better” can be questionable.

New interventions often aren’t tested head to head against their alternatives. Or inadequately tested against them.

So we have no idea if they’re even better, but manufacturers don’t want to risk the embarrassment of evidence that they’re inferior.

Yes. Here's a personal anecdote to illustrate. Years ago I had dental surgery, and the dentist prescribed a new and very expensive antibiotic for me. I was poor at the time, and paying for that was a serious hardship.

The next day I had a severe allergic reaction to it, and the dentist replaced that prescription with plain old penicillin (which cost me $10 for the complete treatment). I was absolutely furious that he didn't go with the affordable and time-tested treatment first.

> They aren't stupid, and know its only marginally better, but they want the best for their patients.

But surely what's best for patients is the treatment with the best cost/benefit ratio, not necessarily the treatment that is marginally more effective but much more expensive.

More and more college bookstores are run by Barnes and Noble. At my last college I wanted to use the edition I had used last year and the bookstore told me they couldn't order that version because there was a new version out. This has happened to a collegue at my current university.

> More and more college bookstores are run by Barnes and Noble. At my last college I wanted to use the edition I had used last year and the bookstore told me they couldn't order that version because there was a new version out. This has happened to a collegue at my current university.

It's not just the choice of the bookstore, but of the publisher. Most scummy academic publishers now (maybe 'scummy' is redundant) simply stop printing or selling the old editions.

Maybe not much. And there the old accusation "the author is just rearranging chapters to make more money" has some merit. But in a lot of sciences things actually do change. My immunology textbooks of twenty years ago would be absurd to teach from today.

In some cases there are very major differences. Dragon Book v2 has entire new chapters on topics that weren't being widely applied when v1 was written. v18-v19 probably isn't a big jump, but that doesn't make all version changes worthless.

The dragon book was updated once, 13 years ago, 20 years after its original publication date. It seems strange to use it as an example within the context of textbooks being updated at a regular cadence.