The comment higher up says A/B testing is used, that may be true for publishers that make use of performance ads. If you're a well known entity then you try to only accept brand ads (CPM - pay per impression/view), and those brand ads come with lots of conditions.
Advertisers want to appear near the top of the page or near the top of the content - that's why ad networks have rules about the placement of ads. Even Google (which is more performance than brand) limits the number of units that are displayed on a single page.
If your ads are at the top then there's a good chance that they will appear on the user's screen. If your ads are at the bottom and the user found the article boring then the browser was closed before seeing the ad, yet the impression was still paid for (hence, much lower rates).
This is the same with paged content, the user is equally less likely to see the adverts because maybe they don't view that page/scroll down that far.
I'm guessing the issue is that if you use paged content you know that the user has looked at that part of the article. If the adverts are all on the same page then an impression is not proof that a user has had that advert appear on the screen. Having said that if bootstrap's scrollspy can detect when you've scrolled content into view then I don't see why you shouldn't be able to track advert views in the same way.
Paged content gives site owners (most often, content farms) the ability to divide what would be a 10-second read into a 2-minute read by way of forcing users to keep clicking through. This can artificially juke pageview metrics, "engagement" metrics, bounce rates, etc., which means the site can a) sell more ads, b) command a higher premium on ads, or c) both.
Typically, it makes for a horrific reading experience. And I believe it will become bad for business, too, if enough high-quality content sites emerge and offer a far more compelling experience to draw readers away.
There's a hurdle preventing that from happening, however, and the hurdle is that users seem fairly site-agnostic in their content preferences. They consume information without a great deal of thought as to where it came from. And it's easier than ever to receive content now, through social channels, and never even have an inkling as to its original source or site. This liquidity and disposability of content means that content farms still have a big advantage over sites that attempt to make themselves into interesting, repeat-visit destinations.
Now, savvier and more intelligent readers -- the kind who probably frequent this community -- tend to avoid content farms, and probably don't click on articles entitled "10 Totally Weird Diet Tricks That Will Burn Away 10 Lbs in One Day!!!," requiring a 10-page click through to read. But sadly, a lot of people do click those things. A lot.
Advertisers want to appear near the top of the page or near the top of the content - that's why ad networks have rules about the placement of ads. Even Google (which is more performance than brand) limits the number of units that are displayed on a single page.
If your ads are at the top then there's a good chance that they will appear on the user's screen. If your ads are at the bottom and the user found the article boring then the browser was closed before seeing the ad, yet the impression was still paid for (hence, much lower rates).