Actually it doesn't. The potential advertising revenue probably exceeds the revenue from subscriptions.
I've seen this before even when subscriptions for a news site was actually very good but potential advertising from page views would be better. (TimesSelect/The New York Times)
I would imagine there are adverts as well. These readers are very valuable to advertisers. 150K well-to-do readers trumps a couple million freeloaders any day.
Do we know this? You would need to see the change in advertising revenue between before this was implemented and now. You also have to factor in the rate of growth in readership to before and after.
I don't know that, but they probably do know themselves. They claim it to be net success, and the only contrary theory out there is they must be rigging numbers. Unless you're Doctorow you'd want some evidence for that.
Well without a way to independently check this kind of thing there is no way to know I guess, it's pretty easy to make the statistics work in your favor to explain later why you could have said something was a success when it wasn't.
Great example is SyFy channel show ratings, they put out press releases focusing on certain measures in which certain eps of shows performed favorably even if overall the ratings aren't that good.
I've seen this before even when subscriptions for a news site was actually very good but potential advertising from page views would be better. (TimesSelect/The New York Times)