Another possible explanation is that the way Quantcast and others count unique visitors is broken: they don't (/can't) take into account multiple computers and multiple browsers.
If the NY Times paywall has user-level tracking (I'm not sure if it does, but sometimes the NYTimes asks me to log in to read articles, so it may do) then a bunch of people who are now hitting their quota browsing across multiple machines and browsers with a single login were previously being counted as multiple uniques.
You can reach your limit on one browser, then read another twenty on another browser. Despite that I have probably cut my reading in half.
I found the NYT had some of the most relevant ads of any site that I visit. As a result I am clicking on fewer ads and resulting in less revenue for them, but I may be atypical.
The whole paywall has encouraged me to check out other sites for news. The BBC I've found to be terrific and offers background articles on American issues not seen in any US paper.
If the argument you're making is "the free market is working because the paywall is driving traffic to other news vendors," keep in mind that the BBC is taxpayer-subsidized by Britain's citizens—and many of them aren't too happy about it.
(I don't know if you're making this argument or not. Just an observation.)
i've not noticed this paywall at all. do anti-tracking plugins (adblock, ghostery, disconnect) zap the cookies or whatever that it uses? or is it perhaps only for americans? (i realise it's not immediate, but i follow links there fairly regularly and have never exceeded any quota).
Adblock does not allow you to dodge the paywall, from personal experience.
There is an easy way to escape the paywall: After a new page opens, wait for the text to load, then after the text is loaded but the rest of the site's infrastructure is still loading, hit Escape to stop loading. Whatever code is responsible for summoning the paywall GUI never finishes loading.
This solution does not work well when trying to open multiple NYT tabs, however.
You might be right that these plugins block cookies or the javascript that informs you that you've hit your limit.
They've tried to make the paywall as permeable as possible. For instance, if you follow a link from twitter or facebook it wouldn't count. But if you went to the NYTimes directly, it would.
The idea is to tax heavy users and those who have more money than time.
It's far from conclusive, but this trend is a disturbing one.
No, the other way is to sell articles individually. Why are you being obtuse?
Why is purchasing by subscription bad, but selling articles individually good? There are benefits to both. Personally, I'd rather pay a subscription because I read a lot of articles and buying them individually would be a pain in the ass.
A possible contribution to the decrease that was not mentioned is that they may be linked to less often because of the pay wall. For general news, a blogger might instead link to some other, more reliably visitable site. When the Montreal Gazette introduced their pay wall at least one blog I follow avoided linking there when there was another source.
Indeed I would think most bloggers would look for alternate sources.
I also find I tend not to follow links if they are identified as NYT, and the loss of the ability to push links to Instapaper has really killed NYT for me.
If the NY Times paywall has user-level tracking (I'm not sure if it does, but sometimes the NYTimes asks me to log in to read articles, so it may do) then a bunch of people who are now hitting their quota browsing across multiple machines and browsers with a single login were previously being counted as multiple uniques.