Until recently, there were two types of newspaper Web sites: those that made you pay to read many of the articles (the New York Times, the Wall Street Journal, and the Financial Times) and those that didn't.
That is changing. The New York Times recently announced that almost all its online material would now be free. FT.com has just moved to a system of free access for occasional visitors. And Rupert Murdoch has strongly hinted that the Journal might do something similar. The theory is that advertising revenue will outstrip subscription revenue.
So, is this the death of subscription-model newspapers? And will the availability of so much free online journalism also doom their pricier print editions to a slow death?
These are hard questions, because the Internet is changing so quickly that first principles are little help. One smart response is to experiment, but such experiments can be pretty expensive if they don't work out.
The New York Times' retreat from a subscription model is widely portrayed by online commentators as a humiliating and belated acceptance of the inevitable. But Matthew Gentzkow, an economist at the University of Chicago, recently published research that suggests that there has been no expensive mistake. Both the subscription model then and the advertising model now were likely to have been reasonable choices. Free online access makes more sense now not because of some fundamental law of online economics, but because the online advertising market has matured.
Armed with a snapshot survey of the reading habits of Washington, D.C., locals, Gentzkow tried to figure out whether the Washington Post was losing print subscribers because of the availability of the free online version at Washingtonpost.com. (Disclosure: The Washington Post Co. owns Slate.)
At first glance, it seems not: Washingtonpost.com and the Washington Post appear to be complements, like peaches and cream. Many people who read one, read both. Other people read neither.
But Gentzkow realized that this was misleading. Just because some people like to get news from lots of sources and other people prefer to read no news at all does not mean that the existence of Washingtonpost.com persuades more people to subscribe to the print edition.
Gentzkow looked for other evidence. He found that people who had access to fast Internet connections were, other things being equal, less likely to read the print edition. He found reasons to believe this was specifically because of access to Washingtonpost.com, not to the Internet in general.
And he also found that while many people read both the print edition and the Web site, they tended not to do so on the same day. That is, on days when you can get to the Web site, you don't buy the paper. And if you have the paper, you don't bother to eyeball those ads on the Web site. The Web site and the newspaper are substitutes after all, and charging for subscriptions helped prevent a modest amount of cannibalization.