Academic study of website impact on newspaper readership

Matthew Gentzhow of the University of Chicago has written a study of the impact of online versions of newspapers on the print circulations of those newspapers, using the Washington Post as an example. It’s the most complete survey of this kind which I’ve seen – there have been some internal studies done (I know News International looked at the impact of the Sun’s website on the paper a year or so ago), but precious little in the public domain.

Gentzhow’s conclusion is that washingtonpost.com reduces the daily readership of the Washington Post by 30,000 (based on a readership of 405,000) – ie, if washingtonpost.com didn’t exist, the readership (not circulation) of the paper would increase by 30,000. Based on 2003 numbers, Gentzhow estimates the estimated $33.3 million in revenues at washingtonpost.com came at a cost of $5.7 million in lost newspaper profit.

He then goes on to suggest that, economically, the optimal model is for a 20 cent daily payment for washingtonpost.com rather than the free model that currently stands. This, says Gentzhow, would provide the biggest overall economic benefit to the Washington Post in terms of compensating for newspaper profit decline with online revenue increases.

Fascinating, and tons of data which I’m sure will be appearing in internal presentations at most English-language newspapers for the next 12 months. Many commentators have already pointed out the basic flaw with the model, though – that it’s based on 2003 online advertising revenues. Gentzhow himself acknowledges this, saying that the zero-cost model for washingtonpost.com makes increasing sense as online advertising revenue increases. Indeed, Gentzhow implies that the rise in ad revenues since 2003 might already have closed the gap. In other words, the Post’s investment in washingtonpost.com is already paying dividends. And online ad revenues are still growing, fast.

One other point – Gentzhow makes an interesting observation that the “overall welfare effect” of having free-to-air newspaper websites is significant. Not being an economist I don’t fully understand this, but if my reading of it is correct, the economic benefit to the readers of having a free washingtonpost.com is double the economic value of the loss to the Washington Post in 2003. For newspapers who see themselves as part of the social fabric of their markets, that seems to me an interesting point for development.