(this is cross-posted from my Globe and Mail blog)
When Industry Canada came out with a study last week that found file-sharing doesn’t lead to reduced CD sales — and in fact may even lead to an increase in sales among those who download a lot — it came as a surprise to many, most of all the music industry, which has been arguing for years that downloads are killing the record business.
It also came as a surprise to Stan Liebowitz, an economist with the University of Texas, who has been studying the impact that file-sharing and other Internet-related technologies have on music sales for several years, and has repeatedly come to the exact opposite conclusion.
Prof. Liebowitz has been studying the impact of technology on copyright since the 1970s, when he did a study for the Canadian government looking at the effect of photocopying on the publishing industry (he concluded that it would not have an overly negative effect). He also wrote a study in 1985 looking at a new technology called the VCR, and has done research that he says shows radio also contributes to lower sales of traditional records and CDs.
On his website, Prof. Liebowitz takes issue with the study done by two researchers at the University of London, who were commissioned by Industry Canada. According to the University of Texas economist, who is also a director of the Center for the Economic Analysis of Property Rights and Innovation, the study has a number of methodological problems and also fails what he refers to as “the laugh test.”
In a nutshell, Liebowitz says, the Industry Canada paper is at odds with well-established research that shows a prominent decline in CD and record sales over the past several years, a period in which the use of file-sharing software has grown dramatically. If downloading either doesn’t affect CD sales or actually has a small positive effect, he says, then how can we explain that large a decline?
Liebowitz says that his own research, including a recent paper to be published in the journal Management Science, shows that “file-sharing is responsible for the entire decline in record sales that has occurred, and that except for file-sharing there would have been an increase in sales since 1999 instead of the strong decline.”
Although the professor admits he is “partial to my own work,” he humbly describes it as “the strongest analysis to date of these issues.” (Other research — including some based on numbers from the music industry itself — has come to different conclusions).
And what about the 2004 study by Felix Oberholzer-Gee of the Harvard Business School and University of North Carolina economist Koleman Strumpf? They came to almost the exact same conclusion as the Industry Canada study — that is, they found the effect of downloading on CD sales was “statistically indistinguishable from zero.” A PDF of the study is here, and there’s a good overview of some of the implications of the study here.
Prof. Liebowitz says that the Oberholzer-Strumpf study’s research methods were also flawed, and that like the University of London researchers they failed to come up with any alternative explanation for why CD sales had dropped so precipitously over the past few years.
The professor might like to think that the matter is settled by his own illustrious research, but there are questions raised by his approach as well. For example, Liebowitz (who refers to “record sales” throughout his research, making it unclear whether he is talking about CDs or not) doesn’t say whether he is including the number of paid-for iTunes downloads when he looks at the decline in music sales over the past decade. Does his research include the sale of CD singles? Ringtunes? That’s unclear.
It’s also not clear whether the professor is talking about sales in dollar terms or the number of actual units that have been sold — since the price of CDs has come down over the past several years, and that has reduced revenues but not the number of units.
It’s natural to assume that downloading music would lead to fewer sales of CDs. But is it solely responsible for the decline in sales over the past decade, or are there other factors at work? And while CD sales have declined, have overall music revenues — concert-ticket sales, public appearances, T-shirts, endorsements, etc. — gone down as well, or have they made up for the drop in CD revenues? The debate continues.