Industry Canada's release earlier this month of an independent study on the impact of P2P file sharing generated considerable public interest and some debate from economists around the world who were provided with complete access to all the raw data. First out of the blocks was Stan Liebowitz, a Texas economics professor who immediately pronounced that "without going into details of the study we can ask whether this result is even remotely plausible" and that "the result is so counterintuitive that I think it fails the laugh test." While those comments generated headlines, once Liebowitz had a chance to actually view the study and the data, he dropped that language and acknowledged that some of the initial criticism was too harsh. His primary criticism is that:
the authors present two sets of results, one for the entire sample and one just for downloaders. It makes little or no sense to look only at downloaders and when they do so the authors find a result that is not only implausible but is actually is impossible to be true, given their data. When the appropriate full sample is used the results are still likely to be biased upward because the authors do not fully account for the impact of music interest, which impacts both downloading and purchasing.
Birgitte Andersen, one of the authors of the study, has now posted a response to Liebowitz.
Andersen responds that Liebowitz's analysis has a weak empirical underpinning, emphasizing that the IC study features micro-economic data rather than Liebowitz's macro-economic data. Andersen notes that:
Macro data can describe a situation or some relationships at the aggregate level, but such data are limited when it comes to shedding light on explanations of the situation. For this we need micro-data, as we need to understand how the micro behaviour underpin the creation of the situation at the aggregate level.
Andersen continues by addressing Liebowitz's criticisms, characterizing some of the content on his site as "hugely misleading."
Attracting somewhat less attention was a response from Australia by Dr. George Barker and Richard Tooth. That response, which a footnote notes was supported by a grant from the Canadian Recording Industry Association (indeed CRIA has cited Barker's work with approval in the past), repeats many of the Liebowitz claims about simultaneity, even though Liebowitz later acknowleged that those concerns were overstated. The authors also argue that:
The most striking problem with the study’s conclusion is that the authors have assumed that the positive relationship observed between CD sales and P2P downloads is because downloading increases CD sales. The more plausible explanation is likely to be that those who download more are more interested in a variety of music and so purchase more CDs.
Andersen did not respond to the Barker/Tooth paper, perhaps because they misstate the conclusions in the study. The study did not say that P2P is the cause of increased purchases, but rather than there is a correlation between the two. Indeed, it seems reasonable to assume that downloaders are more interested in music. The fact that they still purchase more music, notwithstanding the fact that it is available via P2P networks, contradicts CRIA's claims that it cannot "compete with free" and merits the attention it has received.