Back in 2010, the Government Accountability Office (GAO) published a meta-study of the economic effects of intellectual property infringement (including counterfeit goods as well as copyrighted works). The GAO concluded that IP infringement is a problem for the economy, but it’s not possible to quantify the extent of the damage — and may never be. It looked at many existing studies and found bias or methodological problems in every one.
More recently, Michael Smith and Rahul Telang, two professors at Carnegie-Mellon University, published another meta-study that serves as a sort of rejoinder to the GAO study. This was the subject of Prof. Smith’s talk at the recent Digital Book World (DBW) conference in NYC.
Assessing the Academic Literature Regarding the Impact of Media Piracy on Sales summarizes what has been a growing body of studies on the economic effects of so-called media piracy. Their conclusion is that piracy does have a negative effect on revenue — if for no other reason than the vast majority of studies come to that conclusion.
Smith’s presentation at DBW listed no less than 29 studies on media piracy that take actual data into account (as opposed to merely theoretical papers such as this one). Of those, 25 found economic harm from piracy, while 4 didn’t. When the list is restricted to papers published in peer-reviewed academic journals, the ratio is similar: 12 found harm; 2 didn’t. Interestingly, almost half of the cited studies were published after the GAO’s 2010 report.
(When Smith and Telang’s paper was originally published last year, many discredited it instantly because the MPAA helped fund the research. Yet I take the researchers at their word when they say that the funding source had no effect on the outcomes — an assertion bolstered by the paper’s exclusion of the MPAA’s own study from 2006.)
The paper explains why some studies’ methodologies are better than others and discusses shortcomings in some of the studies, such as the Oberholzer-Gee & Strumpf paper from 2007 that showed no harm to sales of music from piracy and therefore has been widely cited among the copyleft.
It’s easy to poke holes in the methodologies of studies that have to rely on real-world data over which the researchers have little or no control. And as someone who wouldn’t know an “endogenous dependent variable” if one bit me in the face, I find it hard to look at criticisms of these studies’ methodologies and determine which ones to believe. Yet it’s obvious that any study on piracy must rely on real-world data in order to have any credibility at all.
Decisions about business and policy have to be made based on the best information we have available. After a certain point, simply poking holes in studies — particularly those whose results you don’t happen to like — isn’t sufficient.
It may indeed, as the GAO suggested, be impossible to measure the economic effects of piracy with a large amount of accuracy. But if dozens of researchers have tried, all using different methodologies, then their conclusions in the aggregate are the best we’re going to do. Put another way, it will henceforth be very difficult to dislodge Smith and Telang’s conclusion that piracy does economic harm to content creators.