My professor at Wharton just published a research paper, quantifying the effects of digital downloads on the sales of CDs - on the topic of the industry debate as to whether illegal downloads are destroying CD sales or whether there is a promotional side benefit from it.

He asked the question: “Would the people who are downloading music buy the music had they not downloaded it?”

In the study, he performed a lot of statistical modeling to determine the relationships and correlations between sales and downloads, using data from tracking peer to peer downloads and weekly CD sales figures.

Some of the results seem quite surprising, while others are expected. According to the study:

1) The top albums (over 600,000 copies, top 25%) were helped by the downloading, those with lower than 36,000 copies had a negative effect.

2) Most people download individual songs, not entire albums. About 90% of the songs on most popular albums are downloaded less than 11 times, and there is a high download concentration toward hits.

3) Music industry marketing still greatly influences people’s habits, both from a sales and downloading aspect.

His press release (.doc) and the full paper (.pdf) are attached.