Presentation on theme: "How Copyright Makes Works Disappear (Study 1) and How Secondary Liability Rules Enable YouTube Help the Problem for Music (Study 2) Paul J. Heald Herbert."— Presentation transcript:
How Copyright Makes Works Disappear (Study 1) and How Secondary Liability Rules Enable YouTube Help the Problem for Music (Study 2) Paul J. Heald Herbert Smith Fellow & Affiliated Lecturer, Cambridge University Professor of Law, University of Illinois Professorial Fellow, CIPPM, Bournemouth University Papers.ssrn.com/sol3/papers.cfm?abstract_id=
Study #1: How Copyright Makes Works Disappear... Random sample of 7000 fiction books on Amazon Only new books available from Amazon [no used titles or titles from Amazon “affiliates”] Gathered via a random ISBN number generator 2317 of 7000 titles located in the Library of Congress catalog Date of earliest LOC edition used as proxy for initial publication date of title [note this biases dates upward chronologically]
We might expect to see this...
What Professor Landes Would Expect to See...
What About Music? Brooks (2006) shows only 14% of famous old tunes digitized, so I Tunes or Amazon CD’s would be age-biased. So, collected data from two movie soundtrack samples Top 100 grossing movies of all-time from randomly selected movies from All from DVD’s currently on sale at Amazon
Why So Many Fewer PD Songs in the Randomly Picked Movies? Unlikely that directors of big budget blockbusters are more price sensitive. But Median release date of 100 highest grossing movies is 1977 and the median release data of the 134 random movies is 2002 [Box Office Mojo skews to newer movies] Oldest random movie was released in 1981.
But why should relatively newer movies have fewer PD songs? Due to copyright term extensions (e.g. 1976), newer movies are farther removed from the public domain reservoir of songs. HYPOTHESIS: The half of the Top Grossing films released before 1977 will rely on more PD songs than the half released after 1977.
Study #2: Uploading Hits on YouTube: How Notice and Takedown Rules Facilitate the Availability of Music Collected the Number One Songs in Brazil and France, from and from US from Searched for each song on YouTube and charted the first 10 uploads containing a version of the song. Tracked 4 types: recorded songs with only a picture of the album or CD cover or a picture of the artist; amateurs performing the song; custom videos with the song in the background; and television and movies clips. Noted uploader identity. # of views, date of upload, monetized or not.
Why is this Non-Monetized?
Preliminaries Almost all of the uploads are unauthorized. Much outright infringement is tolerated – remains up and not monetized. Many uploads are monetized. Average upload date is July ‘09 for non- monetized uploads and September ‘09 for monetized uploads.
Who can take down uploads? Custom videos and new amateur performances: Owner of the composition copyright. Recorded songs: Owner of composition OR sound recording copyright. Television or movie clips: Owner of composition OR video copyright. YouTube requires proof of ownership of all components to monetize!
Questions about U.S. Market Does a coordination effect hamper the ability of to monetized uploads where owners of compositions and videos must cooperate? Do better post-1968 contracts increase the monetization rate? Why do composers tolerate so much infringement in the context of television and movies clips? How bad is the problem of lost television episodes, specials, and newscasts?
France and Brazil in Comparison
Tentative Overall Conclusions Traditional book and music publishing models demonstrate how copyright stands in between works and the consuming public. Non-owners play a very important role in maintaining the availability of public domain and copyrighted works. YouTube, lowers transaction costs, providing valuable information, and facilitating both availability to consumers and revenue streams to owners. Current secondary liability rules look efficient!