How much income does the film industry lose to piracy?
18 Answers
Boris Polania, Behavioral answers
Nick Huber, Thinker, writer, data man
Fundamentally, this is an empirical question. There's only one paper in the literature that I could find that collects the dataset that has both movie performance and piracy metrics that is necessary to rigorously answer this question with any degree of confidence.
Arthur De Vany, Professor of economics at University of California, Irvine, and David Walls, Professor of economics at the University of Calgary, have written extensively about this question. (1, 2)
In “Estimating the Effects of Movie Piracy on Box-office Revenue," they note that predicting movie revenues is notoriously difficult, or what they call the "nobody knows principle.” (2, 3) They argue it’s unadvisable to merely subtract what a movie made from an estimate of what it “could have made” without piracy.
With this in mind, empirical claims on the effects of piracy should be aware of the difficult methodological environment in which they are made.
In particular, the empirical difficulty in answering this question is that piracy affects the supply of a movie, but also might affect demand so its overall effect on producer revenue is theoretically ambiguous.
However, the work of Professors Rafeal Rob and Joel Waldfogel, both at the University of Pennsylvania, seems to indicate that piracy actually reduces paid movie demand in a study designed to measure the impact of being shown free content on undergraduate's paid movie consumption: (4)
Two primary sources of the film industry revenue are DVD sales and box office sales, which may be affected by piracy differently.
First, does piracy affect DVD sales?
Professors Michael Smith and Rahul Telang, both at Carnegie Mellon University, measured the effects of a network broadcast of a movie on DVD sales: (5)
However,
Second, does piracy affect box office sales?
In 2005, the Motion Picture Association of America (MPAA) argued that they were losing $3 billion in box office sales due to piracy. In an industry that takes in less than $10 billion annually, this is a significant claim and, if true, a major concern. (2)
De Vany and Walls use data on one movie’s 14 weeks of box office revenue and associated online piracy activity ("including eDonkey 2000, BitTorrent, Gnutella, FastTrack, Hotline, FTP, Usenet and Internet Relay Chat") during that period to develop a statistical method that is immune to the “nobody knows principle.” (2, 6) In their words,
Their data shows that the availability of pirated copies are meaningfully similar to the availability of paid copies in terms of magnitude and time (i.e. it's both easier to find a movie online and offline on opening weekend).
![]()
They fit a regression to estimate the changes in box office revenue based changes in movie supply and sum the weekly effects driven by the availability of pirated versions of the movie. For the single movie on which they had data, they estimate that piracy directly destroyed $40 million in box office revenue. Overall, the film grossed approximately $61 million or around $600/pirate source. (7)
Conclusion
If the film industry takes in roughly $10 billion per year in box office sales, we could estimate that piracy costs the film industry $3-4 billion annually, which is a range constructed from both academic and MPAA research. The primary mechanism of this lost revenue is through box office sales rather than DVD sales.
---
Arthur De Vany, Professor of economics at University of California, Irvine, and David Walls, Professor of economics at the University of Calgary, have written extensively about this question. (1, 2)
In “Estimating the Effects of Movie Piracy on Box-office Revenue," they note that predicting movie revenues is notoriously difficult, or what they call the "nobody knows principle.” (2, 3) They argue it’s unadvisable to merely subtract what a movie made from an estimate of what it “could have made” without piracy.
With this in mind, empirical claims on the effects of piracy should be aware of the difficult methodological environment in which they are made.
In particular, the empirical difficulty in answering this question is that piracy affects the supply of a movie, but also might affect demand so its overall effect on producer revenue is theoretically ambiguous.
However, the work of Professors Rafeal Rob and Joel Waldfogel, both at the University of Pennsylvania, seems to indicate that piracy actually reduces paid movie demand in a study designed to measure the impact of being shown free content on undergraduate's paid movie consumption: (4)
Employing a variety of cross-sectional and longitudinal empirical approaches, we find large and statistically significant evidence of displacement...These estimates indicate that unpaid consumption, which makes up 5.2% of movie viewing in our sample, reduced paid consumption in our sample by 3.5%. (emphasis mine)
- For my analysis, I will assume that piracy both increases supply and reduces demand, therefore it unambiguously hurts producer revenue.
Two primary sources of the film industry revenue are DVD sales and box office sales, which may be affected by piracy differently.
First, does piracy affect DVD sales?
Professors Michael Smith and Rahul Telang, both at Carnegie Mellon University, measured the effects of a network broadcast of a movie on DVD sales: (5)
With respect to the impact of movie broadcasts on sales, we find that movie broadcasts on over-the-air networks result in an increase in DVD sales at Amazon.com by an average of 118% in the week after over-the-air broadcast.
However,
With respect to the impact of piracy on sales, we use the television broadcast as an exogenous demand shock and find that the availability of pirated content at the time of broadcast has no effect on post-broadcast DVD sales gains.
- That is, piracy doesn’t seem to be affecting the film industry – positively or negatively – through DVD sales.
Second, does piracy affect box office sales?
In 2005, the Motion Picture Association of America (MPAA) argued that they were losing $3 billion in box office sales due to piracy. In an industry that takes in less than $10 billion annually, this is a significant claim and, if true, a major concern. (2)
De Vany and Walls use data on one movie’s 14 weeks of box office revenue and associated online piracy activity ("including eDonkey 2000, BitTorrent, Gnutella, FastTrack, Hotline, FTP, Usenet and Internet Relay Chat") during that period to develop a statistical method that is immune to the “nobody knows principle.” (2, 6) In their words,
We did not have to pose the counterfactual conjecture asking what the movie might have earned but for piracy. Nor did we need to speculate as to how many viewers of the pirated version might have gone to see the legitimate version in a theater. What we did do was to test directly the impact of pirate supply on the rate at which the movie’s theatrical revenues declined during the course of its run.
Their data shows that the availability of pirated copies are meaningfully similar to the availability of paid copies in terms of magnitude and time (i.e. it's both easier to find a movie online and offline on opening weekend).
- Assuming that De Vany and Wall’s data is representative of a typical movie with typical levels of piracy which seems reasonable to me, it’s possible that 40% of box office revenue of a typical film is being lost to piracy.
Conclusion
If the film industry takes in roughly $10 billion per year in box office sales, we could estimate that piracy costs the film industry $3-4 billion annually, which is a range constructed from both academic and MPAA research. The primary mechanism of this lost revenue is through box office sales rather than DVD sales.
---
- De Vany and Wall's papers on Hollywood are available at http://pareto.ucalgary.ca/hollyw...
- De Vany and Walls, “Estimating the Effects of Movie Piracy on Box-office Revenue,” August 24, 2007.
- Screenwriter William Goldman is generally quoted as the origin of this idea: “[P]roducers and executives know a great deal about what has succeeded commercially in the past and constantly seek to extrapolate that knowledge to new projects. But their ability to predict at an early stage the commercial success of a new film project is almost nonexistent.”
- Rob and Waldfogel, “Piracy on the Silver Screen,” Journal of Industrial Economics, Vol. 55, No. 3.
- Smith and Telang, “Competing with free: The impact of movie broadcasts on DVD sales and Internet policy,” MIS Quarterly. June 2009.
- Their specification had an R-squared, or rough overall predictive ability, of .78, or 78%.
- The total gross of the movie is not directly shown in their paper; however, in their summary statistics (Table 3) they have 13 observations of changes in weekly movie revenue with a mean of -4.7 million. Using the statistical identity of mean, the total movie gross is simply 13 * 4.7 million, or approximately $61 million.
Brian Dunlap, I work on a series of tubes.
Anthony Chris Zboralski, Founder/CEO at Belua
Stephen Glynn, Have more than 2,000 DVD's, 1,000 Blu rays and have seen too many movies to have a normal life
Garrick Saito, likes movies
William Petroff, Web Developer
Bert Cattoor, Value Miner