The problem now appears to be that the data after 1991 is junk because registration no longer carried near the benefits it used to. So it doesn't look like we're going to get much in the way of data for the effect of Sonny Bono. Even if it were analyzable, there isn't a real trend after 1991 so studying deviations from the 'trend' would be difficult. I think the closest we can see to a negative effect is after the 1909 extension, which happened in the middle of an economic boom. But, again, 1976 points to no change. If I could find the pre-1870 data I could look at another extension, but it seems that data is nowhere to be found--and believe me, I looked.
--Ari
I have apologized many times for the ads now. I did not realize they were there when I posted the site, due to Mozilla's pop-up blocker. I make no money from them. The free host I use, virtualave.net, does. When I first used starting hosting with them years ago, they merely placed a banner ad at the top of each webpage. I was shocked to find how much worse the situation had become. Again, I apologize for the ads.
As for the trend from 1950 to 1991, I am quite confident it is genuine. The correlation of a quadratic curve fit to those points is most definitely Rsquare>0.99 . I provided the data for you to perform your own analyses if you wish.
For the Anonymous Coward, I would appreciate it very much if you would drop your veil of secrecy and provide specific criticism of my analysis of the data. I will happily post corrections and credit you. I don't believe I ever made any grandeose pretensions about my 'oeuvre.' It is a website, posted among millions upon millions of others, more interesting than some but by no means all. I have remained very open to suggestions and corrections in my methods. The facts I stand by: They are fact, and the sources for the copyright data in particular are beyond reproach.
UPenn is, indeed, one of the best Universities in the world. I am both honored and grateful to be a student there.
--Ari
This neglects the opportunity cost of an extension, namely the lost innovation that can be gained through reducing copyright terms. I have no data to show reducing copyright terms has such an effect--although I suspect, of course, it does--largely because I know of no time when copyright terms were reduced.
--Ari
Indeed. Such is the way of science. I set out to see if x effect really happens as it is claimed. There is no evidence that it does. I have provided the data and an analysis, however crude. If you can look at the decline after copyrights were expanded in 1909 or the zero effect of the 1976 revision and draw a different conclusion, please do. I welcome your arguments. If you have more data I have not considered, I practically begged for it when I published the site. I feel this study is considerably more objective than studies put out by the RIAA. Of course, the RIAA has had a tendency to avoid any sort of analysis at all, instead relying on polemic to achieve their goals. The statistics on their website for researchers are fragmented, showing snippets of data to allow only one conclusion: the conclusion they want. If the data had pointed the other way, believe me I would have concluded that and published in the same manner. It would have intrigued me and caused me to dig deeper to try to find out why, when so many nonstatistical analyses have concluded that increasing copyright so far past the life of the author does nothing to encourage innovation, the opposite effect occurs. But that is not the case.
--Ari
I started to normalize for pop. when I began looking at the data, but it had little effect. Rsquare on the quadratic fit curve from 1950 to 1991 drops from 0.99 to 0.98, but the overall shape of the data is the same. It seems the population growth rate is dwarfed by the copyright registration growth rate.
--Ari
"I really would like to see some analysis on the negative effects (if at all) of copyright extensions on innovation."
There's a negative effect after the 1909 copyright extension; a graph is on the site. However, I refrained from calling it a negative effect because:
a) it's not a very strong/consistent one
b) the 1976 data argues strongly for *no* effect
However, admittedly, the idea that retroactively extending copyright would have a negative effect has some sense to it, though I doubt the effect would be immediate. More likely is if Congress ever begins scaling back copyright durations to see a positive effect, as Steamboat Willie derivatives proliferate;-).
--Ari
"everyone knows that correlation does not mean causation"
Indeed. But a lack of correlation might imply a lack of causation, especially when the extensions of copyright were so dramatic. Look at the 41 years from 1950 to 1991. A quadratic curve of best fit has an Rsquare of 0.99 . Then that amazingly strong trend reverses in 1991--although this could be due to the Berne Convention revision pointed out by an earlier responder...I haven't looked that up yet.
There is definitely more than one Rsquare statistic, most are just included in the graphics of the curves, although CDs shipped vs. New Releases chart definitely has much higher correlation than it should due to the two leftmost points.
--Ari
The bloke that posted the article may not be able to understand that, but he certainly understands that that is not the way the copyright expansion laws have been 'sold' to Congress. Furthermore, the Founding Fathers apparently did not understand the effect you refer to either, as they explicitly state that the purpose of copyright is "to *promote* the progress of science and useful arts." In addition, if insufficient copyright terms were providing insufficient protection of the desire to innovate, shouldn't there have been a jump after 1976 or 1909?
That said, I disagree with the assertion that copyright does much for innovation at all. I, certainly, have never had an idea, be it musical or literary, and decided not to develop it because I couldn't get paid for it. The opposite is also true. With few exceptions, people don't create great art because they get paid to. However, they do have to have the spare time to develop their ideas fully. Copyright is necessary because it allows those who would otherwise create art as a 'hobby' quit their jobs and devote themselves to their expression (assuming it is popular enough). Other functions of copyright include ensuring that forms of art that need significant amounts of capital to succeed (e.g. recording studios to finish an album) get that capital. However, the music world is finding that, as desktop audio solutions improve, an artist no longer needs huge sums of cash to be successful. The music industry will have to redefine its role in the future, I think, just as every business does when faced with technological change. Certainly in the end the winner will be the artist.
--Ari
Apologies for the popups on the site. I tend to forget about them since I use Mozilla. They are an unfortunate consequence of my not having time at the moment to set up my server--Virtualave hosts for free. Site hosting donations are, of course, gladly accepted;-P.
--Ari
At first I had major reservations about making the leap from copyright registrations to innovation, and certainly it does not make sense on an individual level. Gary Paulsen is not a better writer than Hugo, Dick, or Bâ, just because he has written more books then they. In the aggregate, however, the total number of great works per year is approximately correlated with the total number of new works created. Choosing which work should win a Pulitzer is surely harder today than when it was created. Taking an extreme example, clearly more great literary works are produced per year in modern times than in ancient times when few could read. In short, although measuring innovation through copyright registrations is not the absolute ideal, it does provide a fairly strong (and, I think, relevant) way to statistically document the effect of different policies on innovation.
As for the 1991 Berne convention registration change, I was under the impression that it only applied to renewal of copyrights. But, if I am mistaken, as I said many times in the article itself, please e-mail me any data you might have. Furthermore, although when I wrote up the data I was rather fixated on the plunge in 1991, what is truly remarkable is that the trend is so significant--despite one of the largest extensions of copyright in history taking place right in the middle of the trend. From that data alone, it seems clear that Sonny Bono was not justified Constitutionally.
Finally, as for my social life, rumours of its demise are greatly exaggerated. I had an amazing Freshman year and did so many interesting things it blows my mind looking back on it. Of course, those of you whom this comment is directed at likely suspect my mind is already blown, but that's life. Penn is not the sort of place where anyone sits around doing nothing, or even just one thing as an undergrad. Everyone has many passions, and this is merely one of mine. "Spending most of my Freshman year" in the library was perhaps unfortunate wording. I should clarify and say that the project lasted most of my Freshman year, but I spent a sum total of 15-20 hours on the project itself. I find it impossible to prove based on words alone that I have a happy, busy social life, so just trust me, I do. My only regret is that it being at Penn makes summer difficult because my friends (yes, including my girlfriend Tara, who is currently hiking in Alaska) scatter to all parts of the world.
--Ari
Slashdot Mirror
Similar pattern without the bump a little while later. Bandwidth usage in MB by day (the /.ed page was 225KB):
Jul 11 2003 1.35
Jul 12 2003 0.08
Jul 13 2003 0.48
Jul 14 2003 0.3
Jul 15 2003 1.24
Jul 16 2003 3745.91
Jul 17 2003 432.48
Jul 18 2003 38.09
Jul 19 2003 3.29
Jul 20 2003 7.27
Jul 21 2003 3.49
Jul 22 2003 2.04
Jul 23 2003 0.84
Jul 24 2003 1.28
Jul 25 2003 1.29
Jul 26 2003 0.66
Jul 27 2003 0.53
The problem now appears to be that the data after 1991 is junk because registration no longer carried near the benefits it used to. So it doesn't look like we're going to get much in the way of data for the effect of Sonny Bono. Even if it were analyzable, there isn't a real trend after 1991 so studying deviations from the 'trend' would be difficult. I think the closest we can see to a negative effect is after the 1909 extension, which happened in the middle of an economic boom. But, again, 1976 points to no change. If I could find the pre-1870 data I could look at another extension, but it seems that data is nowhere to be found--and believe me, I looked.
--Ari
I have apologized many times for the ads now. I did not realize they were there when I posted the site, due to Mozilla's pop-up blocker. I make no money from them. The free host I use, virtualave.net, does. When I first used starting hosting with them years ago, they merely placed a banner ad at the top of each webpage. I was shocked to find how much worse the situation had become. Again, I apologize for the ads.
As for the trend from 1950 to 1991, I am quite confident it is genuine. The correlation of a quadratic curve fit to those points is most definitely Rsquare>0.99 . I provided the data for you to perform your own analyses if you wish.
For the Anonymous Coward, I would appreciate it very much if you would drop your veil of secrecy and provide specific criticism of my analysis of the data. I will happily post corrections and credit you. I don't believe I ever made any grandeose pretensions about my 'oeuvre.' It is a website, posted among millions upon millions of others, more interesting than some but by no means all. I have remained very open to suggestions and corrections in my methods. The facts I stand by: They are fact, and the sources for the copyright data in particular are beyond reproach.
UPenn is, indeed, one of the best Universities in the world. I am both honored and grateful to be a student there.
--Ari
This neglects the opportunity cost of an extension, namely the lost innovation that can be gained through reducing copyright terms. I have no data to show reducing copyright terms has such an effect--although I suspect, of course, it does--largely because I know of no time when copyright terms were reduced.
--Ari
You are incorrect, sir. 98.57967% of statistics are made up completely on the spot. Mine, however, are not. ;-)
--Ari
Indeed. Such is the way of science. I set out to see if x effect really happens as it is claimed. There is no evidence that it does. I have provided the data and an analysis, however crude. If you can look at the decline after copyrights were expanded in 1909 or the zero effect of the 1976 revision and draw a different conclusion, please do. I welcome your arguments. If you have more data I have not considered, I practically begged for it when I published the site. I feel this study is considerably more objective than studies put out by the RIAA. Of course, the RIAA has had a tendency to avoid any sort of analysis at all, instead relying on polemic to achieve their goals. The statistics on their website for researchers are fragmented, showing snippets of data to allow only one conclusion: the conclusion they want. If the data had pointed the other way, believe me I would have concluded that and published in the same manner. It would have intrigued me and caused me to dig deeper to try to find out why, when so many nonstatistical analyses have concluded that increasing copyright so far past the life of the author does nothing to encourage innovation, the opposite effect occurs. But that is not the case. --Ari
I started to normalize for pop. when I began looking at the data, but it had little effect. Rsquare on the quadratic fit curve from 1950 to 1991 drops from 0.99 to 0.98, but the overall shape of the data is the same. It seems the population growth rate is dwarfed by the copyright registration growth rate. --Ari
"I really would like to see some analysis on the negative effects (if at all) of copyright extensions on innovation." ;-).
--Ari
There's a negative effect after the 1909 copyright extension; a graph is on the site. However, I refrained from calling it a negative effect because:
a) it's not a very strong/consistent one
b) the 1976 data argues strongly for *no* effect
However, admittedly, the idea that retroactively extending copyright would have a negative effect has some sense to it, though I doubt the effect would be immediate. More likely is if Congress ever begins scaling back copyright durations to see a positive effect, as Steamboat Willie derivatives proliferate
"everyone knows that correlation does not mean causation"
Indeed. But a lack of correlation might imply a lack of causation, especially when the extensions of copyright were so dramatic. Look at the 41 years from 1950 to 1991. A quadratic curve of best fit has an Rsquare of 0.99 . Then that amazingly strong trend reverses in 1991--although this could be due to the Berne Convention revision pointed out by an earlier responder...I haven't looked that up yet.
There is definitely more than one Rsquare statistic, most are just included in the graphics of the curves, although CDs shipped vs. New Releases chart definitely has much higher correlation than it should due to the two leftmost points.
--Ari
The bloke that posted the article may not be able to understand that, but he certainly understands that that is not the way the copyright expansion laws have been 'sold' to Congress. Furthermore, the Founding Fathers apparently did not understand the effect you refer to either, as they explicitly state that the purpose of copyright is "to *promote* the progress of science and useful arts." In addition, if insufficient copyright terms were providing insufficient protection of the desire to innovate, shouldn't there have been a jump after 1976 or 1909?
That said, I disagree with the assertion that copyright does much for innovation at all. I, certainly, have never had an idea, be it musical or literary, and decided not to develop it because I couldn't get paid for it. The opposite is also true. With few exceptions, people don't create great art because they get paid to. However, they do have to have the spare time to develop their ideas fully. Copyright is necessary because it allows those who would otherwise create art as a 'hobby' quit their jobs and devote themselves to their expression (assuming it is popular enough). Other functions of copyright include ensuring that forms of art that need significant amounts of capital to succeed (e.g. recording studios to finish an album) get that capital. However, the music world is finding that, as desktop audio solutions improve, an artist no longer needs huge sums of cash to be successful. The music industry will have to redefine its role in the future, I think, just as every business does when faced with technological change. Certainly in the end the winner will be the artist.
--Ari
Apologies for the popups on the site. I tend to forget about them since I use Mozilla. They are an unfortunate consequence of my not having time at the moment to set up my server--Virtualave hosts for free. Site hosting donations are, of course, gladly accepted ;-P.
--Ari
At first I had major reservations about making the leap from copyright registrations to innovation, and certainly it does not make sense on an individual level. Gary Paulsen is not a better writer than Hugo, Dick, or Bâ, just because he has written more books then they. In the aggregate, however, the total number of great works per year is approximately correlated with the total number of new works created. Choosing which work should win a Pulitzer is surely harder today than when it was created. Taking an extreme example, clearly more great literary works are produced per year in modern times than in ancient times when few could read. In short, although measuring innovation through copyright registrations is not the absolute ideal, it does provide a fairly strong (and, I think, relevant) way to statistically document the effect of different policies on innovation. As for the 1991 Berne convention registration change, I was under the impression that it only applied to renewal of copyrights. But, if I am mistaken, as I said many times in the article itself, please e-mail me any data you might have. Furthermore, although when I wrote up the data I was rather fixated on the plunge in 1991, what is truly remarkable is that the trend is so significant--despite one of the largest extensions of copyright in history taking place right in the middle of the trend. From that data alone, it seems clear that Sonny Bono was not justified Constitutionally. Finally, as for my social life, rumours of its demise are greatly exaggerated. I had an amazing Freshman year and did so many interesting things it blows my mind looking back on it. Of course, those of you whom this comment is directed at likely suspect my mind is already blown, but that's life. Penn is not the sort of place where anyone sits around doing nothing, or even just one thing as an undergrad. Everyone has many passions, and this is merely one of mine. "Spending most of my Freshman year" in the library was perhaps unfortunate wording. I should clarify and say that the project lasted most of my Freshman year, but I spent a sum total of 15-20 hours on the project itself. I find it impossible to prove based on words alone that I have a happy, busy social life, so just trust me, I do. My only regret is that it being at Penn makes summer difficult because my friends (yes, including my girlfriend Tara, who is currently hiking in Alaska) scatter to all parts of the world. --Ari