Copyright laws are important because without them there would be insufficient motivation to produce creative works.
However, the problem with copyright laws is the same as any other monopoly: the holders of copyrights create supply less than optimal amounts of content. This is true in two ways: producing fewer new works and charging more than the marginal costs of the works they do produce.
The current US legal code supports copyright 70 years after the author's death, which seems (intuitively) way past the point where the marginal extra year of copyright boosts social welfare.
I have an alternative I am not at all confident in but I think would be an improvement for all parties involved.
The basic solution is to dramatically reduce copyright to just a few years and, in exchange, subsidize income from copyright works to the tune of ~10%.
The basic premise is that the money you can make on a copyright generally declines quite rapidly. We'll get into numbers in a minute, but suppose the money earned from a copyright work has a half-life of 1 year. This would mean that 90% of the revenue a work makes is done in the first 3.32 years. Therefore, we could reduce copyright to, say, 4 years, and the artists would still be better off with just a 10% subsidy.
Another benefit is that it provides a more-or-less objective way to determine copyright duration: the length of time that typical works earn 90% of their revenue. This contrasts with the existing system: however many years Disney can get Congress to agree to.
This proposal just two important questions:
- How much would a 10% subsidy cost?
- How much could we reduce copyright to?
How much would this subsidy cost? Well, US artists produced $71 billion worth of "long-lived entertainment originals" Copyright‐Protected Assets in the National Accounts. That $71 billion represents 0.49% of GDP in 2007 Table 1.1.5. or about $235 per person POPTHM. Subsidizing this by 10% would cost a mere $24 per person per year.
The New Copyright Duration
Estimating how much we can reduce copyright durations is difficult because the half-life of a work's revenue depends a great deal on the work.
Works of music typically earn 90% of their revenue within months. Movies take years. Franchises and logos can take decades. [TODO: cite]
The obvious solution is that the duration of copyright should depend on the medium. Music copyright could last a year, while movie copyright could last a decade.
This naively seems like great, but we run into issues with grey lines. At what point does the copyright for the "Iron Man" soundtrack become the copyright for "Iron Man" the movie, become the copyright for "Iron Man" the character, become the copyright for "Avengers" the franchise, become the copyright for "Marvel" the brand?
I'm going to use this "Iron Man" example going forward.
I'm not certain, but I don't think this is a lethal issue. Even with interpretations pushed to the break points by corporate lawyers, I still think the potential benefits are pretty huge.
For instance, even with the most outlandish interpretation, the literal mp4 file of "Iron Man" would clearly fall under the movie copyright duration rather than the character of franchise one.
It seems to me the main casualty of corporate lawyers is preventing the creation of derivative movies about "Iron Man" as where the line is between derivative-of-movie and derivative-of-franchise is incredibly blurry. I don't really have a solution to this.
But this problem aside, I think the benefits are huge: essentially giving every American legal access to every song after a year every movie after a decade for just $2 per month. I can't really find good statistics on books or physical art, but I'd imagine a similar decade-ish timeframe.