The Digital Opportunity (PDF), the independent review of intellectual property (IP) published last week after being commissioned by the government in 2010, is one of the most important contributions to the IP debate in years. IP laws, it says, have become hostage to special interest groups – “lobbynomics”, in the report’s words – and do not serve the purpose that its advocates claim it does. The balance between protecting past innovators and promoting future ones has not been struck and, says the report, evidence should drive policy – not lobbying by rent-seekers.

IP is one of the toughest areas of policy to get right. The challenge is to protect innovators’ rewards, and thus promoting innovation, without stifling future innovation by making these protections too stringent and long-lasting. The problem is complicated even more by criticisms of the concept of intellectual property itself. Critics argue that IP laws are effective restrictions of private property rights: I should be allowed to arrange my own property in any pattern I like, and bits on a computer or words on a page should be no different. The idea that a non-scarce resource should not have property rights assigned to it is appealing, but if granted it removes much of the incentive for people to write books and create innovations.

I sympathise with the anti-IP argument, but I’m an agnostic. And, for the foreseeable future, it’s clear that only gradual change can take place. And even IP's critics should support reform that makes it less strict. There’s no need to make perfect the enemy of good.

What would evidence-driven copyright law look like? (I won't discuss patents here, although this argument should roughly apply to patents as well as copyright.) The length should be determined by looking at earnings distributions for things like music and books, and cut the copyright protection period to only include, say, the first nine-tenths of the average distribution. Most copyrighted productions follow a power law – the bulk of their earnings from a novel or movie will usually be earned in the first couple of years (see diagram above). It’s the initial high earnings that IP should be aiming to protect, not the “long tail” that comes afterwards. This would reduce the stifling effects that copyright has, without reducing much of the innovation incentive, since most profits would still be protected.

Drastically shortening the copyright coverage period would hardly be perfect – it wouldn’t solve the problem of piracy, and would create some unfortunate outcomes where the creator of a work that only becomes popular after the copyright period misses out. But a big reduction in IP coverage periods, based on the income distribution of copyrighted works, would be a step towards promoting innovation that doesn’t undermine the income incentive to innovate.