Patents Vs Innovation - the Tabarrok Curve

New submitter Optimal Cynic writes "Slashdot likes to argue about intellectual property and patents, and it's clear that both extremes are undesirable. Dr Alex Tabarrok has tackled the question — what is the right level of patent protection? His answer is the Tabarrok Curve, which applies the Laffer Curve methodology to innovation."

Comments

don't help and there's more than innovation

[ciaran_o_riordan]

For software, they don't help innovation, and promoting innovation can't be the only goal. Lots of non-innovative software development is really useful.

* http://en.swpat.org/wiki/Studies_on_economics_and_innovation
* http://en.swpat.org/wiki/More_than_innovation

And it's really crucial that patents be analysed per-domain. The don't affect pharma and the auto industry the same, and they don't affect software the same either. The distribution models are different, as is the profile of who mass produces each thing, as is the complexity (number of ideas) that get added to a product within one lifecycle, and the length of product lifecycles...

* http://en.swpat.org/wiki/Why_software_is_different

Re:don't help and there's more than innovation

[ikaruga]

Another nice feature would be protecting only patents actually used in products made by the inventor or its partners, in order to take out those so called patent trolls that don't produce anything. If in the future the inventor actually decides to use the patent in an actual product and there is someone already using it there could be some sort of predetermined fee. Obviously I'm oversimplifying the huge problem, but something must be done about those parasites.

Re:The Laffer Curve?

[dkleinsc]

Oh, come on! These two curves are perfectly related: They're both attempting to vaguely connect 2 variables that intuitively seem like they ought to have something to do with each other without actually being a remotely accurate description of reality.

For reference, Arthur Laffer said the theoretical relationship between tax rates and government revenue per capita looks like this:
Laffer Curve
A suspicious Martin Gardner then plotted the actual relationship between tax rates and government revenue per capita, and got something that looks like this:
Neo-Laffer Curve

My basic view on the subject:
1. There's absolutely no way to measure real innovation. Some of the problems:
- Discoveries that seem unimportant can turn out to be incredibly important 15 years later, and vice versa.
- Organizations sometimes protect their discoveries by keeping them secret.
- Academics often give away the knowledge they have without patenting it to build their career. However, they can also build their career by giving away nonsense and getting away with it.
- A lot of "innovations" are just tiny variations on things that we already have and don't make much real difference (e.g. the rounded rectangle).
2. There are lots of motivations for innovation, some of which can't be bottled, organized, or turned into policies. For example, the more idealistic scientists are motivated more by the joy of discovery than by the cash they'll get.
3. That means that trying to take a theoretical approach to creating more innovation is just plain unworkable. The one thing that seems to have worked, historically speaking, is (1) put really smart people in contact with each other, (2) make sure they have plenty of cash and whatever they need to do their work, and (3) tell them they can focus on pretty much whatever they feel like working on, just make something awesome happen. That worked in Alexandria 2300 years ago, it worked in Baghdad around 1000 years ago, it worked in London around 200 years ago, it worked in Bell Labs in the last century.

Re:Some fundamental, unchecked assumption here ?

[devent]

It's just his opinion. And his curve is a straw man that he puts out of his ass.
Where is the evidence that some patent protection is good? Ah right there is none. Everybody just assumes that some patent protection is good for innovation. That is just like creationism or like any other theology.

There is evidence that more patent protection is bad. See software patents. But where is the evidence for the other side, i.e. no patent protection as bad?

Don't come to me with some mind experiments like "But without patent protection where is the incentive to innovate". People have innovate for freaking 50,000 years. Our patent system, or our capitalist system is just about 250 years old.

I repeat. There is no evidence that any patent protection increases innovation.

There is plenty of evidence that too strong patent protection hampers innovation. See software patents, patents on DNA, business patents, and so on.

There is also evidence that no patent protection actually increases innovation (I posted already here but I post again)
The only real empirical study that I know of is The patent game.
See http://www.youtube.com/watch?v=q1Pi4w8ddA8
And they come with quite surprising conclusions.

"pure common system" means no patent protection.

1. patents increase innovations?

Average unique innovations is lower in the "pure patent system", and is higher in the "pure common system"
Productivity, aka economic activity, the "pure common system" is almost double the "pure patent system"
Per capita wealth, the total amount of dollars generated in the system: "pure common system" is almost 4 times that of the "pure patent system".

Conclusion: "Despite received wisdom, commons spur innovation better then do patents"