I see here that Nasa has decided to have prizes to pay for accomplishments rather than for proposals for ongoing research.
Prizes are an interesting organizational device. They have some benefits for those who run the system. They tend to work better for innovation, and less well for durable systems. They are open to abuse by those who run the game.
The abuse problem is obvious; though two fold. Like the lottery they tend to draw in a large population of players that don’t appreciate the terms of the game; i.e. the odds. The second aspect is that those that set up the game can often rig it so they capture the long term benefit from innovations done by the winners. Academic research has historically been framed in a way that the innovators gained only prizes and reputation but did not typically get to capture the economic benefits of their work. Of course, in exchange they did get tenure. That can be a fair deal. Particularly when the probability of most academic research leading quickly to economic return is very low.
Microsoft’s developer network is a good example of a game you can play where you might win a big prize but chances are they will own the long term benefit of your work sooner or later. If the players believe that the game has lousy odds and is structured poorly they won’t play. This is one hypothesis about why Microsoft has been forced to create it’s own R&D department. Another example, somewhat less dysfunctional, is the venture capital game. But if you look at the number of cases where a startup was swallowed by a large firm and then comes to naught you might suspect that’s because the startup’s owners didn’t see any reason to design for the long term since that was the problem of the firm that buys them out.
People talk about converting the patent system to a prize system as a way of getting a more functional framework for generating innovations. Innovations, as information, are public goods; but they are a species of public good where you only need one innovator out of a vast pool of contributors. The cure for cancer is the classic example. At the other end of the spectrum are the public goods like safety that require that all contributors remain vigilant. The classic example of this is the maintenance of a river’s levee.
Since I consider Nasa to be a murderous institution that has lost all ability to manage the safety of it’s operations I’m not amused by further organizational moves in the direction toward increased risk taking and reduced collective attention to safety. Prizes are a lousy way to make safe systems – particularly if the innovator doesn’t get to bear the cost or the benefit of the long term. This is at the heart of the generative force that creates a our parade of superfund sites.
Game theory dudes like to treat all activities as a game and there is some validity to that. It breaks down when you need to look at the longterm durablity of the system. Longterm issues lead to trust issues. Game theory isn’t terribly sophisticated about that. When presented with a game look at the odds, look who captures the longterm benefit, look who cleans up the mess, and try to ignore the prize. Oh, and consider: do you trust the house?