Nobel Laureate Paul Krugman once asked, "Who would enter a demolition derby without the incentive of a prize?" (Source: Krugman, Paul. 1998. Soft microeconomics: The squishy case against you-know-who. Slate.)

a. The "demolition derby" he was talking about was the battle over Internet browsers: Many enter the battle, but only one (or two) survive.

But let's take his story literally: If there were two cars in a demolition derby, and each car costs $20,000 to build, and one car will be totally destroyed, how big will the prize probably have to be to get two people to enter if there's a 50-50 chance of losing all your investment?

b. What if we want a really good demolition derby: one where 10 of these cars compete but only one survives. About how big will the prize have to be now?

c. Let's draw the lesson for network goods: Since competition in network good markets is competition "for the market," then it's like winning a prize in a demolition derby.

If there's a fixed price of starting up a new social networking Web site (you need so many computers, so many nerds, so many advertisers), then when would you see a lot of firms competing for the prize: when the prize is large or when the prize is small? Thus, if we want a lot of competition for the market, do we necessarily want to restrict the profits of the winner?