Dinesh Tantri makes an interesting link between the Freakonomics' discussion of incentives and that of incentives for sharing in Moral Versus Economic Incentives For Knowledge Sharing
While incentivizing knowledge sharing and collaboration it is important to find the "golden mean" if you will. Ideally we need to build schemes that rely on peer recognition and self-satisfaction. This needs to be augmented with more tangible incentives that are neither too less or too much.
The topic of incentives comes around frequently, and it is clearly an issue. Good planning has considered this as the KM effort was created. Bad planning says, "Hey, we've built this thing and no one is coming. I know! Let's give them a Starbucks card when they contribute to the repository."
It has been clear to me that reputation (whuffie) encourages knowledge workers far more than do economic incentives. The economic incentive will get people curious and might encourage people to check something out for the first time or possibly take some direct action. (Fill out your profile to get your Starbucks card.) But it won't keep them coming back, nor will it serve to create long term value in contributing to a "knowledge base" or to a KM community.