Cumulative innovation, the process of one innovator building on the efforts of an earlier innovator, is common and important. Because of the work of Suzanne Scotchmer and others, economists recognize that cumulative innovation poses a serious challenge to those who try to design an optimal patent system. When one innovation builds on another, the patent system can be used to divide profits between two distinct innovators. But under normal conditions it cannot be designed to make both innovators full claimants on the flow of value created by their respective innovations.

Several articles have taken up the challenge of uncovering policy concerns that should help to determine the division of profit beyond early and late innovators. Green and Scotchmer note that when different firms contribute innovations to a sequence, total patent-based profits will need to be higher than the profits required to induce a single firm to invest in the same sequence of innovations. Scotchmer and Green observe that a weak nonobviousness standard encourages disclosure that has social value because it speeds further development and reduces redundant innovation. O'Donoghue et al. argue for broad and short patents in a quality ladder model because broad patents reduce duplicative investment. In contrast, Denicolò observes that a weak nonobviousness standard and narrow breadth reduce possibly wasteful racing to achieve a first-generation innovation.