Research and development has an inherent tendency to have a decreasing growth rate. As the pool of things to discover continuously shrinks, it becomes harder to innovate. But we a groundbreaking discovery is made, this opens a lot of new opportunities and one should see a lot of new innovation. But with molecular biology and genomics, the pharmaceutical industry should have seen a burst of innovation, and in particular a jump in innovation productivity. Yet the contrary happened. One argument could be similar to the one that has been made about the productivity slowdown of the seventies, that an groundbreaking innovation like information technology needs time and resources to be understood.
Fabio Pammolli, Massimo Riccaboni and Laura Magazzini claim that this effect is very important. They observe that all the low hanging fruit have been picked in pharmacology and that first have shifted their investment portfolio towards more difficult problems. They suggest that one particular reason to do so is that improving current drugs is not profitable as generics are close substitutes and little rents can be extracted. Thus new classes of molecules are sought.
I would add another development in the field of R&D in general. It has become increasingly difficult and costly to file patents, as the field is littered with "predators" who file vague patents to prevent other from innovating, or to claim royalties. Not only does this increase the cost of innovating, it also increases its uncertainty, as any discovery can be subject to litigation even if it was a genuine discovery. This also encourages laboratories to find new molecules that are much different from existing ones.