Paula Stephan, who is an expert on the labor economics of science, is in Chemistry World (emphases mine):
...Chemists have often had an edge over those trained in the biomedical sciences because of the large number of research positions for chemists in industry. But in recent years, industry hires have lagged – in part because of mergers and acquisitions in pharma and in part because of a fragile economy...
What can be done to solve the imbalance? First, graduate programs should be required to tell students the truth about job placements. Second, the incentive to staff labs with graduate students and postdocs must be altered. Raise the ‘wage’ (that will get the attention of PIs); require faculty to develop alternative training tracks for students. In other words, make faculty understand that there is a serious cost to using graduate students and make them pay for part of it out of their research budgets and their time. Third, put more funds into supporting graduate students on fellowships and training grants and fewer funds into supporting students as graduate research assistants. Fourth, create incentives for faculty to staff their labs with permanent help rather than relying on temporary labour. Finally, if need be, lessen the coupling between research and training. While effective training requires a research environment, effective research can be done outside a training environment. If universities don’t have what it takes to exercise self-control, then turn some of the research funds over to institutes that are not in the training business. Abstinence is, after all, the most effective form of birth control.I think it's the fourth and fifth problems that are problematic for the long-term. How to (de?)institutionalize research (creating permanent staff, moving to research institutes outside of academia) without destroying incentives and creativity? I don't know if I have an answer to that, and I don't know if it's the right question.
Paula Stephan is professor of economics at Georgia State University, US, and a research associate at the National Bureau of Economic Research. She is the author of the book How Economics Shapes Science.