Sir

Your Editorial 'Crisis of confidence' (Nature 457, 635; 2009), about US graduates pursuing careers in biomedical research, hits the nail on the head. The problem is not that junior faculty careers are not fostered. Their crisis is predominantly due to perennial mismatches between supply and demand. The aura of an unstable career permeates the entire profession and has diverted the best students from seeking careers in biomedical sciences.

Strategically, there are two main issues facing the funding of US biomedical research, especially through the National Institutes of Health (NIH).

First, we all wish to see continued growth at this time of exciting developments in biomedical research. This goal seems to enjoy wide public support, and the new stimulus budget reflects a deep commitment to funding science. We need to convince all who will listen of the importance of this goal.

Second, we should be able to work with the hand that is dealt us: annual budgets from Congress. Why can't the NIH evolve mechanisms to regulate the impact of the upswings (increased congressional allocation) and downturns (periods of stagnant growth)? This is a key goal, as ensuring reliable growth is the best mechanism for getting new blood into biomedical research.

Here are some suggestions to consider. One is to modulate the amount of funding and grants depending on funding. Supplements can be used during periods of growth and taken away during downturns, which will tend to directly stabilize the number of grants, and thus the number of applicants and their success rates.

Another suggestion is to adjust the effort a principal investigator can charge to grants. This can be decreased during expansion; universities have to contribute more. This would impose on departments and universities a clear restraint on uncritical runaway hiring. Also, during periods of growth, funding should be diverted to infrastructure and programme activities: instrumentation, renovations and collaborative grants.

Approaches such as these and others, in aggregate, would reduce the incentive for rapid growth during periods of budgetary expansion at the NIH and would redirect funding to the R01 (research project grant) pool during periods of stagnation. Hence, they are likely to result in more 'steady state' growth that will reinvigorate career development in the biomedical and life sciences.