Big Science with little output?

In a recent article entitled Big Science vs. Little Science: How Scientific Impact Scales with Funding (  Fortin and Currie from University of Ottawa provide a quantitative analysis of the impact of recent changes in NSERC funding. They conclude that scientific impact (measured in a variety of ways) was generally a decelerating function of funding. In other words, they found that in Canada, impact per dollar is actually lower for large grant-holders.

Fortin and Currie wonder whether it is more effective to give large grants to a few elite researchers, or small grants to many researchers. The authors argue that large grants are only effective if scientific impact increases as an accelerating function of the grant size. The data shows that this is not the case. This exact same investment argument can easily be applied to the University of Saskatchewan – we have signature areas of research being established.  This is Big Science. Will the impact of the work done in those areas increase as an accelerating function of institutional funding?  Evidence suggests otherwise.  Maybe it is time for our institution to change its assumptions if we want to increase overall productivity?