It seems to me that there are two types of scientist in the world: those who think mainly about doing science and those who think mainly about how to get funding to do science.
The former (let's call them the "Dr Sparks") live and breathe science. They think about little else, whether at work, at home or on holiday. They would probably still do what they do even if they weren't paid. Their passion, bred out of deep insight, provides fertile ground for innovation and breakthroughs of exactly the sort the government wants and believes science funding should deliver.
However, there is a downside to being a Dr Spark. Such scientists are often viewed as brilliant or crazy in equal measure. Few of their peers properly understand their latest work or its implications because the Dr Sparks plough new and difficult furrows. They usually struggle to get funding, partly because they are often so wrapped up in their research that they don't put enough time into their funding proposals, and partly because they often assume that the importance of their work is self-evident. Consequently, for every glowing report that says their research will change the world, there will be one or two others that completely miss the point and brand it "unfundable". These days, only uniformly acclaimed proposals tend to get funding.
The second type of scientist (let's call them the "Dr Plods") is far more common. These people are good enough to do adequate, incremental science. Their niche is based more on getting funding than on pioneering research. The closest they get to originality is when they assiduously scan top scientific journals to identify what is trendy so that they can jump on the latest bandwagon. They follow trends, they do not set them.
This is not to say that the Dr Plods are unsuccessful: many are successful precisely because they play the research councils' game so well. Bureaucracy is their preferred habitat. Such people spend a lot of time at conferences, brown-nosing people of influence, chatting up friends who are likely to look favourably on them when reviewing papers and grant proposals, and getting themselves on to committees that decide where the next funding initiatives will be directed. When the Dr Plods plan their next funding applications, they do so with military precision, canvassing opinions from colleagues to find out exactly what the powers that be are looking for and then executing to best effect. (After all, a Dr Plod is unlikely to get funding on talent alone.)
It is now about 18 months since the research councils introduced the requirement for each funding applicant to explain how they intend to generate and maximise their "impact". What has been the effect of asking these different types of scientist to fill out a two-page form detailing their "pathways to impact"?
Dr Spark realises that good science is unpredictable and that pre-emptive statements such as "I'll make a documentary about my work" are meaningless fantasies. How can you plan a film when you don't know what you will discover? Almost without exception, the Dr Sparks I know reacted to the need for impact statements by saying, "Surely the funding panels will not pay attention to this rubbish?", and then put next to no effort into writing anything convincing. "Well, that's their problem," I hear the funders respond. Wrong. It is the duty of those in charge of distributing public funds to devise processes that stand the best possible chance of directing money to the best possible science.
Meanwhile, Dr Plod spends time judiciously downloading examples of "good" impact statements and asking funded colleagues to see their applications so that they can be copied.
No prizes are awarded for guessing who will end up being funded.
What is the moral of this story? Impact statements may well bring peripheral benefits in terms of encouraging the scientific community to think more carefully about how to make their work more visible. This is a good thing and arguably overdue. However, this could be better achieved by providing support and advice to those scientists who have been funded about how the impact of their work could be increased. To use impact to inform the funding decisions themselves is counterproductive. If you ask for two pages of box-ticking fantasy, you will attract slightly bureaucratic people who are good at spinning yarns. If you want to fund great science, you should judge on scientific ability. Alone.
"Pathways to impact" will actually achieve the opposite of its primary aim: it will radically reduce the chance that vital funding will reach the sort of people who are most likely to discover the next penicillin, figure out a cure for Aids or develop the energy source that ends our reliance on oil. But then it takes one to know one, and the people responsible for the impact agenda are not cutting-edge scientists themselves.
With an eye to the impending Olympics, how about a simple analogy? If you were put in charge of picking the country's running team, would you base your decision on athletic ability alone or rely heavily on a two-page "Pathways to glory" statement, a description of the team's intended celebrations should they manage a victory?