Relations between academic social scientists and policymakers can sometimes seem like a strange dysfunctional dance. As Paul Wiles, Chief Scientific Adviser to the Home Office, puts it: "Government sometimes looks like a conspiracy against problem-solving - since today's structures arose out of yesterday's problems."
As a result, crucial challenges "don't belong to a single department, and so are left as orphan issues. The most important often prove the hardest to address."
At the same time, Professor Wiles said, universities "can also appear to be a conspiracy against knowledge, structured into disciplines that reflect the needs of the past".
He was speaking at a workshop last week run by the British Academy and SAGE Publications - with Times Higher Education as media partner.
The event brought together academics, politicians and journalists to debate "how social research can help improve our quality of life".
Rudolf Klein, visiting professor at the London School of Economics, said academics had to take risks, since "speculating on tomorrow's policy agendas will inevitably produce many misses as well as hits".
Narrow definitions of "impact", which is to be a key factor in allocating cash under the forthcoming research excellence framework, can detract from bold thinking, he said.
Other eminent delegates spelled out in detail the practical applications of their work.
Sir Michael Rutter, professor of developmental psychopathology at the Institute of Psychiatry, King's College London, explained how society has slowly come to understand that "broken homes" are not in any direct sense the causes of crime.
Proper science is needed, he said, because politicians, second-rate academics and the media often put forward "claims based on ideology, small samples or poor analysis".
Meanwhile, Mike Hough, director of the Institute for Criminal Policy Research, King's College London, described work designed to demonstrate that it was cheaper and more effective to try to make people want to comply with the law than to coerce compliance.
Both speakers were asked why their carefully honed insights were often ignored by policymakers. Professor Hough admitted that it can be difficult to influence politicians, adding that non-governmental organisations are often "better at pressing policy levers" than academics.
Malcolm Wicks, Labour MP for Croydon North, argued that "policymakers should bring in academics and challenge them, ask them to tell us what they know and not where we need further research".
He said that when he was Minister for Science, for example, he called on stem-cell experts to discuss the current state of knowledge and imminent developments in front of a group of science journalists.
Other speakers suggested ways in which academics might hone their messages, and how politicians could acquire better listening skills.
Ian Mulheirn, director of the Social Market Foundation, proposed that every new MP should be tested to check if they understood the difference between causation and correlation.