When physicist and astronomer Penny Sackett was appointed to be Australia's chief scientist in 2008, many other scientists thought she was an excellent choice for the job. US-born Sackett is a successful researcher, most notably in the hunt for extrasolar planets, and as head of astronomy at the Australian National University was also an accomplished administrator. But two and a half years into her five-year term as the government's leading science adviser, she resigned. Sackett cited "personal and professional reasons", adding that "institutions, as well as individuals, grow and evolve". It was clear, however, that all had not been well between her and the politicians she was employed to advise.

Sackett's experience highlights some of the tensions inherent in the science advisory process, according to James Wilsdon of the Science Policy Research Unit at the University of Sussex. Scientists and politicians, he says, have a kind of pact they need to honour: politicians having to respect the evidence put before them, while scientists have to steer clear of policy prescriptions. But in 2010 Sackett effectively reneged on this deal, stating publicly that Australia was "not acting with sufficient speed" to combat global warming, following the government's decision to put its emissions trading scheme on hold. Within nine months, Sackett had quit her post.

While Wilsdon says he does not know to what extent Sackett's statement was directly linked to her stepping down, he has no doubt that scientists often are not prepared for the pitfalls of giving advice. "They may have the respect of their peers, while lacking the political and organizational skills needed to navigate complex policy questions," he says.

Avoiding such pitfalls will be one of the central themes of a two-day conference entitled "Science Advice to Governments" that is being held in Auckland, New Zealand, from 28–29 August. Organized by New Zealand's chief science adviser Sir Peter Gluckman on behalf of the International Council for Science (ICSU), the meeting will bring together leading science advisers and policy experts, including Wilsdon, from several dozen countries in order to stimulate "frank and fruitful" discussions about what works and what does not when scientists pass on their expertise to politicians.

Heather Douglas, a philosopher of science at the University of Waterloo in Canada, who will also be attending the meeting, says that such discussions are sorely needed. Douglas points out that several countries and multinational organizations have set up new advisory posts in recent years, and she says that these new positions are leading scientists to "finally think about what the advisory job involves". She maintains, however, that many scientists who agree to take on advisory roles still do not fully understand their responsibilities, particularly when it comes to more complicated aspects of the job such as public communication and whether or not to be an advocate of the science community. "It is only in the past couple of years that these things are starting to be grappled with," she says.

Advice in meltdown

Both the US and the UK governments have had chief science advisers since the 1960s – a presidential adviser in the former case and an adviser to the prime minister and cabinet in the latter (supplemented, since 2011, by an adviser in each government department). A few other countries have since followed suit: Australia set up an advisory post in 1989, India did so in 1999 and Ireland likewise in 2007. In the last few years, however, such posts have become even more fashionable: New Zealand appointed Gluckman in 2009; the EU hired its first chief scientific adviser – Scottish biologist Anne Glover – in 2012; while the United Nations inaugurated a new scientific advisory board earlier this year (the latter having four physicists including Fabiola Gianotti, a former spokesperson for the ATLAS experiment at CERN).

Many of the incumbents will make the trip to Auckland, where they will have plenty to get their teeth into. One recent disaster they are sure to discuss, and the one that prompted ICSU to set up the meeting in the first place, is the tsunami-induced meltdown at the Fukushima Daiichi nuclear power plant in Japan in March 2011 and experts' ensuing failure to properly inform the public about its potential impact. Another episode likely to be on the agenda is the deadly earthquake that struck the central Italian town of L'Aquila in 2009, which led to the trial and conviction on manslaughter charges of six scientists and a government official – full or acting members of a government advisory committee – after they were accused of providing false reassurances to the public.

Observers agree that the controversy in Italy highlights the need for a clear definition of an adviser's role, but opinions differ on exactly what that role should be. Thomas Jordan, an earth scientist at the University of Southern California and chair of an international commission that reviewed earthquake forecasting in the wake of the L'Aquila disaster, believes that an amateur seismologist's baseless alarms "trapped" the Italian experts into downplaying the risk of a major quake, and that as such the roles of science adviser and decision-maker should be cleanly separated in the future. However, Roger Pielke, a social scientist at the University of Colorado, opposes what he calls "the cartoon image" of a wall isolating science from politics. "The challenge is not to keep these things separate but how to effectively integrate them," he says. "Separating them goes against the whole spirit of the enterprise, which is to get expertise in to the decision-making process."

Pielke points out that many scientists provide such expertise by acting as advocates – by backing specific courses of action such as the introduction of quotas on greenhouse gas emissions. Such advocacy, he says, is a central element in a healthy democracy, but he believes it is also vital that some scientists act as "honest brokers". In this role, Pielke explains, they would spell out the science behind a range of policy options rather than narrowing the choice and effectively making policy decisions.

Best placed to play this role, according to Pielke, are learned societies, especially, he says, since effective brokering often requires several specialists working together. Pielke maintains, however, that in practice these organizations sometimes act more as advocates. He cites a 2006 letter written by Bob Ward, a then senior manager at the Royal Society, to the British arm of energy giant Exxon Mobil criticizing the company for funding organizations that "have been misinforming the public" about the science of climate change. "There is no shortage of scientists willing to get into skirmishes," says Pielke, "but we do have a shortage of organizations that will rise above them."

John Pethica, a physicist from Trinity College Dublin and a vice president of the Royal Society, responds by saying that the letter was written by an individual rather than the institution as a whole, and that the many statements on scientific policy made by the Royal Society go through an extensive review process before publication. "The only position we take is: the science must be high quality," he says.

Maintaining trust

The idea that expert advisers should steer clear of backing specific policies is endorsed by many of the scientists who will attend the New Zealand meeting. To illustrate the point, EU adviser Glover takes the example of genetically modified crops. She says that all of the evidence points to such crops being safe. Indeed, last year she was quoted as saying that opposition to the technology on scientific grounds "is a form of madness I don't understand". But Glover recognizes that politicians can have legitimate reasons, such as those based on ethics or economics, for opposing the cultivation of GM crops. "The important thing is that they are in a position of strength so that they know about the evidence beforehand and are transparent about the reasons for rejecting it if they choose to," she says.

To ensure that decision-makers get the information they need, Wilsdon says that scientists need a number of skills beyond those of the researcher. These include communicating complex ideas and scientific uncertainty in simple terms, as well as being able to pull together many different sources of evidence and then present "a set of navigable options" to decision makers. "Advisers are often appointed because they are eminent scientists," he says, "but they also need a sophisticated understanding and sensitivity regarding the policy-making process."

This sentiment is shared by Gluckman, who says that one of the hardest aspects of his job is simultaneously maintaining the trust of the public, the media, policy-makers, politicians and the scientific community. In his experience it is actually the last of these groups that is the toughest to deal with, pointing out that scientists often criticize advisers for "not batting for their interests in public", in part because they confuse the provision of scientific advice with the constant championing of funding for research. "The point about being a science adviser is being an honest broker and leaving the values stuff to the politician," he says. "We don't live in a technocratic world, as much as some scientists would wish we do."

Difficulties notwithstanding, Gluckman believes that it is the one-to-one relationships between a chief scientist and senior politicians that best encourage governments to respect the evidence advisers put before them. However, he acknowledges that cultural factors might make such an approach difficult in some countries. "In the English-speaking world individuals can be put into trusted positions in a number of ways," he says. "But in some European countries that is not so easy. In Germany, for example, there is a tradition of collectivism, which means there is more of a tendency to work through academies."

Japan too tends to prefer committees over individuals, says Reiko Kuroda, a chemist at the Tokyo University of Science. According to Kuroda, Japanese politicians discussed reintroducing science advisers in the wake of the Fukushima disaster (a chief scientist having existed between 2006 and 2008), but in the end no new appointments were made. "We need someone with a broad scientific background who is respected by everyone to advise the prime minister," she says. "But politicians and the general public prefer to have a group deciding. The problem is that the people in the group don't take any responsibility."

Beginning a conversation

Back in Europe, Glover has worked hard to try to bring the continent's science advisers together. After taking up her post, she wanted to get scientists to discuss some of the technical issues underlying EU policies to see where they could reach consensus. But Glover realized that she did not know who to speak to and has since built up a network of individuals from many of the member states – currently 14 – comprising heads of academies, members of science advisory councils and government employees, in addition to the chief scientists of the UK and Ireland. "We had an absolutely outstanding meeting," she says of the group's first get-together at the EuroScience Open Forum in Copenhagen in June. "We can learn a lot from each other."

It is in this spirit that the New Zealand conference will also be held, says Wilsdon. "We hope that this is the beginning of a conversation that improves expertise across different advisory systems," he says. "The object of the exercise is not to promote one particular model, but to discuss the underlying questions at the boundary between science and politics that apply the world over."