Judging experts by the company they keep

Tim O'Riordan discovers why the public trusts some experts but not others.

I and two research colleagues recently asked a cross section of Norwich residents whom would they trust to tell them the truth about risks such as sunbathing, genetic manipulation and being mugged.

The results were revealing. Only 12 per cent trusted the government, 20 per cent trusted businesses, 25 per cent trusted the media, and less than a third trusted religious organisations and trade unions. Some 60 per cent were willing to put their faith in scientists and 78 per cent in their doctor. But the two most trusted sources were their friends (80 per cent) and their family (90 per cent) as well as scientists from environmental organisations, who had a 80 per cent "trust rating''. It appears that the science of protest, precisely because it is believed to be conducted in the public interest and targeted against institutions that cannot be relied upon, is the science that is most wanted to be truthful. No wonder Greenpeace moved so swiftly to admit its error of estimation in the toxicity of the oil/mud residues in the Brent Spar oil storage facility. Public trust is hard to win and easy to lose.

The BSE saga did not appear to enhance the reputation of science or government in securing public trust. According to polls conducted by the MORI organisation the percentage of the public with "none'' or "not much'' confidence in scientists working for government rose from 54 per cent in 1993 to 62 per cent in 1996.

Public faith in scientists working for industry is a little more secure, with about half of the public having "a great deal'' or "a fair amount'' of faith in industry-funded scientists compared to the 32 per cent who felt the same about government funded scientists. But again there is an important element of distrust. This indicates that even when a company like Shell UK seeks to command public support for the final outcome of decommissioning of the Brent Spar oil storage platform, it faces an uphill struggle to command public credibility.

What appears to influence public judgements about the credibility of advice about risks is a combination of social networks - the closeness individuals feel to certain social relationships - and personal efficacy, or how far people feel in control of their lives. This suggests that risk perception is more of a personal and social phenomenon than was earlier thought, and that people in different social settings will have different views about danger, the reliability of science, fairness of treatment, and accountability of government and regulators.

For example, the risks associated with alcohol were interpreted very differently by men and women. Men tended to link alcohol to drunken driving, women to drunken men. By the same token, men regarded mugging as a socially deviant crime, women saw it a personal threat that influenced their sense of safety in any public place.

Our research also showed that people who distrust the institutions of regulation and accountability as unlikely to act in the interests of the vulnerable, are much more likely to be antagonistic to "imposed'' but scientifically ambiguous risks (such as electromagnetic radiation, nuclear power, toxic waste incinerators) than those who accept the social order of regulation, or the vagaries of the market place.

Faced with dangers where the scientific evidence is not sufficiently detailed or unambiguous to link possible cause and effect, people look for a different approach. This may be called a "civic'' science, where it is recognised that the tolerance of any risk is a function of the fairness and the integrity of the consultation process as much as it is of the scientific judgements involved.

Civic science may involve community groups evaluating uncertain futures. In this way decision-making bodies should become more reflective of the social valuations that support difficult choices and science will become a partner in a more trustworthy political process.

Tim O'Riordan is professor of environmental sciences, University of East Anglia.