In 1897, the House of Representatives in the US state of Indiana unanimously passed legislation that redefined the calculation of the value of pi , the ratio of the circumference of a circle to its diameter. Fortunately, the bill died in the state senate.

That historical anecdote might elicit a sardonic chuckle from those who remember their high-school mathematics, but around the world non-experts are increasingly being called upon to formulate public policy that requires an understanding of subtle and complex scientific and technological phenomena.

"How can you tell whether a whale is a mammal or a fish?" a teacher asks her third-grade class.

"Take a vote?" chirps one pupil. This suggestion may be amusing coming from a child, but there's nothing funny when governments apply it, as they increasingly do, to complex policies that involve science and technology.

Britons had their say during the summer, for example, on whether they want gene-spliced, or genetically modified (GM), products in their fields and food. To gauge public opinion in advance of a decision scheduled for later this year on whether to allow commercial planting of GM crops, the British government sponsored (at great expense) a series of public discussions around the country. Local governments and other organizations held hundreds of additional public meetings.

The head of the British debates' organizing committee, Professor Malcolm Grant, called them a "unique experiment to find out what ordinary people really think once they've heard all the arguments."

But the reality argues otherwise. Mark Henderson, science correspondent for the Times (London), offered this view of the UK's half-million-pound initiative: "The exercise has been farce from start to finish. I'm not sure I want the man in the street to set Britain's science, technology, and agriculture policy. One of the six meetings ... spent much of its time discussing whether the SARS virus might come from GM cotton in China. It's more likely to have come from outer space."

Henderson also observed that the meetings were dominated by anti-technology zealots, the only faction sufficiently organized and inspired about the issue to attend. This is consistent with reports that as many as 79% of the 37,000 questionnaire responses were orchestrated by activists.

The urge to base official policy on public opinion about such issues flourishes across the Atlantic as well. The US National Science Foundation (NSF), whose primary mission is to support laboratory research across many disciplines, is funding a series of "citizens' technology forums," at which average, previously uninformed Americans come together to solve a thorny question of technology policy.

According to the NSF's abstract of the project, being carried out by researchers at North Carolina State University, participants "receive information about that issue from a range of content-area experts, experts on social implications of science and technology, and representatives of special interest groups." This is supposed to enable them to reach consensus "and ultimately generate recommendations."

The project, first funded in 2002 to support two panels, calls for eight more panels this year, each comprised of 15 citizens who are "representative of the local population." Their deliberations will be overseen by a research team "composed of faculty in rhetoric of science, group decision-making, and political science." The team will test both "an innovative measure of democratic deliberation" and also "political science theory, by investigating relationships between gender, ethnicity, lower socioeconomic status, and increases in efficacy and trust in regulators."

This is the scientific equivalent of art for art's sake, but getting policy recommendations on an obscure and complex technical question from groups of citizen nonexperts (who are recruited by newspaper ads) is considerably more dangerous. Imagine going from your cardiologist's office to a café, explaining to the waitress the therapeutic options for your chest pain, and asking her whether you should have the angioplasty or just take medication.

The first of these NSF-funded groups tackled regulatory policy toward agricultural biotechnology, and recommended that the government tighten regulations for growing GM crops, including a new requirement that the foods from these crops be labeled to identify them for consumers. These proposals are unwarranted, inappropriate, and contrary to the recommendations of experts, both within the government and in the scientific community.

To be sure, involvement of the public is critical to their understanding of government policy. But it is less useful in the formulation of policy, particularly when complex issues of science and technology are involved. Science is not democratic. The citizenry do not get to vote on whether a whale is a mammal or a fish, or on the temperature at which water boils. Parliaments cannot repeal laws of nature. Even if a billion people embrace a foolish idea, it is still a foolish idea.

The goal of policy formulation should be to get the right answers. Formulating public policy towards science and technology may be difficult, but if democracy is to take public opinion appropriately into account, good government must discount ignorance and prejudice.

The 18th-century Irish statesman and writer Edmund Burke emphasized the government's responsibility to make such determinations. He observed that in republics, "Your representative owes you, not only his industry, but his judgment; and he betrays, instead of serves you, if he sacrifices it to your opinion."