In search of the general public

Sciencewise recently ran an event with the Dana Centre entitled Whose Science, which explored the role of public engagement in science and the possibility of a People’s Panel. Among many of the issues raised by participants was the ‘problem’ of lay panels - once participants become well informed, they cease to be ‘lay’, and lose their representativeness, or ‘go native’ as one blogger has put it. They can no longer be described as ‘the general public.’ However, others at the event argued, that far from being a problem, these informed opinions are exactly what makes lay panels so valuable.

This comment brings to the fore an issue that is often confused in public engagement and dialogue in general. Some methods of deliberation, most notably deliberative polling, rely heavily on representative sampling. The father of this technique, James S. Fishkin, argues that participants should be a demographic cross section of the population, mirroring political views, socio economic background and race. Their conclusions can then be taken as the conclusions that the whole population would reach if they had the opportunity to become more informed and deliberate the issues.

However, while Fishkin sees his deliberative polls as social research, with the aim of understanding public views, some of the best public dialogue is better described as the practice of active, deliberative democracy. The latter is more likely to involve ongoing communication between actors, changes in opinion over time and participants acting alone or in groups on deliberated opinions. These much messier activities, though harder to achieve, can be the most valuable for overcoming the most complex problems. Here, those who participate are those motivated to do so. Representativeness is unattainable and, indeed, is not the point of the exercise. What the comment at the Dana event highlighted, however, is that there too often remains a false expectation that a public dialogue exercise is, should be or claims to be, representative.