“Perry’s recent blog for NESTA on ‘Can small-scale deliberative events have legitimacy?‘ is republished here with permission”

As part of Nesta’s Everyone Makes Innovation Policy programme, Talk Shop and Thinking Box collaborated on a project to get people talking about driverless cars. We created a discussion kit – a deck of cards providing information and raising questions – and used this to facilitate discussions about the topic with different groups of people. We make the kits as game-like and as simple as we can.

By the end of the project, some 600 people had taken part. The smallest sessions had half a dozen participants, say members of a U3A (University of the Third Age) group, meeting in one of their homes. Larger sessions had upwards of 30 people, including older Poles in Birmingham; staff in a holiday company in Hythe; and students at a College for the Blind in Hereford. Events usually lasted about two hours; sometimes more, sometimes less. This type of deliberation fills a gap between citizens’ assemblies, where a small number of people consider a topic for several days, and opinion polls, where lots of people give their views off-the-cuff.

As far as possible, we would like to generate results that are useful to policymakers. This blog is about the legitimacy of those results. I’ll compare our results with both assemblies and with opinion polls, beginning with opinion polls.

“Although you can take a nation’s pulse, you can’t be sure that the nation hasn’t just run up a flight of stairs.” E B White for The New Yorker magazine

E B White, who wrote for The New Yorker magazine, summarised thus the risk with polls. With topics that are unfamiliar to most people, such as driverless vehicles, their opinion can be created by the act of asking for it. One US poll showed that a third of adults had views on the Public Affairs Act of 1975. There is no Public Affairs Act of 1975.

Small-scale deliberative events give participants a chance to form a more nuanced view

But are the views from people at our events more considered, more thought through, than responses to opinion polls? One sign that they do is that people gain a better understanding of the topic. On driverless cars, the proportion of people who reckoned they knew at least a fair amount about the topic went up from a third at the start to two-thirds by the end.

However, the best evidence of this comes from an earlier use of our discussion kits, when I was working for the New Economics Foundation think tank. In 2003 we helped the Human Genetics Commission (HGC) understand how people viewed over-the-counter genetic testing kits. We produced a discussion kit and used it both with the general public and with members of the HGC Consultative Panel, people with direct experience of living with a genetic condition.

Simplifying wildly, when using the discussion kits we saw people’s thinking go through two typical stages. First, their initial views softened. To quote someone from an event held in Trento in Italy:

“At the start [of the discussion] you have some kind of prejudice – either because you have some conviction of your own or because you never made your own opinion on the subject. Then during the discussion you realize that other people think differently and this makes you see how “relative” your own opinion is.”

Secondly, people identified dilemmas or contradictions in their thinking, a prerequisite to developing a more considered view:

“[It made me think about] the contradictions/tension between woman’s right to choose and the validity of human beings with genetic ’disorders’.”

In another project, about stem cells, most of the participants identified dilemmas in their thinking and feeling during the event, and over a quarter resolved them.

Our events also help people give reasons for their views. We found that asking people to consider the impact of a genetic testing kit on the recipient and her/his family prompted many people in our events to support considerable regulation. We were pleased that the Commission followed this line of argument, recommending stricter controls than previously – for example, that genetic tests for which the results were likely to have a big impact on the person being tested should be provided only after a consultation with a medical practitioner.

Small-scale deliberative events don’t try to be representative – instead they give deep insights into the views of different parts of the population

I’ll turn now to comparing what we do with citizens’ assemblies. Assembly members are recruited to be a microcosm of the population at large: as such they are very diverse. Our events are different. The participants are often homogeneous, as in the examples at the start of this blog. The diversity we aim for is across the range of events. With Everyone Makes Innovation Policy, we did pretty well in reaching two seldom-heard groups: students and older people with mobility difficulties. We did less well on two others: ethnic minorities – apart from the Poles – and people from deprived neighbourhoods.

The fact that the people who take part in each deliberative event are often homogeneous raises some challenges: for example, we have to design our stimulus materials to introduce alternative perspectives that might more spontaneously arise with a more varied group. However, they also bring some advantages. We get to see more clearly how different demographic groups see the issue. For example, a group of 16 and 17-year-old students from Hereford Cathedral School were evenly split for and against driverless cars at the start of the event, but they were 2 to 1 against by the end. This was due to concerns about hacking. If I was a regulator thinking about whether and how to allow driverless on our roads, I should want to find out if this view was shared by other young people who are the potential users of such cars.

As I said at the start, this project was under the Everyone Makes Innovation Policy programme. We owe it to our participants to make the best case we can that their views should contribute to policy on innovation. I hope I have achieved that.