As the academic debate over the merits of deliberative democracy and deliberative polling heats up following Professor Lupia's initial criticisms (and the response of deliberative polling masterminds Professors Fishkin and Luskin - and part 2), Lupia continues to expand on the need for representativeness:
On the Tomorrow's Europe website, the representativeness of the sample is claimed as a legitimating factor. Here is an instance where, ex ante, you can commit to credible measures of "representativeness."
As leaders of the DP well know, many people decline the invitation to participate in DPs. In some cases, such as DPs held in New Haven, CT, the rate of decline approached 85%. If those who accept the invitation differ systematically in key attributes (demographic, interest in politics, etc.) from people who reject the invitation, claims about sample representativeness can lose their credibility.
Tomorrow's Europe should commit up front to documenting every invitation made and then to a publicized aggregate analysis of the relationship between those who agree to participate and those who do not.
It is particularly important that they be prepared to conduct short interviews with those who decline the invitation so that such inferences about differences between those who accept and those who reject can be based on evidence. To do otherwise, will be to reduce the basis of "representativeness" claims to speculation.
Alternatively, the credibility of the project may be helped by scaling back the claims about representativeness and simply be transparent about how recruitment will work and then offer the best statistics you have about who agrees to participate. That way observers can draw their own opinions about the quality of the sample.
