Representation in Lottocracy - A trade-off between Obligation and Stratification

In the past years, the use of randomly selected minipublics as a savior of democracy has received much attention. The most radical of these suggestions, lottocracy, argues for completely replacing elected politicians by randomly selected citizens (Guerrero, 2014). In this paper, I analyze which concept of political representation underlies the lottocratic proposal and argue that stratified sampling, which is currently employed to select minipublics, undermines the supposed advantages of lottocratic representation. I begin by explicating that representation in lottocracy is gyroscopic and indicative, as opposed to promissory and responsive representation in elections (Mansbridge 2003, Pettit 2010). The central assumption of lottocracy is that selected citizens do not act upon a mandate they receive from those they are supposed to represent. Instead, the selected are essentially like the represented in relevant regards. Because they participate in a process of information and deliberation, they become better informed proxies of the represented. Thus, they indicate to those who are not selected which policies they would have supported, had they received the same information. However, I argue that stratified sampling undermines these promises of gyroscopic and indicative representation. On the one hand, stratified sampling is based on stereotyped thinking which leads to a misrepresentation of certain characteristics. On the other hand, stratified sampling undermines the equality random selection promises. I point out that the advantages of lottocracy diminish under stratified selection, because it introduces a strong self-selection bias. Using empirical data from the French Citizen Convention on Climate (2019), I show that such self-selection negatively impacts to what extent non-selected citizens accept the recommendations of a minipublic as representing their interests. The promises of lottocracy, and of minipublics more generally, hinge on the selected sample being a truly representative sample of society. However, stratified selection fails to generate such a truly representative sample, because it disregards those who reject participation. But in order to be the savior of democracy lottocracy promises to be, it needs to involve especially those who have turned their backs on political participation. I conclude the paper by pointing out that purely random selection could only be employed if assuming political offices upon being selected was mandatory. To generate a truly representative sample, the selection of participants would need to be fully random and compulsory. While obliging people to take a political office seems to be an objectionable interference with personal liberties at first sight, I provide three reasons why such an obligation might be justified at least as a moral obligation towards other citizens. First, I argue that those who do not accept to participate undermine their fellow citizens chances of being considered in policy-making. Second, I argue that the advantages of living in a democracy generate a moral duty to contribute to this democracy. Third, I suggest a new perspective on political offices and interpret them as unpleasant duties that should be shared equally among all citizens.