Jamie Whyte proposes a "politics should be like a graduate seminar" model of democracy that strikes me as quite curious. The quote that motivates my thoughts is the following:
But how can a bad policy be good politics? What defect in the electoral system can explain this? The most popular explanation these days is the malign influence of “special interests”. Perhaps there is something in this. But a more fundamental defect is always overlooked, presumably because it is mistaken for a virtue of modern democracies. The reason so many bad policies are good politics is that so many people vote: about 62 percent of adults at the last general election, both in Great Britain and in the United States. The best way to get more sensible policies would be to reduce the number of voters to less than 0.01 percent of the population.
The logic of Whyte's call for fewer voters rests on the assumption that ignorance is the primary obstacle to good policy. Consequently, if we raise the probability that each voter is decisive (by having 12 voters), each voter will acquire the knowledge needed to make informed choices. More informed voters select candidates who make better policy.
I believe this reflects a fundamental misunderstanding of politics. Politics isn't about efficiency (the dispassionate search for the optimal policy); it's about distribution. Once one recognizes that politics fundamentally is about who gets what, bad policy ceases to puzzle and ignorance becomes irrelevant. The classical case against democracy isn't that the masses are stupid. The classical case is that an empowered and relatively impoverished majority will redistribute from the relatively wealthy minority. The real puzzle of modern democracy, I would suggest, isn't that it produces bad policy; it is the extent to which massive redistribution of wealth hasn't happened.
The number of people who vote, therefore, is irrelevant because people vote their interest. And one needn't be well informed about the finer points of Ricardo-Viner to know whether trade helps or hurts you. Selecting 12 representative individuals doesn't change this. These 12 voters might better select the candidate most likely to produce their desired distributive outcome. They will not engage in a joint effort to maximize social welfare.