Diversity and representivity

Some question has arisen concerning my proposals for selecting public officials using a multi-stage process of alternative sortition (random selection) and merit selection, similar to the Venetian model, whether the process is only suited for selecting an executive, but not conducive to diversity or representivity in a legislative body.

The value of diversity is to bring arguments and facts into the deliberative process that have merit but might otherwise be overlooked. It is a mistake to conflate it with representivity, which is about allowing minority interests to be heard during deliberation, exercise influence proportional to their numbers, and the perceived legitimacy of decisions made. Any reduction in the number of decisionmakers, necessary to enable deliberation to be efficient, is going to reduce diversity and representivity among decisionmakers to a large extent, but if they are selected wisely, and are wise, they will seek out diverse and representative opinion during the course of their deliberation. That is what hearings and letters from constituents are all about, if attention is paid to merit. (Even nomothetai have constituents.)

There are two  main meanings of representativity: (1) Having a distribution of attributes similar to that of the pool from which representatives are drawn; and (2) Being accountable to constituents for their conduct and in the ways they represent their interests. Election, and re-election, are perceived as critical ways to hold representatives accountable, in a way they would not be if they were selected entirely at random and served for only short periods of time. The desire to get re-elected, or to gain higher office, requires building a reputation for competence and trustworthiness, which in turn enables building influence. On the other hand, election, especially when the result depends in large part on expensive campaigns, is susceptible to undue influence by special interests who focus their financial and other support more than does the general public. A design principle, therefore, is to enable selection of the most reputable without also imparting excessive influence to them. This dictates a multistage selection process in which sortition alternates with merit selection based on things like reputation, professional achievement, or perhaps performance on tests. Accountability could be realized in several ways, such as with retention or recall elections. Let trust go to the trustworthy.

The general population may mostly lack domain knowledge of the issues, but they can do a good job, if asked to deliberate about it, of selecting those among themselves who are wiser and more knowledgeable about policy matters. By having successive rounds of sortition and merit selection, it can be expected that, unless they are all caught up in unthinking passion or devotion to charismatic leaders, there will be an emergence of the wisest and most knowledgeable among them to the final decisionmaking positions, and that is the best we can reasonably hope for, given the limitations of human beings. The only way to overcome that is to turn over decisionmaking to machines, and we are not ready for that as a species or as machinemakers.

My solution for legislative bodies is a bicameral system using different selection processes. One body, call it the Senate, would use the multi-stage sortition/merit process, and the other would use proxy voting http://constitution.org/voting/proxy_voting.htm . The two houses would combine the advantages of each process, allowing diverse and representative voices to be heard, but also serve as a check on one another, to block a rush to judgment.

I would also apply the sortition/merit process to selection of judges, much the way we are already supposed to do for trial and grand juries. Judges might be appointed for life, not to a particular court, but to a general judicial pool from which they would be drawn at random to serve on courts and hear cases.
It is all set forth at http://constitution.org/reform/us/con_amend.htm