Hey! Let’s bring some sanity to choosing our public servants!

The tragic state of American and other political systems should inspire us to look more carefully at alternative ways to select our leaders. Here we consider ten creative ways to use random selection to make our democratic republican systems more competent, efficient, and free of corruption.

Please contribute what you can.

The current US election campaign has not been a good advertisement for democracy-as-usual. I won’t bother here to enumerate why. The more informed you are about American political life, the more reasons you could give about why it’s crazy. Even the ill-informed are filled with righteous dissatisfaction, and rightly so, if not always for the right reasons.

This crazy situation makes it especially important – and potentially very productive – to reflect on how we might do our elections, politics, and choosing our public officials differently.

Most of us are familiar with arguments about the appalling role of money, the superrich, and corporations in elections. Democratic Socialist Bernie Sanders and both billionaire Donald Trump and multi-millionaire Hillary Clinton are speaking out about that issue, which is a good sign but still leaves many of us wondering about how much the situation is going to change when one of them (most likely) gets into office.

In addition to the overriding issue of campaign finance reform, we see proposals to reform the electoral process itself. These range from eliminating the Electoral College and electronic voting machines (“let’s get back to popular votes with verifiable paper ballots!”) to switching to proportional representation and preference voting (i.e., we vote our preferences for multiple candidates, enabling electronic “instant run-offs”) which gets rid of “wasting our votes” and allows third party candidates to make their true level of support visible. This kind of reform also includes better access to voting (including making Voting Day an official holiday) and citizen control of redistricting activities.

If you want to learn more about reforms like that, both FairVote.org and Common Cause offer good resources for specific proposals and advocacy.

I think these are all important reforms. But I just read Simon Threlkeld’s “Should Citizen Juries Choose America’s President, Congress, Governors and State Legislators?”* It reminded me that there is a wildly different approach to choosing who will lead us. Threlkeld presents some pretty compelling arguments for using random selection.

He’s not the first person to suggest this. In fact it has a long history. Political scientists have a word – sortition – for the use of lotteries in politics.

Now, I realize it is easy to dismiss this idea as ridiculous – even dangerous. Thanks to our tendency to equate democracy with elections, we are unlikely to give political lotteries even a moment’s consideration.

But when we confront the tragedy of our political wasteland – a polluted, desiccated fire danger where healthy political life could and should be flourishing – perhaps we might at least pause a moment to consider some mind-expanding alternatives. Sortition is actually quite interesting, as I think you’ll see . . .


Threlkeld first observes that

In Classical Athens, widely considered the birthplace of democracy, a broad range of decisions were made by juries drawn from the citizens by lottery. The Athenian juries kept a great deal of decision-making firmly in the hands of the citizens, and prevented elite rule.

He then suggests that

Ideally, politicians would be chosen in a way that is very democratic, well informed, and independent from moneyed interests and billionaires, with political independents [who he notes constitute 45% of the American electorate] being on a level playing field with party nominees, with no portion of the public being underrepresented, and with candidates not being dependent on the media to get a fair hearing.

Sounds good so far…

In light of these factors, he looks at four proposals that could improve our leader-selection process with random selection:

Proposal 1. An appropriately sized, randomly selected jury could select a public official from the full field of aspiring candidates after studying their written applications, interviewing them and their supporters and opponents and otherwise informing themselves about who each candidate is. This would apply at least for the office of president, but could also be used for governors, Congresspeople, state legislators, etc. These juries would not be permanent bodies: Different juries would be convened for each office, each election cycle. (Think of this as a political version of the hiring process that happens all the time for most jobs, except in this case the boss is “We the People” and the Human Resources Department doing the hiring for the boss is a representative sampling of We the People, convened as a group to whom special support is given to make good decisions.)

Proposal 2. Now we may want a randomly selected jury (or a commission assigned by the jury in Proposal 1) to design the rules and procedures governing how the jury in Proposal 1 would operate. After all, citizen control of this rule-setting activity would be important because we want to protect this powerful leader-choosing institution from the special-interest manipulations that colonize our current political culture and that are inevitably attracted to any system of concentrated social power such as government, especially at national levels (and transnational levels, as our so-called “trade agreements” demonstrate!).

Proposal 3. If we don’t want to risk turning over the selection of our leaders to a jury of our peers, we could at least convene randomly selected juries to establish (and/or oversee) the rules governing popular elections, such as those currently established by state legislators, election commissions, and political parties. Such rules have become controversial in the 2015-2016 U.S. election cycle both in the party primaries, caucuses, and conventions and in state-level battles over voter rights and restrictions. Having a group of ordinary citizens get informed enough to fairly monitor those arrangements might not be a bad idea.

Proposal 4. At a lower level of influence, we could set up randomly selected juries to distribute our public election funds (such as those collected using state and/or federal income tax forms) among the competing candidates for major offices like president.

So those are Threlkeld’s ideas. I see a few other interesting possibilities he doesn’t mention:

Proposal 5. Randomly selected juries could do any needed redistricting to replace the unfortunately common practice of incumbents manipulating legislative districts to favor particular political parties or themselves (a practice known as “gerrymandering”). Such a use of randomly selected juries is one way to ensure citizen control of this arcane but very potent source of political power.

Proposal 6. Using a similar process to Proposal 1 above, we could use a randomly selected deliberative jury – instead of political parties – to decide which two or three candidates for a particular office would be put on the ballot for us to choose among. This would leave the final choice in the hands of us voters, while at the same time improving the quality of candidates presented to us, thereby improving the quality of the ultimate winner. Candidates beholden to special interests and not aligned with our community’s values and the public good would likely have a hard time making it through such an informed, deliberative filtering process.

Proposal 7. Imagine a randomly selected jury doing the process described in Proposal 1, but instead of directly choosing the candidate who will take office, they would tell the voters who THEY think is the best choice and why.

This was done in 1990 by the Jefferson Center and the League of Women Voters using the Citizens Jury process to evaluate candidates for Minnesota state governor.

First, six twelve-person regional juries evaluated the candidates running for governor in the primaries. After the primaries those regional juries picked 18 citizens from among their number to constitute a new jury to evaluate the candidates running in the general election. This final jury provided a LOT of information on those three candidates, including their own preferences.

Sadly, this creative use of ordinary citizens to evaluate candidates on behalf of all voters upset some unknown influential person(s) and suddenly the IRS showed up threatening to withdraw the Jefferson Center’s nonprofit status if they ever organized another such candidate evaluation. So they never did another one. (We could, of course, change that if we wanted to and organized around it…)

In a variation on this theme, professor John Gastil in BY POPULAR DEMAND: REVITALIZING REPRESENTATIVE DEMOCRACY THROUGH DELIBERATIVE ELECTIONS proposed using randomly selected citizen panels to evaluate candidates for legislative and non-legislative positions. His proposals are so intriguing and well thought-out that I want to share summaries with you here. (I recommend his whole book – which goes well beyond these proposals – as a real eye-opener.)

For legislative positions Gastil recommends (a) having a citizen panel examine the most recent legislative session and pick the ten pieces of legislation they feel were most important (whether or not any of those bills passed). (b) Incumbent legislators who voted on those bills would already have their votes on record. Legislative candidates who did not vote on those bills would record how they would have voted had they been in office. (c) Then ten new citizen panels would be convened to evaluate each of those pieces of legislation and decide whether they would support or oppose it. (d) Summaries of all this information would be provided to all voters and (e) the legislative candidates would be listed on the ballot in order of how well their votes (actual or reported) aligned with the preferences of the citizen panelists, with the percentage of alignment printed on the ballot by their name. (Note:  This would be a “we-the-people-general-interest-group” equivalent of the partisan candidate ratings produced by various special interest groups like the NRA or the Sierra Club in most elections.)

For non-legislative positions Gastil suggests (a) having candidates, in the presence of a randomly selected citizen panel, testify and cross-examine each other on what criteria should qualify an applicant for the job they all want to be chosen for. (b) From the criteria discussed, the citizen panel would identify the set of criteria they think should be used. (c) Then the candidates would testify and cross-examine one another about how well they each satisfied the panel’s chosen criteria. (d) Then, using information from the candidates’ debate, the panel would create arguments for and against each candidate. The panelists would also indicate which candidate they each favored. (e) Their analysis of the arguments and their votes for the candidates would then be distributed to all voters, and (f) the candidates would be listed on the ballot in descending order of the votes they received from the panelists, with their percentage printed by their name.

In both of these cases, candidates whose views most closely align with the views of the WELL-INFORMED general public (as represented by the votes of the citizen panelists) would be explicitly favored in the electoral process, without depriving voters of the right to make their own final choice. Rather than feigning neutrality, this approach would generate an explicit electoral bias in favor of the general good, in contrast to the usual implicit bias in favor of the political establishment and/or special interests.

Now let’s look at a few totally different proposals.

Proposal 8. We could randomly select a group of citizens who would – with studies and interviews (as in Proposal 1) – form up a pool of a dozen or more particularly qualified and attractive candidates for a public office, from which the final winner would then be randomly selected – preferably with much fanfare to celebrate and to help the community identify with him or her as their servant leader in that office. It would be like a current politician’s victory celebration, but with the selectee having simply qualified to serve rather than having won a battle.

Proposal 9. Now this one we get to consider the ultimate in sortition: Imagine using random selection not to choose citizen panels or juries, but to directly choose who is going to hold a particular office. In this approach, there would be no intermediate interviews and deliberations. Although this is how 90% of public offices were staffed in ancient Athens for 200 years, societies and governments have grown considerably more complex since then. So if we’re not going to do a formal citizen deliberation, we’d probably be wise to establish certain criteria for a given office so that anyone who was randomly selected would at least satisfy those office-specific criteria. Anyone who fit those criteria could apply to enter an evolving pool of qualified applicants from which the final officeholder could be randomly selected each selection cycle.

To be fair in thinking about criteria for randomly selected public officials, we should consider the qualifications that successful candidates have in this day and age, thanks to the kinds of tests our current system puts aspiring politicians through. To succeed, most modern politicians have to have wealth and/or good fundraising capabilities, a big ego, public charm and charisma, smooth inauthenticity, deal-making skills, and media manipulation expertise – and, at the national level, usually a law degree. Do we actually believe these are the ideal skills for a public decision-maker or administrator? With our answer in mind, do we think random selection from a pool of thoughtfully qualified citizens would do better or worse at producing good public servants?

My current thinking is that some very basic qualifications would generate a good pool from which to randomly select public servants, AND that we don’t have to get too obsessive about criteria in order to do a much better job than our current system does at picking qualified people to serve and govern us.

Now for Proposal 10. Just to complexity things a bit, let’s look at Renaissance Italian city states, where random selection was used by elite families (and later by the emerging merchant class and powerful guilds) to make it harder for any one power center to dominate the others. While this approach was a far cry from the citizen-based (and “voice of the gods”!) logic that motivated ancient Athenian democracy-by-lot or the representative sampling logic that underlies most modern sortition proposals, it opens up a range of other interesting approaches to using random selection in our electoral process. One of the interesting features of the Italian approach was the many iterative layers of random selection and election undertaken within one selection process. For example, a large group of “electors” might be chosen at random to elect a smaller group who did a random selection for a third-level group who would then chose the final leader. So we could imagine (again just for example) political parties using such a process – or any of the previous approaches – to each produce several candidates who are then subject to random selection or to general election for the final job.

In short, there is no end of possibilities. I suspect you could come up with several more.


So my point here is not to advocate for any one of these approaches. My point is to demonstrate that the PRINCIPLE of random selection in politics – sortition – has many applications, any one of which is worthy of consideration as to the role it might play – alone or in connection to elections – in whatever larger systems we use to choose our public servants and political leaders.

I also want to note and stress that ANY of the approaches listed above would almost inevitably enable more informed selections of public servants who, in the vast majority of cases, would be more qualified, responsible, and free from special interest manipulation than those chosen by our current electoral process.

As different as these proposals are from each other and from our current ways of thinking about and doing elections and democracy, I suggest they are worthy of our consideration if we wish to have governance that is more effective, responsive and ultimately wiser than what we have now.

I have to admit that when I look over the current seriously dysfunctional electoral landscape in the U.S. – and in many other countries in a similar or worse condition – I can’t help thinking that this kind of alternative approach could, quite literally, make all the difference in the world.


* Aside from Threlkeld’s main message about sortition, I recommend his section on “Problems with popular elections” as worth a serious read.

* * * * * * *

Summer 2016 Co-Intelligence Institute Fundraising Campaign note: So far 50 people have supported us with $3550. Join them in supporting our work. Our target is $25,000 – and we have been given a challenge grant: If we can raise $10,000 the Kailo Fund will grant us $10,000 more. As you can see, we can use your support. Please donate now. It will make a big difference and your donation is fully tax-deductible in the U.S.

If you wish, you can donate through our crowdfunding site for perks!

Tom Atlee, The Co-Intelligence Institute, POB 493, Eugene, OR 97440
Calling forth the wisdom of the whole for the wellbeing of the whole

Please support our work. Your donations are fully tax-deductible.

Leave a Reply

Your email address will not be published. Required fields are marked *