Finding our way together – through innovations in voting (Collective Coherence Part One)

Groups use various methods to live together, to think together, and to come together to get work done. My new series of essays will explore some leading-edge ways of doing that, starting in this message with some intriguing new ways for people to vote that could make the dynamics of choosing more efficient, more insightful and more fair.

How do we find our way together?

Among the implicit questions underlying my work and the work of thousands of other change agents is this: How does a group of people – from a couple to a community to a country and beyond – develop more or less coherent ways to think and act together so they can successfully tackle shared problems and aspirations?

In most cases this involves some form of decision-making, some way to choose from a set of options. On the surface, the easiest approach to this is for an authority figure to make the decisions. This happens a lot, of course. But when we do participatory or democratic decision-making, we most often use voting.

Some innovative approaches to voting are sprouting up all over the world. Not only that: We’re also seeing different ways of generating group coherence that don’t use voting at all, some of which don’t even look like decision-making. Most of these latter approaches involve developing shared understandings that then shape the behavior of the group’s members. In the process, they usually welcome divergent views and ride the waves of emergent dynamics. Unpredictable, often shifting insights and possibilities show up as such a process unfolds, dancing with reality and folding back into the group’s explorations to generate in an ongoing collective learning adventure.

Other factors intersect with both the decision-making and shared-understanding approaches. These warrant special attention from the wisdom-generating perspective. They include:
* how much information, exploration, creativity, and resources are involved or allowed;
* what kinds of power dynamics are involved;
* how much coherence is expected (and how solid is it); and
* how does the group handle differences and concerns.
These factors play out for the group as a whole and even within and among individual members.

I have not figured out how to chart or model all this in a single coherent frame. But I have noticed the above factors dancing around to various degrees in each of the innovations I describe in this series of posts. They offer interesting dimensions to explore as we expand our understanding of how it all works.

This series of three posts will cover
1. Variations of voting
2. Finding our way through deeper understanding and
3. Scaling up our collective sense-making capacities.

This post will cover the first:


Voting – usually involving majority rule – is considered the fundamental process defining a political system as a democracy. In this post I’d like to highlight some fascinating variations on voting. In later posts I’ll point to some other ways we have of arriving at collective conclusions.

In exploring voting I first want to acknowledge some generic methods that are already widely used, debated and understood. At the end of this essay I’ve noted some of these and some ways to improve them within existing electoral systems. I’m putting them at the end so that we can leap ahead right now to approaches that can creatively shake up our sense of what voting is all about.

So I invite you to consider what approaches and innovations in voting can help us create better systems for collective choice….


Liquid Democracy is a form of electronic democracy where voters can delegate their vote to someone else for a specific choice, or for all their votes regarding a particular subject or issue. Ideally, the person to whom a voter delegates their vote is someone who knows more about the topic or choice than the original voter does. At any time, the original voter can call back their delegation, reclaiming their right to vote on that formerly delegated question or topic. This is called “liquid” because there is a fluidity to the electoral ecosystem that can enable votes to flow towards sources of greater understanding without diluting the individual franchise. That potentially wisdom-supporting dynamic may or may not occur in practice, but the possibility remains, offering the prospect of a wiser ballot-based democracy.

Random Sample Voting (RSV) involves creating a temporary, demographically representative random pool of a few thousand citizens selected to vote on a specific issue or public office. The selected voters could be provided with balanced information and advocacy materials selected or crafted to be understandable to ordinary people. Among the many possible advantages to this approach:

(1) Engagement: The citizens chosen, recognizing the special responsibility they’ve been given and the greater weight of their vote – would likely “rise to the occasion” and make a special effort to learn about the issue or candidates and actually vote, generating far higher and more informed level of participation than happens in most US elections. That means the results would more accurately reflect the informed will of the population than the unwieldily mass participation electoral processes currently used.

(2) Efficiency: It would save massive amounts of resources compared to traditional election processes in which all qualified voters can – but seldom do – participate. This efficient use of resources would permit more and higher quality elections on distinct issues and candidates than our current systems. It would counter the absurdity of presenting millions of individual citizens – whose lives are already crowded with other demands and desires- with dozens of major choices on complex issues they simply don’t have time to attend to.

(3) Integrity: It could be set up so that anyone could readily verify the integrity of the random selection, the voting process, and the outcome, without compromising voter privacy. The random selection would reduce many (if not all) sources of corruption, and the few that remained could be better policed.

(An extensive white paper about the RSV approach deals with a number of objections to the idea.)

An RSV process could theoretically be convened by anyone and be trustworthy to the extent its sponsorship, members, selection process, and voting process were fully transparent and fair. If this were done by a group other than the government, the outcomes could be designed to influence government policy through publicity, lobbying, activism and other forums of pressure to do what the informed public wants to have done.

In a visionary application, RSV could play a major role managing a randomly selected “people’s legislature”. For example, randomly selected panels of voters could shape the rules that govern such a legislature’s operation. RSV panels could democratically define criteria for people to be added to the pool from which legislators get randomly chosen, thereby addressing concerns about whether they were qualified to serve. RSV panels could also review the citizen legislators each year and choose (for example) 25% of them to be replaced (by lottery from the above pool), thus weeding out the most out-of-touch or corrupt(ed) among them, while allowing the most popular and ethical to continue to serve.

Other ways of combining Voting and Random Selection – There is interesting precedent for Random Sample Voting. Most Italian city states during the late medieval and early renaissance period – notably Venice – used the brevia system that combined elections with random selection. The brevia was a randomly selected “electoral college” sequestered like a jury, who then collectively elected people into public office, often through a sequence of nominations and elections until an adequately agreed-on candidate was chosen for each office.

The city of Florence reversed this process, using an approach called scrutiny, in which qualified citizens voted for a pool of candidates from whom public officials were selected at random.

Both these approaches of indirect election – which probably influenced creation of the US Electoral College – were used by elites to maintain overall control without devolving into embattled factions. For a while various regimes also included more working class people in the process, but the elites usually managed to resume some form of control, often by changing who was included in the pool from which electors or officials were selected. But the brevia and scrutiny both used sortition (random selection) to constrain manipulation and factionalism while maintaining the judicious qualities of voting. Forms of them designed to reduce special interest domination could be used today.

The Iroquois Confederacy – indigenously known as the League of the Great Peace – has existed for many centuries. Its women choose the men who make the communal decisions and who can be removed at any time by those women for lack of integrity or for harms done to the community or the world around them. Councils of male clan, tribal, and League chieftains make decisions at the levels of governance for which they were appointed. But clan mothers inherit the power to choose or replace the chieftains of their clan. A chieftain’s authority is usually symbolized by a crown of deer antlers until it is ritually returned to the clan mothers for reassignment.


Voting always involves voters deciding whether they prefer one option over others presented to them. To the extent such decisions involve calculating how MUCH the voters prefer each option, they involve rating. Some voting methods make this rating exercise explicit, and use the results creatively in the final choice process. Here are examples:

Systemic Consensus is a voting system in which each option being considered (for choice or prioritization) is rated according to each voter’s OBJECTION to it. This is, of course, the reverse of most voting systems. A voter rates an option with a 10 if they absolutely don’t like it and with a zero if they really love it. Everyone’s votes are added up and the LOWEST-rated item is the winner, since it has the least resistance. (In the case of voting for priorities, the several lowest-rated items become the priorities.) In a variation imagined by Martin Rausch and myself, this system could be used not to establish a final choice to but sort out which choices could be most efficiently examined to address whatever concerns were felt by those who had objections to the otherwise least objectionable items. Several new versions of each deliberated proposal could be created through that process and then subjected to the systemic consensus process for final decision. (A number of processes that focus on concerns will be covered in Part 2 of this series.)

STAR voting (STAR = Score Then Automatic Runoff) combines option ratings with a virtual runoff election in which computers calculate a winner from the ratings. Here’s how it works: Each voter rates each option on a 0-5 scale, depending on how much they like it, with 0 meaning total dislike and 5 meaning they like it a LOT (the opposite of systemic consensus, above). As noted, this is not *ranking*, as in “I like B better than A and A better than C.” This is *rating*, much as you rate Amazon products, individually expressing how you feel about each one. The ratings for each option are then added up (revealing the electorate’s two most supported options) and a “run-off election” is mathematically held between those two top options, such that each person’s single overall vote goes to whichever of the two main options they rated higher (reflecting the “one person one vote” aspect of democracy). If a voter rated both options the same, of course, then their final vote balances out or is simply not counted, although they will almost always get one of the options they liked most. (Again, if the options are ideas or possibilities rather than candidates, a STAR vote could be followed by deliberations exploring how to address any concerns voters may have had relating to the top-rated options, so they can be improved.)

STAR voting differs from the more familiar Instant Runoff Voting (IRV), in a number of significant ways, notably that STAR voters rate each option separately whereas IRV voters rank the options in relationship to each other. And in STAR voting the runoff is automatic, whereas in IRV the runoff only happens if an option or candidate doesn’t get a majority of top-ranked votes in the first count. But since both STAR and IRV allow voters to effectively cast relative votes for multiple candidates, they both help avoid the problem of “wasting your vote” on “spoiler candidates” in systems that support the dominance of two major parties. And by opening the door to nonpartisan and third party candidates, they also help constrain polarization and partisan gridlock.

Pairwise Comparisons: Participants compare various factors or options in matched pairs, often against how they satisfy some specific or general criteria, and usually in order to generate a ranked list of the best factors or most popular options. Although this is often done systematically in a linear matrix (such as here and here), my experience and interest in this approach – for the purposes of this essay – is with more generalized crowdsourcing applications – like CoDigital and AllOurIdeas – and with group processes like Thirty-five. (Note that although this linked version of Thirty-five is for a special debriefing use, the process can be used to prioritize responses or options around any question, as given in the example below.)

Example of CoDigital: Several years ago the National Coalition for Dialog and Deliberation used CoDigital to ask its members what they wanted to talk about and do in its upcoming biannual conference. Dozens of NCDD members posted ideas. The CoDigital algorithm then repeatedly presented them with pairs of ideas that had been posted by others, asking which one of each pair they thought was most important. This pair-judging process was repeated over and over. Participants were also given the opportunity to revise other people’s ideas and CoDigital would then offer up different edits of an idea to choose between. Over a couple of weeks of collective (but individual!) online posting, editing, and judging, a prioritized crowdsourced list of over 100 ideas emerged which then guided the conference organizers (presumably to the increased satisfaction of the attendees). (GroupMap is another shared workspace which invites collective exploration of an inquiry through brainstorming, rating, clustering, etc. I’m not going into detail about it here since it has many features and I don’t have experience with it.)

Example of Thirty-Five: In the public engagement stage of designing Eugene, Oregon’s Climate and Energy Action Plan, we used a World Cafe to generate and explore possible actions the City could take. After the Cafe, we gave participants 5” x 8” cards and instructed them to clearly write – in no more than a dozen words – THE ONE MOST IMPORTANT THING they thought the City should focus on, stressing that it should be just one thing. I, as facilitator, gave them 5 minutes to do that and then had them mill around for a minute rapidly exchanging cards without looking at them. Then I stopped them and had them find a partner with whom they had one minute to distribute 7 points between the two cards – distributions like 6 + 1, 7 + 0, 3 + 4, etc. – and to write their agreed-on point vote on the back of each card. (The one minute limit required cheerleading, urgings and time warnings from me as facilitator, which kept the energy high!) I then had them repeat the milling-and-exchanging activity followed by the micro-deliberation and pairwise distribution of seven points again, for a total of five times. At the end I told them to turn over whatever card they held and add up the five numbers there. Then I called out: “Does anyone have a 35?” [which is the highest possible vote total, which almost never happens]. “No? Does anyone have a 34?”…. etc. – and went down the numbers until I hit a response: “Ok, there’s a 29! What is on your card?” On a chart pad I wrote the top-voted item the card-holder called out. Then I continued calling out the possible vote totals until we had accumulated about a dozen highly-rated items on our prioritized chart-pad list. The whole Thirty-Five exercise took about 15 minutes, even with about 80 people. Although we had identified priorities, we published all the ideas and their votes online for the whole community to review and comment on afterwards.=

Improvements in traditional electoral procedures

For starters see Wikipedia’s review of electoral systems and initiatives that seek to educate voters about politicians (like Project Vote Smart’s full-spectrum database on candidates, issues and ballot measures) and to make candidates answerable (like FCNL’s questions to ask candidates). There are also many innovations and fixes proposed for (primarily) U.S. electoral systems, such as Fairvote’s electoral solutions and innovations, as well as broad consideration of the logistics of electoral integrity (including issues with voting machines), electoral reforms to benefit people with disabilities and debates about leading-edge technologies potentially useful in elections, such as blockchain (which has environmental/energy concerns that may be eased by newly emerging alternatives).


Just because voting is quantitative and usually majoritarian doesn’t mean it has to be reductionist. We can take into account our deeper individual sense of preferences in ways that add up to wiser versions of collective insight and priority. We can also distribute our power of choice in various ways that help generate greater collective wisdom and less opportunity for manipulation.

But there are also ways to bring forth our latent capacity to move beyond choice itself to co-create powerful shared understandings. I’ll address that in my next message in this series.

* * * * * * *

Summer 2018 Co-Intelligence Institute Fundraising Campaign note: So far 11 people have supported us with $1135. Join them in supporting our work. Our target is $25,000. 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.


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 *