Distributed democracy: a sketch

No elections, politicians, or referendums. This is the promise of distributed democracy.

After the 2016 election, I began asking how to prevent another Brexit or Trump, and realized this would require a new model of democratic participation. It would mean replacing elections—and the partisan divisions on which they thrive—with direct democratic control.

Conventional thought says that direct democracy is only suited to localities; only so many people can fit in an assembly, only so many voices heard at a time. We once thought republics too must be small, but the American experiment proved that assumption false. Our century demands new experiments in democracy at scale.

The model I propose—let’s call it distributed democracy—not only solves the problem of scale, but focuses political life on the search for a common good. Electoral government by its nature emphasizes conflict, ambition counteracting ambition. This has given the human race more freedom than it’s ever known, but has proven inadequate for solving the national and global problems we now face.

We wanted popular government, and got government by popularity. This isn't the only way.

In the brief sketch that follows, I set aside most theoretical justifications. I will not explain how this model preserves liberty, draws on local knowledge, minimizes the tax burden, fosters healthy civil society, or offers forms of legitimacy lacking from modern electoral systems, though it accomplishes all these. Our political moment calls for triage—let us move away from injustice first and theorize along the way.


Distributed democracy is built on a network of delegate assemblies. Assemblies are small; perhaps 150 delegates at most. An assembly legislates for, and draws its delegates from, the geographical region containing most of those who are affected by a given issue. Call these regions polities.

Each polity is defined by a single issue domain—East Coast fishing rights, for example—and the wider the region affected, the larger the boundaries of the polity. These boundaries are fluid and overlapping, so that a citizen is simultaneously the member of many polities. Issues of human rights, necessarily affecting all, would be decided at the level encompassing the widest geographical area (which today would most often mean the nation-state).

Assemblies are provided with all the time and evidence and testimony needed for deliberation, and are asked to decide a single question: does the law allow for harm?

This question does not require that citizens be expert in any given issue. How could they, given the many matters government must consider? It demands only that they attend to the evidence presented and draw on their experience in judging questions of fairness. Rather than asking citizens to understand every nuance of policy, we ask them to engage their sympathies.

When an assembly declares that the law does allow harm, it then charges the civil service with finding the most effective and efficient remedy. That expert recommendation is then approved by the assembly in another majority vote on the same question, and is regularly reviewed to ensure efficacy. This encourages iterative improvements across time, and adds a self-correcting quality to the process.

Assembly service is mandatory, much like jury duty is today. Citizens serve on just one assembly at any given time, and rotate through assemblies every six years with staggered terms to promote knowledge sharing. Given the limited demands placed upon them, I expect assembly meetings to be regular but rare.

Still, this work requires more of citizens than is customary in electoral systems, and instead of adding to the demands placed on them, distributed democracy allows for a new model of citizenship that splits obligations of residency from citizenship.

Let citizenship be annually renewed by explicit consent. Its benefits? Access to social insurance and the power to help decide the course of law. Its cost? Civic participation. Those who forgo the former can avoid the latter, and simply pay for necessary geographic public goods (defense, infrastructure, etc.) through taxes paid by all those in the polity.

The distributed quality of this model allows for infinite scale. It is as practical for neighborhoods as it is for planetary systems. And since polities are defined by issue domains and not arbitrary political borders, it encourages regional and international cooperation. It is a politics for humans, however far the species travels.


Much work remains, of course. Not only in elaborating the various aspects of this model, but showing it to be feasible, effective, and more amenable to justice than the status quo.

But we can start today by interrogating its early form, identifying the challenges, and asking how it might be improved. I welcome your thoughts.