Median Voter Turnout Fell Below 8% in DAOs With More Than Ten Voters

Most DAOs built for collective decision-making show low participation in collective decision-making. Across a census of 10,541 decentralized autonomous organizations on six platforms, built for governance by their members, median voter turnout per proposal was one-third in DAOs with ten or fewer voters and fell below 8% in every size band from eleven voters to a million, and most voters in the larger organizations took part in fewer than 30% of proposals. These were not passive communities. They were built on an explicit promise of open and horizontal governance. And yet the participation was not there.

The reason participation collapses in communities built for it is seldom the people inside them. It has to do with what the architecture is actually asking them to do.

Much of institutional giving is built on transmission. Foundations publish priorities, and nonprofits write proposals to match those priorities. The flow runs in one direction, from institution out to community, which means the deepest design assumption is that the institution already knows what its community needs.

That assumption is the objectivist move, meaning the quiet conviction that the institution holds the question and the community helps answer it. Everything that follows is shaped by that conviction, even when the surface looks participatory. Town halls become listening sessions for predetermined options, and surveys aggregate preferences against fixed menus. Boards then approve proposals against a strategy the community did not help write. The design is doing exactly what it was built to do, which is transmit, and the community is doing exactly what it was built to do, which is receive.

There is another design entirely. It treats the question itself as something the community produces, and it builds deliberative space, meaning structured time where members can reason and disagree together, where members can test each other on the way to understanding none of them held at the outset. That kind of community changes its mind together, which is not the same as voting together.

For nonprofit and foundation professionals, the practical implication is sharp. A participatory grantmaking process that begins with predetermined funding priorities is participatory in execution and objectivist in design. The real participation happens earlier, in the priority-setting itself, because if your community is invited only after the institution decides what to fund, the architecture is transmission no matter how many town halls follow.

For corporate giving and CSR leaders, the same diagnostic applies. Employee giving programs that gather preferences against fixed program menus collect data without producing understanding. The numbers come back, the strategy does not move, and the gap between participation and outcome is an architecture problem rather than a measurement problem. The platform aggregates without deliberating.

For digital and crypto-native giving leaders, the finding cuts closest. A DAO is a structurally ambitious attempt to build participatory governance into organizational form, and even there participation was low, with turnout falling as organizations grew. The census found something specific. In the majority of DAOs with more than 1,000 voters, more than half the voting power sat with around 1% of voters or fewer, a pattern the authors read as the iron law of oligarchy at work, sharpened by the fact that many DAOs allow governance tokens to be purchased. A minority defied it: 39% of the DAOs ran one-person-one-vote systems, and 22 of those had more than 1,000 voters. My reading of the participation gap is architectural. A DAO whose members engage with proposals in open channels before formal votes is built on a different design from one that only aggregates token-weighted votes, and the second design produces the same participation gap that traditional giving infrastructure produces.

For NGOs and international organizations, meaning nonprofits that work outside government, many of them across country borders, this design challenge shows up twice. NGOs answer to funders who want community input collected and presented as data, and they serve communities whose knowledge cannot be captured in a survey. A conventional needs assessment, where staff visit a community and record what residents say they need, is preference collection at the community level. It asks for opinions and aggregates them. A co-design process asks people to reason through the problem alongside the staff who will design the response, which is deliberation rather than data collection. NGOs that build genuine deliberative channels between funders and the community members their programs serve are operating with a structurally different model from the ones running traditional assessments.

There is one more layer worth naming, because much of the hardest new design work in philanthropy sits at the seam between institutional and digital-native communities. When institutional donors and crypto-native participants try to build something together, the design problem is no longer just deliberative space. It is the work of building enough trust that two very different groups can make a decision neither would have made alone. Without that trust, the discourse stalls. With it, the convergence becomes possible. The seam is where the new design work actually lives.

What deliberative design looks like in practice is structural rather than rhetorical, because it is built into the order of operations rather than added to it. It begins before priorities are set, not after, and it gives the community somewhere to meet that neither institution nor community controls alone.

Bigger platforms and better surveys do not close that gap. The fix is deliberative space, built earlier in the decision sequence, with structural commitment that what emerges from the discourse is what the institution acts on.

Donors and family-office advisors who have watched family philanthropy fail at collective decision-making recognize the design diagnosis immediately, because family foundations can replicate the transmission architecture this post critiques, where the senior generation sets the priorities and the rising generation is invited to recommend grants inside a framework they did not help build, which invites the same low participation the DAO census found. The fix is the same: build the deliberative space before the priorities are set, not after, so that what the foundation acts on is a question the whole family reasoned toward rather than a preference the senior generation ratified.

"Participatory design is not a feature you add to a strategy. It is the architecture that determines whether a strategy can be built at all."EILEEN HERNANDEZ, THE STRATEGIC STEWARD

Share this𝕏 / TwitterLinkedIn

This is the principle underneath the Bridge Method. Communities that build for deliberation produce a different quality of collective impact than communities that settle for informed audiences. The work is to design the bridge, not just the platform.