Forty-seven days
In March 2019, Taiwan faced a regulatory crisis over ride-hailing that had paralyzed the legislature for two years. The Digital Ministry deployed vTaiwan. Over four thousand citizens participated. An open-source tool called Polis surfaced consensus that crossed factional lines. The legislature adopted the framework. Total time from deployment to policy: forty-seven days.
We are not theorizing. We have run this at national scale.
Representative democracy is a bandwidth limitation, not a principle. In the 18th century, the only way to aggregate preferences was to elect a few hundred people to argue on everyone’s behalf. The communication problem has been solved. The engineering solution persists because the engineers benefit from it.
Iceland crowdsourced a constitutional draft in 2011. Ireland’s citizens’ assembly broke a forty-year deadlock on abortion. Paris and Madrid run participatory budgets with real allocation authority. In each case, citizens given structured information produced outcomes more moderate and more broadly supported than what legislatures had managed.
The epistocrats see voter ignorance and conclude the public cannot govern. We draw the opposite conclusion: the public cannot govern through a system that gives it no practice. Deliberation produces informed citizens the way practice produces musicians. The polycentric camp’s critique is the sharpest: our tools still assume a central authority implementing the results. vTaiwan worked because a sympathetic minister championed it. That dependency is our next design problem.
Where we concede ground: Participation skews educated and tech-fluent. The gap between who participates and who is affected is unsolved.
What would change our mind: A permanent digital body with binding authority produces outcomes no better than the existing legislature over ten years.
Read the full synthesis: What does governance need to become?