Skip to content
Citizen Infrastructure

vTaiwan and the Uber question (2015)

In 2015 Taiwan faced the same Uber fight as everywhere else: a ride-hailing app pitched against licensed taxi drivers, with regulators caught in the middle. Through the vTaiwan process, the public used Polis — a tool that maps where people actually agree — to converge, over about three weeks, on a coherent set of principles that were then written into law. Uber operates as a legal, licensed fleet in Taiwan to this day.

At a glance

Section titled “At a glance”
Process typeOnline deliberation using Polis (opinion-mapping), feeding a regulatory decision
When2015
WhereTaiwan, online + facilitated meetings
Run byThe vTaiwan process, seeded by the g0v civic-tech community and government reformers
ToolPolis — agree / disagree / pass on short statements, with no reply button
The questionHow should ride-sharing like Uber be regulated alongside existing taxis?
ResultA coherent consensus passed into law; Uber legalised and local taxi co-ops protected

What happened

Section titled “What happened”

Polis works unlike a comment thread. People write short statements; everyone else can agree, disagree, or pass — but there is no reply button, so there is no room for trolls or pile-ons to grow. As people vote, the tool sorts them into opinion clusters and, more importantly, surfaces the statements that win support across clusters — the uncommon ground. Participants watch a live picture of where the whole community is converging, rather than fighting in the comments.

Tellingly, the consultation didn’t open with an abstract prompt like “what is your ideal model for the sharing economy” — almost nobody would show up. It started from feelings: drivers, passengers, and people worried about rural service all centred on what each group actually feared and wanted, and the ideas that took care of everyone’s feelings floated to the top. That is the ethics of care, and it is positive-sum: a 51/49 referendum leaves nearly half feeling they lost, whereas a consensus people built together holds.

Inputs → outputs

Section titled “Inputs → outputs”

In: thousands of citizens’ votes on short statements about ride-sharing, gathered asynchronously over roughly three weeks.

Out: a set of principles with broad agreement — legalising ride-hailing while protecting consumers and giving local taxi co-ops room to compete — which the government adopted into regulation.

Impact

Section titled “Impact”

vTaiwan became a template: a way to take a polarising market fight and produce a legitimate, enacted answer without a winner-take-all vote. Its lineage matters too. Polis’s bridging logic — reward what wins agreement across groups — is the same idea later adopted as the Community Notes algorithm now used by YouTube, Meta and X, and the forerunner of the AI-facilitated Taiwan Alignment Assemblies that followed.

How it went

Section titled “How it went”

vTaiwan showed that asynchronous, software-mediated deliberation could settle real regulatory questions — and that the method travels: Polis is open-source and has since been used from Bowling Green, Kentucky to large national consultations. Its limits are honest ones — it works best on a bounded question with genuine public interest and a government willing to act on the result.

Sources

Section titled “Sources”