It was late in 2015, and things were at an impasse. Some four years earlier, Taiwan’s finance ministry had decided to legalize online sales of alcohol. To help it shape the new rules, the ministry had kicked off talks with alcohol merchants, e-commerce platforms, and social groups worried that online sales would make it easy for children to buy liquor. But since then they had all been talking past each other. The regulation had gotten nowhere.

That was when a group of government officials and activists decided to take the question to a new online discussion platform called vTaiwan. Starting in early March 2016, about 450 citizens went to vtaiwan.tw, proposed solutions, and voted on them.

Within a matter of weeks, they had formulated a set of recommendations. Online alcohol sales would be limited to a handful of e-commerce platforms and distributors; transactions would be by credit card only; and purchases would be collected at convenience stores, making it nearly impossible for a child to surreptitiously get hold of booze. By late April the government had incorporated the suggestions into a draft bill that it sent to parliament.

The deadlock “resolved itself almost immediately,” says Colin Megill, the CEO and cofounder of Pol.is, one of the digital platforms vTaiwan uses to host discussion. “The opposing sides had never had a chance to actually interact with each other’s ideas. When they did, it became apparent that both sides were basically willing to give the opposing side what it wanted.”

Three years after its founding, vTaiwan hasn’t exactly taken Taiwanese politics by storm. It has been used to debate only a couple of dozen bills, and the government isn’t required to heed the outcomes of those debates (though it may be if a new law passes later this year). But the system has proved useful in finding consensus on deadlocked issues such as the alcohol sales law, and its methods are now being applied to a larger consultation platform, called Join, that’s being tried out in some local government settings. The question now is whether it can be used to settle bigger policy questions at a national level—and whether it could be a model for other countries.

Taiwan might not seem like the most obvious place for a pioneering exercise in digital democracy. The island held its first direct presidential election only in 1996, after a century marked first by Japanese colonial rule and then by Chinese nationalist martial law. But that oppressive past has also meant that Taiwanese have a history of taking to the streets to push back against heavy-handed government. In Taiwan’s democratic era, it was a protest four years ago that planted the seed for this innovative political experiment.

The 2014 Sunflower Movement, led by students and activists, derailed an attempt by President Ma Ying-jeou’s government to ram through a trade agreement between Taiwan, which has been locally ruled since 1949, and China, which claims Taiwan as its territory. For more than three weeks the protesters occupied government buildings over the deal, which they felt would give China too much leverage over the Taiwanese economy.

In the aftermath, the Ma government invited Sunflower activists to create a platform through which it might better communicate with Taiwan’s youth. A Taiwanese civic tech community known as g0v (pronounced “Gov Zero”), which had played a leading role in the Sunflower protests, built vTaiwan in 2015 and still runs it. The platform enables citizens, civil-society organizations, experts, and elected representatives to discuss proposed laws via its website as well as in face-to-face meetings and hackathons. Its goal is to help policymakers make decisions that gain legitimacy through consultation.

“I would say vTaiwan is about civil society learning the functions of the government and, to a degree, collaborating,” Taiwan’s digital minister, Audrey Tang, told me during a visit to her office. Tang, a famed hacker who helped the thousands of Sunflower protesters build and maintain their internal communications network, was appointed by the current president, Tsai Ing-wen, who won the 2016 election on a pledge of government transparency.

vTaiwan relies on a hodgepodge of open-source tools for soliciting proposals, sharing information, and holding polls, but one of the key parts is Pol.is, created by Megill and a couple of friends in Seattle after the events of Occupy Wall Street and the Arab Spring in 2011. On Pol.is, a topic is put up for debate. Anyone who creates an account can post comments on the topic, and can also upvote or downvote other people’s comments.

That may sound much like any other online forum, but two things make Pol.is unusual. The first is that you cannot reply to comments. “If people can propose their ideas and comments but they cannot reply to each other, then it drastically reduces the motivation for trolls to troll,” Tang says.

The second is that it uses the upvotes and downvotes to generate a kind of map of all the participants in the debate, clustering together people who have voted similarly. Although there may be hundreds or thousands of separate comments, like-minded groups rapidly emerge in this voting map, showing where there are divides and where there is consensus. People then naturally try to draft comments that will win votes from both sides of a divide, gradually eliminating the gaps.

“The visualization is very, very helpful,” Tang says. “If you show people the face of the crowd, and if you take away the reply button, then people stop wasting time on the divisive statements.”

In one of the platform’s early successes, for example, the topic at issue was how to regulate the ride-hailing company Uber, which had—as in many places around the world—run into fierce opposition from local taxi drivers. As new people joined the online debate, they were shown and asked to vote on comments that ranged from calls to ban Uber or subject it to strict regulation, to calls to let the market decide, to more general statements such as “I think that Uber is a business model that can create flexible jobs.”

Within a few days, the voting had coalesced to define two groups, one pro-Uber and one, about twice as large, anti-Uber. But then the magic happened: as the groups sought to attract more supporters, their members started posting comments on matters that everyone could agree were important, such as rider safety and liability insurance. Gradually, they refined them to garner more votes. The end result was a set of seven comments that enjoyed almost universal approval, containing such recommendations as “The government should set up a fair regulatory regime,” “Private passenger vehicles should be registered,” and “It should be permissible for a for-hire driver to join multiple fleets and platforms.” The divide between pro- and anti-Uber camps had been replaced by consensus on how to create a level playing field for Uber and the taxi firms, protect consumers, and create more competition. Tang herself took those suggestions into face-to-face talks with Uber, the taxi drivers, and experts, which led the government to adopt new regulations along the lines vTaiwan had produced.

vTaiwan’s website boasts that as of August 2018, it had been used in 26 cases, with 80 percent resulting in “decisive government action.” As well as inspiring regulations for Uber and for online alcohol sales, it has led to an act that creates a “fintech sandbox,” a space for small-scale technological experiments within Taiwan’s otherwise tightly regulated financial system.

“It’s all solving the same problem: essentially saying, ‘What if we’re talking about things that are emergent, [for which] there are only a handful of early adopters?’” Tang says. “That’s the basic problem we were solving at the very beginning with vTaiwan.”

But while vTaiwan can bridge gulfs in public opinion, what it can’t always overcome is politics. After the Tsai administration took office in 2016, it withdrew all bills awaiting legislative approval. Observers chalked that up to the new president’s desire to differentiate her agenda from her predecessor’s. The online alcohol sales bill that the Ma government had drafted from the vTaiwan suggestions never saw the light of day.

That the government isn’t required to heed discussions on vTaiwan is the system’s biggest shortcoming. Jason Hsu, a former activist and now opposition legislator who helped bring vTaiwan into being during the Ma administration, calls it “a tiger without teeth.”

Moreover, the Tsai administration has chosen to use it only for issues, such as regulating Uber, that have to do with the digital economy. That’s because people who care about such issues are the ones most likely to be comfortable using a digital discussion platform. But some think it won’t get serious traction with the public unless it is put to use on non-digital issues that matter to more people. C.L. Kao, one of the cofounders of g0v, argues that the government could have applied vTaiwan to two contentious recent issues, pension reform and labor reform, as a way to build its credibility.

In any case, Kao says, if vTaiwan’s recommendations are ultimately ignored, as they were with the alcohol sales law, then the whole process runs the risk of being viewed as “openwashing”—something that creates the pretense of transparency. “The end goal is legislation,” he says.

vTaiwan is one of dozens of participatory governance projects around the world listed on CrowdLaw, a site run by the Governance Lab at New York University. Most of them, says Beth Noveck, the lab’s director, suffer from the same problem: they’re not binding on governments, which means it’s also hard for them to gain credibility with citizens. Still, she says, Taiwan’s experiment is “a step in the right direction.” It’s “far more institutionalized” than what’s been seen elsewhere, she adds.

The platform may be about to get a little more clout. This autumn legislators will debate and vote on a digital communications bill that, among other things, says that “digital-economy issues are to be deliberated in an open, multistakeholder process that the government has the duty to support,” in Tang’s words. But what “support” means—how much weight lawmakers or the government will have to give to vTaiwan’s deliberations—is still up in the air.

Taiwan does have a newer participatory governance system that is getting more traction. Join, also overseen by Audrey Tang, is a platform for hosting and debating online petitions, again using Pol.is to create consensus. She describes it as a vTaiwan within the government—“basically the same … process, but with senior career public servants instead of g0v volunteers” at the heart of the platform.

Although petitions on Join still aren’t legally binding, any government agency that agrees to participate in a deliberation must, if the petition gets more than 5,000 signatures, give a point-by-point response explaining why it agreed to or rejected the proposal. Five of Taiwan’s cities or counties are testing Join; the aim is ultimately to roll it out nationwide, Tang says.

Join tends to attract a broader, older, and less tech-savvy range of users than vTaiwan. The advantage of this, says Tang, is that it doesn’t tackle only digital-­economy issues, as vTaiwan does, but a wide variety of questions, “like whether we should build a hospital in the southmost part of Taiwan, in Hengchun, or whether the first publicly open marine national park should ban fishing.” The downside is that there’s more resistance from the government bureaucracy. Senior public servants “need some hand-­holding,” she says, to be able to see people who comment online as “not protesters or mobs, but actually people with distinct expertise.”

While only 200,000 people have so far taken part in a vTaiwan discussion, nearly five million of the country’s 23 million inhabitants are already on Join. More than 10,000 voted on a recent proposal that advocated caning as a punishment for drunk driving, sexual assault, and child abuse.

Here too, the consensus-building tendencies of Pol.is can lead the discussion in unexpected directions. Initially, opinion on the caning issue was divided into three camps: besides the people who were for and against caning, a third group argued that it was too light a punishment for such offenses.

Eventually, however, the consensus opinions that emerged had nothing to do with caning at all, but were more focused on methods of preventing those crimes. At the time of this writing, proposals being considered for legislation included alcohol locks and confiscating drunk drivers’ cars.

This suggests people had concluded that, in fact, “To cane or not to cane?” was the wrong question to ask. That kind of realization, and solution, wouldn’t have emerged from a traditional online petition that only gives people the option of voting yes or no.

Karen Yu, a legislator in President Tsai’s Democratic Progressive Party, says vTaiwan is “not a huge priority” for the administration and has come “close to death” at times. Join, she points out, at least benefits from the legitimacy of being managed by the government. Wu Min Hsuan, an activist who occupied Taiwan’s Legislative Yuan during the Sunflower Movement protests, says Join has already proved itself much more productive than vTaiwan. The obstacle, he believes, is political will. “The experiment is important and has value,” he says. “But the platform has its limits. It needs real power.”  

Chris Horton is a journalist based in Taipei. An Rong Xu is a photojournalist in New York City.