First, thank you very much for your time this morning. I think we’ve all read about your work in this space, and we’re very grateful that you’ve agreed to come on board and give us your thoughts and views.
I think your email covered everything. Essentially, the leaders report is done by the WPP government practice. It is a report that explores the future of government communications and starts to think about how it’s evolving and where it’s going. What are the key factors that will enable effective government communications?
We did that in 2016. It was launched in Davos. The aim is to now do another round in hopefully 2019 and launch it in Davos. Again, the objective with this is, again, to speak to people, to speak to senior government communicators like yourself, understand how government communication has evolved, and also to start to understand from you what are the big enablers, what are the big barriers that you’ve experienced?
Also, to start understanding how citizen engagement is evolving, because a lot of governments are increasingly talking about citizen engagement, and of course you’ve done a lot of very interesting work in this space.
Just for the purposes of the interview, would you mind -- I mean, we’ve read your profile, but if you could just tell us what your responsibilities are and take us through that, that would be a great start.
I’m Audrey Tang, Digital Minister @ Taiwan. My mandate is youth engagement, social innovation, and open government, which are very closely related. When I joined the cabinet around two years ago, in October 2016, I had a kind of compact.
The three, which I will expand a little bit on, are: radical transparency, voluntary association, and location independence. You’ve already seen my radical transparency protocol, so I won’t elaborate too much on that. Basically, everything I can see, I can publish after checking it here with everyone, of course.
Because of that, I don’t have any access to state secrets. Conversely, regardless of whether a policy ends up being decided to go forward or not, the context before the why, that position, is published. You know, not just the what and how.
By voluntary association, I mean that I don’t give command nor do I take command. That’s where I exist, kind of as a midpoint between the administration and the civil society. Nobody can be forced to say or do anything against my will. Conversely, I will not act in a hierarchical way with the career public service.
There’s kind of this anarchist part, but really, it is just a form of, I would say, consolidative or suggestive for a group of power, or peer-to-peer power, or new power -- that’s a new one now -- or instead of a hierarchical command structure.
The third thing, which I consider the most important, is location independence. It means that wherever I am, I am working in my capacity as Digital Minister. My office is every Wednesday. This is my office in Social Innovation Lab, Taipei. I don’t have to work in an administration building.
From 10:00 AM to 10:00 PM, everybody can just come and talk to me, including rough sleepers, social workers, and anyone individually, as long as they agree to have the transcript published online. I tour, also, around Taiwan to make sure that it is not people who can travel to Taipei.
I actually travel to them, and that’s Tuesday, every other Tuesday. There are 12 different ministries involved in the policymaking of Social Innovation Taiwan seen through my eyes. How is it like in other, like rural, indigenous, and so on places. They all gather here.
Any place where the people in the administration have a policy but it doesn’t turn out to be like exactly they imagined in the local areas, they can speak and the people in Taipei responsible will answer. It is like a parliamentary inquiry except with real local people.
Because 12 ministries are there, everyone can hear the answers and learn this is to be interpreted or reinterpreted like this. The individual team is formed without me having a separate ministry or a silo. This is part of this silo effect.
Of course, I can travel around Asia Pacific or indeed the world to meet people who encounter social injustice or environmental issues and again tell them how I can help to act as a channel. All this is because of location independence, and the fact that my whole working space is an online work space system.
Anyone that has Internet can check in and see whatever my team and other people are working about. This is, again, because of location. The idea is what we call working out loud so that people are not afraid of letting people unrelated to the project know what they are working on. That’s the three principles.
This is how the free software community and the Internet society has always run their decisions. All the three, you can find it in the, for example, the Debian, that’s a Linux distribution, the predecessor of Ubuntu Linus.
You can find all these principles in the Debian constitution and also, of course, the IETF. I’m basically bringing a tradition that’s almost 40 years old by now, but running it in a different operating system within the government systems.
A question for my own, because what you’re bringing is these really fundamental principles of open Internet. How has that been taken from the government side, which can be a bit more traditional and classic?
I think it’s partly because I started helping the cabinet, the career public service, at the end of 2014 Occupy. I literally talked to all the rank 12 officials in Taiwan, of which there is exactly 300 people, over three terms, three lectures over a month or so.
The administration at the time make sure that they all learned about this new way of policymaking. They generally find it complementary, but not reinforcing, of their hierarchical power. There’s an important difference.
Then I started a year or so, working with people supporting the Occupy, and also mediating for the Occupy, to bring them into the National Civil Service Academy. Together we trained, I think, 1,000 or so civil service people into methodology.
That was all before I become the digital minister. There’s a certain level of trust of my predictableness, of the internal coherence of this, of the fact that everybody in the working level also know, instead of just people in the parliament.
I think they now generally perceive this methodology as first, saving time for the junior public service, because they only have to explain once. They don’t have to take 40 different phone calls from different representatives or media, because in the flow, the work is just published online.
The discussion is more in-depth this way, so it saves their time overall, which is why they agreed to have the 1,300 ministerial projects all published online, including quarterly report, the use of budget, which procurement or research they have done, and so on.
They become social objects around which the people have a real discussion about. It’s a time-saver for everybody involved. The second thing is that I’m a risk absorber. The career public service usually gets the blame if things go wrong.
I’m the chief risk absorber. The thing is, because I give no command, plus there is radical transparency at a drafting level, so that when things go right, the journalism, the people generally know who is the innovative career public servant who proposed it in the first place.
I make sure to feature them in my slides and so on. The credit gets spread more evenly this way. That’s it. Usually, they just bring to me cross-ministerial issues that needs resolution and engagement with the public, but I don’t force them to.
You say they bring to you cross-ministerial issues. When you are traveling and speaking to people on the ground, are there any things that you bring in and start to present to them about requiring a policy?
Yes, that, in fact, we do. There is no clear distinction between changing existing policy and generating new policy. After all, the method we’re working on, this whole AI-powered conversation thing, is predicated on the focused conversation method.
This tour is really more about fact finding, people sharing their living experience, but in a group setting, so it can be cross-validated by their peers. Instead of just an anonymous letter to the president, we can have a more ethnographic data or detail about what, inside the context of the local habitation, this issue occurs.
Case in example, for example, there is an e-petition at the time where people in Hengchun, south north of Taiwan petitioned for the Ministry of Interior to station some helicopters to serve as ambulance cars, essentially, because their nearest hospital 90 minutes drive away.
The underlying problem of this is a distrust with the local clinicians and the local equipment, and a shortage of talented nurses and doctors who want to be in that area. Just by touring to that place, and bringing all the ministries available related to this with us, and in a live stream fashion, we get a lot of very significant input from people who were in Hengchun.
Maybe they were stationed in Hengchun. Maybe they have some local ties in it, and so on, but they can also contribute meaningfully. After a while, it became really clear that we should just expand the medical facility there, and if needed, fly doctors to operate instead of flying patients out.
Also, work on international FHIR, F-H-I-R, exchange formats for health, so that when there is a send to medical center situation, everybody is on the same page, so that the local clinicians can still act as a bridge, instead of just being taken out of the picture, and so on.
That results in new policy, because we allocated quite a few budget to essentially build a new large hospital there, which is initially not part of their conditions. It’s far from clear from the initial costs what the usual course or resolution will be, because of the collective fact finding process. It sometimes results in new policies.
I think this methodology is what sets this apart from many other consultative approaches because in other consultative approaches, although facilities usually have time for comparisons or whatever, it is generally seen as a time limit so that they have to steer it eventually back to the course.
When we say we allocate three weeks, four weeks just for people to check their feelings about say autonomous driving, or whatever, and see what everybody else’s positions or feelings are, and in an environment where you cannot reply, or choke other people, you cannot dissuade other people, there you can just post new sentiments for other people to resonate with.
We always see this result where people agree to disagree. A few key issues. That has been far more time on refining the consensus. This is very different if you have to rush to a decision, or if you design a space so that the people control each other in a safe space, and with a set timeframe just for checking people’s feelings.
We always get these powerful resonating feelings which we then use as agenda for the participation of the livestreaming, or just to have a good meeting. That would be the ideation part using this I think technology.
I think what was very interesting also is the fact that you said to create a safe space for everyone to input without enabling, or allowing anyone to discredit or offend anyone else which I think is really important.
Yes. Exactly. We’ve done a few interviews with people across different governments, and we’ve started to have a think about...Are there any policy areas where citizen engagement may, or may not be as effective, or as appropriate? Do you think there are any policy areas where you think citizen engagement should not be used?
Under this methodology, of course, if people cannot agree on facts, then this whole process cannot even start. That happens for example in futuristic scenarios like if people want to talk about, I don’t know, distributed federated identities using zero knowledge proofs distributor ledgers.
This is something that people don’t have a intuitive grasp of unless you’re trained in mathematics, and cryptography. Even if you are, you may not be trained in philosophical, and dependent use of human rights.
It requires very cross discipline conversation in order to even start the sense making. That’s not very effective for consultations if we just put up things like that without a lift end experience that’s shared by all participants involved.
Because of that, we are working on a lot of empathy building projects. For example, using virtual augmented reality to just situate people in a possible future, and have some living experience of the possible future. With that needs strong scientific research.
Like it has to be hard Sci-Fi, and then have a conversation within that virtual environment. That’s one possible solution for this kind of two aspect thing. The other thing is that if the concern’s a local situation that requires almost an anthropological understanding.
For example, the transitional justice process for our 16 indigenous nations. Unless you have a cultural upbringing as that indigenous nation, all those are just going to be long words that cannot meaningfully inform the discussion. We need to even build cultural interpretative services, if you know what I mean, to even start a discussion.
If the cultural context differed too much in the past. The first case, I thought, is too far in the future. If it diverged too far in the past already, then that also needs a process of collective empathy-building before even we start talking about facts, because facts have very different meanings to different cultures.
For some of our e-petition cases, we did...For example, there was one about a really remote island in PengHu, in Pescadores, about the fishing people’s need to fish and also the Marine National Park’s need to ban fishing, and just exactly what fishing means to them culturally is different.
What we did was essentially a listening experience and also using 360 recording devices to make sure that people who visit later can also get that into the empathy space of the four or so hours of town hall and have the fishers people speak their mind about what they view about this.
The space design, which I’ll spend a couple minutes talking about, is there’s essentially two places, a smaller room where the stakeholders are identified and use a ideation process inspired by the open policymaking here, which Fang-Jui can talk about in detail, and then a large town hall, which I am kind of the MC.
...to just analyze, play-by-play, why they made this move, showed this slide. What does it even mean in the local context? If people want to protest, want to shout, want to have a lot of their emotive, nonverbal actions, they can do it to me.
The media, everything, of course was in the town hall. Because it was not reverse-livestreamed, so it doesn’t disrupt the real deliberative action here. Everybody, if they have a constructive suggestion, they can talk to me. They can use Slido, use their phone, and so on. That informs the ongoing agenda of this discussion.
That’s right, but only in a structured way, like how this relates to the mind map already on the table. People can suggest a play, but it has to be in the context of the mind map, which I can explain later.
The point here is that it turns those nonverbal actions into legitimate plays on the shared mind map, while using bandwidth restriction, essentially, lets other disruptive actions have their cathartic value but not disrupt the process.
That’s right. Also, we just use this Slido mechanism, where people get to vote or upvote each other’s sentiments and questions. This is like a simpler, I guess, version of Polis, which I’ve shown earlier with the clustering and so on.
All this does is that people can ask questions and like each other’s questions, but there’s no unlike. It’s anonymous so that people can speak out their mind without feeling oppressed by people with higher rank.
Basically, whenever it’s non-duplicating constructive information, they’ll just highlight it, and just by the very fact of highlighting it within a context of a mind map adds this statement into that. It’s currently a little bit manual but we’re also working on automating this part.
Yeah. That’s how we use Slido. The good thing again is there’s no Reply button, so that the best thing to do when you see questions you don’t agree with is not attack that position but rather bring your position for other people to upvote.
No, personally, I just started using it. It’s freemium, so it doesn’t cost a cent. All the things we use for communication are all free as in free beer. Of course, the tool we use for internal collaboration is free as in free speech.
It’s all free software internally and gratis software externally. All the local people who want to imitate this process, they can freely do so, without even their government’s blessing. It is also how we empower the civil society in the junction, with empowering the participants.
Not at all. We bring the tech to people. There is a lot of words for this kind of technology. I think the earliest is what’s called calm technology in Xerox PARC. Now, we call it, I don’t know, ambient computing, or assistive civic technology.
That’s a lot of words to explain the very simple fact that basically, these are technologies that lets people focus on people more, instead of distract people from each other. They were just computing in the background.
Whenever people say something, utter something, and so on, this role room is essentially a computer that captures the context of the conversation into something that can be visually immediately fed back to the people discussion, so that they know literally whether they are on the same page or not.
Certainly. We do have pre-meetings before each large engagement. Usually, it is through a rolling stakeholder survey thing, where people get to recommend other people. In the vTaiwan process, which is separate from the e-petition process...
In the vTaiwan process, it is essential to have the major stakeholders agree to this timeline, especially the phrasing used as the discussion to be as neutral as possible and as inclusive as possible. For example, the UberX case, which was more widely reporting.
We spent a month, literally, just to agree on the title of the discussion, which is, "People without professional driver’s license carrying passengers and charging them for it." This is as neutral as possible, without mentioning any big words like sharing economy, or whatever.
This process takes time. It needs time, mostly by experts sent by the stakeholder teams. It gives the process legitimacy if this is agreed to be the agenda, and the process that everybody pre-agreed on, regardless of what is the final result.
Of course, the important thing here is the government puts something on stake, saying, "The crowdsourced feelings, we hold ourselves to account to use that, and only that, as the agenda in the stakeholder meeting."
You have people who care about economic talk to the developmental agency. The administration is this rope that makes sure everything is tied together. They make arbitrary decisions from the outside, because this process is not transparent at all.
It is like the old model. The organizational capability and the arbitration capability are both being superseded by peer-to-peer power, because people don’t need a minister or MP to organize. It used to be the case 20 years ago, but now, they don’t.
With the right hashtag, tens of thousands of people just organize themselves. You see a lot of more players. Also, they were single issues that reflects a social or environmental, deep, underlying thing, what we call wicked problem, that requires coordinated action.
Otherwise, they’ve already solved it by now. The curriculum situation is not good, but it’s a Nash equilibrium, usually. If you have new players organize themselves out of nowhere like this, you cannot just have a new agency or a new council set up for each emerging issue.
What we are now saying, the main enabler is that, with these online platforms and also offline civil society partnership, we’re now saying we’re now a space that links together people with different positions.
Our main ask is not how to organize and how to arbitrate, but rather, given the different positions, can we find common value? When we find common value, which is reflected by feelings, when we find common value, are there any innovation that can drive a solution that works for everyone, that everyone can live with?
This is not like fine consensus. People don’t sign their names next to it. This is more like group collective consent, or rough consensus, as we say in the Internet Society. By stating this very clearly, and citing examples of co-creation before, people generally become aware that if they complain, all they get is a ticket to enter the kitchen and to cook together.
Then to co-create real artifacts, like someone will complain a tax filing system being very difficult to use, against the invitation, initially, from participation office, not even two days after they post it.
We put ourself to account, using people’s online feedback in a livestream fashion, and then set up for workshops to essentially co-create the online tax filing experience. Which, I think, it’s 96 percent approval rating this year.
Even the other four percent of the people who have useful feedback in this time, but we did not get time to implement it, they still think this is part of their ongoing work. They’re still willing to be engaged.
There’s many routes, but the most popular one is the e-petition route. In the Join platform, join.gov.tw, as I mentioned, we show the budgets. We show the regulatory pre-announcements 60 days before each regulation and bill draft is sent to the legislators.
People can petition to have any administration deputy or minister give an answer if they reach 5,000 people within 60 days. This is more like a "We the people" kind of way of doing things, so priority setting. Getting 5,000 people isn’t that easy, but because it’s electronic, people can campaign in the midst of online fashion.
What this means for us is because hosting the 23 million people in Taiwan over 5 million is on this platform, so literally a quarter of the population. While not entirely representative, this is a significant audience that we can reach through this platform that we build ourselves instead of relying on social media.
The second things is that once something has 5,000 people, can we use a pro-and-con approach mechanism we took from Better Reykjavik. By the time they get 5,000 people, usually the main arguments has already been posted.
The fact-finding is collective. We have the what we call issue-based mapping that can start off with the information that’s present by this crowd source, 5,000 people, essentially crowd moderating each other. There’s important points to be brought into the issue-based mapping.
It shows it’s not just one or two lobbyists positioned. There are other people, people that are countering of this, like people conditioned to change Taiwan’s time zone to plus eight to plus nine. There’s a counter petition that says we should remain in the same time zone. We invite people from each petition through the same route...
...and they agreed on the same common value, which is they want Taiwan’s uniqueness to be seen more prominently in the world. That’s something that people can agree on, and then start to ideate solutions instead of basing on changing time zone, which may make the international news for the day, and everybody forgets about it, because a country can have different time zones.
How do you reach a broader base of people, because it seems like often these engagement initiatives, the people who are eager to participate, are already the people who have interest in a particular topic, or have something to say, or want to raise concerns to the government.
This is not, strictly speaking consultative, because it’s initiated by the government. Everything I just talked about, the initial wave of mobilization came from the civil society, and sometimes from the social sector, with social enterprise and someone raising the issues.
They know that local domain knowledge in which they posed this discussion. Basically, they’re destroying the administration’s end, rather than the administration trying to draw people in, which is a very different model.
Which is why I always said that our core value is for the government to trust the people and not the other way around. If we trust the people enough, if we learn their language, their story, their lived experience enough, some of them will trust back. We treasure each one of them.
It’s not good to asking people trust the government, without the government trusting back first, because that would be fascism. That’s like the definition of fascism. [laughs] We’re kind of the counter to that.
It’s very interesting because we’ve been within kind of counter public...My boss and I have been kind of thinking about how to measure a city’s engagement. One of the core foundational principles of how we went about it is that trust is a two-way street.
It’s healthy to have the civic sector distrust the government. It’s even healthy to promote that. The whole point is we amplify and connect those mobilizations, essentially turning social outreach into social production. Sometimes it leads to innovation if we discover something that nobody has thought of before.
Without initial outreach and movement, none of this would be possible. People would not want to devote time on this. The civil society organizers are crucial, and I’m merely facilitating it and amplifying it.
That’s right. Also, this has been only possible because, according to the CIVICUS monitor, that if you click Asia, and fully open, and then you only see Taiwan as a...This picture is very powerful. [laughs] This is the relative freedom of association, assembly, and speech around the world, in this whole region.
This is not to say Nordic, or Australia, or in New Zealand is not doing better. They’re probably doing better, but in this region, really, Taiwan is the only place where you can, literally, have any opinion expressed and social movement done in a way that the government doesn’t like.
The government has to join, rather than try to oppress or censor your work. Without this foundation, none of this process would be possible. This rests on the case that...That’s why for example the Reporters Without Borders chose Taiwan as their Asia HQ.
You mentioned just there that the government has to act, and it makes me wonder to what extent is that binding, to what extent do the government ministries feel they have to act, or how do they integrate that into the way they work?
The e-participation, there’s an administrative regulation about it, about the methodologies, how do you contact the petitioner, how the conversation with the petitioner should be public by default, like readily transparent, and using a media form that people are agreed with. Sometimes they don’t want to reveal their true identity or true name.
There’s a lot of fine details in that regulation of e-participation. Also in other regulation participation of office and network of implementation, it also shows this is a capacity-building way of learning facilitative skills, learning translational communicative skills, and also the recording and re-presentating, instead of representative skills, including live stream, cinema graphic transcript, and things like that.
Also, because I don’t give commands, so any commandment they bring to the table, they bring it by their own volition. They have to act in the sense that they have to enter this deliberative space, and listen, and respond point-by-point.
That’s right, and because the PO is like a fractal, it’s recursive. I’m in administration, so I talk with the ministerial POs, but the in the larger ministries, for example Agriculture, Health and Welfare, and certainly Finance and so on, they have their set level, agency-level POs as well. They can also have their lower-level POs.
Each functional agency can have this branch-like structure. When they join working, open, regularly transparent workflow, they’re suddenly reporting directly to the deputy minister. That’s also part of the regulation that we did.
...went off and do that. If they need cross-ministerial support, of course, they surface it on the monthly meeting. We are also now seeing larger ministries PO just engaging directly with people, because it’s a single ministry issue. They don’t have to come to the official minister.
Based on the four core values of open government, which is basically how clear the information is presented, how white is the space and time for people’s input, how easy is it for people to find out what happened to their input, and also how facilitative are we of populations of different backgrounds to join.
These are for core values of open government in Taiwan. I think if we don’t measure, especially these two, it very quickly becomes a, what we call, open washing, meaning that it’s for a selected number of people’s friends and circle, and everything is online, but nobody knows what happens to their inputs.
You see that all the time in other places on the web. We really need to evaluate the whole, we call it, the policymaking account or account trail, essentially, of the policymaking process. Fang-Jui is designing future metrics to generate these measurements as the policymaking accounts gets extended over time on one particular policy.
We’ve currently done these by hand, by people in PDIS who curate the process. This is not done in a very systematic fashion. We do it for larger cases that we would kind of keep track of ourselves. We might as well view the accounts and share it with everybody.
The next moment, I think, is key to have the POs also understand the importance of these interflow measurements that doesn’t require them to work over, over-hard because it’s all measured across the different facilitative systems. It’s just we capture these key metrics as time goes by.
It’s not meant to compare or to measure, because every case is different, qualitatively speaking. It is useful feedback to let us know maybe this case is good on participation but it’s not conclusive enough or...
That’s quite interesting because you say that each case is quite different as it is. How would you define the kind of benchmarks, for want of a better word, or what are your parameters of success for each and how do you define that? How do you see...
I think in the end it is about citizen confidence. It’s like consumer confidence. [laughs] There’s a index for consumer confidence. If there’s one for how people feel, even though maybe at end, they don’t get what they want.
Taiwan’s time zone has not changed, certainly, [laughs] but they feel confident that if they can have a factual point-by-point response from each ministry about the impact of daylight savings time on energy, on tourism, on environment, on everything, then they feel this accountability means that the government trusts people completely to not abuse these numbers and evidences.
We even can co-create the basic facts that we can agree on. If that increases the confidence, not in the government institution, but rather in this process of collective fact-finding, then I think that is the key measurement that we need to measure.