Oceans of Data — g0v.tw and Taiwan’s Open Data Model
Taiwan has been ranked #1 in the Open Knowledge Foundation’s Global Open Data Index for the last two years. Taiwan’s ascendance has been astonishing, rising from #36 in 2013, to #11 in 2014, and claiming the top position in 2015 and 2016. it has managed to surprise data observers across the world. So how was such a meteoric rise possible? And is the ‘Taiwan model’ something that other countries could be supported to replicate?
In reality, there’s a complex array of factors that have influenced the rankings. TH Schee, the ambassador of Open Knowledge Taiwan attributes Taiwan’s success primarily to three core facts :
- A free press — Since the end of martial law in 1987, Taiwan has enjoyed an open media environment and sits at #45 on the Reporters Without Borders Press Freedom Index — the highest ranking for an Asian country.
- Formalized mechanisms on public consultation — In the first half of 2015, the government established dedicated committees in all ministries to promote a discourse around the Global Open Data Index and the open data more generally. This potentially had some negative impacts — the state was very concerned with attaining the highest ranking, and their efforts were very much directed towards this purpose.
- Dataset production — The government has focused on releasing a high volume of datasets. However, in some cases personal information has been published in an open format without prior consultation. One such case is the personal data of the National Healthcare Insurance Program, which was published without agreement from insurance companies. This case subsequently, attracted a class action suit from human rights groups and civil society.
In that context, we decided to talk to Chia-Liang Kao of g0v.tw — a community focused on building open data-driven information tools and platforms to enable greater civic participation, and to allow people to engage with important societal challenges. g0v.tw was one of the leading forces in the 2014 Sunflower Movement, when hundreds of protesters occupied the Taiwanese parliament in opposition to the opaque development of a trade bill between Taiwan and China.
Harsha: Hi Chia-Liang Kao, Thank you so much for taking time to talk to us. Let’s start off by defining what we’re talking about, here — what does open data mean for you?
Chia-Liang Kao: Well, open data is something that should be available freely — like free software — for usage and redistribution without restrictions, to empower citizens to make use of the data for transparency or any kind of economic development. The whole important point of open data is that because the data is open and unlimited, you can expect to discover unexpected things with it.
Harsha: So Taiwan is #1 on the Open Knowledge International, and I’m aware that since 2012 there have been efforts to make Taiwan do really well on those rankings. Can you give us an idea of what’s happening?
Chia-Liang Kao: To put things simply, if you know anything about modern East Asian countries — Japan, Korea, Taiwan and Singapore — they are competitive. Once there is some sort of ranking, it becomes easy for them to compete for the top position. But that said, the OKI rankings are encouraging for all the good work that happens here, not just because the government was chasing that top ranking — but also because there are some aggressive agencies that are making it possible for it to improve Taiwan’s rankings — such as the election commission.
Normally the elections are held on paper, and delivered via mail three days before election. So they opened up all the election results data for the 2016 election. They opened up all the registries such as information about the candidates and their basic information. All of this was made available as a JSON database three weeks before the election. They were so surprised by the possibilities of open data. Instead of being criticised for bad websites and hard to access data, open data lets people focus on what they’re good at — collecting and publishing information. So that’s one good thing that came out of this craze for the OKI rankings. But apart from that, there’s also the public procurement data, where the agency in charge really releases more data than necessary in order to appear open for the sake of the rankings. Agencies often just release data to get some more points on the system.
Harsha: Can you give us an idea of what the position of the government is when it comes to issues of open data? Are the rankings a good proxy for Taiwan’s open data scene to influence government open data policies?
Chia-Liang Kao: My point was that it’s really different from agency to agency. There’s no overarching and fair representation of what they [government policies] are like. There is a strong push within the government for open data, but it varies for each department.
Harsha: Like you said, the election department is a positive example of open data. Have there been instances where this push for open data didn’t go so well?
Chia-Liang Kao: The procurement data guys released some bad data in XML, and said, ‘Look — here’s open data!’. It was really bad. They released only 20 percent of their procurement data and pretended to be open. The whole idea of open data is that you encourage governments to see data as infrastructure. When they achieve that like they did in the election data, it’s good. But what the procurement data agency was doing was not so good.
Harsha: What is the state of freedom of information and expression in Taiwan? How are safeguards for whistleblowers? Elsewhere in some other Asian countries, activists reporting against the government face death threats occasionally.
Chia-Liang Kao: So that’s unheard of in Taiwan right now. You’ve mentioned two different things: freedom of expression and freedom of information. Freedom of Information is protected by the constitution in most parts of the world. I know its implementation is less smoother in countries, but in Taiwan it’s pretty good. You probably won’t get arrested for expressing your opinion unless it concerns public safety or you’re conducting some terrorism-related activities. You can complain if the president is shitty or if their policies are no good, and it’s been pretty safe for a while. Press freedom in Taiwan is pretty good.
What is concerning about the freedom of press is the influence of corporate or Chinese money that tries to cover up some issues, maybe. If a media company’s chairperson has a scandal, their channel will not report on it. There are some very pro-Chinese media outlets, too. I do not know if it’s true, but I hear they play with China’s agenda against Taiwan — like criticising some policies in a way that’s not in line with fair media practices.
Harsha: Have there been any instances of this media trying to find faults or criticise the open data movement?
Chia-Liang Kao: The beauty of open data is that it’s neutral, and does not offend any political affiliations. There are some conservative agencies that try to argue against open data, but the general sentiment even across the public is that transparency is important, and as long it does not invade people’s privacy or threaten national security, then it’s fine.
Harsha: What does Taiwan still have to achieve on the open data scene? What are the challenges you face as an open data practitioner?
Chia-Liang Kao: We were hoping the new government would have a plan to define what crucial data has to be opened — that is, the priorities of the datasets being opened up. But we don’t see that happening. Right now, if someone really wants data they will have to fight for it, and try to put pressure on the ministry to release it… There is no one inside the government that prioritizes what datasets are important to be released. That’s challenging. It’s tough, because it’s not easy to initiate such things inside government because there are a lot of agencies you need to convince about the power of open data. But unless there’s pressure from civil society, that wouldn’t ever happen. The challenges still relate to the government’s determination to actually push the whole thing forward, and develop a clear agenda.
Harsha: On the OKI Rankings, Taiwan loses points for not opening up postal codes, land ownership and government spending? What’s the progress on these measures?
Chia-Liang Kao: I think the land ownership data has been published only for public properties. But privately owned land data is still not open, and you have to pay to retrieve it. This information is published in a document that is useful for land-related transactions. Some people in the government are worried about opening that data because it involves personal identities. The government has to prioritise what information they want to release — if they do not want to disclose the full name of the owner, maybe they could publish corporate or legal owned land data first. I’ve not followed up on where the government has been on this front.
The postcode information is almost a one-to-one mapping to the third level administrative boundaries — so i don’t think there’s official data for the postal codes to be geocoded — that it can be released or derived from whatever information is available.
This concludes the first part of our interview with Chia-Liang Kao from g0v.tw. We’ll be back later in the week with the second half of the interview, where we’ll discuss g0v.tw’s work, and the important role played by civil society in pushing the Taiwanese government to open itself up to public scrutiny.
Open & Shut is a project from the Small Media team. Small Media is an organisation working to support freedom of information in closed societies, and developed the Iran Open Data portal.
Cover image by skp, story edited by James Marchant.
