Open Source Data: Just How Open Is Taiwan’s Government?

When the Global Open Data Index released their latest worldwide ranking on openness of government data earlier this summer, Taiwan topped the list for the second time in a row, earning an “openness score” of 90 percent. In contrast to Taiwan’s success, in small letters underneath the main heading of the Index’s report are the words “11% of dataset entries in this index are open.”

Similarly, when the Open Data Barometer, another world ranking index on open data, released a ranking in May – this one excluding Taiwan – the researchers came to the sobering conclusion that the worldwide open data movement is “at risk,” with 93 percent of government data sets still not open. To fill in the gaps in information, I used the Open Data Barometer’s methodology to test approximately where Taiwan would rank alongside other countries in the world. Using a modified version of the Barometer’s system of analysis to calculate scores for Taiwan,[1] my conclusion is this, as governments worldwide stall, slow down or even backtrack on their commitments to open data, Taiwan remains an oasis of open data.

With the exception of government expenditure, land ownership data and information about the country’s educational and health industry performance, most of Taiwan’s open source data met the Open Data Barometer’s criteria for robustness, and was available online, free of charge and openly licensed, as well as up to date. According to my calculations, Taiwan scores higher than both South Korea and Japan – two of the highest scorers in the world – for the thoroughness of its crime statistics and government budget data. And the nation scored better on nearly all fronts compared to its Southeast Asian neighbors. In a country that only began to embrace true freedom of speech and the press in the last three decades and where textbooks continue to dance around the subject of the White Terror – the oppression of freedom in Taiwan by authoritarian rule from 1949 to 1987 – this is quite extraordinary.

And there continues to be strong movements to continue to not just improve the open source data system, but to ensure that the Taiwanese people know how to take advantage of it. The government somewhat consistently updates their websites to make them easier to navigate, recently making several versions of a detailed map of Taiwan more accessible to the public. There are also civil organizations and online communities like g0v, who have worked relentlessly to better connect members of the public with government information.

Score for Taiwan’s government open data (by Calin Brown)

The Taiwan Advantage

A key component of Taiwan’s New Southbound Policy is building cooperation as well as resource and knowledge sharing capacity between Taiwan and countries in Southeast Asia, South Asia and Australasia. This includes sharing information about anything from best agricultural practices for disaster risk management to ongoing potential security threats. And now, capacity-building for open source data should be included on this list.

In its most recent report, the Open Data Barometer commented that although the East Asia and the Pacific region performed well compared to Latin America and Africa, there are still many parts of Southeast Asia and East Asia that lack information frameworks. Civil organizations (i.e. Open Myanmar, Malaysia’s Sinar Project) have been doing what they could to provide data despite limited resources, but immense barriers persist. According to the results of the recent Open Data Barometer, Vietnam struggles with ensuring that all the information available is machine-readable, nor are their datasets considered robust enough under the conditions listed in the Barometer. In the case of Myanmar, a lot of information is just simply missing online. These are areas that Taiwan has done well in regardless of the type of data. Taiwan’s experience at collecting data and releasing it on relatively accessible platforms is knowledge that Taiwan can share with its neighbors to build stronger networks in the Asia-Pacific region.

Of course, free and open access to datasets, which are central to holding governments and corporations accountable, are not necessarily something many countries in the region are necessarily chomping at the bit for (consider the military junta in Thailand and the various one-party political systems in place right now). In fact, progress in this area has largely remained slow in this region because citizens’ right to information is not accepted by all government officials in each country. For instance, in Indonesia, despite initial progress made in 2014 with a new data portal, Data.go.id, the last two editions of the Open Data Barometer report have indicated a slowdown in progress in the last three years. Malaysia and Thailand, too, have demonstrated reductions in the access to open data over the last couple of years.

However, access to this data is also vital when it comes to information about land ownership so that researchers can conduct research on agricultural practices and land use, access to trade information is important for economists and policymakers, national health and census information is important for combatting diseases. Entrepreneurs can use the data on healthcare provision or education performance and build new services, and all businesses can use the data to make more thoughtful decisions. Many Southeast Asian politicians, who have been responding to widespread criticisms about corruption within the country’s police force and corruption amongst their own ranks, have also recently been winning public favor by promising greater government accountability and transparency. Politicians in Singapore continue to aggressively expunge corruption in the city-state, and leaders in Indonesia, Malaysia and the Philippines have all undergone increasing pressure to cleanse political parties of corruption. Releasing publicly data about elections, government budgets as well as about public contracts, are all necessary to enable a culture of transparency and accountability to prevent corruption. Taiwan (for the most part) shows that this is all achievable, and could provide the expertise to get everyone else on board.

Room For Improvement

However, I’ll add that none of this is to say that there isn’t room for improvement in Taiwan itself. Many of the online databases are clunky and sometimes disorganized, and navigating the websites is not always intuitive. Also, there are some huge gaps in the data.

For example, there is little to no public information with regard to the academic performance of Taiwan’s students. What’s available are information about how many students are enrolled in school in each grade, the average cost of tuition and a comparison of the cost of Taiwan’s schools compared to the cost of education in other countries. One gets the feeling that these statistics are cherry picked to only provide facts about topics that Taiwan relatively excels at. What’s missing is information that could actually give researchers and policymakers an idea of how the education system is doing: student attendance rates, graduation rates, teacher retention rates, or even just a comparison of test scores or grades at different schools. The national health performance data is slightly out of date, and there is no mention of how well healthcare services are serving people’s needs. And, perhaps testimony to Taiwan’s historic lack of investment in welfare services, including elder care, there’s little to no public information about the performance of Taiwan’s welfare services.

But that’s also exactly why greater cooperation between Taiwan and its neighbors is helpful. Strengthened knowledge sharing and sharing experiences and opinions about strategies for better data collection and dissemination is vital if the Taiwan and other Asia-Pacific countries hope to increase their residents’ welfare, and increase their capacity to counteract the myriad shared threats now facing the region.

About the methodology:

Using an adapted version of the Open Data Barometer’s methodology that measures how successful a government has been in implementing their strategy of increasing open data access, I answered a 10-point checklist as the researchers did with respect to the quality of data provided for the following types of data: mapping data, land ownership data, national statistics (i.e. GDP, population), government expenditure, company registration data, legislation data (access to laws and the constitution), public transport timetable data, international trade data, health sector performance data, information regarding educational performance, crime statistics, national environment statistics, national election results data and public contracting data. The Open Data Barometer methodology attached weightings to each type of data in the checklist based on an aggregation logic the team came up so that each column would generate a score between 0 and 100 (for a clearer understanding of this process, refer to the two tables below).

[1] To provide a little background about the Open Data Barometer, in their fourth edition of the Open Data Barometer, the research team created a ranking system based on three kinds of data: a peer reviewed expert survey with a range of questions about open data contexts, including a detailed assessment completed for 15 kinds of data in each country, a government self-assessment through a simplified version of the survey, as well as secondary data selected to complement the expert survey data. Because of labor constraints, I decided to focus my research on only assessing the ease of access to 15 types of data in the country. Knowing full well that the actual research conducted by the Open Data Barometer involved a lot more peer review than what my research team of one was capable of, what I have done is provide a mere sample, or a snapshot.