In 2017, Shannon Ang, then a graduate student in sociology at the University of Michigan, sought access to the Retirement and Health Study (RHS) data from Singapore’s Central Provident Fund. The RHS, a longitudinal study of over 20,000 Singaporeans, was launched in 2014 partly to better understand the needs of a rapidly ageing population. Ang’s request would have been a humdrum procedure in many democracies. But in Singapore, it sparked a six-year journey through the Byzantine bureaucracy.
Initially denied due to supposed data unavailability, he was later told in 2019 to collaborate with government agencies to obtain the data. So he partnered with the Ministry of Health (MOH), secured a grant, and completed all necessary administrative work. In 2021, MOH abruptly terminated the collaboration with Ang—by then in his current position as assistant professor at NTU—without an official explanation. Undeterred, Ang applied for access to the survey data again in 2022 through a Ministry of Education (MOE) pilot program. By 2023, MOE had selected his project for further review and discussed with him the variables he requested and the budget required for the project. Excited about the progress, and after incorporating feedback from the Ministry of Social and Family Development, Ang resubmitted his application. In October of 2023 Ang’s application was denied again, without a clear explanation.
In Franz Kafka’s novel, The Trial, the protagonist Josef K. is arrested and put on trial without warning, with the reason unknown throughout the novel to both Josef and reader. Kafka wrote: “the proceedings were kept secret not only from the public but also from the accused.” Like Kafka’s opaque proceedings, data access in Singapore can sometimes also seem like a black box to outsiders. While the front-end instructions and platforms for data application are clear, transparent, and institutionalised, the back-end processes of decision-making are often obscured.
Over a span of six years, across countless correspondences with multiple ministries, and with their encouragement and guidance, Ang’s data request was repeatedly rejected at the final stages without a clear reason. Ang’s Kafkaesque experiences with data access are symptoms of Singapore’s data governance structure, which follows a closed-by-default—rather than an open-by-default—model. Switching to the latter is crucial for increasing data transparency. That can, in turn, foster an informed citizenry and create robust checks and balances to ensure that no one single societal actor can be the arbiter of truth on salient issues and policy solutions in Singapore. These outcomes will strengthen our democracy.
At a broad level globally, greater data and information transparency can strengthen government accountability and policy evaluation, by providing opportunities for citizens to monitor government performance on various policy issues, such as income distribution, racial equality, carbon/greenhouse emissions, economic growth, and employment levels. Transparency ensures that stakeholders—policymakers and citizens alike—can effectively assess the country’s performance on key policy issues. Moreover, information transparency is fundamental to the functioning of a healthy democracy, enabling the citizenry to make informed electoral decisions with a complete picture of government performance.
As a self-professed global city and smart nation, Singapore regularly proclaims the importance of accountability and transparency in society, as well as data-driven approaches in public discourse and policymaking. Lawrence Wong, prime minister, has also regularly spoken about the need to foster a more open and participatory society, and for government and citizens to “co-create” the future. “Everyone will have a say in shaping our way forward and building our shared future,” he said, at his recent cabinet swearing-in. “For I firmly believe that a more open and engaged society will strengthen, not weaken, Singapore.”
The realisation of these democratic ideals, however, are constrained by our closed-by-default data governance structure, which creates a landscape of limited information transparency that inhibits organising and policy evaluation. We lack accurate, meaningful data on a variety of issues. Several years ago, for example, Sylvia Lim of the Workers’ Party filed a parliamentary question requesting for the ethnic breakdown of the prison inmate population, and noted that analogous information is commonly available in other countries. Her request was denied and, mirroring Ang’s experience, did not come with a clear reason for denial. In the same vein, we also do not know the racial nor socio-economic composition of student enrollment in our autonomous universities and other post-secondary education institutes, although gender breakdown is available. (It appears like 2005 was the last time SingStat published ethnic breakdowns. That year, resident university graduates comprised 81 percent Chinese, 11.5 percent Indian, 5 percent “Other”, and 2.5 percent Malay.) Similarly, the official number of migrant worker deaths in the country remains unknown, even though migrant workers make up almost 20 percent of Singapore’s population.
There also remain gaps in knowledge on the issue of income distribution. According to SingStat, Singapore’s Gini Coefficient in 2023 was 0.371 after accounting for taxes, and 0.433 before. According to this calculation, the Gini Coefficient and by extension, inequality in Singapore, has been declining every year since 2013. However, Singapore’s calculation of the Gini Coefficient diverges from the conventional measure used globally in two important ways. First, while other measures incorporate income from all sources, including investments and property owned, Singapore’s includes only income from work. Put another way, dividends and rental income, among others, are omitted from the measure, perhaps vastly underestimating the true incomes of Singapore’s rich—and rendering cross-country comparisons imprecise, if not pointless. Second, Singapore’s Gini Coefficient excludes the income of shorter-term foreign workers and non-working families, such as retirees. It therefore obscures the actual income distribution across all residents—citizens, migrants with working passes, and non-working households. Moreover, Singapore doesn’t even measure wealth inequality.
In other words, there is a lack of data transparency on the distribution of income and wealth in Singapore, important channels for accumulating intergenerational wealth. This broader opacity also fueled the recent controversy regarding the sales of Good Class Bungalows, where the filing of property caveats allows buyers of these bungalows–some of the most expensive properties in Singapore – to keep their identity concealed. Wealth ownership is increasingly becoming a flashpoint in Singaporean politics. The lack of collective knowledge on these issues has implications for our assessment of equity and equality in Singapore. If people of a certain race are disproportionately imprisoned, if wealth is disproportionately concentrated in the top percentiles, does it not hint at structural inequities that need correcting? We do not know if these are true. However, without the data to even explore potential inequity, we cannot begin to have informed policy discussions on these salient issues.
Currently, data not already available on SingStat or individual ministry websites, can be obtained through written requests to ministries; grant collaborations with ministries and agencies; primary research; or Parliamentary Questions. But several obstacles remain. First, without a full catalog of available datasets, we’re not aware of what sorts of data have been collected and are available. Second, even if a catalog existed, many aspects of the data-request process are unclear: the criteria for approving or denying an application; the decision-makers involved; the timeline for decision; and the appeal process. Finally, data usage can come with conditions, such as a review and approval by the data-holding ministry for the usage of statistics derived from the data in research, or only partial release. The clear and institutionalised front-end of data application contrasts with the opaque back-end processes of decision-making for data release. These issues are born out of a closed-by-default model of data governance. The onus falls on the data user (the citizen) to justify the disclosure of data, rather than on the data holder (usually the government) to justify the withholding of data, as would be in an open-by-default model.
The weak presence of a data transparency movement in Singapore can partly be attributed to how we think about data. The economic value of data is widely celebrated while the political power of data is much less discussed. That is, society thinks of data primarily as an economic resource and only secondarily, vaguely, as a political resource. (A situation that entrenched interests may not mind.) The government has recognised the economic utility of data and has implemented various projects, including Data.gov.sg and Smart Nation initiatives to that end. For Singapore in particular, open data presents outsized benefits. Given our growing embrace of the knowledge and digital economy, the free flow of data is crucial for increasing productivity.
For the private sector, a ready pool of accessible data allows for knowledge transfer and collaborations across sectors, and the repurposing and reusing of data. (It also prevents unnecessary reduplication.) More importantly, accessible data enables fully-informed corporate decision-making. For instance, data on labour wage rates and foreign direct investment flows can help companies optimise their market expansion strategies. Access to high-quality, accurate data is especially important for local small- and medium-sized enterprises (SMEs), who might not be able to afford the datasets and databases that more well-endowed multi-national corporations can. Data transparency can level the playing field, helping them compete better in international markets.
For the public sector, data is also a public good. Data collected by the government through the census, commissioned surveys, individual ministries’ research, administrative data, and data from our built environment helps improve policy-making and public service delivery. In HDB Smart Towns, for example, a network of sensors and smart technologies enable the government to capture real-time information from the built environment to optimise infrastructure maintenance, property maintenance cycles, and community-building according to the living patterns of the residents. Taken together, data has the potential to generate economic growth as well as improve community living. This potential can be further maximised with greater data transparency. Increasing the accessibility and availability of data can help independent and institution-affiliated researchers come up with (alternative) policy solutions backed by data-driven research, or even applications and technologies that better exploit data for society’s good. Hack for Public Good, an annual initiative by GovTech’s Open Government Products, sees its techies “empowered to go into the community to discover public good problems in Singapore, then find and build solutions”. It’s a good example of the value that accessible data can create. The economic potential of data is maximised only when data are made open, transparent, and available—the more minds, the better.
Aside from its economic utility, data is also important for building political power. Data transparency supports socio-political organising and government accountability. In his 1991 paper, “Now out of Never: The Element of Surprise in the East European Revolution of 1989”, political scientist Timur Kuran discussed how no one—not political leaders, not political pundits, and not even the people themselves—anticipated the revolution that heralded the end of the Cold War, symbolised most dramatically by the fall of the Berlin Wall. Kuran was puzzled: why did East European governments, which seemed so impervious and enduring, collapse so swiftly and unexpectedly? He argued that the long-term suppression of information flow disincentivised citizens from expressing their true sentiments toward the regime. However, once a hint of revolt emerged, people felt safety in numbers, were able to coordinate, and revealed their displeasures against the regime. The latter then crumbled. Political scientists like Kuran contend that when information is transparent and allowed to circulate, it can reveal state weaknesses, citizen preferences, and disseminate strategies of resistance. As a form of information, data can similarly reveal public sentiments as well as areas of inequity or injustice in need of mobilisation.
Moreover, within Singapore’s specific political context, institutionalising data transparency as a fundamental political right is also important for levelling a political playing field that’s been titled over generations by information asymmetries that privilege the ruling party, and laws such as the Protection against Falsehoods and Misinformation Act (POFMA). Data transparency can allow every citizen to access the data on which conclusions are drawn, and judge the methodology, analysis, and data for themselves. No single citizen or politician can be the arbiter of truth. Citizens can fact-check the government just as the government can fact-check citizens.
Perhaps more importantly, institutionalising data transparency protects dissidents, activists, and members of civil society from potential persecution. When struck by a POFMA ruling, citizens can draw upon data transparency and freedom of information as an institutionalised right to demand the disclosure of the cache of relevant data that, according to the complainant, produce competing conclusions. Data transparency and freedom of information, given legal grounding through legislations such as a Freedom of Information Act (FOIA), thus acts as a bulwark against potential persecution for individuals who want to use data. When institutionalised, data transparency can create a safe and fair environment in which data usage can truly be enjoyed by all members of society. It might not be instinctive for some to embrace these broadened informational freedoms. Yet we must recognise that if society really wants to tackle the real scourge of misinformation and disinformation, then data transparency is the bedrock of a more enlightened, truthful public discourse.
When I visited Malaysia for research in 2023, one of the first things that I was told was, “Data is there, except for the ones that you need.” The more an issue speaks to the narratives that underpin Malaysia’s socio-political equilibrium—say corruption, cost of living, and racial and religious divides—the harder it is to locate relevant data. But it wasn’t all doom and gloom. In fact, my conversations with various data stakeholders and bureaucrats in Malaysia were filled with optimism—data is becoming more accessible and transparent in the country, due in large part to the joint efforts between civil society and the bureaucracy.
In many ways, Singapore’s and Malaysia’s data governance share foundational similarities. Data transparency in both countries is governed primarily by the Official Secrets Act (OSA) and the Statistics Act. Under OSA 1935, public servants in Singapore can be found guilty of sharing classified pieces of information to unauthorised parties, and the latter can be found guilty of receiving classified information. In Malaysia, under OSA 1972, any minister or chief minister of a State, and/or any public official whom these Ministers delegate their powers to, have the power to classify any government document as “Top Secret,” “Secret,” “Confidential,” or “Restricted”. In and of itself, the spirit of the law is justified—information that can jeopardise national security and societal well-being should not fall into the hands of bad actors. However, the OSA covers such a broad and sweeping range of public sector information, documents, and data that even innocuous information can become classified. Neither country’s OSA sets out clear criteria for determining document classification or declassification. In practice, OSA 1935 and 1972 therefore confer broad authority to the two governments to arbitrate on what sorts of information to share and what sorts to withhold.
Yet despite similar legal structures, Singapore and Malaysia have seen divergent outcomes in terms of information transparency. This is because Malaysia has in recent years been contemplating an open-by-default model.
As I spoke to data transparency activists, journalists, researchers, and bureaucrats over the course of six months in Kuala Lumpur, the possibility of a federal FOIA was brought up repeatedly in excited whispers by various informants. Two states, Penang and Selangor, have already implemented their own versions of FOIA, which creates an open-by-default model of data governance. Under FOIA, individuals can access government records that are not publicly available by submitting a formal request to the relevant agency or ministry, specifying the information sought. The agency is required to search for and provide access to existing records unless they are exempted by law. If a request is denied, the agency must provide a justification, and the requestor has the right to appeal. Appeals are typically reviewed by an independent body or an oversight authority separate from the agency that denied the request. In addition, FOIA requires that exemptions to disclosure be clearly stated within the Act. In this open-by-default model, the onus now falls upon the government to justify any withholding of data, rather than on the data user to provide grounds for the disclosure of data.
To make FOIA a reality, civil society needs to come together to champion for data transparency as an encompassing cause. In Malaysia, there are civil society organisations dedicated to promoting data transparency, most prominently the Sinar Project and the Center for Independent Journalism. These organisations work with the government to organise workshops and focus groups that foster a greater understanding of and support for data transparency. They also take practical steps to make information and data available, including creating depositories of open data datasets, archives of government documents that are no longer available on their original websites, and maps that visualise key electoral data. In a similar vein, leading Malaysian think tanks have also published reports that highlight the issue of data openness in Malaysia. The Center for Independent Journalism also commissioned a report that demonstrates how a proposed FOIA can be reconciled with OSA 1972, recognising the legal and logistical practicalities of the matter. Data transparency is a very tangible social justice movement in Malaysia.
In Singapore, advocacy for data transparency often takes the form of individual civil society groups, such as AWARE or Migrant Death Map, pushing for the disclosure of information and data relevant to their own causes. Recent debates on minimum income standards, prompted by independent reports, have also underscored the need for more and better data on poverty, income, and wages in Singapore.
Encouragingly, there has been more recognition and organisation around data transparency as a social justice cause in recent years. Academia.SG, for one, included a panel on Singapore’s data landscape in their Knowledge Praxis conference last year. Data users have also called out the inaccessibility of government data in various op-eds and essays published in, among others, The Straits Times. Media groups like Kontinentalist have similarly demonstrated the importance of data across numerous issues.
However, this momentum must expand to bring into its fold civil society groups that champion other causes. After all, information and data transparency affects everyone, and the movement to institutionalise it must reflect this reality. With an inclusive and encompassing banner, we can begin to push towards legal and policy options that can institutionalise data and information transparency. Contestations over data are ultimately contestations over the restraint of power. For those intent on centralising power, data is a political threat. For those resolved to redistribute power, data is an equalising force. Data outlines our reality; it tells us what we know, how we know what we know, and what we hold to be true or false. What sorts of data are available and what are not determine which stories are told and which are marginalised, who benefits and who does not. Ensuring data transparency, therefore, is about ensuring that no one person or group has a monopoly on answering these questions. It can empower society to come to these conclusions ourselves through fully informed democratic contestation and participation.
Neo Hui-Yuan is a PhD candidate at Cornell University. Her research examines information and data control in authoritarian regimes, with a regional focus on South-east Asia.
Letters in response to this piece can be sent to sudhir@jom.media. All will be considered for publication on our “Letters to the editor” page.
If you enjoy Jom’s work, do get a paid subscription today to support independent journalism in Singapore.
