Singapore, the social laboratory

Singapore is testing whether mass surveillance and big data can not only protect national security, but also engineer a more harmonious society
By Shane Harris, Published The Straits Times, 9 Aug 2014

IN OCTOBER 2002, Mr Peter Ho, the then Permanent Secretary for Defence in the tiny island city-state of Singapore, paid a visit to the offices of the Defence Advanced Research Projects Agency (DARPA), the US Defence Department's R&D outfit best known for developing the M16 rifle, stealth aircraft technology and the Internet.

Mr Ho didn't want to talk about military hardware. Rather, he had made the day-long plane trip to meet retired Navy Rear-Admiral John Poindexter, one of DARPA's then senior programme directors and a former national security adviser to president Ronald Reagan. Mr Ho had heard that Mr Poindexter was running a novel experiment to harness enormous amounts of electronic information and analyse it for patterns of suspicious activity - mainly potential terrorist attacks.

The two men met in Mr Poindexter's small office in Virginia, and on a whiteboard, Mr Poindexter sketched out for Mr Ho the core concepts of his imagined system, which Mr Poindexter called Total Information Awareness (TIA). It would gather up all manner of electronic records - e-mail, phone logs, Internet searches, airline reservations, hotel bookings, credit card transactions, medical reports - and then, based on predetermined scenarios of possible terrorist plots, look for the digital "signatures" or footprints that would-be attackers might have left in the data space. The idea was to spot the bad guys in the planning stages and to alert law enforcement and intelligence officials to intervene.

"I was impressed with the sheer audacity of the concept: that by connecting a vast number of databases, we could find the proverbial needle in the haystack," Mr Ho later recalled. He wanted to know whether the system, which was not yet deployed in the United States, could be used in Singapore to detect the warning signs of terrorism.

It was a matter of some urgency. Just 10 days earlier, terrorists had bombed a nightclub, a bar and the US consular office in the Indonesian island of Bali, killing 202 people and raising the spectre of Islamist terrorism in South-east Asia.

Mr Ho returned home inspired that Singapore could put a TIA-like system to good use. Four months later, he got his chance, when an outbreak of SARS swept through the country, killing 33, dramatically slowing the economy and shaking the tiny island nation to its core.

Using Mr Poindexter's design, the Government soon established the Risk Assessment and Horizon Scanning programme (RAHS, pronounced "roz") inside a Defence Ministry agency responsible for preventing terrorist attacks and "non-conventional" strikes, such as those using chemical or biological weapons - an effort to see how Singapore could avoid or better manage "future shocks".

Privacy concerns in the US

BACK in the US, however, the TIA programme had become the subject of enormous controversy. Just a few weeks after Mr Poindexter met Mr Ho, journalists reported that the Defence Department was funding experimental research on mining massive amounts of Americans' private data. Some members of Congress and privacy and civil liberties advocates called for TIA to be shut down. It was - but in name only.

In late 2003, a group of US lawmakers more sympathetic to Mr Poindexter's ideas arranged for his experiment to be broken into several discrete programmes, all of which were given new, classified code names and placed under the supervision of the National Security Agency (NSA). Unbeknownst to almost all Americans at the time, the NSA was running a highly classified programme of its own that actually was collecting Americans' phone and Internet communications records and mining them for connections to terrorists. Elements of that programme were described in classified documents disclosed last year by former NSA contractor Edward Snowden, sparking the most significant and contentious debate about security and privacy in the US in more than four decades.

Because of such uproars, many current and former US officials have come to see Singapore as a model for how they would build an intelligence apparatus if privacy laws and a long tradition of civil liberties weren't standing in the way. After Mr Poindexter left DARPA in 2003, he became a consultant to RAHS, and many American spooks have travelled to Singapore to study the programme first-hand.

They are drawn not just to Singapore's embrace of mass surveillance but also to the country's curious mix of democracy and authoritarianism, in which a paternalistic government ensures people's basic needs - housing, education and security - in return for almost reverential deference. Ten years after its founding, the RAHS programme has evolved beyond anything Mr Poindexter could have imagined. Across Singapore's national ministries and departments today, armies of civil servants use scenario-based planning and big-data analysis from RAHS for a host of applications beyond fending off bombs and bugs.

They use it to plan procurement cycles and budgets, make economic forecasts, inform immigration policy, study housing markets and develop education plans for Singaporean schoolchildren - and they are looking to analyse Facebook posts, Twitter messages and other social media in an attempt to "gauge the nation's mood" about everything from government social programmes to the potential for civil unrest.

In other words, Singapore has become a laboratory not only for testing how mass surveillance and big-data analysis might prevent terrorism, but also for determining whether technology can be used to engineer a more harmonious society.

Singapore was the perfect home for a centrally controlled, complex technological system designed to maintain national order.

"Singapore shouldn't exist. It's an invented country," one top-ranking government official told me on a recent visit, trying to capture the existential peril that seems to inform so many of the country's decisions.

But in less than 50 years, Singapore has achieved extraordinary success. Despite the Government's quasi-socialistic cradle- to-grave care, the city-state is enthusiastically pro-business, and a 2012 report ranked it as the world's wealthiest country, based on GDP per capita.

This economic rise might be unprecedented in the modern era, yet the more Singapore has grown, the more Singaporeans fear loss. Singaporeans' boundless ambition is matched only by their extreme aversion to risk.

SARS outbreak in 2003

THE SARS outbreak reminded Singaporeans that their national prosperity could be imperilled in just a few months by a microscopic invader that might wipe out a significant portion of the densely packed island's population.

Months after the virus abated, Mr Ho and his colleagues ran a simulation using Mr Poindexter's TIA ideas to see whether they could have detected the outbreak. It showed that if Singapore had previously installed a big-data analysis system, it could have spotted the signs of a potential outbreak two months before the virus hit the country's shores.

Prior to the SARS outbreak, for example, there were reports of strange, unexplained lung infections in China. Threads of information like that, if woven together, could in theory warn analysts of pending crises.

The RAHS system was operational a year later, and it immediately began "canvassing a range of sources for weak signals of potential future shocks", one senior Singaporean security official involved in the launch later recalled.

The system uses a mixture of proprietary and commercial technology and is based on a "cognitive model" designed to mimic the human thought process - a key design feature influenced by Mr Poindexter's TIA system. RAHS, itself, doesn't think. It's a tool that helps human beings sift huge stores of data for clues on just about everything. It is designed to analyse information from practically any source - the input is almost incidental - and to create models that can be used to forecast potential events.

Scenario planning

THOSE scenarios can then be shared across the Government and picked up by whatever ministry or department that might find them useful. Using a repository of information called an ideas database, RAHS and its teams of analysts create "narratives" about how various threats or strategic opportunities might play out. The point is not so much to predict the future as to envision a number of potential futures that can tell the Government what to watch for and when to dig further.

The officials running RAHS today are tight-lipped about exactly what data they monitor, though they acknowledge that a significant portion of "articles" in their databases comes from publicly available information, including news reports, blog posts, Facebook updates and Twitter messages.

Surveillance starts in the home, where all Internet traffic in Singapore is filtered, a senior Defence Ministry official told me. (Commercial and business traffic is not screened, the official said.) Traffic is monitored primarily for two sources of prohibited content: porn and racist invective. All other sites, including foreign media, social networks and blogs, are open to Singaporeans. But post a comment or an article that the law deems racially offensive or inflammatory, and the police may come to your door.

Singaporeans have been charged under the Sedition Act for making racist statements online, but officials are quick to point out that they don't consider this censorship. Hateful speech threatens to tear the nation's multi-ethnic social fabric and is therefore a national security threat, they say. Singaporean officials stress that citizens are free to criticise the Government, and they do.

Commentary that impugns an individual's character or motives, however, is off-limits because, like racial invective, it is seen as a threat to the nation's delicate balance. Journalists, including foreign news organisations, have frequently been charged under the country's strict libel laws. Not only does the Government keep a close eye on what its citizens write and say publicly, but it also has the legal authority to monitor all manner of electronic communications, including phone calls, under several domestic security laws aimed at preventing terrorism, prosecuting drug dealing and blocking the printing of "undesirable" material.

The surveillance extends to visitors as well. Mobile phone SIM cards are an easy way for tourists to make cheap calls and are available at nearly any store - as ubiquitous as chewing gum in the US. Criminals like disposable SIM cards because they can be hard to trace to an individual user. But to purchase a card in Singapore, a customer has to provide a passport number, which is linked to the card, meaning the phone company - and, presumably, by extension the Government - has a record of every call made on a supposedly disposable, anonymous device.

Privacy International reported that Singaporeans who want to obtain an Internet account must also show identification - in the form of the national ID card that every citizen carries - and Internet service providers "reportedly provide, on a regular basis, information on users to government officials". Perhaps no form of surveillance is as pervasive in Singapore as its network of security cameras, which police have installed in more than 150 "zones" across the country.

Most Singaporeans I met hardly cared that they live in a surveillance bubble and were acutely aware that they're not unique in some respects. "Don't you have cameras everywhere in London and New York?" many of the people I talked to asked. (In fact, according to city officials, "London has one of the highest number of CCTV cameras of any city in the world".) Singaporeans presumed that the cameras deterred criminals and accepted that in a densely populated country, there are simply things you shouldn't say.

This year, the World Justice Project, a US-based advocacy group that studies adherence to the rule of law, ranked Singapore as the world's second-safest country. Prized by Singaporeans, this distinction has earned the country a reputation as one of the most stable places to do business in Asia. But it's hard to know whether the low crime rates and adherence to the rule of law are more a result of pervasive surveillance or Singaporeans' unspoken agreement that they mustn't turn on one another, lest the tiny island come apart at the seams.

In 2009, Singapore's leaders decided to expand the RAHS system and the use of scenario planning far beyond the realm of national security - at least as it's commonly understood in the US. They established the Strategic Futures Network, staffed by deputy secretaries from every ministry, to export the RAHS methods across the entire Government. The network looks beyond national security concerns and uses future planning to address all manner of domestic social and economic issues, including identifying "strategic surprise" and so-called "black swan" events that might abruptly upset national stability.

Gathering public opinion

THE RAHS team has mounted a study on the public's attitude towards the housing system and what people want out of it. The provision of affordable, equitable housing is a fundamental promise that the Government makes to its citizens, and keeping them happy in their neighbourhoods has been deemed essential to national harmony.

Future planning has been applied to a broad variety of policy problems. It has been used to study people's changing attitudes about how kids should be educated and whether it's time to lessen Singapore's historically strong emphasis on test scores for judging student achievement. The Singapore Tourism Board used the methodology to examine trends about who will be visiting the country over the next decade. Officials have tried to forecast whether "alternative foods" derived from experiments and laboratories could reduce Singapore's near-total dependence on food imports.

But the future is one of the things that worries Singaporeans. Last year, the Government issued a so-called "Population White Paper" that described its efforts to grow the country and forecast a 30 per cent population increase by 2030, bringing the number of residents to as many as 6.9 million in the already crowded city-state. Immigrants were expected to make up half the total.

Singaporeans revolted. Four thousand people attended one rally against the population plan - one of the largest public protests in the country's history. The White Paper revealed a potential double threat: Singaporeans were already turning against the Government for growing the country too big and too fast, and now they were turning on their immigrant neighbours, whom they blamed for falling wages and rising home prices.

The protests shook the "nation's mood" at the highest level, and the Government was prepared to take drastic measures to quell the unrest, starting with cutting immigrant labour. The National Population and Talent Division - a kind of immigration-cum-human-resources department - intends to slow the growth of the workforce to about 1 to 2 per cent per year over the rest of the decade, which is a dramatic departure from the more than 3 per cent annual growth over the past 30 years.

With that, GDP growth is likely to retract to an average of 3 to 4 per cent per year. But the Government has concluded that a slowdown is the right price to pay for keeping a harmonious society. The data tells it so.

Singapore is now undertaking a multi-year initiative to study how people in lower-level service or manufacturing jobs could be replaced by automated systems like computers or robots, or be outsourced. Officials want to understand where the jobs of the future will come from so that they can retrain current workers and adjust education curricula. But turning lower-end jobs into more highly skilled ones - which native Singaporeans can do - is a step towards pushing lower-skilled immigrants out of the country.

If national stability means more surveillance and big-data scanning, Singaporeans seem willing to make the trade-off.

Singaporeans speak, often reverently, of the "social contract" between the people and their Government. They have consciously chosen to surrender certain civil liberties and individual freedoms in exchange for fundamental guarantees: security, education, affordable housing, health care.

But the social contract is negotiable and "should not be taken for granted", the RAHS team warns. "Nor should it be expected to be perpetual. Surveillance measures considered acceptable today may not be tolerable by future generations of Singaporeans." At least not if those measures are applied only to them. One future study that examined "surveillance from below" concluded that the proliferation of smartphones and social media is turning the watched into the watchers. These new technologies "have empowered citizens to intensely scrutinise government elites, corporations and law enforcement officials… increasing their exposure to reputational risks", the study found. From the angry citizen who takes a photo of a policeman sleeping in his car and posts it on Twitter to an opposition blogger who challenges party orthodoxy, Singapore's leaders cannot escape the watch of their own citizens.

In this tiny laboratory of big- data mining, the experiment is yielding an unexpected result: The more time Singaporeans spend online, the more they read, the more they share their thoughts with one another and their Government, the more they've come to realise that Singapore's light-touch repression is not entirely normal among developed, democratic countries - and that their Government is not infallible. To the extent that Singapore is a model for other countries to follow, it may tell them more about the limits of big data and that not every problem can be predicted.

FOREIGN POLICY

Shane Harris is a senior staff writer at Foreign Policy and the author of the forthcoming book @War: The Rise Of The Military-Internet Complex, which will be published in November this year.

EDITOR'S NOTE: Shane Harris' trip to Singapore was jointly sponsored by the Singapore International Foundation and the New America Foundation, where he is a fellow. The FP Group, which publishes Foreign Policy, is partnering with Singapore's Centre for Strategic Futures and Peter Ho, who is quoted above, to convene an expert forum on the global impact of rapid technological change. Neither Peter Ho nor the Singapore Government had any control over the content of this article.

RAHS programme not out to spy on Singaporeans

I refer to "Singapore, the Social Laboratory" (yesterday). I am concerned that a combination of hyperbole and speculation in the article will mislead the reader into drawing wrong conclusions about Singapore in general, and RAHS (Risk Assessment and Horizon Scanning) in particular.

I would like to make the following points to correct any misimpression that the Foreign Policy article has created.

First, RAHS is an open system. It is not used in a sinister way to conduct surreptitious surveillance. Our universities have used RAHS to teach students in courses such as how agricultural products are priced, and how to model the risk environment for commodity trading. It is a fantasy that RAHS is part of a larger system of mass surveillance that systematically intrudes into the privacy of individuals, and comprehensively filters Internet traffic.

Second, privacy is respected in Singapore in law and in practice. When RAHS was used in a retrospective study to test whether it could have picked up the SARS outbreak, the article speculates that it trawled through private databases without consent. That is wrong, in fact, and against the law. The reality was more mundane. RAHS looked at old newspaper reports and other open sources of information to see whether it would have picked up enough warning signs to alert the decision-makers that something was amiss.

Third, RAHS is not an omnipotent system that drives "armies of civil servants … to plan procurement cycles and budgets, make economic forecasts …". Neither is RAHS a system that can predict what is going to happen. Instead, it is a computer-based suite of tools that can, if used intelligently together with many other tools and good practice, enhance the quality of planning and decision-making, which ought to be the objective of all governments. The article makes RAHS out to be something that it is not.

Fourth, the speculations in the article extend beyond RAHS to how Singapore is governed. The tight control of mobile SIM cards in Singapore is portrayed as something slightly malevolent. Instead, it is a defensive and entirely legitimate policy against the threat of terrorism, and is a practice in many other countries. The article goes on to say that "no form of surveillance is as pervasive in Singapore as its network of security cameras". This is a gross and mischievous exaggeration. As the article points out, in many parts of the world, such as in the City of London, there are even denser networks of cameras, and for good reason.

Last, I feel that the article has strung together many unconnected pieces of a puzzle that is Singapore to construct something that fits an unfortunately common stereotype of "light-touch repression" in Singapore. Whatever drives government policy in Singapore is neither repressive, nor intrusive, but to serve important objectives of security, safety, and a better quality of life for all.

Peter Ho
The writer, former chief of the civil service, was referred to in the Foreign Policy article.
ST Forum, 10 Aug 2014