Governance of a City-State
Reframing the National Conversation

By Donald Low

The National Conversation, announced by the Prime Minister in his National Day Rally in August, has gotten off to an inauspicious start. Criticisms of it have centred on two main themes. The first is that the conversation so far has not been sufficiently representative and that it has not (yet) made a sincere effort at engaging sceptics and people who might be critical of the Establishment. The second is that based on current evidence, it appears that the scope of the National Conversation would be confined to policy and technocratic issues, and would avoid discussing the political and institutional reforms this country needs. The risk for the government is that a potentially valuable process in trust building and in having difficult but necessary conversations about Singapore’s political future is written off even before it has started.

How might the National Conversation Committee recover from this less-than-ideal beginning? For a start, I believe it is essential for the government to frame the discussion quite differently from what it has communicated so far. Most of the government’s comments on the National Conversation have focussed on the question of what future Singaporeans want for themselves and their country. Government leaders have also been quick to emphasise that the discussion ought to be based on an appreciation of the “trade-offs” that are involved in making a different set of policy choices. So for instance, if Singaporeans want lower immigration rates, they must also be prepared for slower growth, a more modest rate of improvement in their living standards, and fewer exciting job opportunities for their children. Other times, the government has been quick to remind people of the country’s inherent vulnerabilities, which means that certain options are simply not realistic, prudent or sustainable given Singapore’s particular circumstances and that it is critical for Singaporeans to forge a consensus on the way forward.

While not without merit, this framing of the issues is inherently conservative and biased in favour of the status quo. It confines the national conversation to the realm of what policymakers already know (or think they know) and filters our understanding of the issues through the lens of the government’s existing assumptions and ideologies. In a more complex and uncertain environment, such a narrative – based largely on what has worked for Singapore so far – is hardly conducive for an honest, frank and robust conversation about our fundamental choices.

Change, Crises and Conflicts: The Resilience Imperative

I believe it would be useful if the government were to frame the National Conversation less in terms of policy “trade-offs”, and more in terms of how we might build a resilient, future-ready Singapore. Resilience is the “capacity of a system, enterprise, or person to maintain its core purpose and integrity in the face of dramatically changed circumstances.” (Zolli, 2012) Resilience is critical because more so than before, Singapore will face greater volatility and uncertainty – politically, economically and socially. We will be buffeted by more episodes of rapid and disruptive change. We have already had some experience of this: in the last fifteen years, Singapore has had more bouts of economic slowdowns and recessions than in the thirty years before that. Politically, the general and Presidential elections of 2011 suggest that the Singapore polity has become more fragmented, pluralistic and heterogeneous, making good governance harder – but not impossible – to maintain.  Socially, the rapid ageing of our population, rising inequality, wage stagnation and slower social mobility – in the context of a much larger foreign population in Singapore – risk undermining social resilience and cohesion unless key institutions such as our social security system are reformed and adapted for this new context.

In this more complex environment, change, crises and conflicts would become more frequent, varied and unpredictable. The most valuable asset a system can have in such a landscape is not the ability to avoid shocks altogether (as this is unrealistic), or to maintain a superficial harmony and stability even as conflicts bubble below the surface. Rather, the critical asset the system needs to develop is the ability to bounce back, to adapt, and to maintain its core purpose and integrity in the face of rapid or disruptive change.

Thus, the central question which I hope the National Conversation will ask is how in an age of constant disruption and rapid change, we can build better institutional shock absorbers for ourselves. Beyond exploring what policy changes Singapore needs, I hope the National Conversation Committee also explores how Singapore’s governance system can be made more resilient in the face of a more complex and politically contested Singapore, and what norms and habits we need to cultivate – both as a government and people – to enhance social resilience. In short, the exercise should not just be about the substance of government (i.e. government policies), but also the system of governance (i.e. how we govern ourselves).

(This is not to say that far-reaching policy changes are not required. Readers interested in my thoughts on what policy and political changes are needed in the new normal may refer to a piece I had written for the Straits Times last year titled “Governing in the new normal.)

If we are to make resilience the key question in the National Conversation, we then need to identify what makes some systems thrive in the face of adversity and constant disruptions, and what makes others collapse under the slightest pressure. From the work of complexity scientists, ecologists, biologists, and cognitive psychologists, I propose three lessons which I believe are particularly salient for Singapore: cognitive diversity, distributed intelligence, and social trust.

Cognitive Diversity

One of the greatest risks facing well-run and long-established organisations is that success often breeds organisational brittleness and fragility. When the context changes, an organisation’s past successes, its aversion to change, and organisational inflexibility and groupthink often prevent it from adapting appropriately or quickly enough.

The concept of groupthink merits special attention given the long-established practice of elite governance in Singapore. Expounded by Irving Janis in 1972 to explain fiascos like the Bay of Pigs invasion, groupthink is an organisational pathology that can occur in tightly knit group of people which requires a high degree of cohesion to operate. The key signs of groupthink, as the futurist Andrew Zolli points out, are “a strong illusion of invulnerability by key decision makers; a belief in the inherent morality of the group; the stereotyping of those who do not agree with the group’s perspective; and overly simplistic moral formulations that dissuade deeper rational analysis. Self-appointed thought guards prevent alternative views from being aired and place significant pressure on dissenters, leading to an illusion of unanimity, even if dissent is rampant below the surface.”

A related cognitive and organisational blind spot is that of confirmation bias, which is the very human tendency to focus only on information that confirms our pre-existing assumptions, hypotheses and preconceptions, whether these are actually true or not. That some of our government leaders suffer from this is perhaps not surprising, considering their strongly held views on the primacy of economic growth, their aversion to redistribution and welfare, and their belief in the virtues of the Singapore system of elite governance. For instance, during the Budget 2012 debate, a number of PAP MPs argued that Singapore should never become like many European countries where runaway welfare spending has led these countries to the brink of economic and fiscal collapse. Clearly, what these MPs failed to realise (or chose to ignore) is that welfare spending in the heavily indebted European countries are low by rich country standards and that it has not been increasing as a share of GDP in the years before the crisis. Instead, it is the high welfare spenders among OECD countries – Germany, Netherlands, the Nordic countries – which have much lower levels of public debt and are in much better fiscal shape than the low welfare spenders of southern Europe or the US. The point is not that high welfare spending is good for a country’s fiscal health, but simply that it’s dangerous for us to draw conclusions about subjects we are not familiar with, relying only on how well our selective observations “fit” with our underlying beliefs and mental models.

The corrective to all this is to ensure a high degree of cognitive diversity in government, i.e. having different kinds of thinkers in policymaking. This cognitive diversity is, I believe, more important than the representative diversity that many have said was missing from the composition of the National Conversation Committee or the composition of the TV dialogue with the Prime Minister. That we have an ingrained, neurological tendency to seek out views which are consistent or similar to ours, and that we are naturally drawn to people like ourselves is at the heart of the confirmation bias and groupthink phenomena. It is also why the National Conversation Committee needs to make a special effort to engage with and listen to people not like themselves.

Beyond the National Conversation, I would also like to see this practice of actively listening to people not like ourselves embedded and institutionalised in our system of governance. In the US military, this is done through something popularly known as the Red Team University in Fort Leavenworth.  Started in 2004, Red Team University is the US military’s effort to systematically train military devil’s advocates who can bring critical insights to the battlefield and help commanding officers avoid the perils of overconfidence, strategic brittleness and groupthink. In an operating context of greater unpredictability and uncertainty, US military leaders realised that traditional military methods were less effective against the “swarm” methods increasingly employed by insurgents and terrorists. Developing new and more effective ones required not just smart people, but a systematic and institutionalised way for critical and dissenting viewpoints to surface and to challenge the received wisdoms of decision-makers. As Red Team University’s Greg Fontenot points out, “Our underlying assumption is not that people are lazy or incompetent, but that it’s just hard to critique your own work when you’re doing it. People reason by analogy and it’s hard to recognise your own untested hypotheses. If someone doesn’t challenge them, hubris can set in, bred of custom and complacency.” (Zolli, 2012)

Distributed Intelligence

A second important idea that has emerged from complexity science is that of distributed intelligence, which is equally vital for resilience. In trying to anticipate how a system would react to sudden changes and crises, we need to understand the role of networks in transmitting and amplifying shocks. Whether it is an Internet virus, social media, financial contagion or a public health pandemic, networks provide a useful frame of reference for understanding how these things spread and why the effects are often much larger than the initial cause.  Networks per se are neither good nor bad; they simply map the interdependencies and relationships (the links) between the different parts of the system (the nodes).

When systems become tightly coupled – such that the behaviour in one part of the system influences another part in complex and opaque feedback loops – they tend to become brittle and fragile over time. These highly coupled systems also tend to lose their ability to adapt as the survival of each part relies increasingly on the predictable and orderly functioning of the other parts to which it is tightly linked.

The US financial system leading up to the 2008 crisis exhibits many of the characteristics of a tightly coupled system. The financial health and profitability of banks became increasingly linked to the availability of cheap credit, rising house prices and the willingness of home buyers to continually refinance their mortgages, and the stability of the sub-prime mortgage market. Secondly, because financial institutions all pursued broadly similar business strategies, there was insufficient variety and heterogeneity in the system to encourage adaptation. Such highly coupled, monoculture systems, which can exhibit long periods of stability, are extremely vulnerable to shocks and perturbations that disturb key nodes in the network. The collapse of Lehman Brothers had such far-reaching and catastrophic consequences not because it was a particularly large investment bank, but because it was highly connected in the US financial system, and because every other institution was pursuing similar strategies. This “interlocking fragility” (a term coined by Nassim Taleb) made the entire financial system extremely vulnerable to the collapse of a single institution.

The solution to such tight coupling usually involves a combination of decoupling, developing better feedback mechanisms such that decision-makers have advance warning of impending risks and dangers, and ensuring that intelligence in the system is distributed across many nodes rather than concentrated in a few. Andrew Haldane of the Bank of England has argued that the solution to the complexity of the modern financial system – with its hornet’s nest of interconnections between financial institutions – and the homogeneity of those institutions’ strategies bears close similarities to what ecologists would prescribe for the preservation of fragile ecosystems. These include more complete and timely measures of the health of the financial system as a whole (as opposed to just individual institutions), an understanding of the dependencies between the various institutions, and measures to improve the financial system’s biodiversity.

What does all this have to do with governance in Singapore and the National Conversation? A lot, as it turns out. To begin with, Singapore’s continued success is currently too reliant on the PAP regime remaining competent, responsive, adaptive, and amenable to change. This reliance is accentuated by the fact that ours is such a highly centralised state (in network terms, this is represented by a single dominant node to which everything else is linked). In such a system, there is very little scope for the local experimentation (i.e. the biodiversity) that is essential for adaptation. Adaptation works through variety and selection – a variety of strategies and the selection of the best strategies – repeated over and over again. The efficiency gains of a highly centralised polity like ours are offset by the losses in adaptive capacity as a result of insufficient variety. And as the operating environment becomes more complex, the latter begins to matter more than the former.

So what might be done to bolster Singapore’s adaptive capacity? While it would not be realistic to expect the ruling party to (altruistically) engage in efforts to lessen the population’s dependence on it, there are certain things which it can do to expand the space for policy experimentation and innovation. For instance, the government should promote distributed intelligence through a far more open data environment than the current one, the establishment of collaborative platforms with civil society groups, and ways to crowd-source ideas from Singaporeans. By broadening the sources of intelligence and innovation for government, these measures also enhance social resilience.

A more radical idea to promote distributed intelligence would be to institute a proper system of decentralised local authorities. This is not as far-fetched an idea as it may first sound. With a resident population of more than 5 million, a strong case can be made for having three or four locally elected authorities providing not just municipal services, but also having differentiated policies in domains such as social services, environmental programmes, and some parts of the health and education systems (e.g. long-term care facilities, childcare and early childhood education services). Importantly, these local authorities should not just be an arm of the central government delivering services or implementing policies set at the national level. Instead, they should have autonomy to decide on local policies – such as the level of subsidies provided for locally provided services. Certain revenue streams should also be earmarked for financing these local authorities.

I recently had the privilege of spending two weeks in Australia on the invitation of a Melbourne-based think tank to give a series of talks on behavioural economics and to meet government officials at both the state and national levels. As a federation, Australian states and territories have a great deal of policy autonomy in a wide range of areas, including health and education. Much like America, Australians are used to states and cities pursuing quite different policies. The states aren’t particularly large, and the cities are even smaller. While such a decentralised system may be less efficient in that there are less economies of scale in government, it allows for a healthy degree of diversity and experimentation in government. Another model we could look to as we consider how we can decentralise intelligently is Britain, where recent budget cuts have produced a flurry of innovations in local governments in a country that, like Singapore, has traditionally been highly centralised. (See recent articles in The Economist magazine here.)

A more distributed and decentralised system of governance in Singapore also requires a higher level of trust in Singapore – not just trust in government, but also social trust (or trust between citizens). This is the third critical ingredient of social resilience.

Social Trust

Arguably the most important predictor of trust levels in a society is its level of income and wealth inequality. Inequality erodes trust because it increases status differences and social distance between individuals. It also undermines the sense of commonality and shared fate that the nation ought to engender in its citizens.

That Singapore has become considerably more unequal in the last decade is an empirical fact which is beyond dispute. The disagreement is over whether inequality represents a problem that merits policy attention and aggressive government action. Recent government statements suggest that while it is prepared to spend more on lower-income groups, rising inequality as a global phenomenon is inevitable and unavoidable. And as an aspiring global city that is extremely open to global capital and labour flows, it is only “natural” that Singapore will have higher than normal levels of inequality. This intransigence in government’s thinking on inequality is lamentable, and it is incumbent on the National Conversation Committee to address the issue more honestly and with much less ideological baggage than the government has done so far.

I have previously written about the relationship between inequality and social trust in an article for the Straits Times titled, “Beware the inequality trap.” The main argument there was that our current approach of highly targeted (and means-tested) social spending, although fiscally more prudent, is not sufficiently equipped to promote social inclusion and trust. I argued for the approach favoured by the northern European countries, one which relies more on universal programmes – often financed by social insurance – that benefit large majorities of their populations. These systems emphasise the government’s role in redistributing incomes, and in fostering solidarity and social trust. The importance of social trust has become widely accepted in the social sciences in recent years. One reason for this is that generalised trust in society is correlated with a number of normatively desirable things. For instance, people who believe that in general most other people in their society can be trusted are also more inclined to have a positive view of their public institutions, to participate more in their civic and political organisations, to give more to charity, and to be more tolerant towards minorities and people not like themselves.

In a seminal 2005 paper, Rothstein and Uslaner argued that countries with more universal social programmes enjoyed higher levels of social trust for at least three reasons. First, because universal programmes are more redistributive than means-tested ones aimed at the poor, they result in much lower levels of economic inequality once government taxes and transfers are taken into account. Second, since inclusive programmes are based on the principle of equal treatment and minimise bureaucratic discretion, they increase the sense of “equal opportunities” more so than means-tested programmes. The greater use of social insurance – often with the state under-writing the premiums of lower-income and other disadvantaged citizens – also enhances the sense of citizenship and access to opportunities. Third, means-tested programmes often accentuate class and racial divisions within a society, and lead to less generalised trust. By contrast, universal programmes enhance solidarity and the perception of a shared fate among citizens.

Rothstein and Uslaner also contend that societies with an initial level of high inequality are less likely to establish universal social programmes that enhance social trust. The logic of why inclusive programmes are more redistributive than programmes narrowly targeted at the poor is complex and somewhat counter-intuitive. This makes it difficult for the advocates of these programmes to attract support from disadvantaged groups. Second, inclusive programmes also extend benefits to better-off groups. In a society with an already high level of inequality, it is difficult to explain why scarce public resources should go to the middle class. Rothstein and Uslaner predict that societies with high initial levels of inequality are likely to find themselves stuck in an inequality trap characterised by low levels of social trust, a reluctance to embrace more inclusive social programmes, and a continued reliance on targeted programmes that make a sharp distinction between those entitled to benefits and those who are not.

Of course, inclusive universal social programmes have their costs too. Broad-based benefits in childcare, healthcare, pensions and unemployment protection cost more than means-tested ones. In northern European countries, generous benefits have to be financed by a wide range of higher taxes and social insurance contributions. But policymakers should weigh the costs of inclusive universal programmes against their benefits in terms of fostering norms of fairness, and in promoting social trust, citizenship and solidarity.

It would be all too easy for Singaporean policymakers to dismiss the inclusive model of social spending as too costly, too corrosive of our work ethic and competitiveness, and too unique to the cultural contexts of northern European societies. In our multi-ethnic context, given our heavy reliance on foreign investments, policymakers here may argue that we cannot afford the aggressively redistributive model of high taxes-and-transfers practised by the large welfare states of Europe. But notwithstanding the differences in our respective contexts, there are lessons from the northern European experience that are instructive for us. Perhaps the most important is that in considering the design of social programmes, the narrow objectives of efficiency and getting incentives right should be complemented by a nuanced understanding of the wider societal benefits that universal social programmes may generate. In theory, means-tested programmes promise to limit moral hazard and “deadweight funding”. In practice, they often result in high administrative costs, encourage rent-seeking behaviours, and create a sharp distinction between those entitled to subsidies and those who are not. In contrast, a more inclusive approach to social spending, although more expensive, can strengthen norms of fairness, promote social trust and resilience, and foster an egalitarian ethos.

Thus, my final hope for the National Conversation is that it takes a hard, honest – even radical – look at how our social compact should be adapted for a very different socio-political and economic context. This consideration should supported by the necessary data on the full range of fiscal resources available to the state over the next 20 or 30 years. In the absence of these two critical ingredients – ideological honesty and data transparency – the natural instinct of the PAP government is to pontificate on the virtues of a small state and low taxes, and on the evils of a welfare state.


The National Conversation should not just be an exercise to defend and rationalise existing policies and policy assumptions. Neither should it be confined to policy changes, without also looking at the institutional and structural factors that enable a system to thrive in an era of rapid and disruptive change. Above all, it should not avoid conflict and seek only harmony and consensus (which is likely to be artificial in any instance). Instead, the National Conversation should be a platform for honest and generative, and if necessary, heated discussions on how our governance system can be made more resilient and future-ready in the face of increasing complexity, volatility and uncertainty. This resilience and future-readiness cannot be achieved by simply making right policy changes, or by stubbornly asserting current governance principles, or by imposing consensus from above. Instead, it requires us to consider how the current governance system should be adapted. It requires us to move away from the current highly centralised, consensus-seeking system of elite governance to one that aims for greater cognitive diversity, harnesses distributed ways of gathering intelligence and generating innovations, and fosters social trust through a more egalitarian and universalist social compact. That would be a national conversation worth having.



“Most people instinctively avoid conflict, but as Margaret Heffernan shows us, good disagreement is central to progress. She illustrates (sometimes counterintuitively) how the best partners aren’t echo chambers — and how great research teams, relationships and businesses allow people to deeply disagree.

“The former CEO of five businesses, Margaret Heffernan explores the all-too-human thought patterns — like conflict avoidance and selective blindness — that lead managers and organizations astray.” – From TED Talks.


Headline picture by Bernard Oh / Creative Commons license.



• Bhaskaran et al (2012), Inequality and the Need for a New Social Compact, IPS Singapore Perspectives 2012 Background paper

• Tim Harford (2011), Adapt: Why Success Always Starts with Failure

• Andrew Haldane (2009), Rethinking the Financial Network, speech delivered at the Financial Student Association, Amsterdam

• Margaret Heffernan (2011), Wilful Blindness: Why we Ignore the Obvious at our Peril

• Margaret Heffernan (2012), Dare to Disagree

• Steven Johnson (2012), Future Perfect: The Case for Progress in a Networked Age

• Donald Low (2011), Governing in the new normal, Straits Times, 9 Sep 2011

• Donald Low and Yeoh Lam Keong (2011), Beware the inequality trap, Straits Times, 14 Dec 2011

• Jonas Pontusson (2005), Inequality and Prosperity: Social Europe vs Liberal America

• Bo Rothstein and Eric Uslaner (2005), All For One: Equality, Corruption and Social Trust, World Politics, Volume 58, Number 1.

•Andrew Zolli and Ann Marie Healy (2012), Resilience: Why Things Bounce Back


This essay was made possible as a result of my recent visit to Australia. My thanks to John Daley, CEO of the Grattan Institute in Melbourne, and his able team for putting together a wonderfully stimulating programme and for organising the various meetings I had with senior Australian government officials. They provided me with many insights on how the public service can work collaboratively with citizens, experiment with evidence-based policies, and thrive in a politically contested and rambunctious environment. I’m also grateful to my Singaporean friends in Melbourne – Christopher Phoon at the Victoria state government and Louisa Chang at the University of Melbourne – for sharing their ideas and experiences in Australia with me. They provided me a glimpse of the vast possibilities that are available to us if only we open our minds.

  • Tags:

Subscribe to our newsletter

Sign up to our mailing list to get updated with our latest articles!