~by: Donald Low~
Income inequality in Singapore has risen significantly in the last decade. Whether measured by the Gini coefficient, or by the ratio of incomes between the top and bottom quintiles, the evidence points to an incontrovertible fact: Singapore has become a lot more unequal in the last ten years or so. While the fact of increased inequality is beyond dispute, there is little agreement in government that it represents a problem that merits serious policy action.
The Singapore government’s approach to inequality is grounded in a number of implicit, but strongly-held, assumptions about the objectives of economic policy and the relationship between economic growth and social equity. These assumptions have attained almost mythical status among policymakers here. They are reflexively applied in any debate on inequality, and form a common point of reference – an internally consistent paradigm – against which alternative ideas for organising the social security system are evaluated.
The purpose of this essay is to articulate these mythical assumptions and to subject them to a closer examination than they are normally subject to. Like all myths, they contain an element of truth. But it behoves the policymaker to question and assess the validity of these myths in light of economic theory and evidence. Only then can they approach the problem of inequality in a pragmatic way, devoid of dogma and ideology.
Myth #1: Inequality is a necessary counterpart of economic dynamism and competitiveness
What the myth says
The first myth is the belief that rising inequality is an inevitable consequence of rapid economic growth, or more strongly, a necessary condition of economic competitiveness. Since unequal rewards spur individual effort and enterprise, it is argued that policies which reduce inequality invariably reduce incentives for people to work and strive.
Governments which try to mitigate the unequal consequences of globalisation – through more progressive taxation, more redistributive spending and stronger social safety nets – will raise costs for businesses and capital owners, who are now far more mobile than before. In their misguided attempt to prevent inequality from rising, governments will kill the goose that lays the golden eggs. They will sacrifice economic dynamism if they pursue inequality-reducing measures. There is an inescapable trade-off between economic growth and social equity. This is the “Golden Straightjacket” that Thomas Friedman famously talked about in The Lexus and the Olive Tree: when countries join the globalisation system, they find that their economy grows and their politics shrinks. If they want their economies to grow, governments have no choice but to slash income taxes, cut spending on social protection programmes, and roll back state provision.
Prime Minister Lee Hsien Loong also alluded to this trade-off between economic performance and social equity in his recent speech at the Comcare appreciation lunch:
“If you look at America and Europe, they have different models. In America, somewhat less welfare and greater emphasis on self-reliance. In many countries in Western Europe, a very comprehensive welfare state. You can see the result of these different approaches and the way the two societies and two economies feel. America has a more dynamic and competitive economy with fiercer competition. Europe has more generous benefits, more solidarity, not so strong competitiveness but the Europeans believe that they have made a rational choice, a rational trade-off. In return for less growth, they enjoy more welfare, more solidarity and they felt that they were the happier for their circumstances. But it was not entirely as happy as that. In fact, Europe was living beyond its means. It took some time for the problems to show up but I think Judgment Day has been brought forward by the financial crisis. After the huge hole in their budgets because of the rescue of the banks over the last couple of years, the Western Europeans have woken up to a very serious problem and they have been forced to make very tough choices.“
What the evidence says
At first glance, this myth seems consistent with international evidence and economic theory. Income inequality in almost all the rich countries has risen in the last 20-30 years. The reasons for this are well-known. Rapid technological advances have pushed up the wages of highly-educated, highly-skilled workers (what economists call the skill-biased premium), while globalisation and the entry of large numbers of low-wage workers from China, India and other developing economies hold down the wages of the less skilled in rich countries.
Other policy and institutional factors have contributed to the rise of inequality in rich countries too. The collapse of unions in the US for instance has reduced the bargaining power of labour relative to capital-owners. Immigration in some rich countries has brought low-wage competition to the non-tradable sectors of the economy. Meanwhile, reductions in corporate and personal income tax rates in many countries – partly in response to increasing global competition – have reduced the progressiveness of tax systems and accentuated inequality within countries.
But does the relationship between inequality and economic growth stand up to closer scrutiny? It appears not. Despite its intuitive logic, there is little evidence to show that more unequal countries do better economically, or that more equal ones pay a large economic price in terms of competitiveness.
Developed countries that spend less on social welfare – such as the United States – are not necessarily richer than those which spend more. Even in terms of per capita GDP, the US is not the richest country in the world, not even before the financial crisis. The World Bank data tells us the per capita income of the US in 2007 was $46,040. There were seven countries (all European) with higher per capita income in US dollar terms – starting with Norway ($76,450) at the top, through Luxemburg, Switzerland, Denmark, Iceland, Ireland and ending with Sweden. Excluding the two mini-states of Iceland and Luxemburg, this makes the US poorer in GDP per capita than six European countries – all of which spend considerably more on social welfare than the US does, and are much more egalitarian. In short, these European countries provide more generous benefits, have more social solidarity and outperform the US economically.
What about the claim that Europe’s social security systems are unsustainable? True, many of them need to be reformed for their ageing populations and be put on a stronger fiscal footing. But the countries which are facing the most severe fiscal problems – Greece, Italy, Portugal and Ireland – are hardly known for having generous social welfare systems. The ones with the most generous social provisions are the northern European countries – Germany, Netherlands, Denmark, Sweden, Finland and Norway. These are export-oriented, surplus economies with sound fiscal balances and strong social safety nets. The claim that Europe’s fiscal mess is the result of overly generous social welfare systems simply cannot be substantiated.
More generally, we should question whether the economic performance of countries should be measured only in terms of income, or GDP per capita. Surely, what citizens care about is not some abstract notion of aggregate production that does not take into account how that “national income” is distributed or whether the stuff which is produced is valued by society. Various economists have therefore suggested that in measuring economic progress, governments should also take into account other indicators of well-being. These include consumption, health and longevity, leisure time, and yes, how fairly incomes are distributed. By these measures, Singapore does not do as well as its GDP per capita level suggests. Klenow and Jones (2010) for instance shows that while Singapore’s per capita GDP in 2000 was 83% of the US’, Singaporeans’ well-being (measured by consumption adjusted for factors such as leisure, longevity and income inequality) was only 44% of the US’ level.
The conclusion that inequality neither contributes to economic growth nor is a necessary consequence of it should not be surprising at all. Economic growth is a complex process involving a number of technological, policy and institutional factors. It is not apparent that a country’s income inequality has a statistically significant correlation with economic growth or competitiveness. Even if it does, to the extent that citizens care about the distribution of income and other indicators of well-being, and not just the level of per capita income, nothing in economics says that countries should pursue only GDP per capita at the expense of the broader indicators of well-being. That some countries do is a consequence of their own values and politics, not of economic logic.
Myth #2: The best way to help the poor is to help the rich
What the myth says
A common refrain that one hears in the Singapore government is that we first have to grow the economic pie before we can share it. Like it or not, it is the rich and talented who invest, spot and exploit economic opportunities, and create jobs for the rest of us. Depriving them of their just rewards by levying high income or wealth taxes on the rich simply reduces the incentive for them to create wealth, thereby hurting the poor and the rest of society. In the long run, the best way to help the poor is to help the rich grow the pie.
This view, sometimes described as “trickle-down economics”, was articulated by the Finance Minister Tharman Shanmugaratnam at the 2010 Economics Society of Singapore dinner:
“The reality of course is that to create jobs and opportunities for those with lower skills, we have to provide adequate incentive for those with the talents and entrepreneurial abilities and the skills, adequate incentive for them to work, earn better incomes, grow their wealth. That is the reality we face. If we make it difficult for people who are able to seize opportunities in Asian markets and globally to make the most of their talents, if we make it difficult for them to earn their keep in Singapore and to feel that Singapore is a place where they can grow their business, grow opportunities, we will be unfortunately holding down the wages of the rest.”
What the evidence says
Have pro-rich policies led to faster economic growth which then raised the incomes at the bottom? The evidence on this is mixed at best. If one looks at the evidence among rich countries over the last sixty years, the evidence in fact suggests the opposite: countries tended to do grow faster in the years when they were doing more for the poor than in those years when they were trimming social safety nets and cutting taxes for the rich. Following the Second World War, there was rapid growth in progressive taxation and welfare spending in most of the rich capitalist countries. Despite this (or perhaps, partly because of this), the period between 1950 and 1973 saw the highest-ever growth rates in these countries. Before this period, per capita income in the rich capitalist economies used to grow at 1-1.5% per year. During the 1950-1973, it grew at 2-3% in US and Britain, 4-5% in Western Europe, and 8% in Japan. Since then, rich countries have never managed to grow faster than that.
When growth slowed in the mid-1970s onwards, market fundamentalists managed to convince their governments that the reduction in the share of income going to the capital owners was the reason for the slowdown. Across the rich world, and especially in the US and Britain, policies that distributed income from poor to the rich were introduced. There were tax cuts for the rich (top income tax rates were brought down) even as social welfare spending was being reduced. Financial deregulation created huge opportunities for speculative gains as well as astronomical pay cheques for top executives and financiers. Unions were weakened, making it easier for employers to sack their workers. And trade barriers were dismantled, putting downward pressure on low-end wages in the rich world.
During this period, income inequality is the US, already the highest in the rich world, rose to a level comparable to that of some Latin American countries. Much of this was driven by the rise of the super-rich in the US. Between 1979 and 2001, the top 5% in the US saw their share of national income increase from 15.5% to 21%. But what was more astonishing was the increase for the top 1%. Between 1979 and 2004, the top 1% of earners increased their share of national income from 8% to 14%. This was mainly because of the astronomical increase in executive pay, which in the aftermath of the financial crisis appears increasingly unjust and unjustified.
All this upward distribution of income might have been acceptable had it actually accelerated growth. But there is little evidence that these pro-rich, pro-market reforms actually grew the economic pie. According to World Bank data, the world economy used to grow in per capita terms at over 3% during the 1960s and 1970s, while since the 1980s it has been growing at the rate of 1.4% every year.
Trickle-down economics may also be justified if the benefits of growth do in fact trickle down. Again, the evidence from the US suggests that this does not occur if left to market forces. According to the Economic Policy Institute in the US, the top 10% of the US population appropriated 91% of the income growth between 1989 and 2006, with the top 1% taking 59%. In contrast, countries with a strong welfare state find it much easier to spread the benefits of extra growth. Through more redistributive fiscal systems, countries such as Sweden, Belgium, Germany and Netherlands have much more equal income distributions than the US even though their income distributions before taxes and transfers are worse than, or the same as, the US’.
To sum up, there is no reason to presume that trickle-down policies will accelerate growth or benefit the poor. This has not happened in general. Even when there is more growth, the trickle down that occurs through the market mechanism is very limited, as seen in the comparison of the US with other rich countries that have much stronger redistributive fiscal systems.
What about Singapore’s experience? Has trickle-down worked? It is less clear whether the increase in income inequality in Singapore over the last decade has been accompanied by a similarly perverse distribution of income to the super-rich as experienced in the US. Nevertheless, there are some reasons for concern. To begin with, the state may have become less redistributive at a time when its redistribution functions are needed most. Government policies over the decade may have accentuated the rising income inequality wrought by market forces. For instance, the tax system has become less progressive. Corporate and personal income taxes have been reduced significantly while the Goods and Services Tax (GST), which is a regressive tax, has been more than doubled. More liberal foreign worker policies might also have worsened income inequality in Singapore.
If these tax and labour policies had in fact generated more growth, and if the government had been aggressively redistributing the fruits of growth to large segments of the population affected by wage stagnation, trickle-down economics may not be all bad. But while the Singapore government has increased it spending on lower-income segments of the population, through Workfare and discretionary fiscal transfers, its redistribution has simply not been aggressive enough. This is demonstrated by fact that the income distribution after taxes and transfers has worsened at about the same rate as the income distribution based only on market incomes.
Myth#3: Inequality is not really a problem as long as there isn’t extreme poverty and incomes are rising across the board
What the myth says
The third myth says that policymakers should not worry about inequality per se. As long as there are opportunities for all to a good education, social mobility will dampen people’s demands for a fairer distribution of income. Furthermore, as long as everyone’s incomes are rising, we do not need to worry about the fact that incomes at the top are rising much faster than those at the bottom. As long as the rising tide is lifting all boats, we need not worry about the fact that some boats (the yachts for instance) are being lifted up much faster than the rest. Meanwhile, extreme poverty can be addressed with targeted measures such as social assistance. These limited welfare programmes for the indigent and those who cannot work and have no other means of financial support are affordable so long as they are strictly means-tested. There is no need for measures that redistribute incomes significantly since the problem – poverty – is a relatively limited one that can be surgicallyaddressed.
The underlying assumption behind this myth is that people care only about their absolute, and not relative, levels of income. So long as my income is rising, I should be happy and not begrudge my neighbour whose income is rising at a faster rate than mine. To do so is irrational, and governments should not pander to my green-eyed envy by redistributing income from my neighbour to me. As parents, we also teach our children to be satisfied with what they have and not to compare themselves with others who have more material possessions.
What the evidence says
To begin with, conventional economics does not prescribe that distributional concerns should be subordinate to growth objectives. Even if one takes a purely utilitarian view, there is a case to be made for aggressive redistribution. Since an additional dollar is worth more to the poor person than it is to the rich person, a utilitarian perspective says that any growth in incomes should accrue to the poor and that this should continue until people’s marginal utilities are equal. Furthermore, to the extent that increasing inequality reduces society’s well-being, it hurts the rich as much as it does the poor, and redistribution would enhance overall welfare. For instance, inequality has been shown to increase crime, which hurts the interests of everyone, not just the poor.
Behavioural economics provides additional reasons for redistribution. A number of behavioural experiments have suggested that people care jut as much about fairness and relative income as they do about absolute gains. In the ultimatum game for instance, people routinely reject offers that they consider too low even though they are better off accepting whatever offer that is made to them. This suggests that people believe that windfall gains which other people receive should be shared with others in society.
We should also try to understand the effects of inequality and why people care about them. One line of argument emphasises the role of positional goods, or goods in which people’s utility depends on how much they own relative to others. The point about positional goods – and of fashions and brands in general – is their relative attractiveness. Owning a better car or the latest branded fashion item gives me more utility when others do not have it, much like how buildings are valuable because of their location. With such goods, a rising tide does not lift all boats. I yearn to be not merely richer, but richer than my neighbours. The more important brands, fashion, houses, cars and other positional goods are in a society, the more relative income and inequality matter.
Rising inequality causes people to be more conscious of their status and to channel more of their spending to positional goods. Because the rich’s incomes are rising faster than everyone else’s, they can afford to spend more on such goods. Their spending causes “expenditure cascades” that induce others lower in the income ladder to also spend more just to keep up. But in an era of rising inequality, the incomes of the middle class and the lower-income groups are not rising as fast and they can only spend more on positional goods by diverting resources from non-positional goods (such as leisure time, or having babies) or by taking on credit.
What about the claim that equality of opportunity ensures social mobility and so governments need not worry about rising income or wealth inequality? Again, while this has some intuitive appeal, it is not borne out by empirical evidence. Countries which are more unequal, such as the US, also tend to be less mobile (as measured by how much of a person’s income is predicted by his parents’ income) than countries that are more equal, such as Canada or the welfare states of Western Europe. Why should this be? It turns out that equality of opportunity cannot be easily separated from equality of outcomes. Unequal resources easily translate into unequal access to opportunities, say to quality education. Families with more resources have greater means to ensure that their children have a better education. A more unequal society therefore finds it harder to achieve genuine equality of opportunity and social mobility than a more equal one.
To sum up, there are sufficient reasons – both theoretical and empirical – for policymakers to start taking inequality seriously even when incomes across the board are rising. They should also reject the glib and empirically false dichotomy between equal opportunities and equal outcomes. Being an opportunity society also means accepting the need for active government redistribution of incomes if market forces are producing more unequal outcomes.
Myth #4: Since pay is tied to ability, rising inequality is simply the result of increasing differences in people’s ability
What the myth says
In a market economy, people are paid according to their marginal productivity. If the worker in the top 20th percentile earns ten times more than the worker in the bottom 20th percentile, it is simply because the former is ten times more productive than the latter. While we find it difficult to accept this outcome, it is nevertheless the case that there are wide differentials in people’s ability and that these are reflected in wide income disparities. Attempts to reduce these pay differences – say by introducing minimum wage legislation – lead only to inefficient and rigid labour markets. Consequently, it follows that the best way to increase the incomes of middle- and low-wage Singaporeans is to increase their productivity. That their wages have been growing slowly in the last decade is the result of their stagnant productivity levels.
What the evidence says
As with the first three myths, this one also contains a kernel of truth. There is some evidence to suggest that Singapore’s workers in the service industries are less productive than their counterparts in other rich countries. Cleaners, bus drivers and construction workers are probably more productive in Sweden than they are in Singapore.
But the question still remains: why has inequality increased at a time when quality education in Singapore has become widely available? The democratisation of education suggests that differentials in productivity should have narrowed, which in turn suggests that wage differentials should also have been reduced. Singapore prides itself in having one of the world’s most successful education systems that not only enables bright children to do well, but also the rest. If so, why should lower-wage Singaporean worker, who has benefitted from the state education system, be less well-paid than his Swedish or Swiss counterpart doing a similar job? Is it possible that something else other than individual productivity determines our wages?
In reality, wages are not just a function of one’s productivity levels. The other important determinant of market wages in rich countries is immigration. Low-skilled workers in many European countries earn more because of tight immigration controls. If these countries were to import large numbers of low-skilled workers from poor countries, it is hardly conceivable that their low-skilled workers can earn the wages they do now, their relatively higher productivity levels notwithstanding. A Swedish bus driver for instance earns 50 times his Indian counterpart. If Sweden were to allow Indians and other immigrants from poor countries to enter its labour market, simple economics tells us that the wages of bus drivers in Sweden will be immediately depressed.
If the above analysis is correct, it suggests that to raise the wages of our low-skilled workers, the emphasis should not be on raising their productivity, but on reducing our intake of low-skilled foreign workers. All societies have limited capabilities to absorb immigrants. How open a society wants to be to immigrants is a not just an economic choice, but also a political choice with social and economic consequences. Too rapid an inflow leads not only to more competition for jobs and reduced wages, but also stretches the country’s physical and social infrastructures.
What about the high incomes of top-earners in Singapore? Aren’t their high incomes the result of their higher ability and productivity? Yes, but only to a point. It is not only, or even mainly, because they are cleverer and better-educated that the rich in Singapore earn many more times their counterparts in poor countries. They achieve this because they live in an economy that has better technologies, better infrastructure, better organised firms, better institutions, and better government – all things in large part products of collective actions taken over generations. As Warren Buffet once said in a television interview in 1995:
“I personally think that society is responsible for a very significant percentage of what I’ve earned. If you stick me down in the middle of Bangladesh or Peru or someplace, you’ll find out how much this talent is going to produce in the wrong kind of soil. I will be struggling thirty years later. I work in a market system that happens to reward what I do very well – disproportionately well.”
Why do these myths matter?
The myths matter for public policy because they shroud almost every discussion of inequality, and of how our social security system should be enhanced, in a thick and unquestioned set of assumptions. They act as an ideological blinker, and cause the policymaker to respond reflexively to any suggestion to redistribute incomes and reduce inequality with the argument that doing so will compromise the efficient working of markets. The myths matter because they reduce the ability of the Singapore government to pursue pragmatic and creative solutions to the challenge of inequality. And like many other myths and ideologies, they prevent a comprehensive and objective assessment of the policy alternatives successfully pursued by governments elsewhere.
As cognitive psychologists have shown, having a strong predisposition to a particular world-view deeply influences our assessment of the evidence. People – including policymakers – suffer from confirmation biases. This means that we seek and pay more attention to evidence that supports our existing world views rather than revise our world-views constantly in light of new evidence. When confronted with contrarian evidence, or evidence that goes against our world-views, the result is cognitive dissonance. This is uncomfortable and so we react by denying the evidence altogether (e.g. “the northern European countries are not as successful as they are made out to be”) or by discounting the veracity and credibility of the evidence (e.g. “your data is suspect”, or “your methodology is flawed”). That policymakers also suffer from such cognitive biases and blindsides is well-documented.
Questioning these myths opens up a number of possibilities for public policy in Singapore. At the broadest level, it throws up the question of whether the object of economic policy should be to maximise growth or GDP per capita regardless of its distributional consequences. Even if we assume that there is a trade-off between economic growth and social equity, the implication is not that societies should maximise the former at the expense of the latter. A trade-off means that governments should try to find the rightbalance of economic growth and social equity that reflects society’s preferences.
In Singapore’s context, given that GDP per capita is already one of the highest in the world (higher than the US’ by some estimates), the marginal gains in society’s well-being from further increases in income at the expense of social equity are likely to be very small. Put another way, Singaporeans are likely to be happier with a more equitable distribution of income (at the margin) than with further increases in GDP per capita that are not equitably distributed. The old adage – which says that we should concentrate on growing the economic pie before worrying about how it is distributed – which was valid and relevant when Singapore was relatively poor is no longer adequate as a guide for policymaking today.
Another implication is that policymakers should try to unlock the apparent mystery that the more successful European economies pose. Once these myths are jettisoned, policymakers will be liberated to ask: Why have these countries been able to combine high levels of economic performance and productivity with high levels of equality? What economic and social policies have enabled them to achieve this? Why do their citizens appear willing to accept the high rates of taxation that are necessary to finance their generous social provisions? How do they deal with the problems created by moral hazard and free-riders? What lessons can we draw from them in designing our own social safety nets?
My own study of the issue suggests that what matters more for economic growth and competitiveness is not how generous a country’s social protection system is, but how it isdesigned. It is not the level of social welfare spending that determines whether a country is on the efficiency-equity frontier but the way its social welfare programmes are structured and the incentive effects they create. Good design means paying attention to individual incentives, structuring the programmes in a way that encourages work, and redistributing incomes without reducing the incentive for investments. In short, the question governments should be asking themselves is not whether they can grow the economy without reducing social equity. It is how they can achieve this.
Governments, including Singapore’s, have more choices than they think. They can expand the range of choices they have by first rejecting these myths about inequality.
The following article was first published on Donald Low’s Facebook. TOC thanks him for giving us permission to republish it in full here.