for richer, for poorer - paris school of economics economoist... · est man is a mexican (carlos...

October 13th 2012

S P E C I A L R E P O R T

W O R L D E C O N O M Y

For richer, for poorer

WorldE.indd 1 25/09/2012 16:17

IN 1889, AT the height of America’s �rst Gilded Age, George Vanderbilt II,grandson of the original railway magnate, set out to build a country es-tate in the Blue Ridge mountains of North Carolina. He hired the mostprominent architect of the time, toured the chateaux of the Loire for in-spiration, laid a railway to bring in limestone from Indiana and em-ployed more than 1,000 labourers. Six years later �Biltmore� was com-pleted. With 250 rooms spread over 175,000 square feet (16,000 squaremetres), the mansion was 300 times bigger than the average dwelling ofits day. It had central heating, an indoor swimming pool, a bowling alley,lifts and an intercom system at a time when most American homes hadneither electricity nor indoor plumbing.

A bit over a century later, America’s second Gilded Age has nothingquite like the Vanderbilt extravaganza. Bill Gates’s home near Seattle isfull of high-tech gizmos, but, at 66,000 square feet, it is a mere 30 timesbigger than the average modern American home. Disparities in wealthare less visible in Americans’ everyday lives today than they were a cen-tury ago. Even poor people have televisions, air conditioners and cars.

But appearances deceive. The democratisation of living standardshas masked a dramatic concentration of incomes over the past 30 years,on a scale that matches, or even exceeds, the �rst Gilded Age. Includingcapital gains, the share of national income going to the richest 1% ofAmericans has doubled since 1980, from 10% to 20%, roughly where itwas a century ago. Even more striking, the share going to the top 0.01%�some 16,000 families with an average income of $24m�has quadrupled,from just over 1% to almost 5%. That is a bigger slice of the national piethan the top 0.01% received 100 years ago.

This is an extraordinary development, and it is not con�ned toAmerica. Many countries, including Britain, Canada, China, India andeven egalitarian Sweden, have seen a rise in the share of national incometaken by the top 1%. The numbers of the ultra-wealthy have soaredaround the globe. According to Forbes magazine’s rich list, America hassome 421 billionaires, Russia 96, China 95 and India 48. The world’s rich-est man is a Mexican (Carlos Slim, worth some $69 billion). The world’slargest new house belongs to an Indian. Mukesh Ambani’s 27-storey sky-scraper in Mumbai occupies 400,000 square feet, making it 1,300 timesbigger than the average shack in the slums that surround it.

For richer, for poorer

Growing inequality is one of the biggest social, economic and

political challenges of our time. But it is not inevitable, says

Zanny Minton Beddoes

1

A C K N O W L E D G M E N T S

CONTENT S

This special report bene�ted from

the time and insight of many people

in addition to those mentioned in

the text. The author would like to

thank in particular: Daron Acemoglu,

David Coady, Ashwini Deshpande,

Francisco Ferreira, Sanjeev Gupta,

Himanshu, Peter Lanjouw, Nora

Lustig, Pratap Bhanu Mehta, Marcelo

Neri, Changyong Rhee, Mônica

Rodrigues and Dinesh Tiwari.

A list of sources is at

Economist.com/specialreports

An audio interview with

the author is at

Economist.com/audiovideo/

specialreports

SPECIAL REPORT

WORLD ECONOMY

2 The Economist October 13th 2012

4 History

As you were

5 Sizing the gap

Like a piece of string

7 Measuring social mobility

Like father, not like son

8 United States

The rich and the rest

9 America’s tax system

Makers and takers

11 Asia

Crony tigers, divideddragons

12 Inequality in an Indian

village

Lessons from Palanpur

14 Latin America

Gini back in the bottle

16 Sweden

The new model

17 Trade-o�s

Having your cake

19 Policy prescriptions

A True Progressivism

% change ininequality1980-2010†

-1 to -20

0 to 20

>20 No data

<-20

More equal

More unequal

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Brazil

China

United States

Britain

India

Germany

South Africa

Sweden

1980 or earliest 2010 or latest

More or less unequalIncome inequality, Gini coefficient*

Sources: Forbes;IMF; The Economist

*0=perfect equality, 1=perfect inequality,inequality of disposable income; where

unavailable consumption or expenditure data†Or closest year available ‡September 2012 §March 2012

Russia18.696

India10.948

China2.695

Germany7.255

Britain4.036

United States10.5421‡

Brazil6.237

Net worthas % of GDP

Number ofbillionaires§

The concentration of wealth at the very top is part of amuch broader rise in disparities all along the income distribu-tion. The best-known way of measuring inequality is the Ginicoe�cient, named after an Italian statistician called CorradoGini. It aggregates the gaps between people’s incomes into a sin-gle measure. If everyone in a group has the same income, theGini coe�cient is 0; if all income goes to one person, it is 1.

The level of inequality di�ers widely around the world.Emerging economies are more unequal than rich ones. Scandi-navian countries have the smallest income disparities, with aGini coe�cient for disposable income of around 0.25. At the oth-er end of the spectrum the world’s most unequal, such as SouthAfrica, register Ginis of around 0.6. (Because of the way the scaleis constructed, a modest-sounding di�erence in the Gini ratio im-plies a big di�erence in inequality.)

Income gaps have also changed to varying degrees. Ameri-ca’s Gini for disposable income is up by almost 30% since 1980, to0.39. Sweden’s is up by a quarter, to 0.24. China’s has risen byaround 50% to 0.42 (and by some measures to 0.48). The biggestexception to the general upward trend is Latin America, long theworld’s most unequal continent, where Gini coe�cients havefallen sharply over the past ten years. But the majority of thepeople on the planet live in countries where income disparitiesare bigger than they were a generation ago.

That does not mean the world as a whole has become moreunequal. Global inequality�the income gaps between all peo-ple on the planet�has begun to fall as poorer countries catch upwith richer ones. Two French economists, François Bourguignonand Christian Morrisson, have calculated a �global Gini� thatmeasures the scale of income disparities among everyone in theworld. Their index shows that global inequality rose in the 19thand 20th centuries because richer economies, on average, grewfaster than poorer ones. Recently that pattern has reversed andglobal inequality has started to fall even as inequality withinmany countries has risen. By that measure, the planet as a wholeis becoming a fairer place. But in a world of nation states it is in-equality within countries that has political salience, and this spe-cial report will focus on that.

From U to N

The widening of income gaps is a reversal of the pattern inmuch of the 20th century, when inequality narrowed in manycountries. That narrowing seemed so inevitable that Simon Kuz-nets, a Belarusian-born Harvard economist, in 1955 famously de-scribed the relationship between inequality and prosperity asan upside-down U. According to the �Kuznets curve�, inequalityrises in the early stages of industrialisation as people leave theland, become more productive and earn more in factories. Onceindustrialisation is complete and better-educated citizens de-

mand redistribution from their government, it declines again. Until 1980 this prediction appeared to have been vindicat-

ed. But the past 30 years have put paid to the Kuznets curve, atleast in advanced economies. These days the inverted U hasturned into something closer to an italicised N, with the �nalstroke pointing menacingly upwards.

Although inequality has been on the rise for three decades,its political prominence is newer. During the go-go years beforethe �nancial crisis, growing disparities were hardly at the top ofpoliticians’ to-do list. One reason was that asset bubbles andcheap credit eased life for everyone. Financiers were growingfabulously wealthy in the early 2000s, but others could also bor-row ever more against the value of their home.

That changed after the crash. The bank rescues shone aspotlight on the unfairness of a system in which a�uent bankerswere bailed out whereas ordinary folk lost their houses and jobs.And in today’s sluggish economies, more inequality oftenmeans that people at the bottom and even in the middle of theincome distribution are falling behind not just in relative but alsoin absolute terms.

The Occupy Wall Street campaign proved incoherent andephemeral, but inequality and fairness have moved right up thepolitical agenda. America’s presidential election is largely beingfought over questions such as whether taxes should rise at thetop, and how big a role government should play in helping therest. In Europe France’s new president, François Hollande, wantsa top income-tax rate of 75%. New surcharges on the richest arepart of austerity programmes in Portugal and Spain.

Even in more buoyant emerging economies, inequality is agrowing worry. India’s government is under �re for the lack of�inclusive growth� and for cronyism that has enriched insiders,evident from dubious mobile-phone-spectrum auctions anddodgy mining deals. China’s leaders fear that growing dispari-ties will cause social unrest. Wen Jiabao, the outgoing prime min-ister, has long pushed for a �harmonious society�.

Many economists, too, now worry that widening incomedisparities may have damaging side-e�ects. In theory, inequalityhas an ambiguous relationship with prosperity. It can boostgrowth, because richer folk save and invest more and becausepeople work harder in response to incentives. But big incomegaps can also be ine�cient, because they can bar talented poorpeople from access to education or feed resentment that resultsin growth-destroying populist policies.

The mainstream consensus has long been that a growingeconomy raises all boats, to much better e�ect than incentive-dulling redistribution. Robert Lucas, a Nobel prize-winner, epito-mised the orthodoxy when he wrote in 2003 that �of the tenden-cies that are harmful to sound economics, the most seductiveandpoisonous is to focus on questions of distribution.�

The Economist October 13th 2012 3

WORLD ECONOMY

SPECIAL REPORT

2

1

2


WORLD ECONOMY

SPECIAL REPORT

1

Disposable incomeMarket income

Gini coefficient*, late 2000s

0.50

0.55

0.60

0.65

0.70

1820 50 70 90 1910 29 50 70 92 2008†

Global inequality, Gini coefficient*

0

10

20

25

15

5

30

1913 20 30 40 50 60 70 80 90 2000 10

US SwedenFranceBritain Germany

Top 1% income share‡, %

Sources: The World Top Incomes Database; IMF; OECD; World Bank; “Inequality among World Citizens: 1820-1992", by Bourguignon & Morrisson, The AmericanEconomic Review, 2002; "A short history of global inequality: The past two centuries”, by Branko Milanovic, Explorations in Economic History, May 2011 ‡Includes capital gains, except Britain and France

*0=perfect equality, 1=perfect inequality †Estimate

0 0.1 0.2 0.3 0.4 0.5UnitedStates

Britain

Japan

Sweden

Germany

OECD-29average

But now the economics establishment has become con-cerned about who gets what. Research by economists at the IMF

suggests that income inequality slows growth, causes �nancialcrises and weakens demand. In a recent report the Asian Devel-opment Bank argued that if emerging Asia’s income distributionhad not worsened over the past 20 years, the region’s rapidgrowth would have lifted an extra 240m people out of extremepoverty. More controversial studies purport to link widening in-come gaps with all manner of ills, from obesity to suicide.

The widening gaps within many countries are beginning toworry even the plutocrats. A survey for the World Economic Fo-rum meeting at Davos pointed to inequality as the most pressingproblem of the coming decade (alongside �scal imbalances). Inall sections of society, there is growing agreement that the worldis becoming more unequal, and that today’s disparities and theirlikely trajectory are dangerous.

Not so fast

That is too simplistic. Inequality, as measured by Gini coef-�cients, is simply a snapshot of outcomes. It does not tell youwhy those gaps have opened up or what the trend is over time.And like any snapshot, the picture can be misleading. Incomegaps can arise for good reasons (such as when people are reward-ed for productive work) or for bad ones (if poorer children do notget the same opportunities as richer ones). Equally, inequality ofoutcomes might be acceptable if the gaps are between youngpeople and older folk, so may shrink over time. But in societieswithout this sort of mobility a high Gini is troubling.

Some societies are more concerned about equality of op-portunity, others more about equality of outcome. Europeanstend to be more egalitarian, believing that in a fair society thereshould be no big income gaps. Americans and Chinese put moreemphasis on equality of opportunity. Provided people can moveup the social ladder, they believe a society with wide incomegaps can still be fair. Whatever people’s preferences, static mea-sures of income gaps tell only half the story.

Despite the lack of nuance, today’s debate over inequalitywill have important consequences. The unstable history of LatinAmerica, long the continent with the biggest income gaps, sug-gests that countries run by entrenched wealthy elites do not dovery well. Yet the 20th century’s focus on redistribution broughtits own problems. Too often high-tax welfare states turned out tobe ine�cient and unsustainable. Government cures for inequal-ity have sometimes been worse than the disease itself.

This special report will explore how 21st-century capitalismshould respond to the present challenge; it will examine the re-cent history of both inequality and social mobility; and it will of-fer four contemporary case studies: the United States, emergingAsia, Latin America and Sweden. Based on this evidence it will

make three arguments. First, although the modern global econ-omy is leading to wider gaps between the more and the less edu-cated, a big driver of today’s income distributions is governmentpolicy. Second, a lot of today’s inequality is ine�cient, particu-larly in the most unequal countries. It re�ects market and gov-ernment failures that also reduce growth. And where this is hap-pening, bigger income gaps themselves are likely to reduce bothsocial mobility and future prosperity.

Third, there is a reform agenda to reduce income disparitiesthat makes sense whatever your attitude towards fairness. It isnot about higher taxes and more handouts. Both in rich andemerging economies, it is about attacking cronyism and invest-ing in the young. You could call it a �True Progressivism�. 7

JANE AUSTEN’S �PRIDE AND PREJUDICE� is a story aboutlove. It is also about inheritance and income gaps. The hero-

ine, Elizabeth Bennet, comes from a well-o� family, the secondof �ve daughters. But her �nancial future is dark because, in theabsence of sons, her father’s estate will pass to a cousin. Eliza-beth’s suitor, the brooding Mr Darcy, is fabulously wealthy. Toher mother’s horror, Elizabeth at �rst rebu�s him.

All ends happily when Elizabeth decides that Darcy is rav-ishing after all. But her mother’s reaction is a rational response tothe realities of income distribution and social mobility in Aus-ten’s time. In an entertaining analysis of inequality, �The Havesand the Have Nots�, Branko Milanovic works out that by marry-ing Mr Darcy, Elizabeth would increase her income 100-fold.Without him, she would have the same income as a merchantseaman. With him she would be catapulted into the top 0.1%.

Before the industrial revolution, wealth gaps betweencountries were modest: income per person in the world’s tenrichest countries was only six times higher than that in the tenpoorest. But within each country the distribution of income wasskewed. In most places a small elite lorded it over a mass of peas-ants. There was little social mobility except, as Elizabeth found,through marriage. Colonial America was an exception to thisfeudal sclerosis. Research by Peter Lindert and Je�rey William-son shows that on the eve of the American revolution incomes

History

As you were

After a period on the wane, inequality is waxing again


WORLD ECONOMY

SPECIAL REPORT

2

1

in the 13 colonies that formed the United States were more equalthan in virtually �any other place on the planet�.

The industrial revolution widened the gaps both betweencountries and within them. As incomes accelerated in westernEurope and then America, the distance between these countriesand others grew. So, too, did internal income disparities. Onestudy suggests that England’s Gini coe�cient shot up from 0.4 in1823 to 0.63 in 1871. Mill workers were more productive andearned more than rural labourers. The great industrialists reapedthe rewards of building railways, steel mills and other transfor-mative technologies. Their fortunes were also boosted by mo-nopolistic power and crony capitalism.

The growth of the industrial workforce brought increasingpolitical pressure for redistribution. Communism was the mostdramatic result. But capitalist economies changed profoundlytoo. In response �rst to the formation of workers’ unions and therise of socialist parties and then to the Depression, politicians onboth sides of the Atlantic introduced progressive taxes, govern-ment regulation and social protection. In Germany Bismarckpioneered pensions and unemployment insurance in the 1880s.In America Theodore Roosevelt’s Square Deal broke up mono-polies (�trusts�) in the �rst decade of the 20th century. In the1930s the New Deal introduced Social Security (pensions), dis-

ability and unemployment insurance. In Britain Lloyd George’sPeople’s Budget of 1909 raised income taxes and inheritance tax-es at the top to fund basic pensions as well as unemploymentand health insurance for workers. This spartan social safety netwas transformed by the Labour government after 1945 with a Na-tional Health Service and a system of cradle-to-grave bene�ts.

Of the three levers used to narrow inequality�taxation,government spending and regulation�the tax system changedthe fastest. Until the late 19th century tari�s and excise taxes werethe main source of revenue. By the 1930s governments reliedheavily on progressive income taxes to fund their (much larger)spending. Britain’s tax take in 1860 was some 8% of GDP; by 1927it had risen to almost 20%. America changed its constitution tointroduce an income tax in 1913. In 1944 the top rate reached apeak of 94%.

Punitive rates of taxation did not, by themselves, transformthe income distribution. Many fortunes in the early 20th centurywere destroyed by wars, hyperin�ation and the Depression;France, for instance, lost a third of its capital stock in the �rstworld war and two-thirds in the second. But high tax rates madeit much harder for fortunes to be built up again. In most countriesthe share of the top 1% fell persistently from the 1920s until thelate 1970s.

ECONOMIC INEQUALITY CAN be measured inmany ways�by the distribution of wealth,income or consumption, or between races,sexes, regions or individuals. The resultingpicture can vary a lot. In America, for in-stance, the income gap between blacks andwhites, and men and women, has narrowedover the past 30 years, even as that be-tween individuals has widened. Disparitiesin consumption are always smaller thanthose in income because people save andborrow to smooth their living standards.The distribution of wealth is usually lessequal than that of annual incomes. Gaps inpre-tax income are larger than those indisposable income after taxes and govern-ment transfers.

The main measures of economicinequality used in this special report are theGini coe�cients for disposable income andconsumption derived from householdsurveys. These surveys are now conductedin almost all countries. In the rich worldand in Latin America, o�cial Gini coe�-cients are usually based on income. In Asiaand Africa consumption-based �gures aremore common.

Cross-country comparisons can betricky. Inequality in India, for instance, isoften said to be lower than in China. ButChina’s Gini coe�cient of 0.48 measuresinequality of income, whereas India’so�cial Gini of 0.33 measures consumption.

Peter Lanjouw and Rinku Murgai of theWorld Bank calculated an income Gini forIndia which, at 0.54, is much higher thanChina’s and close to Brazil’s.

Another problem is that there areseveral international databases, all slightlydi�erent. Nor are household surveys goodat capturing inequality at the very top, notleast because it is all but impossible to getthe ultra-rich to take part in them. The bestinformation on the highest incomes comesfrom tax returns, thanks to work pioneeredby two French economists, Emmanuel Saezand Thomas Piketty, together with a Briton,Anthony Atkinson, and an Argentine,Facundo Alvaredo. These four have built ahuge database of top incomes which nowincludes 26 countries. Their statistics goback much further than household surveys(in America’s case, to 1913).

Gini coe�cients and the top incomeshare can paint di�erent pictures. Argenti-na’s Gini, for instance, has fallen sharplyover the past decade even as the share ofincome going to the top 1% has risen.Germany’s Gini has risen by 32% since theearly 1980s, but the share of income goingto the very top has barely budged. Onereason is that the statistics cover di�erentpeople; another is arithmetic. The Giniaggregates all disparities, so it is a bettersummary measure, but it does not tell youwhere the gaps are growing.

Like a piece of string

Sizing the gap


SPECIAL REPORT

WORLD ECONOMY

2

1

Taxes rose across the advanced world, but the ways thatgovernments spent them varied greatly. In America, whose gov-ernment was more interested in equality of opportunity than ofincome, the most transformative shift was to bring in mass edu-cation. Starting around 1910, America made huge investments inpublic high schools in pursuit of universal secondary education.After the second world war the GI bill o�ered all returning sol-diers the chance of higher education.

Claudia Goldin and Larry Katz, two economists at Harvard,see this dramatic boost to education as the main cause of the nar-rowing of inequality in America in the mid-20th century. It alsoboosted social mobility. Daniel Aaronson and Bhashkar Ma-zumder of the Federal Reserve Bank of Chicago found that as col-lege enrolment surged in the 1940s, the relationship between par-ents’ and their children’s relative earnings notably weakened.

In Europe the emphasis was on ensuring egalitarian out-comes with big government transfers, particularly after the sec-ond world war. Governments in Europe were slower than inAmerica to invest in mass education, but many continentalcountries built even bigger welfare states than Britain, with gen-erous jobless bene�ts, child subsidies and income support. Invirtually all rich countries other than America such bene�ts(rather than progressive tax systems) became the most impor-

tant instruments for reducing inequality. The third leg of the state’s response to inequality was regu-

lation. Roosevelt’s trustbusting weakened America’s robber bar-ons, and other legal changes protected workers’ rights to organ-ise and, especially in Europe, to conclude binding national payagreements. Union power soared and minimum wages en-shrined in law narrowed the gap between workers and manag-ers. Banking, a big source of wealth in the early 20th century, washeavily regulated after the Depression.

The Great Compression

All this meant that for decades incomes at the bottom andin the middle of the distribution grew faster than those at the top.The exact timing and scale di�ered. In America disparities de-clined fastest in the 1930s and 1940s, in Europe after the secondworld war. America’s Gini coe�cient reached a low of around0.3 in the mid-1970s, and Sweden’s hit 0.2 at about the same time.In most advanced economies the gap between rich and poor inthe 1970s was a lot narrower than it had been in the 1920s. Thiswas the era now widely known as the �Great Compression�.

Income gaps between countries, however, continued towiden as the advanced industrial economies pulled ever fartherahead of less developed ones (with a few notable exceptionssuch as post-war Japan and then Taiwan and South Korea). Bythe 1970s average income per person in the ten richest countrieswas around 40 times higher than that in the ten poorest. This di-vergence among countries outweighed the compression withinthem. As a result, the �global Gini�, as measured by Messrs Bour-guignon and Morrisson, rose.

But around 1980 both these trends went into reverse. Glob-ally, poorer countries began to catch up with richer ones, andwithin countries richer people began to pull ahead. The surge inemerging markets began with Deng Xiaoping’s 1978 reforms inChina. By the 2000s the large majority of emerging economieswere growing consistently faster than rich countries, so much sothat global inequality at last started to fall even as the gaps with-in many countries increased.

The coincidence of timing suggests that the reversals are re-lated. The huge changes that have swept the world economysince 1980�globalisation, deregulation, the information-tech-nology revolution and the associated expansion of trade, capital�ows and global supply chains�narrowed income gaps be-tween countries and widened them within them at the sametime. The modern economy’s global reach hugely increased thesize of markets and the rewards to the most successful. New tech-nologies pushed up demand for the brainy and well-educated,boosting the incomes of elite workers. The integration of some1.5 billion emerging-country workers into the global marketeconomy boosted returns to capital, ensuring that the �haves�would have more. It also hit the rich world’s less educated folkwith unaccustomed competition.

Politicians in search of a scapegoat �nd it easier to blameglobalisation than technology for the widening wage gaps inrich countries, and some studies of America’s wage dispersionconclude that around 10-15% of the widening wage gap can be ex-plained by trade. One analysis, by David Autor at MIT and col-leagues, suggests that in manufacturing the impact of trade withChina could be much bigger. But most economists reckon thattechnological change plays a far bigger role. The OECD, in a bigcross-country analysis, concludes that �skill-biased technologi-cal change� is one of the main determinants of the rich world’swage inequality. On average, it �nds, globalisation�as measuredby a country’s trade exposure and �nancial openness�has nosigni�cant impact.

Whatever the exact breakdown, these two factors are in-


WORLD ECONOMY

SPECIAL REPORT

2

IN HORATIO ALGER’S famous story, �RaggedDick�, a plucky boot shiner improves his lotthrough hard work, honesty and learning his�three Rs� (reading, ’riting and ’rithmetic).The marks of his success are a suit, an o�cejob and a new name, �Richard Hunter, Esq�.

These days economists use more soph-isticated gauges. They measure mobility overa lifetime (rags to riches, or the reverse),between generations (how children dorelative to their parents), in absolute terms(whether children are richer or poorer thantheir parents) or in relative ones (whetherchildren are higher or lower on the incomeladder than their parents).

When countries are growing fast thereis a lot of absolute upward economic mobil-ity. In most emerging economies childrenalmost invariably earn more than theirparents. Even in America, despite slowgrowth and widening income gaps, mostpeople do better than the generation abovethem: a recent study by the Pew CharitableTrusts found that 84% of adult Americanshad higher real incomes than their parents.

The more important gauge of a meri-tocracy, however, is relative mobility, partic-ularly between generations. In a society withbroad equality of opportunity, the parents’position on the income ladder should havelittle impact on that of their children. Eco-nomic historians use clever techniques tomeasure this. Gregory Clark at the Universityof California, Davis, and Neil Cummins of CityUniversity of New York, for instance, havetracked families with rare surnames. Lookingat English census records since 1800, theypicked out names such as Bazalgette and

Leschallas and compared them with recordsof students at elite institutions such asOxford and Cambridge universities. Theirresults show that even over 200 years socialmobility has been rather limited. The wealthand social status of people with rare sur-names in 1800 is strongly correlated withthat of their descendants today.

Individual families’ fortunes over timecan now be tracked by statistical surveys.This allows economists to measure howmuch parents’ position has in�uenced theiradult children’s relative income or educa-tion. The resulting coe�cient, the inele-gantly named �inter-generational elasticityof income�, is today’s main measure of socialmobility. The higher the coe�cient, the lessmobility there has been.

This technique shows Scandinaviansocieties to be very mobile. Only around 20%of parents’ relative wealth (or poverty) ispassed on to their kids. China, in contrast, is

fairly immobile: 60% of income di�erencespersist between generations. The big sur-prise is the United States, where parentalincome explains around half of the di�er-ences in adult children’s income, much morethan in Canada, and more than in any Euro-pean country except Italy and Britain.According to this measure, social mobility inAmerica now is lower than in most of Europe.

Another way to measure economicopportunity is to tease out what share ofinequality can be explained by factors overwhich people have no control: race, gender,birthplace, parents’ education and occupa-tion. The smaller that fraction, the greater acountry’s equality of opportunity.

Such an �Inequality of OpportunityIndex� was pioneered by Francisco Ferreiraof the World Bank and now exists for 40countries. At one extreme lies Norway,where only 2% of the�already low�inequal-ity can be explained by accidents of birth. Atthe other extreme, in Brazil a third of thehigh income inequality is due to people’sbackground. America is closer to Brazil thanto Norway (see chart 1).

Economists also gauge equality ofopportunity by measuring disparities inchildren’s access to basic services that willin�uence their prospects, such as educationor running water. The World Bank is devel-oping indices which adjust overall access tosuch services by a measure of the inequalityin that access. South Africa, for instance,has the same overall rate of access to san-itation as Nicaragua. But once you adjust forrace disparities, its �Human OpportunityIndex� for sanitation is much lower.

Like father, not like son

Measuring social mobility

1Give me a chance

Source: World Bank *From period 1996-2010

Inequality of opportunity as % of total inequalityLatest available*

0 5 10 15 20 25 30 35

Brazil

Guatemala

Britain

United States

Germany

Sweden

Norway

creasingly hard to separate. The IT revolution has allowed moregoods and services to be traded across borders, and it has fuelledthe integration of the global capital market. At the same timeemerging economies are now often the source of innovation.Technology accelerates globalisation, and globalisation acceler-ates technological progress.

At the same time technology is undermining some of the20th century’s equalising institutions. Assembly-line manufac-turing, for instance, was conducive to union organisation. That ismuch less true of many of the cognitive jobs of the digital era.Many social transformations are also making inequality worse,particularly the rise of single parenthood and �assortative mat-ing� (the tendency of educated people to marry each other).

Does all this mean that ever widening inequality is inevita-ble? The history of inequality suggests it need not be, and o�erstwo lessons. The �rst is that market and social forces do not oper-ate in a vacuum. For good or ill, the mix of tax reforms, welfareprogrammes and regulatory interventions pursued in the 20th

century combined to reduce inequality. Those policy choicesmatter just as much today. If they did not, changes in income dis-tribution would have been much more uniform across coun-tries. Instead, much like a century ago, sweeping global forceshave been muted, or exacerbated, by government policies andsocial institutions.

The second lesson is that governments can narrow in-equality without large-scale redistribution or an ever growingstate. The 20th century’s most dramatic reductions in incomegaps took place when governments, by and large, were smallerthan they are today. Large, rigid welfare states proved unsustain-able. But there was also a successful progressive prescription forreducing income gaps and boosting mobility by attacking cronycapitalism, investing in the young (especially by broadening ac-cess to education) and creating a safety net for the poorest (partic-ularly through unemployment insurance and pensionschemes). Worryingly, governments in some of the countrieswhere inequality has risen most seem to have forgotten that. 7


SPECIAL REPORT

WORLD ECONOMY

1

THE HAMPTONS, A string of small towns on the southshore of Long Island, have long been a playground for

America’s a�uent. Nowadays the merely rich are being crimpedby the ultra-wealthy. In August it can cost $400,000 to rent a fan-cy house there. The din of helicopters and private jets is omnipre-sent. The �Quiet Skies Coalition�, formed by a group of angryresidents, protests against the noise, particularly of one billion-aire’s military-size Chinook. �You can’t even play tennis,� moansan old-timer who stays near the East Hampton airport. �It’s likethe third world war with GIV and GV jets.�

Thirty years ago, Loudoun County, just outside Washing-ton, DC, in Northern Virginia, was a rural backwater with a richhistory. During the war of 1812 federal documents were kept safethere from the English. Today it is the wealthiest county in Amer-ica. Rolling pastures have given way to technology �rms,swathes of companies that thrive on government contracts andpristine neighbourhoods with large houses. The average house-hold income, at over $130,000, is twice the national level. Thecounty also marks the western tip of the biggest cluster of a�u-ence in the country. Between Loudoun County and north-westWashington, DC, there are over 800,000 people in exclusivepostcodes that are home to the best-educated and wealthiest 5%of the population, dubbed �superzips� by Charles Murray, a lib-ertarian social scientist.

The Hamptons and Washington’s chic suburbs o�er twosnapshots of the most striking characteristic of American in-equality: the surge in wealth at the top. Washington’s superzipsare full of the rich: people in the top 5% of the income distribu-tion (which means an annual income of at least $150,000) andthe top 1% (those earning more than $340,000 a year). The heli-copter passengers in the Hamptons epitomise America’s ultra-wealthy, the 0.1% of the population whose annual household in-come is at least $1.5m, and the top 0.01%, with an annual incomeof $8m or more. Over the past 30 years incomes have soaredboth among the wealthy and the ultra-wealthy. The higher upthe income ladder, the bigger the rise has been. The result hasbeen a huge, and widening, gap��nancially, socially and geo-graphically�between America’s elite and the rest of the country.

How this happened is a story in three acts. During the 1980sthe least-educated Americans fell behind those in the middle. As

the computer revolution increased the demand for skilled work-ers and old manufacturing industries crumbled, those with justa high-school degree or less saw their relative earnings sink.Over the past decade the squeeze moved to the middle of the in-come distribution, to those who attended college but did notearn a degree. Incomes at the top, meanwhile, rose smartly dur-ing the whole period.

The result was a dramatic divergence in fortunes. Between1979 and 2007 (just before the �nancial crisis) the real disposableincome after taxes and transfers of the top 1% of Americans morethan quadrupled, a cumulative rise of over 300%. Over the sameperiod the bottom �fth’s income rose by only 40%. The middleclass shrank, both as a share of the population and geographi-cally. Only 40% of American neighbourhoods now have an aver-age income within 20% of the national median, compared with60% in the 1970s.

The recession temporarily upended this trend. America’swealthiest fared poorly in 2008 and 2009, largely because thetanking stockmarket ravaged their bonuses and share options.The government safety net prevented a collapse at the bottom.But the sluggish recovery has brought back the old pattern. Morethan 90% of all income gains since the recession ended havegone to the top 1%.

What lies behind these widening gaps? A big reason, partic-ularly in the bottom half, is education, or rather the lack of it. Justas the information-technology revolution demanded moreskilled workers, the continuous improvement in Americans’education stalled. High-school graduation rates stopped climb-ing in the 1970s, for the �rst time since 1890. College completionrates also slowed. Many Americans were failing to match thenew technologies with better skills. According to Harvard’s MsGoldin and Mr Katz, this explains 60% of America’s wideningwage inequality between 1973 and 2005.

College and/or bust

Both the soaring cost of a college education and the short-comings of America’s schools system played a part. In the 1970s ayear’s tuition at a public university cost 4% of a typical house-hold’s annual income; at a private university it took about 20%.By 2009 tuition fees had jumped to over 10% of median incomefor a public university and around 45% for a private one. Evenwith the surge in subsidised student loans, many potential grad-uates were priced out or dropped out early without a degree.

In primary and secondary schools the problems are partly�nancial but mainly organisational. America spends a lot on itsschools, but that funding comes largely from state and local gov-ernments. Richer neighbourhoods can a�ord better schools,which reinforces the growing geographical gap between di�er-ent social groups. According to the OECD, America is one of onlythree advanced countries which spends less on the education of

The United States

The rich and the rest

American inequality is a tale of two countries


WORLD ECONOMY

SPECIAL REPORT

2

1

AMERICANS ARE ENGAGED in a furiousargument about redistribution. In nowinfamous comments at a fund-raiser in May,Mitt Romney, the Republican presidentialcandidate, wrote o� the 47% of Americanswho pay no income tax as people whoconsider themselves �victims�, entitled togovernment handouts. Conservatives liketo point out that 40% of all income taxescome from 1% of taxpayers. America’sgovernment, they argue, redistributes fartoo much from a shrinking pool of �makers�to a vast number of loafers. Those on theleft peddle the opposite view: that thegovernment redistributes far too littlebecause the tax system is skewed to bene�tthe rich and America’s welfare state is theskimpiest in the developed world.

Both sides are wrong. Because Ameri-ca relies mainly on (progressive) incometaxes, whereas other rich countries raise abigger share of their revenue from (regres-sive) consumption taxes, its tax system is,in fact, one of the most progressive in therich world. But it is riddled with deductionsand loopholes, most of which favour thewealthy, so it is both less progressive andmuch less e�cient than it could be. On thespending side of the budget, Americaallocates far less than other rich countriesto cash transfers, such as unemploymentinsurance or income support. But it doesspend a large and growing share of itsbudget on social services, particularlyhealth care for the poor and the old. Theresult is a welfare state that is skewedrather than skimpy. America’s governmentraises revenues ine�ciently and redistrib-utes them oddly: too much from young toold, too much in the form of health care,and ever less from rich to poor.

Start with taxes. It is true that in 2011only 54% of Americans paid federal incometaxes. That is partly because, since Ronald

Reagan, the government’s main form ofassistance for the working poor has beenthe Earned Income Tax Credit, a kind ofnegative income tax. It is also the conse-quence of America’s high jobless rate andtemporary tax credits to boost the econ-omy. In more normal times around 40% ofAmericans pay no federal income tax. Butmore than 60% of those who don’t payfederal income taxes do pay payroll taxes.And if you include taxes raised at state andlocal level, such as on property and sales,virtually all Americans pay some tax.

One study suggests that when youconsider all taxes, the share paid by thewealthiest 1% falls to 21.6%, close to theirshare of pre-tax income, whereas the poor-est quintile pay 2.1%, not much below theirshare of pre-tax income. In other words,America’s Byzantine tax system does hardlyanything to redistribute income. It looksprogressive relative to other rich countriesonly because consumption taxes elsewhereare regressive.

America’s system could be moreprogressive and much more e�cient if itspoliticians were less wedded to �tax prefer-ences�. These exemptions, which includeinterest paid on mortgages up to $1m andcontributions to gold-plated health insur-ance, are now worth some $1.3 trillion, or8% of GDP. Most are hoovered up by thewealthy and the upper middle class. Morethan 60% of all tax preferences �ow to thewealthiest 20% of Americans, with only 3%going to the bottom quintile. Successfulprofessionals do not see themselves asbene�ciaries of government largesse, butthe government in e�ect subsidises theirbig houses, expensive health care andretirement savings.

If America’s tax system represents amissed opportunity to redistribute incomewhile improving e�ciency, it is its spending

system that makes its overall policy far lessprogressive than that of other rich coun-tries. Its cash transfers are stingy. For allthe conservatives’ insinuations of loafersliving on handouts, America spends lessthan half as much as the average OECD

country on cash transfers for people ofworking age. At the same time bene�ts inkind, such as state provision of education,health care and housing, gobble up a largeand growing share of America’s budget. Butaccording to the Centre on Budget andPolicy Priorities, over half of all entitlementspending �ows to the elderly and around40% is spent on health care. The poor donot get much of a look-in. Around 10% ofthe total goes to the richest �fth of Ameri-cans, almost 60% to the middle three-�fthsand only 30% to the poorest �fth.

If you combine tax expenditures andentitlements, America’s e�orts at redistri-bution look even more perverse. The gov-ernment lavishes more dollars overall onthe top �fth of the income distribution thanthe bottom �fth. As Irwin Gar�nkel, LeeRainwater and Timothy Smeeding point outin �Wealth and Welfare States�, a bookcomparing America’s safety net with thoseof other countries, the federal government�spends� four times as much on subsidisinghousing for the richest 20% of Americans(via the mortgage-interest deduction) thanit spends on public housing for the poorest�fth. It also short-changes the young byspending far less on pre-school educationand far more on old people’s health carethan other rich countries.

Some of this is unavoidable. Govern-ments are bound to spend more on the oldas societies age. But America takes this toextremes, propelled by an ine�cient taxcode and the rapid rise in health-care costs.The combination is �scally unsustainable,bad for growth and not very equitable.

Makers and takers

America’s government redistributes, but not well

poorer children than richer ones. And unlike most OECD coun-tries, America does not put better teachers in poorly performingschools, where teachers’ unions often obstruct reform e�orts.

Not everything can be pinned on schooling. Americanwomen (like women almost everywhere) are better educatedand earn more than they did 30 years ago. It is less skilled menwho have fallen behind. Almost uniquely among rich countries,American men now aged between 25 and 34 are less likely thantheir fathers to have a college degree. The damage from this hasbeen compounded by institutional changes, such as the weaken-ing of unions and, particularly, the erosion of the minimumwage. But the main culprit is educational slippage.

This poor performance has a racial tinge. High-school drop-outs are disproportionately black or Hispanic. America’s habitof locking up large numbers of young black men does not helptheir employability. But the decline increasingly a�ects the whiteworking class too. Ever more low-skilled white American menhave left the labour force, many moving onto disability rolls.Even before the recession, only around two-thirds of white menwith nothing more than a high-school diploma were working.

This decline of work among less skilled white men has hadprofound social consequences, which in turn have exacerbatedincome inequality. Marriage rates have fallen, divorce has in-creased and the share of children born to single mothers has


WORLD ECONOMY

SPECIAL REPORT

2

1

soared. Mr Murray calculates that fewer than 30% of children inthe poorest third of white America live with both parents by thetime their mother turns 40. Among the most a�uent �fth,around 90% of children live in a household with both parents.Marriage has become a fault-line dividing American classes.

Tax and bene�t changes have also had an e�ect, but a sub-tle one. Most Americans below the median income level pay nofederal income tax (and, thanks to the Earned Income Tax Credit,the working poor get substantial rebates). Poorer Americans arehit disproportionately by payroll taxes, which are regressive andhave grown in importance. But the biggest hit is on the bene�tside. Although America’s social spending has rocketed (it is nowworth some 16% of GDP), it is becoming less redistributive asMedicare, the universal health plan for the elderly, swallows upever more (see box, previous page). According to the Congressio-nal Budget O�ce, in 1979 over half of all federal social spendingwent to the poorest �fth of households. Now it is only 36%.

Part of the trend at the top of America’s income ladder issimply the mirror image of that at the bot-tom. The rising skill premium has reward-ed those with lots of education, and socialshifts have reinforced the income concen-tration. Not only are the well-o� and well-educated far more likely to marry and staymarried than poorer folk, they tend tomarry each other. In 1960 American cou-ples with two college-educated partners accounted for only 3%of the total. Today that �gure is 25% and in the top 5% of the in-come distribution it is 75%. Apart from the cleaning lady, it is hardto �nd an adult without a degree (or two or three) in super-ziphouseholds.

But if educational di�erentials and assortative mating liebehind much of the gap between those in Loudoun County andpoorer Americans, they do not explain the Hamptons phenome-non. America’s top 0.1% are no better educated than the top 1%.Income gaps at this level have less to do with the skills-bias of themodern economy and more to do with its global reach.

In a classic paper published in 1981, the late Sherwin Rosenof the University of Chicago pointed out that the very best in a�eld, be they entertainers or textbook authors, earned vastlymore than the next best. Modern communications, he mused,would transform the market for superstars. And so they have, asChrystia Freeland, a journalist, points out in her new book, �Plu-tocrats�. Moreover, in the past three decades the potential markethas become dramatically bigger, whether for Hollywood block-busters or celebrity dentists.

Celebrities do not account for a large share of America’s ul-tra-rich. But the same factors�winner-takes-all economics cou-

pled with an incomparably bigger global economy�explain partof the rise in the incomes of the chief executives who make up abigger share of the very wealthy. During the 1980s CEO paysurged more in America than anywhere else. Until the early1980s American chief executives, on average, earned 40% morethan their next two lieutenants. By 2000 they earned two-and-a-half times as much.

This rise is widely put down to failures of corporate gover-nance and a collapse in social norms which once set an informallimit on the earnings gap between bosses and workers. There istruth to both explanations, and it is not hard to �nd chief execu-tives earning tens of millions of dollars despite lacklustre perfor-mance. But these e�ects should not be exaggerated. In a recentpaper Steven Kaplan, of the University of Chicago, �nds thatCEO pay has fallen in recent years and that, contrary to myth,CEOs who perform badly get paid less and are �red more oftenthan successful ones.

There is also a less bothersome explanation for CEO paythat is based on superstar economics. America is home to a lot ofthe world’s biggest companies, and globalisation has mademany of them a lot bigger. In a global market for the best CEO tal-ent where winner-takes-all economics prevails, the gap betweenthe top and the rest is bound to be vast.

For all the attention focused on CEO pay, the numbers ofchief executives among America’s ultra-rich are neither particu-larly big or growing. The very richestAmericans�those who feature in theForbes list of billionaires�tend to be entre-preneurs, from the icons of the tech era(Bill Gates, Mark Zuckerberg) to manywhose money has more prosaic roots(Sara Blakely, America’s youngest femalebillionaire, made her fortune from wom-en’s underwear).

A disproportionate, and growing,

chunk of the very rich, however, have made their money in WallStreet rather than Main Street. An analysis by Mr Kaplan andJoshua Rauh, now of Stanford University, shows that the shareof investment bankers among the top 0.1% is larger than theshare of senior executives. America’s top 25 hedge-fund manag-ers make more than all the CEOs of the S&P 500 combined. The�nancial industry’s outsize pay partly reects its growth. Forgood or ill, �nance’s share of American GDP soared between1980 and 2007. Capital markets have globalised faster and morecomprehensively than any other part of the economy, enablinghedge funds and other asset managers to deploy ever biggerpools of funds. According to Thomas Philippon of New YorkUniversity and Ariell Reshef of the University of Virginia, �nan-ciers also have higher skill levels than they did a generation ago.

These fundamental economic shifts explain part of the risein Wall Street incomes, but not all of it. Messrs Philippon andReshef argue that between a third and half of Wall Street’s higherpay is unjusti�ed, deriving from rents rather than productivity.But what explains these rents? Luigi Zingales of the University ofChicago points out that one source is the implicit subsidy(through lower borrowing costs) that banks enjoy by being toobig to fail. He reckons this subsidy is worth some $30 billion a

In a global market for the best CEO talent wherewinner-takes-all economics prevails, the gap betweenthe top and the rest is bound to be vast

2The rich get richer

Source: Congressional Budget Office

US, average real income after taxes and transfers% change since 1979

50

0

50

100

150

200

250

300

350

+

–

1979 85 90 95 2000 05 09

Top 1%

Bottom 20%

ar7kf

Highlight


SPECIAL REPORT

WORLD ECONOMY

2

1

year, enough to fund a fair few bonuses. Others point to a broad-er cronyism between Wall Street and Washington over the past30 years which has allowed �nanciers to tilt rules in their favour.The �nance industry (along with property and insurance) em-ploys more lobbyists than virtually any other industry, aroundfour per Congressman.

Financiers have also been among the biggest winners fromchanges to America’s tax code. The country’s top rate of incometax has been repeatedly slashed since 1980, from 70% to 35%. Byitself, that reduction has not greatly a�ected average tax burdensat the top (since there have been enough loopholes to ensure thatfew people paid the top rate). America’s richest have gainedmore from reductions in the capital-gains tax, which is now only15%. Private-equity moguls have done particularly well, since thetax code allows them to classify their income as capital gains.

Scratching each other’s backs

The combination of tax loopholes, bank bail-outs and mas-sive lobbying has led many observers to conclude that America’sgrowing inequality has political roots. The wealthy, in this logic,control the political system and rig it to their advantage. In an in-�uential book, �Unequal Democracy: The Political Economy ofthe New Gilded Age�, Larry Bartels of Vanderbilt Universityshowed that senators’ votes are in�uenced by the preferences oftheir rich citizens but not their poor ones. As money plays anever bigger role in politics, goes the argument, so the clout of theultra-wealthy grows, particularly to block things they don’t like.

This claim is hard to prove, but circumstantial evidence forit seems to be mounting, particularly since the Supreme Court’s2010 �Citizens United� decision lifted any restrictions on politi-cal spending by individuals or �rms. That opened the way forthe rise of �super-PACs�, privately funded organisations set upto in�uence election outcomes. These have now raised hun-dreds of millions of dollars. The sources of this money are highlyconcentrated: one analysis suggests that 80% of the total comesfrom fewer than 200 donors. America is still a long way from the�rst Gilded Age, when the robber barons openly bought unelect-ed senators’ loyalty by giving them shares in their companies.But it is hard to believe that this surge of cash from the richest willhave no impact at all.

Whatever its causes, the strati�cation of American societyis having profound consequences. A country that prides itself onits social mobility is already less mobile than most people thinkand is almost certainly becoming even less so. As the box withthe previous article showed, standard measures of inter-genera-tional mobility in America are lower than in Canada and muchof Europe. Most of this has to do with the di�culty of escapingfrom the bottom rungs of America’s income ladder. According toMarkus Jantti, a Finnish economist who has studied mobilityacross countries, more than 40% of the sons of the poorest 20%of Americans stay in that quintile, compared with around 25% inNordic countries. The evidence is mixed on whether social mo-bility has lessened or simply stayed the same over the past 30years. But it is clear that there has been no improvement in mo-bility to compensate for widening inequality.

And even the most recent studies of social mobility look atthe earnings of people who were children over two decades ago.Since disparities in income, education and social behaviour nowstrongly reinforce each other, future mobility might be a lot low-er still. A study by Sean Reardon of Stanford University suggeststhat the gap in standardised test scores between schoolchildrenfrom high- and low-income families is roughly 30-40% bigger to-day than it was 25 years ago. Bob Putnam, of Harvard University,puts it starkly. Put away the rear-view mirror and look at futuresocial mobility, he says, and �we’re about to go over a cli�.� 7

THE SUMMIT OF Songshan mountain, some 60 miles(100km) from China’s capital, marks the boundary be-

tween Beijing municipality and the neighbouring province ofHebei. It is also a study in contrasts. On the Beijing side themountain road is wide, freshly surfaced and �anked by a solidsafety wall. A Lycra-clad cyclist sweats his way up on a fancymountain bike. A large car park is under construction for visitorsto hot springs in the nearby village of Bangongqu. Enterprisinglocal families can make 100,000 yuan ($16,000) a year catering toBeijing tourists, not far o� the city’s average white-collar wage.The Beijing provincial government provides pensions and othersocial bene�ts.

Hebei is a much poorer province. On its side of the moun-tain the road narrows and the tarmac deteriorates. Half a milefrom the summit is the village of Yanjiaping, where some 50 fam-ilies scrape a living growing cabbages. No one has a car, no onegets a pension, and the nearest primary school is 12 miles away.Farmers are barred from grazing cows on the mountainside so

Asia

Crony tigers, divideddragons

Why Asia, too, is becoming increasingly unequal


WORLD ECONOMY

SPECIAL REPORT

2

1

that trees can grow to stem sand storms from Inner Mongolia.Shen Zhiyun, a gnarled man in fake US army fatigues, says a vil-lage family makes 4,000-5,000 yuan a year, nowhere near Indi-an levels of poverty, but a far cry from the living standards only afew miles away. �We live in a di�erent country,� he says.

The transformation of China’s economy over the past 30years is the most spectacular growth story in history. Less no-ticed, China has also seen the world’s biggest and fastest rise ininequality. China has not o�cially published a Gini coe�cientsince 2000, but a study by the China Development ResearchFoundation suggests that it has surged from less than 0.3 in 1978to more than 0.48. In little more than ageneration Mao’s egalitarian dystopia hasbecome a country with an income distri-bution more skewed than America’s.Asia’s two other giants, India and Indone-sia, have also seen disparities rise sharply,though less dramatically than China. In-donesia’s Gini is up by an eighth, to 0.34.

Part of this rise was both inevitableand welcome, a natural consequence ofthe end of Maoist communism in Chinaand Fabian socialism in India. The threeeconomies, particularly China’s, are farricher and more dynamic than they were30 years ago. Just as Kuznets suggested, ur-banisation and industrialisation havebrought widening gaps. As people haveleft subsistence agriculture for more pro-ductive work in cities, inequality has ris-en along with prosperity.

But that cannot be the whole expla-nation, if only because the experience oftoday’s Asian tigers is in striking contrastto that of an earlier pack. In Japan, HongKong, South Korea and Taiwan growthrates soared in the 1960s and 1970s andprosperity increased rapidly but incomegaps shrank. Japan’s Gini coe�cient fellfrom 0.45 in the early 1960s to 0.34 in 1982;Taiwan’s from 0.5 in 1961 to below 0.3 bythe mid-1970s. That experience launchedthe idea of an �Asian growth model�, onethat combined prosperity with equity.

Education, again

Today’s Asian growth model doesthe opposite. One explanation is that thebig forces driving modern economies�technological innovation and globalisa-tion�bene�t the skilled and educated inemerging markets much as they do in therich world. Narayana Murthy, the billion-aire co-founder of Infosys, an Indian soft-ware giant, or Robin Li, the creator ofBaidu, China’s most popular search en-gine, have harnessed technology muchlike Bill Gates has done. Senior lawyersand bankers in Mumbai or Shanghai arepart of a global winner-takes-all market,able to command salaries similar to thoseof their colleagues in New York or Lon-don. And as Ravi Kanbur of Cornell Uni-versity points out, the o�shoring of tasksthat has hit mid-level workers in America

and Europe often bene�ts people higher up the skills ladder in re-cipient countries. Call centres in Bangalore are manned by well-educated Indians.

As in the rich world, these fundamental economic forcesare not the only drivers of income distribution. Governmentpolicy has also played a big role. One problem is cronyism. As inthe Gilded Age in America, capitalism in today’s emerging mar-kets involves close links between politicians and plutocrats. In-dia is a case in point. From spectrum licences to coal deposits,large assets have been transferred from the state to favoured in-siders in the past few years. Many politicians have business em-

AT FIRST SIGHT Palanpur is a powerfulreminder of the stubborn persistence ofIndia’s rural poverty. The village is notparticularly remote. It is next to a railwayline, only a few miles from a big highway,and less than 120 miles from Delhi. It issurrounded by some of India’s most fertileagricultural land, well suited to the culti-vation of sugar cane, groundnuts andmenthol. Yet Palanpur’s residents arecrowded into sparse dwellings along mudpaths, with no running water, no drainsand only intermittent electricity.

Spend a day in the village, and thepicture becomes more nuanced. You hearhow life has improved. Even the poorestvillagers now have brick rather than mudhouses; only a couple of years ago manywere still made of mud. From marble-polishing to brickmaking, more jobs out-side agriculture are becoming available.The government’s rural employment-guarantee scheme has put a �oor underwages. The roads have got better. Allchildren now attend the village school,when only a few years ago many childrenfrom the lowest castes were not in school.

There are obvious gaps betweenwealthier and poorer folk. MahendraMorya, head of the richest family in thevillage, recently bought a second tractor.Some households now have pit latrines. Acouple even have a television on which towatch DVDs. Many of the most visiblywealthy are members of the upper castes.Mr Morya is a Murao, a high caste of culti-vators. But some further down the socialpecking order seem to be doing well too.Nanhe, the head of a Muslim family, start-ed out repairing bicycles in the 1990s. Nowhe has a menthol-processing facility andplans to branch out into mustard oil.

Most surprising is the success ofRamjimal, a Jatab, the group at the bottomof Palanpur society. He is a skilled bricklay-

er, travelling around neighbouring villagesbuilding houses. One of his brothers hasbecome a lawyer and moved to Chandausi,the nearest town.

A long-running study at the LondonSchool of Economics provides statistics tocon�rm these impressions. Its researchershave spent over 50 years conducting de-tailed surveys to track the fortunes ofPalanpur’s residents. The most recent one,in 2008-09, shows considerable changefrom the previous one in 1983. Real incomeshave doubled (which, over 25 years, trans-lates into modest average annual growth ofjust under 3%), and income disparities havebecome much wider. Palanpur’s Gini coe�-cient in 2009 was 0.4, 30% higher than in1983. But social mobility has increased too,and a disproportionately large number ofthe winners came from the bottom of thesocial heap. Half of the families thatclimbed most were Jatabs.

A study by Viktoria Hnatkovska andcolleagues of the University of BritishColumbia suggests that Palanpur is not anisolated case. It shows that the inter-generational mobility of India’s Dalits (orScheduled Castes, the most disadvantagedgroup) has improved and is now similar tothat of other groups. Ms Hnatkovska’s�ndings remain controversial, but mostIndian academics agree that caste rigiditiesare loosening, mainly thanks to the growthof non-agricultural employment and im-proved access to basic education.

There is still a long way to go. Second-ary-school attendance among the Sched-uled Castes generally remains shockinglylow. And with income gaps widening, thereis a danger that those at the bottom will getstuck there. But in a country where forcenturies the disadvantaged had no chanceof improving their prospects, more socialmobility, even amid wider inequality, is abig step forward.

Lessons from Palanpur

More inequality in an Indian village is balanced by greater mobility


WORLD ECONOMY

SPECIAL REPORT

2

1

pires of one kind or another. Rich businessmen often becomepoliticians, particularly at the state level. Raghuram Rajan, an In-dian-born economist at the University of Chicago who recentlybecame chief economic adviser to India’s government, haspointed out that India has the second-largest number of billion-aires relative to the size of its economy after Russia, mainlythanks to insider access to land, natural resources and govern-ment contracts. He worries that India could be becoming �an un-equal oligarchy or worse�.

In China cronyism is even more ingrained. The state stillhas huge control over resources, whether directly through state-owned enterprises, monopoly control of industries from rail-ways to mining or the distorted �nancial system, where interestrates are arti�cially depressed and access to credit is in�uencedby politics. The importance of the state means that the bene�cia-ries tend to be close to state power.

Moreover, inequality in China could be higher than the o�-cial statistics suggest because rich people often understate theirincome and hide it from the taxman. A lot of money is investedin property, where soaring prices have reinforced inequality.Wang Xiaolu, of the China Reform Foundation, caused a stir acouple of years ago with a study that tried to measure this �grey�income. His results suggest that the income of the richest 10% ofurban Chinese is some 23 times that of the poorest 10%. O�cialstatistics say the multiple is nine.

Cronyism is the most obvious way inwhich Asian governments make inequali-ty worse, but it is not the only one. Broadergovernment strategies have distortedcountries’ growth paths in a manner thatincreased income gaps. In India a big pro-blem is the lack of job creation. UnlikeChina, where the surge in factories assem-bling goods for export brought millions ofmigrant workers into the formal urban la-bour force, India’s formal workforce hasbarely grown since 1991. More than 90% ofIndians are still employed in the informalsector. Even in manufacturing, most peo-ple toil in one-room workshops ratherthan big factories. Productivity is lower,workers �nd it hard to improve their skillsand their incomes rise more slowly.

India’s failure to become a powerhouse of labour-intensivemanufacturing owes much to its appalling infrastructure. Just-in-time delivery is hard to achieve when power supplies are so pre-carious. Another reason is the country’s rigid labour laws, whichdiscourage the formation of big �rms. Between the federal gov-ernment and the states, India has around 200 di�erent laws, allsetting detailed rules and making it virtually impossible to �repeople. That deters employers from hiring workers and widensthe gap between the lucky educated few and the rest.

We know where you live

In China the regulations that contribute most to inequalityare the remnants of the country’s hukou system of householdregistration. This hails from Mao’s era, when China’s rural sectorwas punitively taxed to �nance the development of heavy in-dustry. To ensure a stable supply of workers in agriculture de-spite the appalling conditions, people were barred from leavingtheir province of origin. The restrictions on mobility were dis-mantled in the 1980s, permitting millions to become migrantworkers. But they still retain the rural hukou of their birth, as dotheir children. From housing to schooling, this puts them at a bigdisadvantage compared with holders of urban hukou.

Migrants’ children must take the gaokao (the all-importantstate college-entrance exam) in their place of origin, not wherethey and their parents might be living at the time, so lots of mi-grants send their children home for schooling. Since education is�nanced largely by local governments, these schools tend to beless well-funded and of lower quality. Hebei has far worseschools than Beijing. In Shanghai municipality, spending per stu-dent in rural areas is only 50-60% that of urban areas. As a result,the education system reinforces income disparities rather thanmitigating them.

Along with disparities in infrastructure, the hukou systemis a big reason for China’s vast urban-rural gaps, which explainabout 45% of the country’s overall inequality. Other Asian econ-omies do not su�er from a hukou problem, but there, too, govern-ment social policies have often made inequality worse becausemost social spending, from public housing to health insurance,has traditionally been con�ned to the formal, urban workforce.Moreover, many Asian governments spend a lot on universalsubsidies, especially for energy. These are highly regressive. In-donesia, for instance, lavished 3.4% of GDP on fuel and electric-ity subsidies last year, more than it spent on infrastructure. Ac-cording to the Asian Development Bank, 40% of that largesse�owed to the richest 10% of Indonesian households and as muchas 84% to the top half.

Things are beginning to change. Across emerging Asia polit-ical concerns about rising inequality are prompting reform, of-ten in ways that echo the changes of the Progressive Era a cen-tury ago. In China the �Great Western Development Strategy�has poured vast sums into infrastructure in the western prov-inces. More recently the government has made a big e�ort to im-prove rural social services. Almost 100% of China’s rural popula-tion now have basic health insurance (including the villagers ofYianjiaping), and a majority have basic pensions. Inequality be-tween urban and rural areas has recently stabilised and that be-tween regions has begun to fall slightly, but from an extraordi-narily high level.

In the past couple of years several Asian economies, fromThailand to Vietnam, have introduced, or expanded the reach of,minimum wages. China’s minimum wage, which is set at theprovincial level, rose by an average of 17% last year. Some coun-tries have introduced public-work schemes for the poorest. In-dia’s NREGA scheme, for instance, guarantees 100 days’ work ayear to the country’s rural households and now covers 41m peo-ple. Others have experimented with targeted subsidies to thevery poorest that have helped reduce inequality in Latin Ameri-ca (see next article).

AcrossemergingAsiapoliticalconcernsabout risinginequalityarepromptingreform


WORLD ECONOMY

SPECIAL REPORT

2

1

By introducing a more e�cient, and progressive, social safe-ty net, Asia’s governments will go some way towards mitigatingtheir growing income gaps. But there will be no big break-throughs until the bigger problems of informality (in India), dis-crimination against migrants (China) and cronyism (every-where) are dealt with. And the longer that takes, the greater thedanger that today’s disparities will become entrenched.

Thanks to remarkable economic growth, almost all Asiansare rapidly becoming better o�. In India, old caste rigidities arebeing broken down (see box earlier in this article). But wideningincome gaps threaten to harm future social mobility. Using amethodology developed at the World Bank, a study by ZhangYingqiang and Tor Eriksson found that the rise in China’s incomeinequality is mirrored by a rise in its inequality of opportunity.Parents’ income and their type of employer explain about two-thirds of China’s inequality of opportunity, a much bigger sharethan is explained by parental education.

The stakes are high. Yu Jiantuo of the China DevelopmentResearch Foundation argues that China’s inequality is now hurt-ing its growth prospects. Sustained cronyism could turn Asia’sbig economies into entrenched oligarchies rather than dynamicmeritocracies. Ironically, in that sense they might become morelike Latin America just as that continent appears to be moving inthe opposite direction. 7

MICHAEL JACKSON BROUGHT Santa Marta a moment offame. In February 1996 the King of Pop landed by helicopter

at the top of one of Rio de Janeiro’s most notorious favelas. Poli-ticians tried to stop him, but Mr Jackson had permission from thedrug barons who ruled the slum. He danced down the steeppaths between shacks clinging precariously to the mountain-side, surrounded by a cheering crowd of Rio’s poorest citizens,and belted out his hit single �They don’t care about us�. The mu-sic video was played around the world. It trained a spotlight onRio’s poverty and inequality.

Sixteen years later Santa Marta is once again a showcase,but of a better sort. It was the �rst favela to be �paci�ed� under agovernment plan to wrest control of Rio’s slums from the druglords. The place was stormed by the army in 2008. It now has apolice station, and is peaceful. It is a thriving example of theboom at the bottom of Brazilian society.

Meet Salete Martins, a bubbly 42-year-old, whose familymoved to Santa Marta from Brazil’s north-east when she waseight. By day she works as a trainee tour guide, showing visitorsaround her neighbourhood for a city-�nanced non-pro�t groupcalled Rio Top Tours. At night she studies tourism at a local col-lege. At weekends she sells Bahian food from a bustling stall nearthe favela’s entrance. And in between she �ogs a popular line ofbeauty products. Her monthly income is around 2,000 reais($985), four times as much as she made selling sandwiches threeyears ago and more than three times the minimum wage. Sheplans to launch her own tour-guide company before the end ofthis year.

Ms Martins’s success is striking, even in Santa Marta. But it

mirrors a trend that has swept the wholeof Latin America. Poor people’s incomeshave surged over the past decade, leadingto a big drop in inequality. In most LatinAmerican countries the Gini coe�cient in2010 was lower than in 2000. The region’saverage, at 0.5, is down from almost 0.54 adecade ago, and lower than at any time inthe past 30 years (see chart 3), though stillhigh relative to other regions. Judging byevidence from Argentina, the only coun-try in Latin America to publish statisticson tax returns of top earners, the richest 1%are still pulling ahead of the rest. But thatconcentration is more than made up for bythe narrowing of gaps further down theincome scale.

Both shifts are re�ected in popularculture. �Mulheres ricas� (�Rich women�)is a new reality-TV show about Brazil’s ul-tra-wealthy (�I bathe in mineral water ev-ery day,� said one woman in an early epi-sode). But the country’s most popularprime-time soap is �Avenida Brasil�, which documents lifeamong the newly minted middle classes. Although Latin Ameri-ca saw only half the average GDP growth of emerging Asia overthe past ten years, its poverty rate fell by 30%. Around a third ofthe decline is due to improvements in income distribution.

How did a continent that had been egregiously unequalsince the conquistadores’ land grab suddenly change course? Notbecause of radical nationalisation and redistribution. LatinAmerica has a few asset-seizing hard-left governments, notablyArgentina and Venezuela, but inequality has also fallen in coun-tries following a more orthodox economic course, such as Chileand Colombia. Nor is the turnaround just a side-e�ect of thecommodities boom. Inequality has fallen in countries that relyheavily on exports of commodities, such as Peru, but also inthose where manufacturing plays a bigger role, such as Mexico.Nor can demography be the main cause. Poorer Latin Americanfamilies have become smaller, which reduces inequality, butthese changes were well under way in the 1980s and 1990s.

According to Nora Lustig, an economist at the University ofTulane and one of the �rst to document the narrowing of the re-gion’s income gaps, two things have made a big di�erence. First,the premium for skilled workers has been falling: a surge in sec-ondary education has increased the supply of literate, reason-ably well-schooled workers, and years of steady growth haveraised relative demand for the less skilled in the formal work-force, whether as construction workers or cleaners. Second, gov-

ernments around Latin Ameri-ca have reinforced the narrow-ing of wage gaps with socialspending targeted at peoplewith the lowest incomes. Theseinclude more generous pen-sions and conditional cashtransfers�schemes that o�erpayment to the poorest familiesin return for meeting speci�cconditions, such as making suretheir children go to school.

The most striking changehas been in education. In thepast Latin American govern-ments lavished cash on univer-

Latin America

Gini back in the bottle

An unequal continent is becoming less so

3That’s better

Sources: Gasparini et al(2012) based on SEDLAC(CEDLAS and the World Bank)

*0=perfectequality,

1=perfectinequality

Latin America’s income inequality Gini coefficient*

0.490.500.510.520.530.54

1980 85 90 95 2000 05 10


SPECIAL REPORT

WORLD ECONOMY

2 sities. State primary and secondary schools were underfundedand of appalling quality. That bias in favour of tertiary educa-tion, perversely, most bene�ted the children of the rich, who hadattended private primary and secondary schools. But since theearly 1990s education spending has become much more progres-sive, with a huge expansion in public secondary educationamong the poor. According to Karla Breceda, Jamele Rigolini andJaime Saavedra, three economists at the World Bank, LatinAmerican governments, on average, now spend a larger share ofGDP on education for the poorest 20% of children than does theUnited States.

More progressive spending has produced results. Somecountries have seen an increase of 20 percentage points in theshare of children �nishing secondary school. Another study forthe World Institute for Development Economics Research in Hel-sinki by Guillermo Cruces, Carolina García Domench and Leo-nardo Gasparini showed that the gap between rich and poor insecondary-school enrolment has fallen in all countries except ElSalvador, Honduras, Guatemala and Nicaragua.

Many Latin countries are also championing pre-schooleducation. Rio’s city government, for instance, has dramaticallyincreased its network of nursery schools since 2009, building 74new ones in the past three years. Any child from a family belowthe poverty line is guaranteed a free place in a nursery from theage of six months.

A nudge in the right direction

Conditional cash transfers (CCTs) reinforce this focus onschooling. These stipends cost relatively little (typically 0.2-0.8%of GDP) but in�uence the priorities of many. About a quarter ofBrazil’s population now gets some money from Bolsa Família,the country’s CCT scheme. State and local governments piggy-back on top. In Rio, for instance, the city supplements Bolsa Fa-mília payments for 700,000 of its poorer families. If children doexceptionally well in exams, a bonus is paid. If they miss school,the payment stops. Ms Martins realised her 14-year-old was skip-ping school only when her monthly stipend was docked. Severalacademic studies in Mexico show that kids in CCT schemes stayat school longer.

Better education is boosting social mobility. Historically,the link between parents’ and children’s education has beencloser in Latin America than anywhere else. In Peru, for instance,

almost 70% of a child’s educationalachievement can be predicted from its fa-ther’s schooling. But a forthcoming reportfrom the World Bank suggests that the cur-rent generation of Latin American chil-dren are both better educated than theirparents and moving relatively faster upthe education ladder. And, like India’spoorest castes, disadvantaged indigenouspeople have made big gains.

These newly educated workers en-joy far better prospects in the formalworkforce than their parents did. Statepensions have become more generous.Countries from Argentina to Bolivia haveintroduced non-contributory pensionschemes�in e�ect, a promise of govern-ment support for the elderly. Minimumwages across the continent have soared.Brazil’s has risen by more than 50% in realterms since 2003. And since pension bene-�ts are linked to the minimum wage, thetwo trends reinforce each other.

The precise contribution of better education, better oppor-tunities for less skilled workers and bigger social spending dif-fers by country. An analysis by Ms Lustig, Luis López-Calva of theWorld Bank and Eduardo Ortiz-Juarez of the United Nations De-velopment Programme suggests that narrower wage gaps ex-plain most of the reduction in inequality throughout the region.According to calculations by Marcelo Neri, of the Institute for Ap-plied Economic Research, government transfers explain aboutone-third of the drop in inequality in Brazil.

So far, so good. But will these gains last? In education, thebig challenge is to complement quantity with quality. LatinAmerica has now reaped the bene�ts that come from simply get-ting more children into school for longer. But most of the stateschools are still much less good than their private equivalents.Virtually all middle- and upper-class children still go to privateprimary and secondary schools. Until those gaps in quality havebeen eliminated, educational inequities will persist. They are be-hind the recent wave of protests over education in Chile.

The more immediate challenge is how to pay for all this.Latin American states have traditionally not been progressive inoutlook. Put crudely, governments raised revenue from the morea�uent, then spent it on generous public pensions for thosesame people. Even now, 60% of transfer spending in Bolivia, forexample, goes to people who are not poor. Mr Saavedra calls it a�fragmentary social contract�. Governments fail to provide goodpublic services, and middle-class people rely on private educa-tion and health care. But they do get generous pensions in returnfor their taxes.

The long boom of the 2000s allowed a painless change tothis social contract. Sustained growth brought in enough tax rev-enue to boost both education spending and transfers at the bot-tom without pushing up tax rates. The boom also allowed hugeincreases in minimum wages without apparent damage to em-ployment. But as growth slows and the real value of minimumwages rises, that combination is becoming unviable.

If the improvements in inequality are to be maintained, letalone continued, tough choices will have to be made. Middle-class entitlements will need to be squeezed. Much like the Un-ited States, many Latin countries will have to decide whether toinvest in poorer kids or continue to pay generous pensions toricher old people. In both places the social contract needs to beremade. For evidence that this is possible, turn to Sweden. 7


WORLD ECONOMY

SPECIAL REPORT

1

SALTSJÖBADEN, A CHARMING seaside town on the out-skirts of Stockholm, has an iconic place in Swedish eco-

nomic history. The �Saltsjöbaden Accord�, signed there betweenunions and employers in 1938, ushered in the consensus systemof labour relations that remains a pillar of Sweden’s economicmodel. Nowadays the town is famous for a di�erent reason. It isone of Stockholm’s fanciest suburbs, and the setting for �SunnySide�, a popular television comedy thatpokes fun at the country’s new rich. In theshow, Saltsjöbaden’s yuppy residents fretover how to get their babies into the bestnursery. A badly behaved child is threat-ened with banishment to Fisksätra, a poorenclave a few train stops away, where im-migrants from 100 countries cram into di-lapidated blocks of �ats.

The most equal country in the worldis becoming less so. Sweden’s Gini coe�cient for disposable in-come is now 0.24, still a lot lower than the rich-world average of0.31but around 25% higher than it was a generation ago. That riseis causing considerable angst in a nation whose self-image isstaunchly egalitarian. A leftist group caused a media hubbubearlier this year by organising a �class safari� bus tour of Saltsjö-baden and Fisksätra. Opposition leaders insist that the rulingcentre-right party is turning Sweden into America.

Anders Borg, the �nance minister, vehemently disagrees.Sweden, he argues, has gone from being a stagnant bene�t-based society to a vibrant modern economy with a remarkablysmall rise in inequality. Its experience, he says, shows that dyna-mism and egalitarianism do not need to be at odds.

The facts bear him out. Thanks to deregulation, budget dis-cipline and an extensive overhaul of the welfare state, Sweden’s

economy has been transformed in the two decades since itsbanking crisis. The new Swedish model is quite di�erent fromthe leftist stereotype.

Capitalism in Sweden is not inherently a lot more egalitari-an than in other countries. Before the government steps in, thecountry’s Gini coe�cient for the working-age population is 0.37,close to the OECD average and higher than Switzerland’s. Wagedisparities are narrower than in Anglo-Saxon countries, thanksto centralised bargaining between unions and employers thatsets minimum wages in di�erent sectors. Top CEO pay has notrisen nearly as dramatically as in America. But in other waysSweden has been in the vanguard of many of the social changesthat have boosted inequality in other countries, such as the de-cline of marriage.

The main source of egalitarianism in Sweden (and else-where in Scandinavia) is redistribution by the state. Under theold welfare model people paid high tax rates and got lots of so-

cial services and big transfers in return. The new model, broadly,retains most of the services but has cut the taxes and transfers.

In the early 1990s Sweden introduced a �dual income tax�system, which combined a �at tax on capital with a higher, pro-gressive income tax. More recent reforms went further. The in-heritance tax was eliminated in 2005, the wealth tax in 2007 andtaxes on residential property in 2008. Thanks, in part, to thesetax changes, capital income has soared, particularly at the top ofSweden’s income scale. That has not always improved e�cien-cy. For instance, Sweden’s tax code now favours residential prop-erty over more productive investment.

More recently, and more sensibly, Sweden has cut taxes onlabour, especially for the low-skilled. The Earned Income TaxCredit, which o�ers strong incentives for lower-skilled people towork, marks the biggest change. Other innovations, such as a

credit for hiring household help, are de-signed to spur demand for low-wageworkers. Union membership dues, incontrast, no longer qualify for tax relief.Bene�ts have been reformed at the sametime as taxes. All handouts, from joblessaid to disability bene�ts, have becomeless generous, more short-lived and hard-er to qualify for.

All this has brought about palpablechanges. Notice boards at Stockholm’ssuburban railway stations are �lled withadvertisements for cleaners, once an un-heard-of luxury. The Iraqis, Somalis andother low-skilled foreigners in Fisksätra,unlike migrants a generation ago, can nolonger count on a drip-feed of govern-ment support. The combination of lowertaxes and fewer bene�ts is intended to en-courage people to work. And getting moreof them to take jobs, argues Mr Borg, is thekey not only to faster growth but also tokeeping inequality low. His ministry reck-

Sweden

The new model

A bit more unequal, a lot more e�cient

Thanks to deregulation, budget discipline and anextensive overhaul of the welfare state, Sweden’seconomy has been transformed in the two decades sinceits banking crisis

1

ons that in the long term Sweden’s reforms will raise the coun-try’s employment rate by 5%.

Critics on the left fear that inequality will surge, for at leasttwo reasons. Trade unionists worry that the reforms will reduceunion membership, undermining the consensual system of la-bour relations. Ola Pettersson, chief economist of Sweden’sTrade Union Confederation, says the government is �undercut-ting� the Swedish labour model. That seems an exaggeration.With more than seven out of ten workers still members of un-ions, Sweden’s collective bargaining model looks safe for now.

Others suspect a poverty trap in the making, with peoplestuck with low skills in low-wage jobs. That seems a more seri-ous risk, particularly for Sweden’s recent migrants who, by andlarge, are poorly educated and speak little Swedish. Pernilla Lan-din, a social worker who runs a multi-faith community centre inFisksätra, already sees dangerous signs of social exclusion. �Peo-ple don’t have enough money to buy a train card,� she says, �sothey can’t get out to look for work.�

But the danger is vastly reduced by Sweden’s all-envelop-ing public services. Although government spending has shrunkin recent years, the Swedish state is still large (51% of GDP lastyear), and it spends much more than Anglo-Saxon countries doon everything from early-childhood education to job search andtraining. According to the OECD, more than 70% of the childrenof the poorest �fth of Swedes are in state-�nanced child-care andeducation schemes, compared with fewer than 30% in America.

Sweden’s government has also experimented more boldlythan others with boosting public-service e�ciency. Manyschools are now independently run, and in health care privatemanagement is a growing trend. Public services have not entire-ly escaped cuts, but they started high and were designed to pro-tect the poor. Once you allow for the progressivity of public ser-vices, the OECD reckons, Sweden’s Gini drops to 0.18. That stillleaves it as the world’s most equal place, as well as one of thefastest-growing and �scally stable countries in the rich world.

It would be naive to think that its model can simply be cop-ied elsewhere. Sweden’s citizens are strikingly committed to so-cial cohesion, and willing to pay for a large state. A revival ofAmerica’s union movement would be likely to lead to growth-destroying rigidities. Equally, it is hard to see Americans accept-ing the taxes that would go with government spending of morethan 50% of GDP. Sweden’s remake of the welfare state is mostrelevant in Europe, where in the aftermath of the �nancial crisismany countries are now struggling with unsustainable public �-nances, as Sweden did 20 years ago.

A place to look for ideas

Nonetheless, there are broader lessons. Sweden’s experi-ence suggests that the welfare state can be trimmed by cuttingtransfers and maintaining progressive investment in social ser-vices, without allowing inequality to surge. And a revamp of thewelfare state that encourages employment can boost growthwhile keeping income gaps to a minimum.

The most important conclusion, however, comes from con-sidering Sweden’s experience alongside the recent record of theUnited States, emerging Asia and Latin America. All these casestudies indicate that the geography of contemporary inequalityhas as much to do with government policy as with underlyingeconomic forces. But it has not been a simple tale of tax and re-distribution, nor is there a simple trade-o� between e�ciencyand inequality. Sweden’s economy has become much more e�-cient while still keeping inequality low. America’s system of tax-es and transfers is less progressive than it was in the 1970s, yet thestate is no smaller. That suggests there is room for reforms thatboth counter inequality and improve economies’ e�ciency. 7

MITT ROMNEY, AMERICA’S Republican presidential can-didate, caused a kerfu�e earlier this year when he dis-

missed concerns about inequality as the result of �class warfare�that had no place in America’s public discourse. Rather than an�envy-oriented� debate about distribution, he argued in favourof creating a �merit-based� America, with policies that focus oneconomic growth.

Mr Romney’s nonchalance about income gaps is controver-sial, even in America. But he is not alone in assuming that distri-bution and dynamism do not go together. The predominantview among economists has long been that there is a trade-o�between prosperity and income equality.

A century ago inequality was deemed an essential condi-tion for investment and growth because rich people save more.Keynes wrote in 1919 that it was �precisely the inequality of thedistribution of wealth which made possible those vast accumu-lations of �xed wealth and of capital improvements which dis-tinguished [the Gilded Age] from all others�. More recently thefocus has been on its incentive e�ect. Milton Friedman arguedthat greater inequality would spur people to work harder andboost productivity. Gary Becker, of the University of Chicago,thinks that inequality encourages people to invest in their educa-tion. Redistribution, in contrast, brings ine�ciencies as highertaxes and government handouts deter hard work. The bigger thestate, the greater the distortion of private incentives.

That logic remains as powerful as ever. Economic freedomand better incentives boosted growth in China, India and else-where. Sweden’s experience shows that deregulation, lower tax-es and fewer bene�ts increase economic dynamism even as theyreduce equality. Yet the analysis in this special report suggeststhat logic is incomplete. Some of today’s inequality may be inef-�cient rather than growth-promoting, for several reasons.

First, in countries with the biggest income gaps, increasinginequality is partly a function of rigidities and rent-seeking�be itlabour laws in India, the hukou system and state monopolies inChina or too-big-to-fail �nance in America. Such distortions re-duce economies’ e�ciency. Second, rising inequality has not, by

Trade-o�s

Having your cake

Less inequality does not need to mean less

e�ciency

4The Great Gatsby curve

Source: “Inequality from Generation to Generation: The United States in Comparison”, by Miles Corak, 2012

Late 1990s, early 2000s

Inequality, Gini coefficient (1=perfect inequality)

Inte

r-ge

nera

tion

al e

arni

ngs

elas

tici

ty(1

=no

inte

r-ge

nera

tion

al m

obili

ty)

0

0.2

0.4

0.6

0.8

0.2 0.4 0.60.3 0.5

Denmark

United States

Brazil

China


WORLD ECONOMY

SPECIAL REPORT

2


WORLD ECONOMY

SPECIAL REPORT

2

1

and large, been accompanied by a smaller (and hence less distor-tive) state. In many rich countries government spending has ris-en since the 1970s. The composition has changed, with moremoney spent on the health care of older, richer folk, and relative-ly less invested in poorer kids. Modern transfers are both lessprogressive and less growth-promoting.

Third, recent experience from China to America suggeststhat high and growing levels of income inequality can translateinto growing inequality of opportunity for the next generationand hence declining social mobility. That link seems strongest incountries with low levels of public services and decentralisedfunding of education. Bigger gaps in opportunity, in turn, meanfewer people with skills and hence slower growth in the future.

It is not easy to distinguish between e�cient and ine�cienttypes of inequality. The development of big cross-country statis-tical databases in the 1990s allowed economists to compare Ginicoe�cients and GDP growth in lots of countries over manyyears, but the results were mixed. Some studies found that wideincome gaps were associated with slower growth. Others foundthe opposite. In a 2003 paper entitled �Inequality and Growth:What Can the Data Say?�, Abhijit Banerjee and Esther Du�o ofMIT concluded that the answer was �not very much�.

More recent studies, however, support the idea that in-equality can be ine�cient. In an in�uential analysis in 2011 twoIMF economists, Andrew Berg and Jonathan Ostry, looked at thelength of �growth spells� rather than simply comparing growthrates. They found that growth was more persistent in more equalcountries, and that income distribution mattered more for thelength of growth spells than either the degree of trade liberalisa-tion or the quality of a country’s political institutions.

Other researchers have tried to isolate the �unhealthy�types of inequality using the two indices of inequality of oppor-tunity �rst developed by the World Bank and described earlier inthis special report. Two Spanish economists, Gustavo Marrer0and Juan Gabriel Rodríguez, built an index of economic oppor-tunity for individual American states. They found that states’GDP growth was inversely correlated with their inequality ofopportunity, but not with overall inequality. In a forthcomingWorld Bank working paper, Ezequiel Molina, Jaime Saavedraand Ambar Narayan �nd that countries with lower educationalequality, as measured by the Human Opportunity Index, growmore slowly.

This line of research is in its early stages, but a second strandof evidence, which examines the link between inequality andsocial mobility, is more developed. There are now plenty of stud-ies which use the inter-generational elasticity of income to mea-sure social mobility in di�erent countries. Miles Corak, a Canadi-an economist, �rst plotted the results of these studies on a singlegraph. It is known as the �Great Gatsby Curve� (see chart 4, previ-ous page), and suggests that countries with higher Gini coe�-cients tend to have lower inter-generational social mobility.

Perpetuating advantage

In some ways the link between wider income gaps andlower social mobility is unsurprising. From violin lessons to tu-tors for tests, richer parents can invest more in their children, im-proving their chances of getting into the best universities. Themeritocratic assumption is that public provision of basic ser-vices, particularly education, does enough to counter this advan-tage to give everyone a reasonable start. That was never true inpoor countries with rudimentary social services. Increasingly, itdoes not seem to be true in rich ones either, particularly America.But the link between inequality and declining mobility is not in-evitable. Countries such as Sweden that invest heavily and pro-gressively in public services are more likely to prevent widening

income inequality from reducing opportunity. And Latin Ameri-ca shows that investing more in education at the bottom can im-prove social mobility even in the most strati�ed places.

Lower growth rates may not be the only symptom of eco-nomic damage from inequality. Another could be macroeco-nomic instability. In an in�uential recent book, �Fault Lines�,Raghuram Rajan pointed to inequality as the underlying causeof America’s 2008 crash. As less-educated Americans saw theirincomes fall, he suggested, politicians encouraged reckless mort-gage lending so that poorer folk could keep up their living stan-dards by borrowing. This argument echoed John Kenneth Gal-braith, who wrote in the 1950s that �bad distribution of income�was the main cause of the Depression.

The thesis seems plausible. There is evidence that widen-ing income gaps in America pushed less a�uent people tostretch their �nances, particularly to buy pricier houses. RobertFrank, an economist at Cornell University, has documented �ex-penditure cascades� where rich people’s spending patterns af-fect those of the near-rich. (One reason is that the less a�uentwant their children to go to the best schools, and house prices of-ten re�ect the quality of the local school.) Other scholars havespotted a link between inequality and �nancial distress. DavidMoss at Harvard Business School, for instance, found that therate of American bank failures was highly correlated with thelevel of inequality.

But the link is not ubiquitous. In Germany and, especially,in China, higher inequality has encouraged saving rather thanspending. Nor are �nancial crises always preceded by wideningincome gaps. Michael Bordo of Rutgers University and Christo-pher Meissner of the University of California, Davis, looked at 14�nancial busts in rich countries between 1920 and 2008 andfound that these crises were typically preceded by credit booms,but only occasionally by rising inequality. In the most compre-


SPECIAL REPORT

WORLD ECONOMY

2

1

hensive analysis Anthony Atkinson and Salvatore Morelli of Ox-ford University looked at �nancial crises in 25 countries over thepast 100 years and concluded that there was no systematic rela-tionship between inequality and macroeconomic disaster.

Since both the levels and the origins of inequality varywidely, it is hardly surprising that there is no established relation-ship between income gaps and �nancial crises. That does notmean inequality never aggravates macroeconomic instability,but unfortunately critics of inequality often exaggerate theirclaims. A case in point is �The Spirit Level�, a book by two Britishepidemiologists, Richard Wilkinson and Kate Pickett, publishedin 2009. They claimed that higher levels of inequality were asso-ciated with higher murder rates, lower life expectancy, more obe-sity and all manner of other ills. Their explanation was a medi-cal one. Inequality literally gets �under your skin� because thestress of keeping up with the Joneses raises cortisol levels.

�The Spirit Level� caused a sensation when it was �rst pub-lished in Britain, probably because it re�ected the post-crash Zeit-

geist. Its conclusions, however, have been largely debunked. In adevastating critique, published by the Democracy Institute,Christopher Snowdon showed that Mr Wilkinson and Ms Pick-ett made highly selective use of statistics. Other, more carefulstudies show that although there is a strong relationship be-tween individual income and health (richer people tend to behealthier and live longer than poorer ones), the link betweencountries’ income gaps and their citizens’ health is weak.

Exaggerated claims of the damage from inequality havethemselves done damage by reinforcing caricatures in an al-ready highly charged debate. Quite legitimately, di�erent peoplehave di�erent notions of what is fair, and what is the right bal-ance between fairness and e�ciency. But whatever their views,there is a reform agenda which both sides should embrace, onethat both boosts e�ciency and mitigates inequality. 7

ON AUGUST 31ST 1910 Theodore Roosevelt, by then Ameri-ca’s ex-president, addressed a crowd of 30,000 at a civil-

war commemoration in Osawatomie, Kansas. In one of Ameri-ca’s most famous political speeches, he laid out his progressivephilosophy. The federal government had a responsibility to pro-mote equality of opportunity and attack special privilege andvested interests. �In every wise struggle for human betterment,�he argued, �one of the main objects, and often the only object,has been to achieve in large measure equality of opportunity.�

A century on, many emerging economies face circum-stances not unlike those of Roosevelt’s era. In the rich world gov-ernment has become bigger than he ever imagined. But both richand poor, in their e�orts to boost growth and mitigate inequality,could draw inspiration from the spirit of the Osawatomiespeech. Three broad reforms stand out.

One is to curb cronyism and enhance competition, particu-larly in emerging markets. Just as Roosevelt broke up America’strusts (monopolies) and cracked down on political corruption,China, India and many other emerging economies need to dosome trustbusting and graft-attacking of their own. In China,freeing monopoly sectors, from mining to railways, would reori-ent the economy towards domestic consumption and reduce in-come gaps. A freer �nancial sector, with market-driven interestrates, would remove a potent source of income concentrationand economic distortion.

In advanced countries, removing subsidies for too-big-to-fail �nancial institutions should also be high on the new progres-sive agenda. That, too, would result in more balanced economiesand remove the rents that lie behind a lot of the surge in wealthat the top. Rich countries also need more competition in tradi-tionally mollycoddled sectors such as education. Governmentshave a responsibility to invest in the young, but also to ensurethat teachers have incentives to do their best.

The sooner the better

A second priority is to attack inequality with more targetedand progressive social spending. In emerging economies, espe-cially in Asia, that means replacing expensive universal subsi-dies for energy with tailored social safety nets. It means wideruse of conditional cash transfers. Latin America’s models aregradually being copied elsewhere, but there is much farther togo: rich countries would do well to adopt the idea of tying socialassistance to individuals’ investment in skills and education.

Both rich and emerging economies must bring about a shiftin government spending�from transfers to education, and fromolder and richer people to younger and poorer ones. Even if in-equality were irrelevant, developed countries would need to re-form their pension and health-care systems because today’spromises are simply una�ordable. Concerns about distributionand its e�ect on future growth add impetus: the longer that gov-ernments prevaricate about reforming entitlements, the morewill be squeezed from investment in the young and poor.

These days, public investment in education needs to go be-yond primary and secondary school. Giving the less advantageda leg up means beginning with pre-school and includes retrain-

Policy prescriptions

A True Progressivism

Bold moves are needed to tackle inequality and boost

growth at the same time

Parts of this agenda aretaking shape, particularly inemerging economies. Brazil hasbegun a pension overhaul. Chi-na has boosted social servicesin rural areas. Indonesia andmost recently India have cutfuel subsidies. But in the richworld the decibels of the in-equality debate have beenmatched by the inadequacy ofthe reform e�ort. In continentalEurope there is nothing muchbeyond a clamour to raise toptax rates. Britain’s coalition gov-ernment has taken on the wel-fare system but balks at gettingrid of free bus passes for a�uentold folk.

The most shocking short-comings are in America, the richcountry where income gaps arebiggest and have increased fast-est. The Republicans are right tosay that Medicare, America’shealth-care system for the old,must be overhauled. But byslashing government spendingon basic services such as educa-tion and advocating yet moretax cuts at the top, they under-mine equality of opportunity.

The Democrats are little better. Barack Obama gave his ownspeech at Osawatomie last year, wrapping himself in Roosevelt’smantle. Inequality, he said, was the �de�ning issue of our time�.But his response, from raising the top income-tax rate to increas-ing college-tuition subsidies, was just a laundry list of small ini-tiatives. Roosevelt would have been appalled at the timidity. Asubject of such importance requires something much bolder. 7

Offer to readers Reprints of this special report are available. A minimum order of five copies is required.Please contact: Jill Kaletha at Foster Printing Tel +00(1) 219 879 9144e-mail: [email protected]

Corporate offerCorporate orders of 100 copies or more are available. We also offer a customisation service. Please contact us to discuss your requirements.Tel +44 (0)20 7576 8148e-mail: [email protected]

For more information on how to order special reports, reprints or any copyright queries you may have, please contact:

The Rights and Syndication Department26 Red Lion SquareLondon WC1R 4HQTel +44 (0)20 7576 8148Fax +44 (0)20 7576 8492e-mail: [email protected]/rights

Future special reports

Previous special reports and a list of forthcoming ones can be found online: economist.com/specialreports

Information technology andgeography October 27thFrance November 17thMexico November 24thObesity December 15th

SPECIAL REPORT


2

WORLD ECONOMY

ing for the less skilled. In both areas America, in particular, isfound wanting. Its government spends barely more than 0.1% ofGDP on �active labour-market policies� to get the less skilledback to work, one-�fth of the OECD average. Only half of Ameri-can children attend pre-school. China plans to have 70% of itschildren in three years of pre-school by 2020.

The third priority is to reform taxes, to make them a lotmore e�cient and somewhat fairer. Critics of inequality oftentout higher marginal taxes on the rich. Yet in most countries oth-er than America, government spending is a much more impor-tant tool for combating inequality than the tax system. Tax rev-enue is better seen as a way to fund the state, not a tool to punishthe rich. Economists argue about the disincentive e�ects of high-er tax rates. (Messrs Piketty and Saez, the economists who havetransformed analysis of income concentration at the top, reckon,controversially, that the optimal top income-tax rate could be ashigh as 80%.) But no one doubts that there are trade-o�s.

In countries where the state is already large, rebalancinggovernment spending should take precedence over raising morerevenue. But given the mess that public �nances in most coun-tries are in, more tax revenue is likely to be necessary, particular-ly in less highly taxed countries such as America. Even there,though, higher marginal income-tax rates should not be the �rstchoice. Instead, the focus should be on eliminating distortionsthat reduce both progressivity and the tax system’s e�ciency.

The �carried-interest� loophole, which allows private-equ-ity managers to pay (low) capital-gains rather than (higher) in-come tax on their earnings, is one such sore. So are many tax de-ductions, from those for charitable contributions to mortgageinterest, most of which disproportionately bene�t the wealthy.An overhaul of the tax code to reduce corporate tax rates andnarrow the gap between individuals’ tax rates on capital and la-bour income would improve its e�ciency and make richer peo-ple pay higher average tax rates. Higher property taxes would bean e�cient and progressive source of revenue. Inheritance taxcould be reformed so that it falls on individual bene�ciaries rath-er than on the estate as a whole, as it does in Germany. Thatwould encourage the wealthy to distribute their wealth widely,thereby making a hereditary elite less likely.

In mostcountriesother thanAmerica,governmentspending isa muchmoreimportanttool forcombatinginequalitythan thetax system

for richer, for poorer - paris school of economics economoist... · est man is a mexican (carlos...

Documents