A Perfect Crime

A Perfect Crime: Inequality in the Age of Globalization Author(s): James K. Galbraith Source: Daedalus, Vol. 131, No. 1, On Inequality (Winter, 2002), pp. 11-25 Published by: The MIT Press on behalf of American Academy of Arts & Sciences Stable URL: http://www.jstor.org/stable/20027734 Accessed: 27/10/2009 11:18 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=mitpress. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The MIT Press and American Academy of Arts & Sciences are collaborating with JSTOR to digitize, preserve and extend access to Daedalus.

http://www.jstor.org

James K. Galbraith A perfect crime: inequality in the age of globalization

of the world's leaders IVlost political economic have embraced globalization on two : that open markets and grounds are transnational networks production unstoppable surely flow

Genoa.

countries, globalization and for imperialism synonym the word evokes colonialism ;generally common rather than collapse gain. All in

of deregulation, all, if the imposition pri free trade, and free capital has in fact raised living stan mobility

;and that the benefits will out to all the world's people,

vatization,

and rich and poor. Leading economists the logic journalists agree, convinced by of laissez-faire, comparative advantage, and technology transfer. Accordingly,

is devilishly dards worldwide, gratitude hard to find. So what are the facts? Has globaliza tion hurt or helped? Oddly, researchers not ask. do not know; mostly they do as it is For the doctrine of globalization

they declare, class conflict and competi tive struggle are obsolete. Yet, outside elite circles, conflict and to disappear. In rich struggle refuse and pressure electorates countries, and proreg groups remain protectionist in parts of Europe ulation, even socialist there iswide sympathy for the (nonvio of and Seattle, Davos, lent) protestors

In poor

is the new

understood

in elite circles

contains

the that the global mar assumption ket is itself beyond reproach. The formu lae for success in that market from curious

;

to sound and transparency openness to investment in education finance remain matters

James K. Galbraith is a professor at theLyndon B. Johnson School of Public Affairs, theUniversity of

Texas

at Austin,

and

a senior

scholar

at the

Levy Economics Institute of Bard College; in the early 1980s, he served as executive director of the congressional Joint Economic Committee. His recent academic work emphasizes practical issues of measurement and data quality as the key to progress on important questions of economic poli cy.His books include "Created Unequal" (1998) and "Inequality and Industrial Change: A Global View" (2001).

of national responsibili countries that fail have only their ty ; own deficiencies to blame. In line with this view, most research has focused on national

conditions and national poli or the cies, and not on global conditions as such. effects of globalization Whether such a national focus is or whether a appropriate, global view a would be better, is question of great we To resolve would it, importance. need new efforts to measure economic and social progress across development countries in effect, around the world

D

dalus

Winter

2002

11

James K. Galbraith on inequality

we would

a global report in this vein, initiatives

to write

need

carcj# gut official theWorld

notably

Bank's Human

Devel

opment Report, are caught in contradic tions between the cheerful predictions of globalization theory and what the evi unem of epidemics, illiteracy, is actu and poverty suggests ployment, that contradictions ally happening have driven several senior figures (most former chief economist famously, Joseph dence

the Stiglitz) out of the Bank. Meanwhile, International Monetary Fund (imf) and World Trade Organization (WTO) are sealed off from today closed societies, most

forms of serious critical discussion. can independent research con what So, of the state tribute to an understanding of the global economy today? Not much, evi the fragmentary perhaps, beyond dence of case studies and field reports. usually lack new bring together on a information scale. global Still, one broader area has attracted attention: economic inequality and its

Truly independent to the resources

scholars

to economic In this growth. relationship as we shall see area, data (collated, below, from many sources) independent are already available. A rich (but inex can be to econometrics brought pensive) on the findings. hinges economic and inequality

bear. And much

For, while are hardly the only issues in growth litera world development life, health, are more the and cy, peace important these between relationship perceived two economic opment

policy

variables

underlies

in profound

devel

ways.

economist Simon -L/ong ago, the great to Kuznets linked (1901 -1985) equality ar Kuznets the development process. as industrialization began, it gued that at lead to an increasing inequality might first: would

the rough parity among farmers an emerging urban give way to 12 D

dalus

Winter

2002

But as industrialization

rural divide.

the center

of economic gravi shift to the cities. To live in

deepened, ty would

cities, to work for wages, requires free labor and that every family have some access to cash. The larger the share of basic consumption goods provided the more equal through the marketplace, incomes must become. Eventual money and social-welfare systems ly, democracy would toward the emerge ;progress social democratic of West frameworks ern was a and North America Europe as well as an ideal. pragmatic possibility Kuznets offered an optimistic vision, to that of John Maynard Keynes more austere in though style. Properly could be civi development managed, similar

lead to the misery and that Karl Marx had earlier fore upheaval seen. But in recent years, as the Marxian lized;

it need not

threat disappeared, Kuznets's vision also unions in disorder and receded. With welfare

states

in disrepute, the 'Kuznets serves mainly as a

now

hypothesis' research whipping boy of development at length, ers, to be raised, sometimes as discon but usually to be dismissed tinued by modern data, generally by economists who see no contradiction

between and development. inequality In a 1993 report entitled the East Asian to as EAM), Miracle referred (hereafter economists

at the World

Bank offered

an

alternative

theory. They argued that of land redistributions, early especially were precondi and primary schooling, tions for industrial

success

in Japan, Ko

rea, Taiwan, China, and Thailand. This widely cation

Singapore, Malaysia, argument has been a case for edu cited to underpin as a development tool, to support

in the early stages, redistributive policies to argue that development and especially can be market-friendly, the provided in and "right" pattern of endowments centives

exists

at the outset.

pre By emphasizing market-friendly conditions for growth, the EAM team undermined the presumption that devel and rising equality normally - even together though the Asian evidence did suggest that this was in fact the case inmost places. The team also the policy activism of many downplayed Asian states, especially their commit ments to planning, industrial policy, opment occur

financial

control, and the development - all cornerstones of social welfare of a as normal development under process stood by Keynes

and Kuznets

in their

day. The EAM's striking hypothesis also to in called attention, the indirectly, of in about information completeness around

the world. While efforts to measure

equality had been many at the national equalities

there in

level, no one measurements to had brought those a gether in single global data set. As a one result, it remained unclear whether

could generalize perience. Were valid on awider

from

the East Asian

the apparent scale ?Was

that redistributive

policies

ex

lessons it really true set the stage

for growth? To help answer

that question, Klaus and of the World Deininger Lyn Squire to mine Bank decided the economic de literature for surveys of in velopment come of thousands inequality. Finding in scores of studies, such measures they evaluated each data point on three crite ria. Did the study focus on households, rather than persons ?Did it attempt to measure all forms of income, including in-kind incomes ? Did it attempt to cov er all parts of the society, including rural as well as urban areas ? In 1996 re they the data these three published satisfying criteria as a "high-quality" data set on income inequality household since 1950. The Deininger-Squire data set is now a standard

reference

;a recent version

of

A perfect cnme

fers nearly 750 country/year observations on which dozens of papers have been based. The

an unintended

result has been

scent

into confusion.

As

de

scholars

sought in between relationships in and the equality, income, growth World Bank's data, no consistent pat tern emerged. Some seemed to confirm systematic

the Kuznets

Others hypothesis. argued that inequality first falls and then rises with rising income :the oppo instead

a rela finding of low inequality and first supported, and on the that the ground seemed to rest on conti

site pattern. The tionship between later growth was then questioned,

EAM

relationship differences nent-specific America and Asia. Meanwhile, came under associated servative

Latin

between

viewpoints pro-equality from a quarter

challenge, with older

policy

views.

theories

and con

In Victorian was

nomic

eco

itself the

thought, inequality spur of growth. Growth required capital and itwas the accumula accumulation, tions made possible by concentrating incomes that justified an unequal class structure. The Victorian system worked - so in practice, well enough as, long occur did and the class growth working es enjoyed the benefit of a steadily rising living standard. But, as Keynes particu all this had died larly well understood, with World War I. And then, after sixty years, the very same idea was born again. Let the rich - so wrote econo rule the supply-side mists who came to power behind Ron in America ald Reagan and Margaret Thatcher in Britain. Tax cuts would

im

the incentives

to of the wealthy prove and invest." High interest "work, save, rates would reward saving and quash inflation. And a fetish of the entrepre neur ness

and busi spread through political that implicitly culture, a doctrine D

dalus

Winter

2002

13

James K. Galbraith on

inequality

the concept of value in innova t^on an(j ieacjership, rather than in labor . . r i i or consumers. or even in the happiness rooted

a similar doctrine finds expres Today, sion in models that emphasize syner returns, technological increasing in part Two good educations, are more than twice as good as nership, gies,

change.

should concentrate one, so the educated into enclaves. Technical development of the the relative productivity the skilled well trained therefore, pay more and the unskilled less. In these as clusters cases, inequalities expressed will foster of privileged opportunity raises

more

rapid growth. an empirical basis for such Seeking Kristin Forbes uses the Dein claims, and inger Squire data set to find that in are followed by in creases in inequality creased growth rates. If valid, these find reverse

the EAM idea without

lost luster

to Simon Kuznets.1

ings would restoring

matters have i\s stand, economists four different possibilities, broached for the each with different implications : economic of strategy development The redistributionist view holds that egal are a precondition itarian social policies for growth and points to the Asian mira cle before the crash of 1997 as prime land This view emphasizes but tends to resist and education, once inmarket processes intervention have been successfully preconditions

Inequality may well rise, but the of a growth strategy makes the sacrifice worthwhile.

change. success

The Kuznets and Keynes view implies that is the normal out increasing equality come of a process of rising incomes whether fast or slow and that social are normal welfare policies outgrowths to an urban industrial of the transition This perspective that redistribution

economy.

suppose cede growth equality will

; it only decline

matures.

process

not pre

does

should

pre implies that in as the development

But if Kuznets

and

were

Keynes right, then strategies of im industrial diversification, planning, financial and substitution, control, port state are a legitimate part of the welfare tool kit; they do not the development contradict the fundamental project. is The Scotch verdict "not proven" to that development be unrelated may in any inequalities is This the fallback posi systematic way. tion of some who argue for growth as the sole development objective. and economic

social

is complex. Many could be true. But how

the world

Plainly, different

things of these four conflicting

many ses

are,

correct?

actually,

At

hypothe one.2

most,

one ?A great deal rides on the

Which answer.

evidence. reform

met.

The should sources

ports,

neoliberal view is that policymakers re go for growth, concentrating on

comparative

and the fostering

advantage,

ex

of technological

v_>Jne reason why the question is that the evidence unresolved

2 To

see

14 D

dalus

Winter

2002

the

contradiction

hypothesis consider hypothesis, identical levels with If both

i See Kristin of the "A Reassessment Forbes, and Growth," Between Relationship Inequality Review American Economic 90 (3) (September 869-887. 2000):

against have been test - is data set

views which the modern ed the Deininger-Squire unreliable.

Kuznets

then Kuznets

both.

But

country inequality

the

of

income

and

start

inequality.

income level, higher in decline should inequality

the Forbes can

between

and, say, the Forbes two countries that

to a

grow

under

remains

raise in the

model

its growth short run.

suggests rate by It follows

that

one

increasing that later

cannot

One

-

are

measurements

sparse,

Africa, West Asia, and Latin America. And where observations exist, they are : falls for about questionable inequality in this exercise, in a half the countries decade marked by wide protests against data and com rising inequality! When mon perception clash so sharply, which

sepa

rated in time by many years or even dec some of the rank ades. Second, while - one seem reasonable expect ings might low inequality readings from Eastern years Europe during the communist others are, to put itmildly, implausible.

is to be believed?

in India, Pakistan, and In Is inequality donesia really in the same general league as in Norway? in Spain Is inequality Is inequali really lower than in France? in Australia the States and United ty or the to, say, Nigeria comparable really Sudan? It is doubtful that any Sudanese thinks so.

One

on

that

country

will income

higher

have, and

other

things

at least

equal,

the EAM model reducing

is correct,

inequality

in the

a country run will

then short

across

income

case,

if growth

level then

as one

that

the

transition,

and

it may

be

be measured

growth

specify what accelerates.

happens

to

inequality

easily and accurately countries. Pay is, of course,

if

many

for a large

be

In this

re

stored at all ; the EAM hypothesis does not clearly

countries.

of house Inequalities widely neglected. hold income the focus of Deininger - are to meas and Squire very difficult we measurements and the do have ure, come from unofficial often surveys.3 Levels of pay, on the other hand, may

be re may be observed not

per

so turns out to be possible, long to narrow the focus and iswilling to return to official sources of informa - sources tion that are very rich, but

in the redistrib

accelerates

the Kuznets country, pattern uting not stored after a time. But it will during

not.

does

basis,

year-to-year

This as one

observed at lower inequality but initially the same

a

of changes mitting comparison to changes in inequality in GDP. It should be based on data that are reason consistent ably accurate and reasonably

being or per

The downward-sloping haps higher, inequality. no relation Kuznets will longer be observed, in time series or panel either data. Similarly,

on

detailed

is rapidly reduced to a data set are from half of all observations

equal,

V Ve clearly need more and better data. A new data set should, ideally, approach coverage of the global comprehensive economy

these re The thin coverage on which amore serious sults are based becomes one tries to compute still when problem in theWorld Bank's inequality changes time period. data over some consistent where

of the

for Economic

Cooperation (this is true of the Forbes study, for example). A simple effort to compare changes from the the decades for 1980s to the 1990s which the Bank reports the most obser vations shows no data for most of Organization and Development

data set quickly reveals "high-quality" fundamental problems. in First, in many parts of the world Africa most of all, but also in Latin America, Asia, and even in parts of Eu rope

countries

the affluent member

but admire the effort of some or and Deininger Squire to bring der to the chaotic history of income-in Still, a compari equality measurement. son of coefficient values from their

3 The only consistent formal definition of comes

have

from

the

income

tax,

It is tax

the precise allowable specifies of each type of inflow and outflow. law that specifies and that wages

code

salaries

are

treatment

not, after

we

income whose

should vary,

that

but

income,

and

be deducted, the

are gifts received of business expenses

that

reimbursements

of

concept

D

so on.

income

dalus

Winter

Since

tax

is therefore

2002

laws nation

15

A perfect cnme

James K. Galbraith inequality

of income. Levels of manufactur - an subset of all pay important ingpay have been measured with reasonable a as matter of official routine in accuracy subset

most

countries

around

the world

for data nearly resulting sector have payrolls by manufacturing indus been placed in a single systematic trial accounting framework by the Unit forty years. The

ed Nations

International

Development which makes

(UNIDO),

Organization

both

cross-country comparison relatively reliable. if one

In short,

easy and

into focus.4 of the University of has been (UTIP) Inequality Project to compute consistent measures of man contribution

Texas

this

from

pay inequality

ufacturing

set permits us to generate large of inequality measures that can across countries be compared and data

numbers

through time. The year 2000 release of the UNIDO data set alone contains suffi cient information for nearly 3,200 coun between 1963 and try/year observations than four times the coverage 1998 more of Deininger and Squire.

to look at the

iswilling

relationship through growth-inequality the narrow lens of pay rates and earnings to get the picture it is possible structures,

xhe

an using inequality metric in the 1960s by the late econo developed metrician Henri Theil.5 The application to UNIDO's of this technique industrial information,

main

4 The siveness

trade-off

measurement

are

conventions

of

effects

trade,

accounting laws. Differences across

ments

in measured

tax

in accounting and countries will produce In countries

incomes.

treat

differences

where

sub

or are unrecorded, parts of income or hidden in kind, from delivered tax, or where are vague, of standards accounting problems valuation and aggregation rapidly mount. stantial

is no

be

reason

to to expect measurements time or across surveys. through the project of comparing facts make across countries and years inequality

There

consistent

These income

and difficult, doubly at moderately reliable meticulous

even

in practice

comparisons of micro-level

examination

to arrive requires data.

Such a project has been undertaken with in recent Studies, wealthiest

countries. on

researchers,

The World

the other

their

of data judgments about work information

tury

in far-flung

surveys. pendent is not their fault are

coefficients

l6 D

quality done

to the Bank's had

hand,

on over

to base summary a half cen

often inde locales, through and It cannot be surprising that the resulting inequality

dalus

Winter

2002

the

the broader which

distribution, effects

of

and

taxes, Second,

composition. a strong correla apparent our measures of manufacturing

changing there is nevertheless tion

of concept includes

transfers,

household

between

pay inequality and the (we believe, reliable)

but inequality reported of countries range only by the Income This Studies. is, of course, Luxembourg : countries with highly plausible strongly egali are com to have both tarian values likely and strong welfare pay structures pressed

measures

of

income

for a limited

states.

we

Thus

a very

provides mate rankings

believe

that

the UTIP way

inexpensive of income

to

inequality and reliable

where detailed countries, data are not available.

5 To

be precise,

we

groups component a set of industrial two-

to

three-digit

compute of Theil's

approach approxi for many micro

inequality entropy mation

the between T

statistic

categories, generally standard international

industrial classification statistic

desirable

problematic.

than

income

skill

Income

years by the Luxembourg but it is restricted mainly

itself,

hypothesis of pay, rather

confounding tax

and the change, concern the distribu

technological

Kuznets tion

in the

codified

comprehen accurate

emphasize

manufactur domain, over measure ing pay inequality, implausible ment of a comprehensive income one, inequali pay inequalities ty. However, manufacturing are in their own interesting right. First, many economic the important problems, particularly

household

ally specific,well defined only where precise

is between

: we accuracy of a limited

and

(isic)

across at the

levels. Theil's T

of a class example as "generalized on infor based measures," originally measures have all the Such theory. is the best-known measures

mathematical

known

characteristics

of

the

of

A map averaging values of an inequali from these data ty coefficient computed

(see figure 1) graphically displays the key as findings. Over the 1963 -1998 period, these data reveal, manufacturing pay in was lowest in the social democ equality racies of Scandinavia and in Australia the communist

and under

regimes of and Cuba (be changes, data for

Eastern

Europe, China cause of the boundary Russia here exclude the Soviet

years). Southern Europe and North America form a second group of countries with

low levels of pay inequality. relatively countries The wealthier of Latin Ameri ca (such as Argentina, and Venezuela, and West Asia (Iran) form a Colombia) middle group; Russia (after 1991) and

South Africa rank slightly higher. The are found regions of highest inequality a in broad equatorial belt, from Peru and Brazil through central Africa and south ern Asia, the largest gaps be reflecting tween city and countryside, between oil more

Gini known widely are additive addition they that inequality (meaning

plus total

inequality

between

and

coefficient,

in

and

decomposable a set of groups within sums to the groups

for the combined inequality population). This makes tool for work them a very flexible data, and we have ing with semi-aggregated meas in that changes shown between-group ures are often a very robust of indicator

changes in the whole distribution the unobserved Use

of

es of

industrial international

innovation cussion technical Maureen

within-groups

of of

category

comparison the UTIP project.

technical

(including

component). for purpos schemes

appendices Berner, eds.,

Inequality

the and

and Industrial

Change (Cambridge :Cambridge University I thank 2001). sub Kum, and George

Press,

Pedro Purcell

Concei?ao, especially,

Hyun from

among the UTIP team, for diligent help with and with calculations ; the present concepts world data set in particular is Kum's handi are at The UTIP measurements work.
://utip.gov.utexas.edu>.

A perfect cnme

tion tend to go hand in hand. the Kuznets hypothesis Although to levels of levels of inequality relating or is income corroborated pay broadly some doubts about his by these findings, views

do remain.

example, Kuznets

It is debatable, for if inequality must increase,

as

in the earliest

stages Yet that so far places, as modern times and data are con are past the ear cerned: most countries

supposed, of industrial development. ismoot inmost question

ly stages. Itmay also be that inequality rises slightly in a few of the very richest as income grows, due countries partly to - a sectors in capital gains technology pattern of interest for students of the United States and the United Kingdom, but not broadly relevant to the study of economic development. On

is the principal For further dis

consult details, please in James K. Galbraith

and food, and the weakest development and the producof mass manufacturing tion of capital goods. are in These findings striking accord with Kuznets's basic hypothesis. Higher are incomes and lower pay inequality is Because associated. this true, strongly there is no reason to expect a systematic between relationship today inequality and growth later, and none can be found. Redistribution in either direc or not a tion is down up apparently economic for ; in precondition growth and redistribu stead, successful growth

within

the whole,

of pay inequalities to tend be lower manufacturing

in rich countries than in poor. That means that inequality almost surely de clines as industrialization and deepens as incomes rise. This is consis finding tent over the globe, with limited excep run of an tions, over a thirty-five-year in the early 1960s. nual data, beginning that Kuznets's -Besides confirming view of the relationship between growth remains basically and inequality correct, D

dalus

Winter

2002

17

James

K.

Galbraith inequality

i

Figure

Global Inequality 1963

-

1997 of manufac

Inequality

turing pay computed by the University of Texas from Inequality Project 2000 edition the UNIDO of Industrial

Statistics over 1963 and averaged ranked 1997 ; countries

as in six quantiles the high 3. Note and geographic coverage into

Figure

of inequality consistency in the OECD patterns and across the develop for ing regions. Data are

Russia only start

; those

post-Soviet for China

in 1979. Data the Czech Republic,

1

for

3 0.036-0.052 4 D 0.052-0.074

and the post Slovakia, states begin Yugoslav with the formation of those

states

in the

Least

inequality

Germany Netherlands Poland

0.002510 0.002604 0.002916 0.004644 0.005988 0.006639 0.007344 0.007539 0.008529 0.009058 0.009170 0.009634 0.010957 0.011076 0.011777 0.012012

Luxembourg Hong Kong Hungary Slovakia Malta Britain Taiwan Slovenia France Austria

0.013181 0.013624 0.014684 0.015137 0.015548 0.015610 0.015858 0.016014 0.016278 0.017799

Cape Verde Latvia Cuba Sweden Czech Republic Denmark Seychelles Romania Macau Norway Australia Finland

dalus

Winter

i^^^^^^^^B^^*. ^^^^^^^^^^^^^^^B| ^HF ^^

5 I 0.074-0.099 i^^fl^^^Hr 6 H 0.099 -0.893, most inequality ? No data available ISB^^Kr

1990s.

China

l8 D

(Theil Inequality <=O.Ol8,

Canada Bulgaria Italy Algeria Croatia Nicaragua Afghanistan New Zealand Ireland Belgium United States Mexico Iceland Bangladesh South Korea Ethiopia Egypt Bosnia & Herz. Iraq Japan Namibia Moldavia Spain Malaysia Ukraine Greece

2002

0.018428 0.019515 0.019734 0.020126 0.020982 0.022988 0.023208 0.025219 0.025589 0.025629 0.025646 0.027479 0.028083 0.028631 0.028825 0.029650 0.029919 0.030493 0.030966 0.031031 0.031425 0.031845 0.033788 0.034413 0.034697 0.035561

^^^^^^^v ^^^^^^?

^^^^^^^?^

^#^?ff|5f Gambia Colombia Burkina Faso

Azerbaijan Costa Rica Portugal Nigeria Libya Turkey Iran Burma Senegal Madagascar Cyprus Macedonia Israel

Fiji Ecuador North Yemen Pakistan Uruguay Argentina Sudan Somalia El Salvador Venezuela

0.037424 0.037912 0.038420 0.038456 0.038469 0.039110 0.040251 0.040600 0.040697 0.040782 0.041626 0.042025 0.042821 0.043173 0.043215 0.044112 0.045027 0.045361 0.045504 0.045698 0.045894 0.048402 0.048610 0.048952 0.051577 0.051952

~*** s"^^

*&t--

A perfect

' "5*s=^^_

crime

Most

4 Haiti Zimbabwe Botswana Sri Lanka

0.052814 0.054305 0.055032 0.056491 0.058459 0.058507 0.058517 0.058699 0.061265 0.061392 0.061593 0.061596 0.062061

Singapore Chile Russia South Africa Zambia Tonga Neth. Antilles Philippines Suriname Panama Barbados Ivory Coast Syria Cen. African Bolivia Nepal Am. Samoa U. Arab Emir. Mauritius Eritrea Albania

Rep.

0.062894 0.063116 0.064147 0.065215 0.066105 0.066374 0.068082 0.069093 0.069918 0.071762 0.072629 0.073632

Benin India Yugoslavia Brazil Rwanda Domincan Tanzania Lithuania Tunisia Burundi

Rep.

Jordan Peru Indonesia Kyrgystan Liberia Morocco Kenya Togo Papua N. Guinea Honduras Eq. Guinea Mauritania Thailand Bhutan Malawi Guatemala Bahamas

inequality

Lesotho Belize Gabon Swaziland Armenia South Yemen

0.074386 0.075959 0.076278 0.077607 0.078161 0.079278 0.079455 0.080079 0.080430 0.082687 0.082940 0.082944 0.084083 0.085131 0.085607 0.085982 0.086183 0.086519 0.086716 0.086866

0.105516 0.105935 0.106911 0.106922 0.108136 0.109596 0.109652 0.113724 0.118893 0.121705 0.129300 0.136999 0.145823 0.149679 0.184693 0.188703 0.194245 0.201428 0.206362 0.230317 0.253880 0.312738 0.403546 0.404105 0.442336 0.892605

Uganda Oman Ghana Puerto Rico Cameroon Congo Trinidad

& Tobago Mozambique Saudi Arabia Niger St. Vincent

& Gren.

Angola Cambodia Kuwait Sierra Leone

0.089228 0.092261 0.094014 0.095370 0.095639 0.097286 0.098721

Jamaica Bahrain Qatar Mongolia Paraguay

D

dalus

Winter

2002

19

Figure 2 A regression income and

James K.

Galbraith on inequality

on of inequality time, UTIP data a reflects regression

set. The

sloping Kuznets as a as well global

downward relation toward

drift

sharply higher inequali time. (Color scales have

ty over a similar

but are not gradient to previous graphs.)

matched

o.?

&*

Year 63

65

67

69

71

73

75

77

79

81

83

85

87

89

91

93

97

95

O.OO -0.05 11

-O.lO

I ?

-0.201

UUilU

-o.30?J? -0.35

IMI|||

-0.40

Figure Panel

5 estimates

of the world

of rising in for coun controlling effects the ef and try-specific in per capita re fect of changes of al income levels. The method

wide

time pattern

equality,

panel lation

estimates

permits of a year-to-year

in changing

20 D

dalus

Winter

2002

inequality.

calcu pattern

A perfect crime

1981 to 1987 Change -42.4 to -1.09 -1.09 to 0.65

?

to 3.42 3.42 to 9.14 9.14 to 24.74 [~] N.A. 0.65

S

/

s (above) in the Age of Debt. in inequality from 1981 Changes to 1987. Dark gray indicates the

Figure 4 (below)

Figure

Inequality

in largest increases (notably Latin America and among oil at this time of col producers ;blue indi lapsing oil prices) cates declines ; light gray is "neutral."

Almost

the only

cas

ed from

Patterns

of rising inequality the age of globalization. in inequality from Changes

the global financial sys tem (China, India, Iran). Greece and Turkey showed very large increases their con following over in the frontation Cyprus 1970s may

;declines be a return

1988 to 1994. The rise is extreme. The only

in the

including policy the end of military

in Russia significant

is region of declining inequality in the boom countries of South - more east Asia evidence of

1980s to normality, in Greece after

the Kuznets

rule.

effect.

es of in this declining inequality are in countries insulat period

n ?

1.77to 5.5 5-5 to 11.82 %. 11.82 to 78.83 N.A.

in

* Wr (f'

D

dalus

Winter

2002

21

James K. Galbraith inequality

the UTIP data also permit us to detect in changes of inequality, global patterns to take a fresh look at the New World

of the oil slump that followed. During rises most that time, inequality rapidly cone of Latin America in the Southern

Order.

East. Second, and in parts of the Middle there was the collapse (in part induced, as in and Poland, by intract Yugoslavia able debts) of the communist world; the focus of ris these countries become

This result.

exercise For when

lated, we find

a

produces disquieting the global trend is iso that in the last two de

cades, inequality has increased through out the world in a pattern that cuts across the effect of national income

in the late 1980s. By the ing inequalities as mid-1990s, figure 4 reveals, almost the countries with declining inequality only were the countries of Southern booming Asia (even if the crash after 1997 almost

the decades that hap changes. During pen to coincide with the rise of neoliber of al ideology, with the breakdown and with the end sovereignties, in the global debt of Keynesian policies rose crisis of the early 1980s, inequality curve In effect, the Kuznets worldwide. national

to income shifted inequality the upward slope This finding upward. of the plane in figure 2 points to influ ences on of a global order. inequality that there is a common The finding global upward trend in inequality pro for one of two vides strong evidence relating

raised inequality independ universally ent of income changes but for what itmay be that reason? Alternatively, economic alone do not national policies and cannot entirely control national pay and that there is a common, structures, global

element

rested their hopes fifty years ago. Keynes Even so, the figures do not fully capture

(though they do reflect) the deepening ex whose dissolution of nation-states, treme cases lead to war, as in Bosnia, the eastern Croatia, Kosovo, Macedonia,

is that na One possibility propositions. almost have tional economic policies

pernicious,

some of them in another surely took direction). - as as any This pattern shows clearly one indicator to do the ser might hope ial failure of the global development to permit the build process as awhole democracies advanced of industrial ing states on which Kuznets and welfare and

Aceh, southern Co Congo, Chechnya, even all of them and lombia, Chiapas of their the among parent poorest parts states. But as inequality deepens, more of this is surely on the way.

in the global passed out of fashion, fallen into have generally two camps :those who believe that redis on the one tribution fosters growth, hand, and those who believe on the con

Oince

economy.

In the latter case we have

to ask, what some Is there, perhaps, is that element? laissez the of process global thing about faire itself

that created

this outcome

?

if so, is it an inherent feature of Or is it only an artifact "globalization?" of the particular policies under which

And

the global market recent years?

has been

liberalized

3 and 4 illustrate the regional of patterns increasing inequality during two key episodes. First, there was the early 1980s, the years of debt crisis and Figures

22 D

dalus

Winter

2002

Kuznets

economists

in

is a price trary that rising inequality These worth paying for development. statements national about are, however, UTIP with the data conditions. Working a set to isolate different set, it is possible in of factors :those factors that co-evolve the world

economy through of the movement

pendently income. The

annual

pattern

time, inde of national of these

time effects, presented in figure 5, gives us an essential so clue far lacking: the at which

precise turning point in inequality element and started to rise.6

ceased

Nor

the global

As figure 5 shows, the common global in pay inequalities declines element slightly through the late 1970s. It then turns around in 1981 -1982, just as Ron

the of real interest rates brought latter from near o to 5 percent or higher for completely riskless assets and much higher for most countries with currencies. The result was depreciating a to precipitate global debt crisis in the course of which many poorer nations were forced first to cut imports and capi climate

to

fare policies. The years since 1980 have test of the second thus seen an empirical an of view: system point extraordinary, atic increase in inequality. It has not followed by any increase in the glo bal rate of economic expansion. For a cause of worldwide rising in one must to events look that equality, been

not before.

the period after 1980, but of trade will not do. Growth

Worldwide

trade grew very rapidly " the devel through period of stabilizing (the Mexican term) that began opment" in 1945 and ended in the 1970s ; it is not a af peculiar feature of the environment ter

1980.

6 Our fects

technique data

panel

here

is a two-way which

estimation,

fixed entails

ef

estimates

common

course

economy income

over

ed

changes. in figure 5

of of

the

time

dummies

reveal

in the global for controlling

inequality

the years, The time

effects

are present

1970s, long before the personal comput er revolution. And after 1980, inequality rises more sharply in poorer countries, where of course new technologies a spread the least. In country like Fin land, a leader

in Internet penetration, rose at all. inequality hardly can a sequence of events What explain that affects an almost universal spec trum of poorer countries after 1980, ex cluding only India and China in that de eco cade and a handful of the booming nomies of South Asia in the 1990s? The evidence of timing points toward the effect of rising real interest rates and the debt crisis. For this, the stage was set in 1973 of the Bretton by the dissolution Woods

framework

of fixed-but-adjust rates and international

able exchange

of capital flow. As figure 5 supervision framework shows, the collapse ofthat ended a period of relatively stable structures. and stable pay growth a short There then followed period of the oil and commodities

boom, with commer fueled declining inequality by cial debt. But this was unsustainable, and it came to a crashing end in the worldwide financial shock that was ini in 1980 -1981 by the United States, as the U.S. Federal Reserve interest rates past 20 pushed nominal tiated

cre

for each country and variables ating "dummy" The coefficient estimates year in the sample. on the dummies the pat then reveal country tern of national while the coeffi institutions, cient

A perfect cnme

ty in theUnited States beginning in the

ald Reagan took office in the United States. At this time, a shift in the global

characterize

technological accelerating change explain the pattern. The story case is often hinted at for the American that the rapid spread of computers after "skill-biased 1980 made technological a change" driving force behind rising But the UTIP (and all pay differentials. data other) clearly show rising inequali

declining

and then were pressured tal spending, abandon trade and wel long-standing

can

the

percent. The rise in interest rates pro cuts in duced dramatic and continuing with results for the imports, devastating of poorer coun prospects development tries. Many

of them have never

recov

ered.

D

dalus

Winter

2002

23

James K. Galbraith on

inequality

Indeed, matters were made worse by the concurrent triumphr of neoliberal ism in the United

Russia

States and the United same years. Following the rich countries

was

chitecture

created

from

ar

the wreck

control.

It is not,

that the then, by accident at a global level of neoliberalism resemble those of a coup d'?tat at a

effects

national In an

investors.

in Latin honing these policies were after America, 1989 in they applied After

before,

during, : the case

coups the general pat chart on the bot

and

tern. The tom

level.

early analysis using UTlP's set, George Purcell and I calculated

data the

6

Inequality and after Chile

in

and then

in China, in Northern notably Europe, and in the United States itself after the did growth continue and mid-1990s remain reasonable under pay inequality

banks. In age left by the commercial Fund stead, the International Monetary and then financial preached austerity, and privatization sale of deregulation state assets at fire-sale prices to foreign

Figure

Europe,

crisis ensued. Only where Everywhere, countries resisted the neo successfully - most liberal policy prescriptions

in these

Kingdom the debt crisis, the "magic of the market preached place" to the poor. No new financial

and Eastern

Asia.

averages

the change for up to five

Inequality in Chile 300

of

o o

of

years up to five years after a coup d'?tat ; the aver across twen age is calculated cases. historical ty-seven

inequality before and

s;

200 150

50

63 65 67 69

71 73 75 77 79 81 83

85 87 89 91 93 95

Year

Change in Inequality: 27 Coups D'Etat

I I I 1before 1 after 3 before 3 after 5 after 2 before 2 after 4 before 4 after Coup year

$before

24 D

dalus

Winter

2002

coups average effects of twenty-seven of pay in d'?tat on our measurements a pattern of equality. We found striking After consistency. rising four and five years before the coup, inequality would decline sharply in the two years immedi In the year of the coup ately beforehand. in inequality would itself, the decline stop. And in the five repressive years that followed from (coups, as distinct are almost revolutions, invariably right occur sys rising inequality would wing), in each year, until overall tematically stood far higher than in the inequality before the coup. Figure 6 presents period case of Chile our data for the canonical and the curve that emerges from averag cases. ing effects over the twenty-seven Viewed

from a global perspective, the of time effects observed world

pattern wide after 1975 strongly resembles this curve. Global characteristic inequality

fell in the late 1970s. In those years, poor countries had the benefit of low interest

trade as such In sum, it is not increasing we that should fear. Nor is technology the culprit. To focus on "globalization" as such misstates the issue. The problem is a process of integration carried out since at least 1980 under circumstances in which of unsustainable finance, has flowed upwards from the to the rich, and mainly to poor countries strata of the richest the upper financial

wealth

countries.

In the course toward

tolerable

of these events, progress levels of inequality and

sustainable

virtually development Neocolonial patterns of center

stopped. and of debt dependence, periphery were but with reestablished, peonage, out the slightest assumption of responsi for the fate of bility by the rich countries the poor. It has been,

itwould appear, a perfect crime. And while statistical forensics can a small role in this out, play pointing no mechanism to reverse the policy

rates and easy credit,

and high commod for oil. Indeed, in ity prices, especially the 1970s, the UTIP data shows that it was the lower-income in the workers

still less any that might repair the countries have damage. The developed the pretense abandoned of attempting to foster development in the world at

countries who made the largest in But in 1980 -1981, the age of pay. gains low interest rates and high commodity

to substitute the rheto large, preferring ric of ungoverned for the hard markets work of stabilizing The prog regulation. nosis is grim :a descent into apathy, disaster, and despair, disease, ecological wars of separatism and survival in many

poorer

prices ended. - a In 1982, the repression took hold to be sure, but not financial repression, less real for having taken that form. And the debt crisis was not accompa nied by overt violence coups are, in in their overt often limited deed, very violence the effects were soon felt

while

and with a savage intensity worldwide, that has continued for two decades.

A perfect crime

exists,

of the poorest parts of the world. of course, the wise spirits of Unless, Kuznets and Keynes can be summoned back to life, to deal more constructively with the appalling disorder of the past twenty years.

D

dalus

Winter

2002

25