For Exceptional Proff 2

2 3 Is Globalizatton Reducing Poverty and Inequalityl Robert Hunter Wade

The neoliberal argument says that t}re distribution of income between all the world's people has become more equal over the past two decades and the number ofpeople living in extreme poverty has fallen, for the first time in more than a century and a half. It says that these progressive trends are due in large Part to the rising density of economic integration between countries, which has made for rising efficiency of resource use worldwide as countries and regions specialize in line with their comparative advantage. IIence the combination of the "dollar-Wall Street" economic regime in place since the breakdown ofthe Bretton Woods regime in the early 1970s, and the globalizing direction of change in the wodd economy since then, serves the great majority of the world's people well. The core solution for lagging regions, Africa above all, is freer domestic and international trade and more open financial markets, leading to deeper integration into the world economy.

Evidence from the current long wave of globalization thus confirms neoliberal economic theory * more open economies are more ProsPerous, economies that liberal- ize more experience a faster rate of progress, and people who resist further economic Iiberalization must be acting out ofvested or "rent-seeking" interests. The world econ- omy is an open system in the sense that country mobility up the income/wealth hier- archy is unconstrained by the structure. The hierarchy is in the process of being flanened, the North-South, core-periphery rich country-poor country divide is being eroded away as globalization proceeds. The same evidence also validates the rationale ofthe World Trade Organization (WTO), the World Bank, tie International Monetary Fund (IMF) and other multilateral economic organizations as agents for creating a global "level playing" field undistorted by state-imposed restrictions on markets. This line of argument is championed by the more powerful of the centers of "thinking for the world" that influence international policy making, including the intergovernmental

Original publication details: Robert Hunter Wade, ftom "Is Globalization Reducing Poverty and Irequality)" Wo tt Dnel|qment,32,4,2l04,pp.567-8,571,572,574-5,576,577-8,579 40,581. Reprinted with permission of Elsevier.

Tbe Globdlizdtiox Readrl, Fourth Edition. Edited by Frank J. Irchner and Iohn Boli. Editorial material and organization O 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

188 Globalization and, the Wo d. Econowy

organizations such as dre World Bant, the IMF and the WTO, also ttle US and UK Treasuries, ard opinion-shaping media s uch x Tl.te Financinl Tinaes and. The Econornist.

The standard Left assumption, in contrast, is that the rich and powerfi,rl countries and classes have litde interest in greater equity. Consistent with this view, the "antiglobalization" (more accurately, "anti-neoliberal") argument asserts that world poverty and inequality have been rising, not falling, due to forces unleashed by the same globalization. The line of solution is some degree of tightening of public policy limits on the operation of market forces; though the ,,anti-neoliberal,, camp. embraces a much wider range ofsolutions than the liberal camp.

The debate tends to be conducted by each side as ifits case was overwhelming, and only an intellectually deficient or dishonest person could see merit in the oiher's case. For example, MartinWolf of The Financial Titnes claints that the ..anti-globalization" argument is "the big lie." Iftranslated into public policy it would cause more poverty and inequality while pretending to do the opposite. [...]

The Regional Collage

The growth rate ofworld GDP, measured in US dollars and at current exchange rates, fell sharply from around 5.5% in 1970-80 to 2.3o/o rn 1980-90 to l l% in f 990 2000. This is bad news, environmental considerations aside. But it still grew a little faster than world population over the past two decadesl and the (population-weighted) GDp of developing countries as a group grew a litle faster thal tllat of the high-income countries. On the other hand, regional variation within the global South is large. [...] During 1960-90 the per capita incomes of sub-Saharan Africa, Latin America, and West Asia alrd North Afiica fell as a fraction of the core,sl South Asia,s remained more or less constant; East Asia's (minus China) rose sharply; China's also rose sharply but from a very low base. The most stuiking feature is not the trends but the size of the gaps, testimony to the failure of "catch-up." Even success-story East Asia has an average income only about 13% of the core's. It is a safe bet that most development experts in 1960 would have predicted much higher percentages by 2000. [...]

How does the collage - positive world per capita growth and wide divergence of economic performance between developing regions - net out in terms of global trends in poverty and inequalityf

Poverty

The standard poverty numbers - the ones normally used in discussions about the state of t]-le world - come from the World Bank's data set. This is the source of the claims that, in the words ofPresident lames Wolfensohn, "Over the past 20 years the number ofpeople living on less than $t a day has fallen by 200 million, after rising steadily for 200 years" and "the proportion of people worldwide living in absolute poverty has dropped steadily in recent decades, from 29yo n \990 to a record low of 2\o/o in 1998." The opening sentence ofthe Bank's World. Dewlopntent Irudicators 200) says, "Of the world's 6 billion people 1.2 billion live on less than $l a day,', the same number in 1987 and 1998.

l u":

Fir Po' l s { bri Re 2 0

I suI qu

Pu.

ha,

ur( twl po cor po suI dif Th Par Th evi mi

Po M C

l.rar

in

cor n.f shr

nu th: po

Iivi CI ilrc

Is Gkbalization Reducing Poterty? I89

No ifs or buts. I now show that the Bank's figures contain a large margin oferror, and tie errors p/0, ably flatter :Jte resalt in one direction. [...]

There are several reasons to exPect a large margin oferror, regardless ofdirection. First, the poverty headcount is very sensitive to the Precise level ofthe international poverty lines. This is because the shape ofincome distribution near the poverty line is such that, in most developing countries, a given percentage change in the line brings a similar or larger percentage change in the number of people below it. Recelrt research on China suggests that a 10% increase in the line brings a roughly 207o increase in the poverty headcount.

Second, the poverty headcount is very sensitive to the reliability of household surveys of income and expenditure. The available surveys are of widely varying qualiry and many do not follow a standard temPlate. Some sources of error are well known, such as the exclusion of most of the benefits that people receive from publicly provided goods and services. [...]

Third, China and India, the two rnost important countries for the overall trend, have PP? adjusted income figures that contain an even bigger component of guess work than for most other significant countries. The main sources ofPPP income fig- ures (the Penn World Tables and the International Comparison ?roject) are based on two large-scale international price benchmarking exercises for calculating purchasing power parity exchange rates, one in 1985 in 60 countries, the other in 1993 in 110 countries. The government of China declined to ParticiPate in both. The purchasir.rg power parity exchange rate for China is based on guestimates from small, al /roc price surveys in a few cities, adjusted by rules of thumb to take account of the huge price differences between urban and rural areas and between eastern and western regions. The government oflndia declined to participate in the 1993 exercise The price com parisons for India are extrapolations from 1985 qualified by later ad hoc prtce surveys. The lack ofreiiable price comparisons for China and India hence the lack ofreliable evidence on the purchasing power of incomes across their distributions - cornpro mises any statement about levels and trends in world poverty.

Fourth, the often-cited comparison between 1980 and 1998 - 1.4 billion in extreme poverty in 1980, 1.2 billion in f998 - is not valid. The Bar.* introduced a new methodology in the late 1990s which makes the figures noncomparable. The Bank has recalculated the poverty numbers with the new method only back to 1987. [...]

We can be fairly sure that the Bank's poverty headcount has a large margir.r oferror in allyears, in the sense that it may be significandy different from the headcount that would result from the use of PPP conversion factors based more closely on the real costs ofliving ofthe poor (defined in terms of income needed to buy enough calories, micronutrients and other necessities in order not to be poor). By the same token we should question the Bank's confidence that the trend is downward.

Wc do not know for sure how the late 1990s revision ofthe med.rod and *re PPP numbers alters the poverty headcount in any one year and the trend. But it is likely that the Bank's numbers substantially underestimate the true numbers ofthe world's population living in extrerne PoYerty) and make the trend look brighter.

On the other hand, it is quite plausibl e *\at tlle prlplrtilz of the world's population living in extreme poverty has fallen over the past 20 years or so. For all the problems wid.r Chinese and Indian income figures we know enough about trends in other variables -

including life expectancy, heights, and other nonincome measures - to be confident that

I90 Globalization a.nd tbe World. Econom.y

their poverty headcounts have indeed dropped drarnatically over the past 20 years. Ifit is the case (as some experts claim) that household surveys are more likely to miss the rich tha.rr the poor, their results mvy |yerttdte th.' proportion of the population in poverty. The magnitude of world population ir.rcrease over the past 20 years is so large that the Bank's poverty numbers would have to b e buge underesnmates for the world poverty rate not to have fallen. Any more precise statement about the absolute number oftie world's people living in extreme poverty and the change over time currendy rests on quicksand.

Inequality

The world poverty headcount could move in one direction while world inequality moved in the other. The neoliberal argument says drat they have both dropped. But in the past several years world income distribution has become a hot topic of debate in international economics and in sociology (much hotter than trends in world poverty). Disagreements about the overall inequality trend should not be surprising given the variation in regional economic performance - different ways of measuring emphasize different parts ofthe collage.

The only valid short answer to the question, "S4rat is the trend of world income distributionf" is, "It depends on which combination out of many plausible combina- tions ofmeasures and countries we choose." [...]

Proposition L. World. incorne distribation bas becotne rapid.ly wore unequal, when 'incornes are measwred at warket exchange rates and, expressed in US d.ollars.

No one disputes tlfs. The dispute is about what the figures rnean. Most economists say that excl.range-rate-based hcome measures are irrelevant. GDP incomes should always be adjusted by PPP exchange rates to take account of differences in purchasing power, they say This makes a big difference to the size of the gap between rich and poor. As noted, the PPP adjustment is made by computing the relative prices for ar average bun- dle ofgoods and services in different countries. The PPP adjustment substantially raises the relative hcome ofpoor countries. India's PPP GD! for example, is about four times its market exchange rate GDP The PPP adjustment thus makes world income distribu- tion look much more equal than the distribution of market-exchange rate ir.rcomes. [... ]

Proposition 2. World PPP-incowe polarization bas increased, witb polarizntion nrqsurrd as richest r0 po0r6't dc(il(.

The broad result is hardly surprising: the top 10% comprises almost entirely people living in the core coultries ofNorth America, western Europe, and Japan, where incomes have grown over the past 2G-30 years, while a large chunt ofthe bonom I0% is comprised of African countries where ilrcomes have stagnated or fallen. According to one study, the trend ofrichest to poorest decile goes like this: 1970-92,1980-109, 1990-104, 1999-104. Another study firrds a jump in the ratio of25% over 1988-93. The change is made up of the top decile pulling sharply up from the median and the botrom decile falling away from the median. The polarizing trend would be much sharper with the top I% rather than the top decile.

Propos hnst ll coelrtcit

Of cou relative of obse growth the like their tr perforn

Propos falling ,

Tl.fs is probler also ex lnequ&t combin,

With and ine, if the fi1 rose fio in only

We cr China a cauttou even m dreir nc lations r lysts ha governr authorit t . . . l

Over 6-8% tt that \vo

W e h income oversaar offsets r gro\\.th trvo col than be regions

: : :

t r :

:r'lc =

lori

l\_e

i o f d t , n.r

cm P E I

Is Globalization Redwcing Powrty? 191

Proposition 3. Beween-colnmy worhl PPP-incowe inequnlity bas increa.sed' since fit Ieast 1980, using per cttpitlz GDPS, eqwal cowntry weights (China = Ugandu), and o cofficient like tbe Gini for tbe wbole rlisnibution.

Of course, we rvould not weight countries equally if we were interested simply in relative Well-being. But we would weight them equally treat each country as a unit of observation, analogous to a laboratory test observation - if we were interested in gro\\,th theory and the growth impacts ofpublic policies, resource endowments, and the like. We might, for example, arrange (unweighted) coulrtries by the openness of their trade regime and see whether more open countries have better econornic performance. [... ]

Proposition 4. Between-czantry worLl PPP'incowe inequality hz's been clnsto'nt 0r falling since nround. 1980, with cowntries weigbted by population.

This is the result that the neoliberal arBument celebrates. There are just two problems. First, exclude China and even this measure shows a widenir.tg since 1980; also exclude India and thc widening is pronounced. Therefore, falling income ineqaelity is ru,t a. general featwre of the world economy, even asing tbe ruost fatorable corubination of m.easures. 1...1

With 38% ofthe v'orld's population, China and India shape world trends in poverty and inequality. They have grown very fast over the past decade (India) or tr, 'o (China), if the figures are taken at face value. China's average purchasing porver pariry income rose from 0.3 ofthe world average in 1990 to 0.45 in 1998, or 15 percentage points in only eight years.

We can be sure that world poverty and inequality are less than they rvould be had China and India grown more slowly. About ar.ty stronger conclusion we have to be cautious. First, recall that China's and India's purchasing power parity numbers are even more questionable dran those for dle average developing country, because of their nonparticipation in the international price comparisons on which the PPP calcu- lations rest. Second, China's growth in the 1990s is probably overstated. Many ana Iysts have recently been revising China's growth statistics doli'nward. \\4rereas government figures shorv annual real GD? grorvdr of 7-8% ill 1998 and 1999 one authority on Chinese statistics estimates that the economy may not have grown at all.

Over the 1990s Cl.rina's annual growth rate is more likely to have been around 6-8% than the 8-10% ofthe official statistics. This one change lowers thc probability that world interpersonal distribution has become more equal.

We have to be cautious about going from China's fast growth to falls in world income inequality not only because China's gror,th rates and income level may be overstated but also because the rise in inequality within both China and India pardy off'sets the reduction in rvorld incorne inequality that comes from their relatively fast growth ofaverage income though careful calculations ofthe relative strength ofthe two contrary effects have yet to be made. China's surging inequaliry is now greater than before the Communists won the civil u'ar in 1949, and inequality between regiolrs is probably higher than in any other sizable country. The ratio ofthe average

192 Globolization and the Worhl Econowy

income ofthe richest to poorest province (Guangdong to Guizhou) rose liom around 3.2 inI99l (current y-ran) to 4.8 in 1993, and remained at 4.8 in 1998 2001.The corresponding figure for India in the late 1990s was 4.2, the United States, 1.9. [... ]

The evidence does support the liberal argument when inequality is measured with population-weighted countries' per capita PPP-adjusted incomes, plus a measure ofaver- age inequality, tfing Chir.ra's income statistics at face value. On the other hand, polariza- tion has clear$ hcreased. Moreover, several studies that measure inequality over the whole distribution and use either cross sectional household survey data or measures of combined inequality between countries and withh countries show widening inequality since around 1980. The conclusion is that world inequality measured in plausible ways is probably rising, despite China's and India's fast growth. The conclusion is reinforced by evidence of a quite different kind. Dispersion in pay rates within manufacturing l.ras becone steadily wider since the early 1980s, having remair.red roughly constant from 1960 to the early 1980s. Meanwhile, absolute income gaps are widening fast.

Globalization

I have raised doubts about the liberal argument's claim tiat (a) the number ofpeople living in extreme poverty worldwide is currently about 1.2 billion, (b) it has fallen substar.rtially since 1980, by about 200 million, and (c) that world income inequality has fallen over tlte same period, having risen for many decades before then. I-et us consider the other end of the argument - that the allegedly positive trends in poverty and inequaliry have been driven by risir.rg integration ofpoorer countries into the world ecoromy, as seen in rising trade/GD?, foreign direct investment/GDP, and the like.

Clearly the proposition is not well supported at the world level if we agree that globalization has been rising while poverty and income inequality have not been falling. Indeed, it is striking that the pronounced convergence of economic policy toward "openness" worldwide over the past 20 years has gone with divergence of economic performance. But it might still be possible to argue that globalization explains differences between countries: that more op€n economies or ones that open faster have a better record than less open ones or ones than open more slowly.

This is what World Bank studies claim. The best known, Globalization, Gtowth and. Poyerty, disttngurshes 'hewly globalizing" countries, also called "more globalized" countries, from "nonglobalizing" countries or "less globalized" countries. It measures globalizingby cbangesin the ratio oftrade to GDP over 1977-97. Rarrking developing countries by tie amount ofchange, it calls the top drird the more globalized countries, the bottom two thirds, the less globalized countries. It finds that the former have had faster economic growth, no increase il equaliq', and faster reduction ofpoverty than the Iatter. "Thus globalization clearly can be a force for poverty reduction," it concludes.

The conclusion does not follow. First, using "change in the trade/GDP ratio" as the measure of globalization skews the results. The globalizers then include China and India, as well as countries such as Nepal, C6te d'Ivoire, Rwanda, Haiti, and Aigentina. It is quite possible that "more globalized" countries are las open than many "less globalized" countries, both in terms of trade/GDP and in terms ofthe rnagnitude of tariffs and nontariff barriers. A country with high trade/GDP and very fiee trade policy would still be categorized as "less globalized" if its increase in

trad.e/C the glol relativel lust the relativeI with mu or even

Exclu of more commo( their ecc trade. Il questron hand, ir and rela by popu

Conch

It is plar exffeme for deca Bant's 1 downwa bution, r Past twc same cor at marke adjusted power, 2 yield the weighter take out thus not ure. Finr widenin

Is Globalization Redacing Powrty? 193

trade/GDP over 1977-97 put it in the bottom two-thirds ofthe sample. Many of the globalizing countries initially had very low trade/GDP in 1977 and still had relatively low trade/GDP at the end of the period in 1997 (reflecting more than just the fact that larger economies tend to have lower ratios oftrade/GDP). To call relatively closed economies "more globalized" or 'globalizers" and to call countries with much higher ratios oftrade/GDP and much fteer trade regimes "less globalized" or even "nonglobalizers" is an audacious use oflanguage.

Excluding countries with high but not rising levels oftrade to GDP from the category of more globalized eliminates many poor countries dependent on a few natural resource commodity exports, which have had poor economic performance. The structure of their economy and the low skill endowment ofthe population make them dependent on trade. If they were included as globalized their poor economic perforrnance would question the proposition that the more globalized countries do better. On the other hand, including China and India as globalizers - despite relatively low trade/GDP and relatively protective trade regimes - guarantees that the globalizers, weighted by population, show better performance than the nonglobalizers. [...]

Conclusion

It is plausible, and important, that the proportion of the world's population living in extreme poverty has probably fallen over ttre past two decades or so, having been rising for decades before then. Beyond this we cannot be conident, because the World Bank's poverty numbers are subi€ct to a large margin of error, are probably biased downward, and probably make the trend look rosier than it really is. On income distri- bution, several studies suggest that world income inequality has been rising during the past two to three decades, and a study of manufacturing pay dispersions bumesses the same conclusion from anottter angle. The trend is sharpest when incomes are measured at market-exchange-rate incomes. This is less relevant to relative well-being than PP?- adjusted incomes, in prhciple; but it is highly relevant to state caPacity, interstate power, and the dynamics of capitalism. One combination of inequality measures does yield the conclusion tl.rat income inequality has been falling - ?P?-income per capita weighted by popu.lation, rneasured by an averaging coefficient such as the Gini. But take out China and even this measure shows widening inequality. Falling inequality is tllus not a gener6.Azed. feature of the world economy even by tlle most favorable meas- ure. Finally, whatever we conclude about income inequality, absolute income gaPs are widenins and will continue to do so for decades.