Is Globalisation Good or Bad for Poor People?
Reviews by Helen Hughes
Click here for PDF version

Imagine ThereÕs No Country: Poverty, Inequality and Growth in the Era of Globalization
Surjit Bhalla
Washington D.C., Institute for International Economics, 2002,
288 pp, US$28.00, ISBN 0 88132 348 9

Globalisation, Living Standards and Inequality: Recent Progress and Continuing Challenges,
D. Gruen, T. OÕBrien and J Lawson (eds),
Sydney, Reserve Bank of Australia and Australian Treasury, 2002

The publication of Surjit BhallaÕs Imagine ThereÕs No Country has created a furore in Washington. He argues persuasively that the World Bank has been using its monopoly of data to double the estimated extent of poverty in developing countries. Bhalla, of Oxus Research in New Delhi, is a highly respected participant in development debates. His book is published by the Washington D.C. Institute of International Economics. He shows that the target of halving developing country poverty from 1.1 billion to 650 million people by 2015 (used by James Wolfensohn to urge industrial countries to increase their aid efforts at Monterrey in March 2002), was already achieved in the year 2000. The World Bank poverty estimates were reflected in papers at the Reserve Bank-Treasury Conference on Globalisation, Living Standards and Inequality in May 2002 1 and the 2:1 ratio of the World BankÕs estimate of poverty to actual poverty is mirrored in the work of the Luxembourg Income Study, also presented at that Conference.2

Bhalla and the Reserve Bank-Treasury Conference were concerned with three issues. How many poor people are there in the world? Has inequality decreased or increased? And what is the impact of globalisation/liberalisation on poverty and income distribution? The Reserve Bank-Treasury Conference also looked at the impact of globalisation/liberalisation on income distribution in industrial countries.

Bhalla unequivocally argues that globalisation/liberalisation has led to falling poverty and inequality. The scholars the Reserve Bank-Treasury chose to address these issues gave very different answers: David Dollar weakly supported the positive influence of globalisation on the reduction of worldwide poverty and inequality in the 1990s; Robert Wade and Nancy Birdsall argued that globalisation/liberalisation increased poverty and made worldwide incomes less equal; and Timothy Smeeding claimed that globalisation/liberalisation was inversely related to income equality in industrial countries.

Poverty is necessarily a subjective concept that does not have much validity beyond a given country and time. Asking how much poverty there is worldwide is like asking how long is a piece of string. But poverty exists. It is still a terrible scourge. Evaluating the research that underlies the various measures of poverty is important in determining where poverty is located so that national policies can alleviate it.

Income distribution is determined by economic and social policies. In democratic societies, these are chosen by voters. It is thus important to establish how globalisation/liberalisation has affected income distributionÑthat is, whether national reform agendas should be accelerated, modified or even abandoned.

How many poor people are there in developing countries?

Estimates of numbers living in absolute poverty are generally based on the minimum incomeÑaround $US1 a dayÑneeded to sustain people in developing countries. These are presented as head count ratios, that is, the number and percentage of a population whose incomes are estimated to be below this level of income.

Defining poverty has a long history. Attempts to establish a poverty line based on calorie counts have been abandoned for estimates of the income needed to feed, clothe and shelter individuals and families. The World Bank developed a head count of global poverty in the 1970s from national income accounts on the basis of the minimum income needed to feed an individual.3 Purchasing power parity estimates that determined the cost of a ÔbasketÕ of goods in various countries were then becoming available and were used for international comparability. The World Bank initiated a Living Standards Measurement Study Program in developing countries to supplement official income and consumption surveys (to which it had access, but which were often not available to other researchers, including nationals of the countries carrying out the surveys). In 1991 the World Bank published new global developing world poverty estimates.4 The absolute poverty income line continued to be around $US1 a day. These estimates have been elaborated and updated to 1.1 billion poor people in developing and transitional (former Communist East European and Central Asian) countries.

Bhalla argues that the World BankÕs erroneous use of income survey data and its poverty model greatly exaggerate the number of poor people in developing and transitional countries. He agrees with those who contend that income surveys have serious problems. ÔInformalÕ income is underestimated. Consumption (expenditure) surveys, available for fewer countries, are typically also underestimated, but consumption is, nevertheless, always markedly higher than income for low income groups. Over time this is, of course, nonsense. In developing countries poor people without sufficient income starve and die. It is widely accepted that developing country income and consumption surveys have many additional problems, including sampling errors and low and erroneous response rates. It is common for respondents to regard income surveys as sources of information for the private and public use of officials and to answer accordingly. In more than 30 statesÑa substantial proportion of sub-Saharan AfricaÑgovernment, let alone censuses and surveys, do not exist. The World BankÕs Standard of Living Survey Program suffers from all these difficulties. It is also sporadic, depending on the variable quality of consultants and modestly trained and paid local staffs who often overcome very real access difficulties by filling in survey forms with information they think they ought to contain.

Bhalla supplemented survey data by national income data. He also used expenditure data from national accounts. His historical purchasing power parity (PPP) estimates are those relating to income as used by Angus Maddison, by far the best compilation of historical income and PPP series.5 Bhalla argues that the (hitherto unpublished and hence untested) World Bank consumption purchasing power parity estimates increase the poverty count egregiously by comparing them with the Maddison series. Another of BhallaÕs major innovations is the breaking down of the population from quintiles or deciles into percentiles for a better focus on individuals. Bhalla has used a US $1.50 a day poverty cut-off to account for inflation and for a more generous poverty estimate. The resulting poverty count shows poverty numbers, though not percentages, rising from 1959 to 1980 as population growth exceeded poverty reduction, but then falling sharply to 2000.

Reduction in poverty began in East Asia in the 1960s with the Asian Tigers liberalising and opening their economies to the world. The numbers of poor people increased, but the proportion of the population living in poverty fell. Malaysia, Thailand and Indonesia followed in the 1970s and China from the 1980s. South AsiaÑmainly IndiaÑalso liberalised, accelerating its pace of reform from the 1980s. Progress has been limited in Latin America. Poverty numbers have increased slightly in the Middle East (though the percentage of people living in poverty has fallen). In Sub-Saharan Africa the numbers in poverty have trebled and the percentage living in poverty has scarcely fallen. The AIDS epidemic contributes to high poverty levels.

BhallaÕs conclusions are strongly supported by the unequivocal rise of life expectancy (except in Sub-Saharan Africa) that indicates improving nutrition, housing, education and health. This is the only reliable indicator of well being for international comparisons.6 David Dollar, a leading World Bank poverty economist, led the discussion at the Reserve Bank-Treasury Conference with a paper entitled ÔGlobal Economic Integration and Global InequalityÕ.7 He avoided quoting World Bank poverty numbers directly, although he referred to several poverty studies based on World Bank data. He also emphasised the reduction of poverty since liberalisation in India and China, while pointing to increasing poverty numbers in Sub-Saharan Africa.

BhallaÕs and DollarÕs analyses and estimates, however, differ markedly in country detail. Dollar, in keeping with World Bank practice, accepts inflated official figures for ChinaÕs growth. Madison, and hence Bhalla, reflect the real advances made in China during the past 20 years, but take a more realistic, if still perhaps overly generous, view of ChinaÕs growth. But even limited liberalisation, when steadily pursued, has a major impact on poverty as the more reliable Indian data show. Some of DollarÕs analysis is reminiscent of the World BankÕs past love affairs with countries that were its Ôflavour of the monthÕ. Romania was described as having Ôthe highest growth rate in Eastern Europe and perhaps the highest in the worldÕ in the 1970s when standards of living were falling as the population was fiercely exploited by the Ceausescus.8 The World Bank is currently being romantic about Uganda. Dollar extols its performance even as child soldiers swarm on the northern borders and the Ugandan army rampages in the Congo, exacerbating poverty and creating fiscal crises at home. In Asia Dollar picked Vietnam as a model despite its failure to reform that enables communist cadres to continue to exploit the Vietnamese people at the cost of employment and social and economic progress.

How is income distributed in the world?

When Robert Wade, a Professor at the London School of Economics, argued in The Economist in April 2001 that world income distribution was becoming less equal,9 former Australian Commonwealth Statistician Ian Castles showed the weakness of WadeÕs conclusion, which was based on a paper by another World Bank economist, Branko Milanovic (later published in The Economic Journal).10 Ian Castles wrote in a letter in The Economist: ÔMr WadeÕs confidence that he has drawn upon Òprobably the most reliable dataset on world income distributionÓ is absurdly misplacedÕ. He added (in a portion of his letter not published by The Economist): ÔAll that Mr Milanovic has done is to lump together the results of hundreds of household surveys for different countries, compiled using widely different concepts, measures of income, units of observation and interview procedures. His analysis relies on household surveys alone. Far from being a virtue, this is a massive deficiency. One effect is that such important elements of national product as public spending on goods and services have been left entirely out of account. The relative size of these components differs greatly between countries and over time, and can be particularly important in the level and growth of real living standards in the poorest countriesÕ.11

Disregarding Castles, Wade continued to use the same data and arguments in ÔGlobalisation, Poverty and Income Distribution: Does the Liberal Argument Hold?Õ at the Reserve Bank-Treasury Conference. He thought the World Bank poverty estimates too low without adding new evidence for this view. MilanovicÕs and WadeÕs conclusions, that income distribution is worsening, only hold if countries rather than individuals are used as the basis of comparison. Thirty failing Sub-Saharan states thus far outweigh improvement in India and China. The title of BhallaÕs bookÑImagine ThereÕs No CountryÑexposes this critical weakness which follows the UN model of ÔdemocracyÕ that gives any five Pacific Island ÔstatesÕ greater voting weight than India, China, the United States and Russia. It is not surprising that MilanovicÕs paper was used by the organisers of the anti-WTO demonstrations in Sydney in November 2002 to claim that globalisation harms poor people,12 but it is surprising that WadeÕs data sources and methodological weaknesses were not critiqued in the Conference discussions.

Wade also uses official exchange rates rather than purchasing power parity for international comparisons. He argues that consumption should reflect the value of goods in international markets. But poor people eat little caviar or other internationally traded goods. They buy local goods at local prices. The underestimation of the purchasing power in low income countries by official exchange rates was so clearly recognised that in the early 1970s it led the World Bank and the UN Statistical Office to initiate the International Comparison of Real Gross Product Program.13 Sporadically pursued since the initial effort in the 1970s, purchasing power parity estimates provide more realistic international comparisons than official exchange rates.

Dollar was ambivalent about global trends in income distribution, quoting an array of studies using World Bank data to show that there was steep income distribution deterioration from the 1950s, and in some authorsÕ views, until the end of the 20th century. He conceded that while global income distribution had been worsening for some hundreds of years, there was, albeit weak, improvement in the 1990s. At the same time he reproduced in his paper a steep decline of inequality postulated by Sala-i-Martin for the 1990s that agrees with BhallaÕs conclusions.14

Bhalla, with his greatly improved database, argues that since the 1980s, increasing equality of income distribution has become very marked, again as a result of progress in Asia. He concludes: ÔNot only has inequality not increased. It has actually fallen, and by the end of 2000 was at its lowest level in 50 years. Moreover, by the end of this decade, the level of inequality is likely to be equal to that prevailing 100 years ago.Õ15

What impact has globalisation had on poverty and income distribution?

The Reserve BankÕs and the TreasuryÕs principal interest in holding the Globalisation Conference was presumably to draw positive linkages between globalisation, liberalisation, growth and improved standards of living. David Dollar, with Paul Collier, had been the principal author of the World BankÕs widely promoted Globalisation, Growth and Poverty which came to this conclusion,16 and author (with Aart Kraay) of two papers with positive messages about the impact of growth and trade on poverty.17 DollarÕs Conference paper concluded that globalisation led to a (weak) improvement in income distribution in the 1990s. How seriously can these results be taken? They are based on the same 100 income surveys of dubious accuracy used by Milanovic. The econometric models employed make further heroic assumptions about data. Elegant techniques often make up for a lack of sturdy information.18

The weaknesses of DollarÕs work enable Wade to find DollarÕs models unconvincing. Wade persists in believing that globalisation, because of its reliance on economic liberalisation, does not lead to poverty reduction or greater income equality. His Marxist persuasion is evident in his argument that government direction, following the ÔJapanese modelÕ, was the source of growth in East Asia.19 The many studies of the conflicts between liberalisation and dirigisme in East Asia, and the high costs of government intervention in the 1970s and 1980s that led to a decade of stagnation in Japan and to the East Asian 1997 breakdown, have done nothing to shake his belief. Nancy Birdsall, formerly a World Bank staff member and now President of the Washington Center for Global Development, an organisation closely linked to NGOs and their development agendas, supported WadeÕs high poverty and growing inequality views. She revived some very old chestnuts by claiming that commodity exporting developing countries have been injured by globalisation. She followed a World Bank study by Easterly, Islam and Stiglitz that showed that export price fluctuations correlated with GDP growth fluctuations,20 but did not provide evidence to counter several studies that showed that export price fluctuations did not lower long term growth. BirdsallÕs underlying argument, reflecting the Prebisch thesis that led to import substitution strategies that impoverished many developing countries for years, was that falling relative prices of commodities harm developing country commodity exporters. But rapidly rising productivity is leading to falling prices for all products worldwide. There is no consensus on what has happened to terms of trade for commodity exporters, many of whom are also major grain importers and hence have benefited from improving terms of trade. Declining barter terms of trade need not, in any case, result in low growth. East Asian countries worked hard to reduce their export prices to increase their exports. They deliberately worsened their barter terms of trade. The critical export indicators are the income terms of trade, and these have improved for those countries, no matter what they export, that have taken advantage of globalisation. BirdsallÕs sample consists of countries too incompetent or ideologically unwilling to diversify and increase their exports.

Bhalla ascribes the remarkable decline in global poverty to economic growth following liberal, outward oriented policies and the subsequent increase in trade, capital flows and more productive employmentÑthat is, globalisation. He concludes: ÔThere is no welfare indicator for which the world economy has not done better in the past 20 years. And poor people do better, much better than the average with globalisation . . . No matter what statistic is used, the revealed truth is that we have just witnessed the 20 best years in the history of poor peopleÕ.21

Income distribution within industrial countries

Poverty is much less severe in industrial than in developing countries, so that in comparison to developing countries it is relative. But in the context of the resources available to deal with it, remaining pockets of poverty are even more reprehensible. Policies that alleviate poverty are essential to industrial countriesÕ social stability. Democratic countries, however, disagree about what income distribution is appropriate. Some countries are apparently willing to trade off the level of income and income growth for more equal income distribution. Others prefer higher income levels and more rapid growth so that even low income groups are better off, even if this means less equal income distribution.

Two studies of economic freedom and income levels, the Fraser InstituteÕs Economic Freedom of the World and the Heritage Foundation/Wall Street JournalÕs Index of Economic Freedom, suggest that internal and external openness stimulates growth, and show that economic freedom is correlated with levels of incomes.22 Timothy Smeeding, the Director of the Luxembourg Income Study, disagrees. The Luxembourg Income Study collects household income surveys across 25 member countries. Smeeding claimed at the Reserve Bank-Treasury Conference that in industrial countries inequality of income distributionÑmeasured by the gap between low and high income individuals in each countryÑwas inversely related to liberalisation. Relatively liberal Economic Freedom Index economiesÑthe United States, the United Kingdom, Canada and AustraliaÑhad the highest income gaps according to Smeeding. The countries with the least gap between high and low income individuals had relatively low Economic Freedom Index rankings (in brackets): they were Sweden (19), Finland (11), Norway (24), Luxembourg (13), the Czech Republic (38), Netherlands (8) and Denmark (13); three of the countries with the largest gaps between high and low income individuals had high Economic Freedom Index rankings: they were the United States (3), United Kingdom (4), and Australia (8).23

The Luxembourg Income Study is, like World Bank poverty estimates, income survey based and thus an unreliable source of poverty levels and inequality trends. Its income survey data suffer from all the shortcomings identified by Castles and Bhalla. Peter Saunders and Kayoko Tsumori have noted additional problems.24 There is no account of the movement of people in and out of poverty over time. Some of those recorded, correctly, as Ôlow incomeÕ, are students working low skilled jobs part time while studying, but who may go on to become medium or high income earners. Not all of those who fall into poverty through unemployment or other misfortunes remain poor. Two further factors lead to misleading international comparisons. First, countries measure social security contributions in different ways. European social insurance systems mean that the gap between wage earners and those receiving no market income as measured is reduced. Second, the high welfare, high tax and low economic freedom states have a compressed income structure, albeit at a lower level of per capita income than the liberal United States, Canada, United Kingdom and Australia. That is their democratic choice. But a larger gap between higher and lower economies does not mean ÔworseÕ income distribution or greater poverty.

By ignoring income levels, the Luxembourg Income Study also abstracts from actual living standards. The Czech Republic has low inequality according to Smeeding, but the bottom 5% or 10% of its people have lower living standards than the bottom 5% to 10% in the United States. According to Maddison, in 1999 per capita income in the United States was (in purchasing power parity terms) US$27,300 and in Czechoslovakia it was US$8,600. The rate of income growth is also an important factor in living standards. Low income people in Scandinavian countries are losing out on income growth as their countries grow very slowly, compared to faster growth in the United States, Canada, the United Kingdom and Australia.

Smeeding argues that in industrial countries Ôincome inequality (was) falling in the 1960s (few comparable observations) and early 1970s, but then rising from the late 1970s and 1980s into the 1990s . . . though the rate of increase in inequality slowed markedly in the United States and United Kingdom in the late 1990sÕ.25 The evidence, however, suggests that real poverty (taking into account current living standards) in Australia, Canada, the United States and the United Kingdom was lower than the Luxembourg Income Study claims and that income distribution was not becoming more unequal.

Australian poverty levels and income equality trends have been carefully examined by Saunders and Tsumori. They concluded that a ÔguessestimateÕ of 5% of the population in poverty was more likely to be realistic than the Luxembourg Income Study estimates of 19.2% (below half the mean income) or 14.3% (below half the median income) in 1994. They have noted that the Australian Bureau of Statistics has made it clear that the Australian income surveys (on which the Luxembourg Income Study Australian estimates are based) are faulty in several respects and so inaccurate for the bottom 10% of the population as to be unusable. Saunders and Tsumori also concluded that there is no basis for the view that income equality deteriorated in Australia in the 1990s. The evidence for Canada, where a recent direct countrywide estimate of poverty is available, as well as the United States and the United Kingdom, similarly disprove the conclusions of the Luxembourg Income Study.26

Conclusion

Ken Henry, in giving the closing address at the Reserve Bank-Treasury Conference, sensibly disregarded the dubious data and econometrics used to question the impact of globalisation on growth and poverty, focusing on the core issue of how reforms in the direction of openness, even if small and partial, helped countries to catch up with the living standards of the democratic West. Following basic principles of comparative advantage, he urged continuing reform in developing and industrial countries to free up trade for further advances in living standards.

Exaggerating poverty levels nationally and globally does create a culture of guilt, but it is doubtful if it increases charitable and aid contributions. ÔCrying wolfÕ takes attention away from really poor people and results in donor fatigue.

Endnotes

1 D. Gruen, T. OÕBrien and J Lawson (eds), Globalisation, Living Standards and Inequality: Recent Progress and Continuing Challenges (Sydney: Reserve Bank of Australia and Australian Treasury, 2002).

2 T. M. Smeeding, ÔGlobalisation, Inequality and the Rich Countries of the G-20: Evidence from the Luxembourg Income Study (LIS)Õ, in D. Gruen, T. OÕBrien and J Lawson (eds), Globalisation, Living Standards and Inequality.

3 Hollis Chenery et al (eds), Redistribution with Growth (New York: Oxford University Press, 1974).

4 M. Ravallion, G. Datt and M. van de Walle, World Development Report 1990: Poverty and Development (New York, Oxford University Press, 1991).

5 A. Maddison, The World Economy: A Millennial Perspective (Paris, OECD Development Centre, 1991) was referred to peripherally in P. Harper, P. McCarthy, L. Pietsch and K. Woolford, ÔImproving our Knowledge and Analysis of Changes in Poverty and Inequality: The International Statistical ArchitectureÕ at the Reserve Bank-Treasury Conference, but was not used by any of the principal authors.

6 I. Castles, ÔThe Mismeasurement of Nations: A Review Essay on United Nations Development Programme ÒHuman Development Report 1998ÓÕ, Population and Development Review 24:4 (1998) led to some improvements in the Human Development Index, but there is still has a long way to go before the Index can be taken seriously.

7 Gruen, OÕBrien and Lawson, Globalisation, Living Standards and Inequality.

8 World Bank, Romania: The Industrialisation of an Agrarian Economy under Socialist Planning (Washington D.C., The World Bank 1979).

9 R. H. Wade, ÔWinners and LosersÕ, The Economist (28 April 2001), 73-75.

10 B. Milanovic, ÔTrue World Income Distribution, 1998 and 1993: First Calculations Based on Household Surveys Alone, Economic Journal 112:476 (2002), 51-92.

11 Quoted from the full letter sent to The Economist (26 May-1 June 2001, p.16), personal communication from Ian Castles (4 November 2002).

12 Letter from Tom Bramble, West End, Queensland, The Australian (19 November 2002).

13 I. B. Kravis, A. W. Heston and R. Summers, World Product and Income: International Comparisons of Real Gross Product (Baltimore: Johns Hopkins University Press, 1982), was the first of a series of reports which struggled, and still struggles, against bureaucratic neglect to provide a basis for international comparisons of incomes and expenditures.

14 X. Sala-i-Martin, ÔThe Disturbing ÒRiseÓ of Global Income InequalityÕ, NBER Working Paper No. 8904 (2002).

15 Bhalla, Imagine ThereÕs No Country, p. 1.

16 World Bank, Globalisation, Growth, and Poverty (Washington D.C. and New York: World Bank and Oxford University Press, 2002).

17 D. Dollar and A. Kraay, 2001, ÔGrowth is Good for PoorÕ and ÔTrade, Growth and PovertyÕ, World Bank Research Working Papers No 2587 and No 2199.

18 As an indication of the modelling problems, the specifications of DollarÕs (and his colleaguesÕ) models muddy the waters between dependent and independent variables. Such variables as Ôgood rule of lawÕ, Ôopenness to international tradeÕ and Ôdevelopment of financial marketsÕ are well intentioned but highly subjective, vary in validity from country to country and are open to wide margins of measurement error. Elasticities and other coefficients are derived from a great variety of national sources. For countries as large as China, India, Brazil and Indonesia they cover very considerable internal differences. They are often not available for small countries so that such coefficients are based on ÔneighbouringÕ or Ôlike countryÕ data. ÔDummiesÕ are freely used when all else fails.

19 R. H. Wade, Governing the Market : Economic Theory and the Role of Government in East Asia (Princeton, New Jersey: Princeton University Press, 1990)

20 W. Easterly, R. Islam and J. Stiglitz, ÔShaken and Stirred: Explaining Growth VolatilityÕ (World Bank, 2000), http://www.worldbank.org/research/growth/padate.htm

21 Bhalla, Imagine ThereÕs No Country, pp. 201-202.

22 J. Gwartney, R. Lawson et al., Economic Freedom of the World: 2002 Annual Report (Canada: The Fraser Institute, 2002) and G. P. OÕDriscoll Jr, K.R.Holmes and M. Kirkpatrick, 2001 Index of Freedom (Washington D.C. and New York: The Heritage Foundation and the Wall Street Journal, 2002).

23 Smeeding in Gruen, OÕBrien and Lawson; for the income gap see p.185 and Gwartney, Lawson et al. for the Economic Freedom Index at p.11.

24 P. Saunders and Kayoko Tsumori, Poverty in Australia: Beyond the Rhetoric, Policy Monograph 58 (Sydney: The Centre for Independent Studies, 2002).

25 Smeeding in Gruen, OÕBrien and Lawson, p.197

26 H. Hughes, 'The Politics of Envy: An International Phenomenon', Policy 18:2 (Winter 2002), http://www.policymagazine.com

Helen Hughes is Professor Emeritus, The Australian National University, and Senior Fellow at The Centre for Independent Studies. Endnotes to this article may be accessed at www.cis.org.au

Policy is the quarterly review of The Centre for Independent Studies. For more information on subscribing to Policy, click HERE

If you are interested in the Centre's activities and publications, why not subscribe to e-PreCIS, our regular email update on the latest news and events.

(e-PreCIS requires html capable email facilities, such as Microsoft Outlook Express or Netscape Messenger)