Widespread poverty and excessive inequality
remain the principal challenges to the legitimacy of the process of globalization
that has been underway during the last two decades. Even as economies
and governments adjust to afford a larger role for markets and a smaller
role for the state in development, the importance of public action to
deal with poverty and vulnerability has increased. It was for this reason
that the 1995 World Summit for Social Development called upon countries
not just to set "time-bound goals and targets" to substantially
reduce overall poverty and eradicate extreme poverty, but to implement
national anti-poverty plans to achieve these targets.
Not surprisingly, as anecdotal evidence of the inequalizing effects of
globalization accumulates, a charged academic debate has arisen on the
issue of globalization and its impact on world incomes. Neoliberal economists
are prone to argue that with the lifting of trade barriers between countries
and freer movement of capital across borders due to globalization, there
is a tendency towards the narrowing down of cross-country income differentials.
Inasmuch as this narrowing or homogeneity is the result of a faster increase
in per capita incomes in the poorer countries, and so long as the income
inequality in these countries is not worsening, the global decline in
income differentials should be accompanied by a decline in the incidence
of poverty, as measured by the head-count ratio.
Such narrowing is a must in the present context, although differences
in ability, in resources and in many other factors—less justifiable
or obvious—are likely to remain. Income, which is the most basic
indicator of economic well-being, must clearly reflect the apparent advantages
of globalization. Open market policies, that are advocated by the Washington
Consensus as the route to economic prosperity that delivers benefits to
all and specially the poorer nations, must ensure that income disparities
across the world, as also within countries, do not increase, and that
all human beings are included in the distribution of the gains that arise
out of this new system. This is especially necessary since liberalization
is being thrust on many who are as yet unwilling to have it. The opposite
cannot be justified on moral grounds, and even the most cynical must admit
that it poses many practical problems, which include according to some,
the creation of conditions that engender terrorism.
But whether or not these contentions forwarded by neoliberal economists
capture actually the reality of the world today remains to be tested.
The question is in the first instance empirical. A number of studies in
recent years have examined the trends in inter-country and intra-country
inequalities during the years of globalization. As is to be expected,
the debate has increasingly given rise to disagreements with regard to
the measurement of inequality and poverty, and about cross-country and
inter-temporal data comparability and methodology of analysis relating
to these. Clarification at this level is a minimal prerequisite to establish
causal links between and test statistically the relation between income
inequality and globalization.
What kind of income inequalities are we talking about and what impact
does globalization have on these? Inequality is multidimensional and we
must study and treat each aspect in detail.
Intra-Country Inequality
The first and most basic concept of inequality that we are familiar with
is within-country or intra-country inequality. It indicates the disparity
between the incomes of individuals within a particular country. This kind
of inequality has existed for a long time though the degree has varied
across countries.
In clear evidence against the reduction-in–inequality hypothesis,
Cornia and Kiiski (2001) found that over the last two decades, inequality
has risen in 48 out of the 73 countries that they found high-quality data
for. These countries accounted for 73 per cent of the total GDP and 59
per cent of the total population of the 73 countries put together. Of
the rest, 16 countries experienced constant and 9, decreasing inequality.
At a regional level, the African economies faced rising inequalities,
Latin America saw declining inequalities of the 1970s reverse into a rise,
while Russia and the Eastern European transition economies experienced
a collapse of the middle class that made inequality soar. China experienced
rising inequality, especially between its urban coastal areas and the
rural interiors. Even the developed countries experienced rising inequality
as a result of 'greater disparities in market income', the effect of which
has been compounded recently by changes in the tax system, public services
and income transfers.
A noteworthy feature of this study is that it records adverse trends in
income distribution not just for poorly performing economies, as in Latin
America during the "lost decade" of the 1980s, but also for
economies that experienced remarkable growth after liberalization, as
was true of China. Thus, even when liberalisation was followed by growth,
the benefits of that growth did not seem to accrue in full measure to
the poorer sections of the population. In fact, in a World Bank paper,
Lundberg
and Squire (1999) found a negative correlation between greater openness
to trade and income growth among the poorest 40 per cent of the population,
but a strong and positive correlation with income growth among the remaining
groups. This was for a sample of 38 countries between 1965 and 1992. Hence
it seems that the adjustment costs of greater openness are borne exclusively
by the poor, regardless of how long the adjustment takes.
However, in a paper entitled, 'Growth is Good for the Poor', World Bank
economists David
Dollar and Art Kraay (2000a) found a positive relation between
the incomes of the bottom fifth of the population and per capita GDP that
holds in a sample of 80 countries covering four decades. On the other
hand, they did not find any significant impact of openness to trade, measured
as exports plus imports as a share of GDP, on poverty. They argued that
this shows that globalization or openness had no adverse impact on incomes
of the poor. However, globalization, given its promises, must do better
than that.
This result stands directly opposed to the Lundberg and Squire result.
The question, therefore, remained unanswered, forcing attention on the
nature of the sample used by the two sets of authors and on issues such
as the correctness of the "trade openness" measure used by Dollar
and Kraay. Apart from this, a major problem with the Dollar and Kraay
study is that, given the collapse of the large middle class in Russia
and the East European countries which has increased the extent of income
inequality, taking into account only the income of the bottom 20 per cent
of the population leaves out a significant section from the income inequality
scenario.
Many studies have found results showing increased intra-country inequalities
during the era of globalization, and interestingly, some such results
have come precisely from studies that otherwise show, or attempt to show
that overall global inequalities have actually declined over this period.
In one such attempt, Xavier
Sala-i-Martin, in a recent discussion paper (2002) of Columbia University,
measured inequalities both within and across countries. He found that
within-country inequalities have increased over the last two decades,
and that this is more so when differences in incomes within quintiles
of population are included in the analysis. This has mainly been driven
by income inequalities in China. The thrust of his argument is that the
larger part of global income inequalities is explained by inter-country
inequalities and these have actually come down over this period. But his
paper does show that as far as within-country inequalities are concerned,
the results run contrary to the propositions forwarded by the neoliberal
school. Similar results are also found in Quah (2001).
In a paper by Weller
and Hersh (2002), the authors analyse the short-term and long-run
effects of global liberalisation on the poor in developing economies.
The results indicate that more current and capital account liberalization
hurt the poor. "This is not because trade is directly harmful for
the poor, but because of the institutional design under which trade is
conducted. In particular, trade in a more deregulated environment lowers
the income share of the poor, whereas trade in a more regulated environment
raises the share of the poor." As far as income growth is concerned,
the results indicate that global deregulation has no measurable, robust
impact on growth rates. Therefore, the short-term effects on the income
share of the poor are not offset by significant faster income growth in
the long-run.
In addition their results also indicate that trade and possibly capital
flows may have a beneficial effect on growth in the long-run, and no systematic
adverse effect on the income share of the poor in a regulated environment.
To quote, "hence, greater trade and capital mobility in a regulated
environment, as was the case for the majority of countries for most of
the sample period, appears to be a preferable development choice".
However, the authors point towards the need for detailed research to find
out the exact nature of regulations that can help reap maximum benefits
from trade and capital flows and simultaneously let the poor share in
these gains.
Inter-Country Inequality
Inter-country inequalities accounting for a large part of world inequality
is a result derived also by Milanovic
(2001). He shows that rising income inequality in the world is mainly
due to between-country or inter-country (or cross-country) inequality
(75 to 88 per cent) and not within-country inequality. Let us move onto
a more detailed discussion of such inequality.
Inter-country inequality refers to the disparities in average or per capita
incomes between countries. Analytical treatment of this kind of inequality
involves an inherent assumption that all individuals within a particular
country earn close to that average income i.e. within-country inequalities
are not taken into account. With the dilution of economic borders between
countries, the question of inter-country disparities has been in the forefront
of discussions recently. Inequalities between nations, although present
for a long time, show up as being lower the further we go back in time.
Economic historian Angus
Maddison (2001), on the basis of GDP per capita, has estimated that
the richest countries were about three times richer than the poorest countries
in 1820. This difference has grown over time. However, in today's age
of increasing integration of the world economy with the intention that
countries should jointly share benefits, further increases in inequalities
between nations can have no possible justification.
But sadly, as Marc
Lee (2002) points out, "the UN's Human Development Report (1999)
adds that in 1960, the top 20 per cent of the world's people in the richest
countries had 30 times the income (in terms of total GDP) of the poorest
20 per cent. This grew to 32 times in 1970, to 45 times in 1980, and to
59 times in 1989. By 1997, the top 20 per cent received 74 times the income
of the bottom 20 per cent." This would seem to imply that the globalized
world we are living in today has seen a dramatic rise in inequality.
The attempt to show that policies of globalization have been accompanied
by decreasing poverty and income inequalities is contained in a report
by Dollar
and Kraay (2000b). They identify the globalized nations on the basis
of two variables, namely, increased trade to GDP ratio and reduced tariff
barriers. Their finding is that the globalizers have overtaken the non-globalizers
in terms of their joint GDP growth rates. Apart from serious problems
of bias in the selection of sample countries in each variable category
and the measures of openness used, e.g. measure of trade volume rather
than trade policy, as pointed out by Dani
Rodrik (2000), the study has certain other major drawbacks.
First, the inclusion of India and China as globalizers does pose a major
problem since both countries are known for their reluctance and sluggishness
in adopting the new policies prescribed by the IMF, the World Bank and
the WTO. This is true also of many other economies included in the 'globalizer'
category, for example, Malaysia, Thailand and Brazil, many of which have
followed limited forms of globalization. On the other hand, many of their
'non-globalizers' in Latin America and Africa display remarkably open
trade and structural adjustment policies as prescribed by the IMF and
the World Bank.
Second, since India and China have huge populations that jointly account
for a third of the world total, the weighting of country GDP estimates
by population results in an upward bias of growth rates of the so-called
globalized nations, as India and China have both had high GDP growth rates
in the recent decades.
Another argument on lines similar to that of Dollar and Kraay has been
advanced by Martin
Wolf, in an article in the Financial Times (2000). He argues that
China's growth, which has been a major factor responsible for the reduction
of cross-country inequalities, has been driven by the increased opening
up of its economy. He also argues that many of the backward countries
have remained backward because they have not opened up enough. Again,
the same criticism holds.
In another set of results favouring the advocates of globalization, Sala-i-Martin
(2002), using a number of measures including the Gini coefficient
and the Theil Index, has found that cross-country inequalities have came
down during the period 1978-98. His argument is that the growth of incomes
in China, and to a lesser extent in India, contributed largely to the
reduction in inequality. On the other hand, the African economies have
shown a huge disparity in incomes as compared to the developed nations,
and if their incomes remain stagnant then world inequality will rise again
in the future. We discuss his finding in more detail in later.
Branko Milanovic, in
a 1999 World Bank research paper that starkly contradicts the above
results, found that the Gini coefficient measuring across-country inequalities
based on per capita GDP actually rose from 55 to 58 between 1988 and 1993.
His use of population weighted per capita GDP figures as opposed to the
per capita country GDP figures used previously, yielded results that showed
a decline in income inequality. This latter result can easily be explained
by high GDP growth rates in the hugely populated countries of China and
India. Dropping China from the sample threw up a constant inequality,
and dropping India in addition, again showed an increasing Gini for between-country
inequality.
World Inequality
Combining the concepts of intra-country and inter-country inequalities
give us a definition of global or world inequality, which calculates income
disparities between individuals as if they belonged to the same nation.
This is calculated as a distribution of total world incomes among the
total population of the world. Study of such inequalities has emerged
recently, and has concentrated on developing different measures of calculation.
Sala-i-Martin's
2002 paper finds that global income inequality has closely followed
the pattern of inter-country inequalities and come down over the two decades,
since 1978. In addition, worldwide poverty numbers have gone down sharply,
signifying an increase in the economic well-being of the poor.
Milanovic
(1999), on the other hand, finds that the world or global Gini coefficient,
which takes into account both inter-country and intra-country inequalities,
increased from 63 in 1988 to 66 in 1993. If purchasing power differences
are not taken into account, i.e. if one looks at the differences in simple
dollar incomes, the world Gini increased from 78 to 80. In a later study
(2001), Milanovic confirms the above results using both the Gini and the
Theil index. He cites three reasons for this growth in inequality. The
first is the slow growth of incomes, especially rural, in Asian countries
relative to OECD countries. The second is the pulling ahead of urban China
relative to rural China and India. The third is the 'hollowing out' of
the middle class in Eastern Europe.
In a study from the London School of Economics by Robert H. Wade ('Is
globalization making world income distribution more equal?', May 2001),
similar factors have been cited as the cause for the rise in global and
between-country inequalities. Rapidly widening income distribution within
the biggest countries (India and China) has also been a major contributory
factor. While the gap, worldwide, between the average income of the top
quintile of people (top 20 per cent) and the average income of the bottom
quintile within each country is about 5:1, the gap between the average
income of the top quintile of states and that of the bottom quintile is
of the order of 25-30:1.The report takes the World Bank and the IMF to
task for their failure to use their formidable research capacities to
look at the trends and causes of world income distribution. Wade also
points to "the need to remember that East Asian states achieved economic
success by creating national economic space (partially separate from the
world economy) and setting conditions on the entry of foreign capital"
and that "China's current dirigiste strategy, similar to that of
pre-liberalization Japan and South Korea, is more likely to succeed than
the World Bank's model".
If we look at asset distribution across the world, financial assets held
by 7.2 million individuals in the top income group were valued at US$27
trillion in 2000, almost as high as the world 's total GDP ($31 trillion
in 2000). The assets of the top 200 richest people amount to more than
the combined income of 41 per cent of the world's population. However,
detailed analysis of wealth distribution, which is definitely an important
indicator of economic well-being, has been lacking.
Globalization, China and India
Throughout the debate on income inequality in the globalized world of
today, China and, to a lesser extent India, emerge as key movers of the
results. Some interesting points emerge from a closer analysis of their
roles.
World Inequality, Globalization, China and India
Robert.
J. Barro, in a recent article, cites the Sala-i-Martin results
to argue that inequality has decreased in the globalization era. He states
that the 1999 UN human Development Report, 'should base their assessments
of world poverty and inequality on a better understanding of the facts'.
Two points need to be noted here.
First, when China is dropped from the sample, Sala-i-Martin's Mean Logarithmic
Deviation index (the only one he reports) shows a relatively flat pattern
with no consistent increase but no clear decline either. In Milanovic's
analysis, dropping China and India seem to re-establish the rising inequality
scenario. Second, as pointed out earlier, China and India, on whose performance
the Sala-i-Martin results are heavily dependent, cannot be taken as model
examples of globalizers whereas many of the African countries have actually
opened up much more. So while these two countries are doing well, it is
not because they are relatively more globalized than the countries which
are doing badly. In another example of such misused definitions, Dollar
and Kraay (2000b) take China and India as globalized and more open, and
find the growth rates of globalized nations as a whole to be higher. As
argued earlier, this classification cannot be acceptable to the discerning
economist.
That including China and India as examples of successful globalizers is
a conceptual mistake, has also been pointed out by Wade (2001) and Rodrik
(2000). In the words of the latter: "the main trade reforms followed
a decade after the onset of higher growth" in these two countries.
In the Chinese case, "the increase in growth started in the late
1970s with the introduction of the household responsibility system in
agriculture and of two-tier pricing", whereas the trade reforms did
not begin till the second half of the eighties. As for India, its growth
rate increased substantially during the 1980s, whereas serious trade reforms
were not in place before 1991-93. The tariff averages displayed in the
chart show that tariffs were actually higher in the rising growth period
of the 1980s than in the low-growth period of the 1970s. Rodrik further
argues that the fact that China and India participated in foreign trade
does not mean that their growth followed increased foreign trade. One
needs to look at proper sequencing of events here.
Income Inequality within China and India
In the literature (D. Quah, Cornia and Kiiski, Sala-i-Martin, Milanovic,
Wade), there is ample evidence of rising income inequalities within China
and India, as also in many other countries including some in the developed
world. Most studies find that rising global inequalities stem from differences
in income growth between the urban and rural sectors in both India and
China, though this has coincided with rises in overall income per capita.
Milanovic and Wade have both identified this as one of the major factors
explaining rising international income inequality. Even with a reduction
in absolute poverty, rising income inequalities between sectors, and rising
or even constant income inequalities in general should still be a matter
of concern.
So the argument that if one-third of the world's population is apparently
doing well, it indicates that the world is also doing well (Sala-i-Martin),
is compelling but not conclusive. Firstly, not all of that one third may
be doing that well. Secondly, the rest of the world, especially the African
countries, the erstwhile USSR and the East European countries are doing
much worse and that has to be a matter of great concern. As pointed out
by Barro and Sala-i-Martin themselves, stagnancy in African incomes may
be a major cause for a future rise in world income inequality. As poverty
estimates show, the largest number of the world's poor are now in Africa.
Two countries, which happen to have large populations, cannot change the
reality of the huge disparities that exist between a few rich countries
and numerous small, backward ones.
The Three Concepts of Inequality
While it is useful to measure and analyse all three concepts of inequality,
for the individual in a particular country it is the concept of within-country
inequality that is the most relevant. And in that regard, the world is
definitely worse off today.
With the increasing integration of economies, it is perhaps also pertinent
to ask what is happening to the inequalities between countries, which
It are supposed to be the major source of world income inequalities.
The first problem in undertaking this kind of analysis is conceptual.
The fact is that individuals within a country do not earn the average
per capita income by which the country is represented on the world income
distribution map. Wide disparities within the country can render such
representation meaningless. And most studies have proved that this is
so.
The second problem remains that of data. All studies focusing on this
issue must resort to empirical evidence. However, it is very difficult
to construct data-sets across countries that are consistent, clean and
comparable.
- Disaggregated household-level income and consumption data for a long
enough time period are hard to get even for a single country. Most countries
do not conduct intensive surveys at the levels required for such analyses.
India's National Sample Survey (NSS) data is one among a few examples
of systematic and detailed data that can be used for such purposes.
This poses a problem for measuring intra-country inequalities too, but
is much more acute in the case of inter-country analysis.
- Even detailed sources of data, like the NSS, are now being contaminated
and tampered with in order to arrive at the 'desirable' results. A major
problem here is that conducting wide and intensive surveys to get the
required data is very expensive and can be pursued by very few agencies,
sometimes not even by the government of a country. If the few that do
have the resources have their own agenda to push in this regard, then
one cannot but question the veracity of such data. Recent history tells
us that such fears are very real today.
- Unless data-sets can be kept robust across countries, the error margin
for a study involving a large number of countries would be huge. From
base-level information (if available), such data are generally estimated
first at the district level, followed by the state/province, national
and finally the international level. At each step, these estimations
involve a lot of assumptions, many of which can also be biased. Most
studies, including that used by Milanovic and Sala-i-Martin, involve
wide extrapolations and interpolations that puts into question the veracity
of the final data-set that is constructed.
- Comparing data-sets that specifically relate to individual countries
raises another major problem. Despite using the notion of purchasing
power parity, such comparisons are very difficult. And, as discussed
later, the World Bank standards of purchasing power equivalence, which
are commonly used in such studies, have problems in themselves.
- The concept of world inequality is an attempt at resolving the first
kind of problem with between-country inequalities. This is supposed
to take into account both the first two concepts of inequality. However,
though an interesting analytical concept, it has limited relevance in
reality. The phenomenon of globalization notwithstanding, individuals
in the present day do not function as members of a single economic community.
They exist under varying economic and political regimes. Therefore,
to cite their economic positions and income ranking as if they did live
in a homogeneous society does not say much about the actual state of
the world. Even if world inequality were down, would it really mean
that the poor in Africa were better off? Arguably not. Also, as we have
seen, the measurement of world inequality can get biased by the presence
of a few large nations. Again, studies like those by Milanovic and Sala-i-Martin
take inter-country inequality as a higher contributor while calculating
overall world inequality. This, apart from being fraught with the data
problems discussed above, is of limited relevance for a specific country.
Even if world inequality is shown to be declining, driven in large part
by higher average per capita incomes in China and India compared to
other countries (the between-country factor), it does not really mean
that the poorer people in China are better off unless inequalities within
China are also declining. Ultimately, therefore, it is within-country
inequality that we need to take into account.
Two more issues need to be addressed before concluding our discussion.
Economic Growth, Globalization and Income
The first concerns the role of overall economic growth in increasing incomes of the poor and income inequality. Economic growth by itself cannot be a measure of proportionately increasing incomes and well-being, if there is a simultaneous rise of income inequality within the country. Martin Wolf's (2000) argument is an example of this kind of serious misconception.
Many scholars, Dollar and Kraay (2000a) and D. Quah (2002) for example, have tested a relation between overall growth of an economy and poverty reduction. Quah, in a major study on China and India, finds rising and constant income inequality, respectively, in the two countries (over a period of rising growth in per capita incomes). But his argument is based on poverty numbers. He argues that growth definitely led to a decrease in poor people in absolute terms in both countries. The Gini coefficient would have had to rise phenomenally to compensate for this fall in poverty numbers, of which there is no evidence. So there is an overall fall in poverty resulting from growth. Dollar and Kraay (2000a) found a similar positive relation between growth and mean income of the poor (as discussed earlier). But they did not find that growth had any impact on the distribution of income. Weller and Hersh have also tested the impact of economic liberalisation on growth and as mentioned earlier, the results indicate that global deregulation has no measurable, robust impact on growth rates. At the same time, in a regulated environment, trade and possibly capital flows may actually have a beneficial effect on growth in the long-run. On the basis of this, the authors argue for more trade and capital flows in a regulated environment.
There are many questions that arise here. First, even if growth leads to reduction of poverty, increasing within-country inequality may lead to a less-than-proportionate distribution of this growth across different segments of the population. The increasing rural-urban income gap in China and India is an indicator of this fact. Second, even if growth does lead to reduction of overall poverty, the crucial question is, 'does globalization lead to growth?' This crucial link has not been tested by Dollar and Kraay, but simply assumed. The growth argument is not enough if globalization is advanced as an explanation of poverty reduction (though not by Quah). Again, as mentioned earlier, Kraay and Dollar find no significant impact of globalization on poverty reduction.
Measurement Errors in World Bank Estimates
The second issue involves possible errors in measurement methods that might have contaminated all measures of inequality and poverty which are routinely used by economists in measuring disparities in these variables.
For example, in a recent paper, Sanjay G. Reddy and Thomas W. Pogge argue that "the estimates of the extent, distribution and trend of global income poverty provided in the World Bank's World Development Reports for 1990 and 2000/01 are neither meaningful nor reliable." First, the Bank uses an arbitrary international poverty line unrelated to any clear conception of what poverty is, and one that will allow identification of the commodities that must be commanded in order to avoid being poor. Second, it employs a misleading and inaccurate measure of purchasing power "equivalence" that creates serious and irreparable difficulties for international and inter-temporal comparisons of income poverty. Finally, it extrapolates incorrectly from limited data and thereby creates an appearance of precision that masks the high probable error of its estimates. "The systematic distortion introduced by these three flaws likely leads to a large understatement of the extent of global income poverty and to the false appearance that it is falling." Since poverty estimates are widely quoted to contradict or support income-inequality effect arguments (D. Quah, Robert J. Barro, Sala-i-Martin, Dollar and Kraay, Martin Wolf), it has a great relevance for the latter and for the study of the impact of globalization on global income movements.
Conclusion
It is undeniable that the world needs more studies testing the links between specific aspects of globalization and income inequality. From the available analyses, the following points emerge. The movement of a few large, high-growth, developing countries like China and India may sometimes throw up statistics showing that world inequality has fallen, but given the evidence of rising income inequality within them and the stark and growing inequalities in a large chunk of the less-developed countries like Africa, Latin America, Russia and East Europe, we cannot be complacent. The latter countries have also experienced rising disparities in average incomes compared to those of the richer developed countries. Rising inequalities have been noted within some developed countries too. Further, globalization cannot be claimed as the definitive factor that has driven growth in countries like China and India. In fact the very opposite may be true. And there are enough examples of African and Latin American countries that have actually been badly damaged after following a more open policy. Finally, there is need for conceptual clarity and intellectual honesty to register and acknowledge facts, and then to prescribe or uphold the necessary policy. This holds for the statistics and data that are generated and published widely, for methods of analysis that are resorted to, and interpretations that need to look at results more critically and sometimes beyond the outer covering.