Vi Digital Library - Text Preview
Global Inequality: From Class to Location, from Proletarians to Migrants
Working paper by Milanovic, Branko, 2011
Download original document (English)
Policy Research Working Paper 5820
From Class to Location,
from Proletarians to Migrants
The World Bank
Development Research Group
Poverty and Inequality Team
Produced by the Research Support Team
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.
Policy Research Working Paper 5820
Inequality between world citizens in mid-19th century
was such that at least a half of it could be explained by
income differences between workers and capital-owners
in individual countries. Real income of workers in most
countries was similar and low. This was the basis on
which Marxism built its universal appeal. More than 150
years later, in the early 21st century, the situation has
changed fundamentally: more than 80 percent of global
income differences is due to large gaps in mean incomes
This paper is a product of the Poverty and Inequality Team, Development Research Group. It is part of a larger effort by
the World Bank to provide open access to its research and make a contribution to development policy discussions around
the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be
contacted at email@example.com.
between countries, and unskilled workers’ wages in rich
and poor countries often differ by a factor of 10 to 1.
This is the basis on which a new global political issue
of migration has emerged because income differences
between countries make individual gains from migration
large. The key coming issue will be how to deal with this
challenge while acknowledging that migration is probably
the most powerful tool for reducing global poverty and
Global inequality: From class to location,
from proletarians to migrants
JEL Classiification: D31, I3, N3
Key words: global inequality, inequality between nnations, migration, labor in 19
Resesarch Department, World Bank. The paper was written while I was a Visiting Fellow at Universidad Carlos
III in Madrid.
1. Global inequality in the mid-19
In the spring of 1848, as the pan-European revolution was nearing its crescendo, Karl
Marx and Friedrich Engels wrote probably the best known political pamphlet of all time: The
Communist Manifesto. When they surveyed the situation of the world, they persuasively and
repeatedly insisted on the fact that people in all ―civilized‖ societies were divided into two large
classes: that of the owners of the means of production (the capitalists), and those that were
selling their labor for a living and held no property (the proletarians). It was an almost self-
evident division at the level of each and every country. Having capital meant being rich; having
only labor power meant being poor. There were not too many people in-between, with middling
levels of income, whether those who owned some capital and yet had to work with their hands
like artisans, or peasants who toiled their own land. Moreover, even they—the logic of capitalist
development was implacable—were doomed to extinction or irrelevance, as they would
gradually ―dissolve‖ mostly into proletariat with perhaps only a few making it to the capitalist
class. The division into two or three main classes (the third being landowners, who in Marx and
Engels’ view could be assimilated to capitalists) was not, of course, introduced by Marx and
Engels. It has been present in contemporary political economy, and it hailed back to Adam
Smith, and even to François Quesnay. It was used by David Ricardo in his Principles, published
in 1817, as a key feature, motivating his entire economic analysis.
So evident seemed the class division in all societies that Marx and Engels concluded their
Manifesto by writing that ―…modern industrial labor, modern subjection to capital, the same in
England as in France, in America and in Germany, has stripped [the proletarian] of every trace of
national character‖ (1978, p. 482). Being proletarian was thus a global condition, they held, and
being global, it presented an ideal basis on which international solidarity of the working class
could be built. Proletarians were equally poor and exploited everywhere and they could liberate
themselves and usher in classless societies only in a common effort that knew no national
borders. This was expressed in Marx and Engels’ famous statements ‖[t]he proletarians have
nothing to lose but their chains. They have a world to win‖ (ibid, p. 500). Local emancipation
and international solidarity were thus intertwined, part of the same struggle.
This same view of workers who have no true homeland because they are everywhere
equally destitute and powerless was reflected, only a year earlier, in a speech delivered to the
Free Trade Congress in Brussels by Engels. He defended his and Marx´s pro-free trade stance as
Under freedom of trade the whole severity of the laws of political
economy will be applied to the working classes. Is this to say that
we are against Free Trade? No, we are for Free Trade, because by
Free Trade all economical laws, with their most astounding
contradictions will act upon a larger scale, upon a greater extent of
territory, upon the territory of the whole earth; and because from
the uniting of all these contradictions…will result the struggle
which will itself eventuate in the emancipation of the proletariat.
Engels believed that by letting capitalist rules of the game embrace ever greater portions
of the globe, the outcome of such a capitalist domination would be similarity of economic
conditions among workers. The similarity in economic conditions would, in turn, lead to the
concordance of economic interests and to the emergence of solidarity among workers of different
countries. Ultimately, it would culminate in a worldwide revolution.
But did similarity of economic conditions among workers exist then? Did Marx and
Engels depict the reality of that era correctly? Today we have more data than Marx and Engels
possessed at the time. Yet, their insight is confirmed by what we know today, both as regards the
ubiquitous split between the two major classes and similarity in the economic position of the
laborers, or, more broadly, the poor across countries.
Consider class and income compositions of a couple of societies for which we have the
data from the mid-19
century. In R.D. Baxter’s social table for England and Wales for 1867, 3
incomes of the wage-earning classes, representing 72 percent of the population and divided into
eight subgroups, start at the bottom of the distribution, just above paupers, with an estimated per
capita income equal to one-third of the national mean, and end at the richest point, with an
income 10 percent below the mean. The propertied classes (a little over 6 percent of the
population) covered the upper part of the income distribution. The income ratio between the two
classes was greater than six to one.
See Karl Marx and Frederick Engels, Collected Works, vol. 6, New York, International Publishers, 1976, p.
290. Quoted from Moellendorf (2009, p. 94).
Available at Peter Lindert´s Website ―Global Price and income History Group‖ , see http://gpih.ucdavis.edu/.
England was the prototypical and the most advanced capitalist economy. But we have the
data for Chile in 1861 which reveal the same, or even sharper, polarization.
from the poorest (female fisherman) to the richest (shoe-makers), covered the range of income
that goes from less than one-seventh of the national mean to two-thirds of the mean. Only the
artisans, who mixed ownership of some capital with their own labor, and possibly even hired a
few workers, achieved an income higher than the national average. At the high end of the income
distribution were owners of manufactures and mines with incomes between 20 and 55 times the
mean, and land owners and large-scale merchants with respectively 35 and 9 times the mean.
These top groups comprised only 2 percent of the population while working classes (including
peasants) accounted for more than 90 percent.
But if workers filled in most (or perhaps in all) countries, that part of the income
pyramid which ran from the subsistence to just somewhat under the mean, their real incomes in
various countries could still have differed a lot if country means were very different. But, as we
shall see, around the 1850s, they were not. Angus Maddison has estimated that around 1850, the
mean income in the poorest countries in the world (Ceylon and China) was around $PPP 600.
At the top were the Netherlands and the United Kingdom with a GDP per capita of about $PPP
2,300. Thus, the ratio between the top and the bottom (of country mean incomes) was less than 4
to 1. Consequently, the better-off workers who earned incomes close to the national means,
could not, in terms of their standard of living, differ from each other by more than the ratio of 4
to 1. And the bulk of workers who lived at less than their countries’ average income and closer to
the subsistence, could not have incomes that differed by more than 2 to 1—with many of them
living at approximately the same subsistence level. Indeed, Broadberry and Gupta (2006, Table
6, p. 17) show that in the period 1800-1849, the wheat-wage of an unskilled daily laborer in India
(among the poorest countries in the world then) was about 30% of the wage of a similar worker
in England. And comparing the Netherlands with the Yangtze valley, two relatively developed
areas sharing a number of similar geographic features, Li and van Zanden (2010, p. 21) conclude
The data are derived from an Occupational Census, were consolidated by Javier Rodriguez Weber, and can be
found on Peter Lindert’s Website ―Global Price and Income Histrory group‖ (see the previous footnote).
In 1990 international prices.
that in the 1820s, real wages in the Netherlands were about 70% higher than in the Yangtze
Thus, similarity in the economic position of workers across the world, implicitly assumed
by Marx and Engels, had a firm basis in the reality of the time. And, as we have seen, this is
crucial because it is the similarity in the economic positions which allowed Marx and Engels, as
well as later Marxists, to derive the principle of international solidarity of the working class.
We can reach the same conclusion that the main income cleavage was the one between
classes, and not between countries, from a slightly different vantage point, if we consider global
inequality, that is inequality between world citizens. In a pioneering work, François Bourguignon
and Christian Morrisson (2002) have used a collection of their own and other authors’ estimates
of income distributions for 33 country-groups and mean incomes from Angus Maddison to
construct worldwide income distributions, at approximately twenty-year intervals, for the period
1820-1992. They estimate global inequality, measured by the Gini coefficient, to have been
about 53 Gini points in 1850, and to have been composed in almost equal proportions of
between-country and within-country inequalities. The former (between-country inequality) refers
to that part of inequality that is due to the differences in country mean incomes while the latter
(within-country inequality) is that part of total inequality which is due to inequalities existing in
each individual country. To fix the ideas, we can call the between-country inequality ―locational‖
because it depends on the differences of mean incomes between various places (countries), and
the within-inequality, ―class‖ inequality because it depends on different individuals, living in the
same country, having different incomes and belonging to different social groups.
Bourguignon and Morrisson estimate that the global Gini in 1850, amounting to 53.2 points, can
be broken down into 25.9 Gini points (49 percent) due to location, and 27.3 Gini points (51
percent) due to class. Thus, around mid-19
century, one-half of inequality between individuals
in the world was explained by unequal development of countries and another half by income
Li and van Zanden´s results favor the view that appreciable difference in wages existed as early as 1800s in
opposition to other writers (especially Pomerantz (2001) and to some extent Bairoch, 1997, p. 111) who believe
that the gaps were less.
Between-inequality is inequality that would exist if everybody in every country had the mean income of his/her
country, or in other words, if inequality within each nation were zero.
differences between social classes—that is, essentially between workers and capitalists. How
does it compare with the situation today?
2. Global inequality in the early 21
If we use the same decomposition between location and class today, when our data are
much better than for the past, we find that of the global Gini, which amounts to 65.4 points, 56.2
Gini points or 85 percent is due to differences in mean country incomes, and only 9.2 Gini points
(15 percent) to ―class‖. 8 Not only is the overall inequality between world citizens greater in the
century than it was more than a century and a half ago, but its composition has entirely
changed; from being an inequality determined in equal measures by class and location, it has
become preponderantly an inequality determined by location only. This fact is of great political
and economic significance.
Figure 1 helps us visualize this new reality. On its horizontal axis we draw the population
of a given country divided into twenty equally-sized groups, called ventiles, each including 5%
of the population, ranked by their average per capita incomes. Thus, value 1 on the horizontal
axis corresponds to the poorest 5% of the population in a given country, and value 20, to the
richest 5% of the population. On the vertical axis we show the global percentile position of each
national ventile. The vertical height, corresponding for example to the bottom ventile in the US,
is y=60 and it indicates that the poorest 5% of Americans have an income that places them at the
In other words, they are better-off than 60 percent of world population.
The same interpretation applies to any other national ventile. Figure 1, using the example of
BRIC countries (Brazil, Russia, India and China) and the US, illustrates vast differences in
incomes which exist between countries, and in particular between the poorest ventiles of the
population. While the poorest Americans are (as we have just seen) at the 60
the poorest Brazilians and Indians are at the 3
global percentile, that is among the poorest
people on the planet. The poorest Chinese are around the 16
global percentile, the poorest
Russians around the 37
percentile. Even more striking is the comparison of the income of the
All 2005 global data derived from nationally-representative household surveys conducted around year 2005 (see
World Income Distribution (WYD) database available on http://econ.worldbank.org/projects/inequality). In order to
keep comparability with the precedent calculation for mid-19
century, I use price levels and PPPs that are
commensurate with 1990 PPPs rather than the most recent version of PPPs (year 2005). The latter gives an overall
higher global inequality (70 Gini points rather than 65.4; see Milanovic 2011, and Milanovic 2012). The share of
the between-country component is also about 85 percent if we use more up-to-date PPPs.
The position of the poorest ventile of the rich country is indicated by a broken line in this and following graphs.
poorest groups of Americans with the richest Indians: the second poorest ventile of American
population is approximately at the same level of income (just 1 global percentile lower) than the
richest 5% of Indians.
Figure 1. Importance of location vs. class: United States compared to BRIC countries, 2005
Note: Based on national household surveys; people ranked by per capita income or per capita consumption adjusted
for price differences between the countries using the most recent PPPs. BRIC is an acronym denoting the emerging
market economies of Brazil, Russia, India and China.
The figure could be made both more dramatic or less dramatic without losing anything
of its essential message: namely, that most of global income differences today depend on
location. We can make the figure more dramatic by contrasting incomes of people living in a
Note that if class alone determined one’s global income position, and all country means were the same, the line
for every country would be a straight 45 degree line.
1 5 10 15 20
very rich and egalitarian country like Denmark with people living in various poor African
countries (Figure 2a). Danish poorest ventile has an income that places it as the 82
percentile while in many African countries even the income of the richest ventile hardly exceeds
global percentile—implying thus that the two distributions do not overlap at all.
In other words, if we lined up all individuals from these countries by their per capita income,
Denmark’s income distribution would only start at the point at which many African countries’
distributions end. The richest Malians are poorer than the poorest Danes.
The picture of global location-induced inequality could be also rendered a bit less
dramatic, if we present it in a more finely-grained form, viz., if instead of using ventiles on the
horizontal axis, we used percentiles of national income distributions, or even individuals lined up
and ranked one by one. For sure, we should then be able to find some Malians who are richer
than some Danes, and the two distributions would indeed display some overlap. However, that
overlap would be, in a statistical sense, minimal: there may be one-half of a percept or 1 percent
or even 2 or 3 percent of Malians who are richer than the poorest Danes, but this does not in any
way invalidate the main message from our graph. In effect, when we contrast US and India, and
instead of ventiles divide their populations into percentiles, the overlap is only 4 percent: that
many Indians are better off than the poorest Americans (see Figure 2b).
Figure 2. Income gaps between various countries’ populations; early 21st century
a. Denmark vs. selected African countries (ventiles) b. US vs. India (percentiles)
Note: Based on national household surveys; adjusted for price differences between the countries using most recent PPPs. Year 2005.
1 5 10 15 20
1 10 20 30 40 50 60 70 80 90 100
Figure 3 displays broadly the same information in a different way. It shows the
distribution of per capita $PPP incomes received by the lowest and highest ventiles in all
countries in 2005. Two points stand out. First, the distributions are quite wide indicating that
people who are nationally poor (or nationally rich) receive vastly different incomes, depending
on what country they live in. Second, the two distributions do intersect, implying that there are
countries such that people who are poorest there are yet better-off than the richest people in other
countries. The link between being nationally and being globally poor is severed.
Figure 3. Income of people belonging to the poorest and richest country ventiles
(year 2005, 114 countries)
Note: Based on national household surveys; adjusted for price differences between the countries using most recent
The income gap between countries’ average incomes is much greater today than around
1850. Again, using Maddison’s data, in order to keep comparability with the 1850 results, we
find that the top-to-bottom ratio in 2007 was in excess of 100 to 1 (as opposed to 4 to 1 that it
was in 1850). The increase in the gap is easy to understand: while the poorest countries today are
poorest ventiles richest ventiles
2 4 6 8 10 12
log of per capita income in 2005 international dollars
not richer than the poorest countries in the past, the richest countries are immensely better-off.
Instead of Britain and the Netherlands that were the richest countries in 1850, today, we have
USA, Singapore and Norway, all with incomes around $PPP 30,000, that is, 13 times richer than
Britain or the Netherlands in 1850.
At the bottom of per capita income ―league‖, we have
Congo, Burundi, Niger and Central African Republic, all with incomes just above subsistence,
some $PPP 350-700 per capita, not at all different from the level of the poorest countries in
1850. The world today presents a peculiar picture where some of its parts are immensely richer
than ever in history while other parts have an income level about the same as it was 150 or even
500 years ago. Thus to speak of an average global income is correct in an arithmetic sense but
otherwise lacks much meaning.
Even when we contrast the fast-growing ―emerging economies‖ of China and India with
the rich world, the gap in the first decade of the 21
century is greater than it was around 1850.
To keep to the comparison from the 1850s, consider the GDP per capita of the United Kingdom
(not any longer the richest country in the world) against GDPs per capita of India and China
(among the fastest growing economies in the last two decades)—all GDPs adjusted for the
differential price levels between the countries. The gap between UK and India in 2009 is in
excess of 10 to 1 while it was only 5 to 1 in the mid-19
the gap between UK and
China is 5 to 1 today while in 1850 it was (as we have seen above) less than 4 to 1.
Indeed, the number of countries today is much greater than it was some 160 years ago,
and it may seem that this factor alone would lead us to conclude that inter-country income
differences, and the gap between the top and the bottom, must have widened anyway. But the
widening is not an artifact of the number of countries. We can see this if we limit our
observations to the same 63 countries as included by Maddison in 1870,
and whose populations
In 1990 international prices. We leave out some outliers, small oil-producing economies and Luxemburg, with
even higher incomes.
Maddison does not give GDP per capita estimate for India in 1850, so we use the next proximate year (1870). It is
believed that Indian GDP per capita was stagnant throughout much of the 19
century, so the exact year should
make little difference.
Data from World Bank World Development Indicators, year 2009, expressed in 2005 PPPs.
The number of countries included by Maddison in 1850 (twenty-four) is too small for a valid comparison.
today accounts for some 5 billion people. In 1870, the gap between the richest countries
(Australia and Great Britain) and the poorest (Nepal and Ghana) was 8 to 1; in 2007, it is 31 to 1
(United States and Norway vs. Nepal, North Korea and Ghana).
Moreover, the very logic of widening income disparities alluded to before shows that the
effect of the number of countries cannot be decisive. Because the logic is that some people,
living in specific geographical locations, earn today approximately the same—very low--level of
income as their ancestors did more than a century and a half ago, while other people, at other
locations, live at income levels that are manifold greater than in the past and some 100 times
above the subsistence. Thus, the ratio between top and bottom must rise, almost regardless of
how we decide to ―slice‖ the population of the world (that is, in how many countries they live).15
We obtain the same conclusion if we compare real wages in rich and poor countries.
Table 1 shows wages for three occupations in five selected countries (or rather their major
cities): two rich (London and New York) and three poor (Beijing, Delhi and Nairobi). The data
refer to after-tax nominal wages which are then deflated by the food cost of living in the same
cities to obtain real food wages. The advantage of these data, collected for 74 cities and 14
occupations and published by Union de Banques Suisses, UBS (2009), is their almost full
comparability. In effect, the wage data refer to net wages earned by the same, very narrowly
defined, professions (see Notes to Table 1); they are adjusted for the effective number of hours
worked, and are geographically limited to large cities for which UBS publication also calculates
a food cost of living index.
In Table 1, I have selected three occupations with increasing level of skills: building
laborer, skilled industrial worker, and engineer. The real wage gaps are the greatest in the case of
unskilled workers: for them, the ratio of real wages in rich and poor countries is almost 11 to 1.
The caveat ―almost‖ is necessary because in the extreme case, when the world population were divided into (say)
two countries and in such a way that each half would contain the same proportion of locations that have not seen
any progress since 1850 and locations that are today much richer, the gap between these two halves need not be
greater, and may be smaller, than it was in 1850. But obviously, this is an extreme case which has nothing in
common with the real processes of state-formation and real state of affairs both now and in the past.
For the skilled industrial workers and engineers, the gaps are respectively 5.8 to 1 and 3.3 to 1.
If we contrast this result with the estimates of the ratio between wheat wages of English and
Indian unskilled laborers around 1850 (Broadberry and Gupta, 2006, quoted above), we note that
the UK/India unskilled wage gap has increased from around 3.3 to 1 in 1850 to more than 9 to 1
today. The advantage of this particular comparison is that it keeps the level of skills constant
across time, and focuses on the very low skills representative of those who are at the bottom of
the income pyramid.
It is the gap between the poorest members of rich and poor societies that we found to be
the widest, both in income and wage data. Or, to translate this finding in terms of political
economy, the commonality of interests between the poor and unskilled workers in the rich world
and the poor and unskilled workers in the poor world is hard to detect if we focus on their
economic conditions only. This is of course a major departure from a situation which existed a
century and a half ago.
A decreasing rich-poor country wage gap with greater skills of workers is also found by Warner (2002). Warner
attributes it to greater global competition which exists for more skilled workers whose wages reflect international
conditions while the wages of low-skilled workers behave like non-tradables. It can also be explained by greater
relative scarcity of more skilled workers in poorer countries.
Table 1. Nominal and real (food) hourly wages for several occupations
Annual after-tax wage divided by the number of effective annual hours of work, March 2009
Building laborer 1/ Skilled industrial worker 2/ Engineer 3/
wage (in $)
wage (in $)
wage (in $)
New York 16.6 16.6 29.0 29.0 26.5 26.5
London 9.7 15.4 19.0 30.4 22.1 35.2
Beijing 0.8 1.3 2.3 3.8 5.8 9.5
Delhi 0.5 1.7 2.1 6.9 2.9 9.1
Nairobi 0.6 1.5 2.0 4.7 4.0 9.2
20.4 10.9 11.0 5.8 5.8 3.3
Note: Food prices are estimated from a basket of 39 food products with weights reflecting West European
consumption patterns. New York food prices are set equal to 1. Real food wage (in New York food prices) is
estimated by dividing the nominal after-tax dollar wage by the food price index (not shown here).
Annual number of hours worked is equal to the weekly number of hours of work given for each profession and
country separately (UBS, 2009, pp. 34-35) multiplied by 52 weeks, and reduced for the number of official and paid
vacation days per year for each country (UBS, 2009, p. 30).
1/ Unskilled or semi-skilled laborer, about 25 years of age, single.
2/ Skilled worker with vocational training and about 10 years of experience, working in a large company in the
metal working industry, approximately 35 years of age, married, two children.
3/ Employed in an industrial firm in the electrical engineering sector, university or technical college graduate with at
least 5 years of experience, about 35 years of age, married, two children.
4/ Rich are New York and London; poor are Beijing, Delhi and Nairobi.
3. From ―permanent revolution‖ to ―fortresses Europe and America‖
The fact that the position of the poorest members of different societies is vastly different-
--the fact, for example, that the poorest 5% of Americans, adjusted for price levels, earn 35 times
more than the poorest Zambians or 12 times more than the poorest Malians—has global political
implications. We have seen above that the entire construct of workers’ (poorest people)
solidarity was based on some objective conditions, that is on similarity in their living conditions.
But if that similarity in living and working conditions has evaporated, would not the
commonality of interests and ―class solidarity‖ similarly vanish? And indeed it is not easy to
find any examples of shared interests between the poorest classes in rich countries and poorest
classes in poor countries. More likely, their interests conflict.
A second implication of the world where location determines to a large extent one’s
income is that it must be a world of huge migratory pressures because people can increase their
incomes severalfold if they migrate from a low mean-income location to a high mean-income
location. Only if knowledge of these income differences could be somehow hidden, and not
widely-shared, could we expect that people would not want to migrate. But this is most patently
not so in the era of globalization, instant communication, and broad access to TV, movies and
These two factors thus shape the key global political issues. While in the years between
the heady days of the pan-European ―Springtime of peoples‖ in 1848, and probably the second-
half of the 20
century, the conflict between capital and labor was the main political issue that
influenced several generations of thinkers, politicians, social activists, and ordinary people, this
is no longer the case today. Globally, the issue has receded in importance as the objective
conditions that gave rise to it have changed. This was already even if dimly becoming apparent
in the last decades of the 19
century when the term ―workers’ aristocracy‖, denoting this
divergence of living conditions among the ―exploited‖ classes internationally, was coined. To
quote Engels again—but now ten years after The Communist Manifesto: ‖…the English
proletariat is actually becoming more and more bourgeois.‖ 17
In a letter to Marx in 1858, Marx and Engels: Selected Correspondence, p. 132. See also Hobsbawm (1996, pp.
This somewhat derogatory term reflected a real process of betterment of the standard of
living among the working classes in the most advanced capitalist countries, or if one prefers,
their ―embourgeoisement‖. The process continued, and accelerated, most famously during the
―Glorious Thirty‖ years of almost uninterrupted and widely shared economic growth in Western
Europe and the United States. In the second half of the 20
century, the prospect of permanent
revolution, that is of a worldwide revolution that would bring working classes’ parties to
power—a prospect so eagerly called forth by Trotsky in the 1920s, became clearly unrealistic.
The bourgeoisies of the western world needed no longer fear proletarian revolutions spreading
from one country to another. The fact that in 1968, during another pan-European uprising,
superficially similar to that 120 years earlier, workers’ organizations in the most ―revolutionary‖
country (France) were the last to join the protest, unable to formulate any coherent demands, and
in fact never made any moves towards the abolition of capitalist relations of production, showed
very clearly that the times had changed. The specter of Communism, eloquently evoked by Marx
and Engels in the opening sentences of the Communist Manifesto, was exorcised.
The new problem which is likely to dominate the present century is different: it is the
problem of uneven development between the countries and, associated with it, the pressure of
migration emanating from poor countries. It is no less complicated problem than the earlier one,
and involves similar fears of loss of power and income among those who are richer. It is a
problem born out of importance of location for one’s lifetime earnings no less than the earlier
fear of Communist revolution was born out of income differences between the social classes
within the same societies.
4. The key policy issue in this century: Better living standards for the poor people of the
The specter of Communism disappeared because poor people in what are today rich and
upper middle income countries have experienced substantial and sustained increases in their real
levels of incomes. Analogously, the problem of migration will disappear, or become
manageable—in the sense that migration will be just a result of people’s preferences (e.g. those
who prefer to live in a warmer country may relocate there, as is broadly the case today among
the group of rich European countries) and not a problem of massive exodus, prompted by large
gains in real income—only when differences in mean incomes between countries become much
smaller. This implies that the best ―solution‖ to the problem of massive, economically-motivated
migration lies in the accelerated development of the poorest countries of the world and that it is
also in the well-understood (self-enlightened) interest of rich nations to help this process by
increased aid and economic benefits given to the poor countries. The self-enlightened interest
should be particularly obvious in the case of countries that have problems managing large
migration flows whether because of economic reasons (possible downward pressure on wages)
or social reasons (difficulty of accepting a different style of life or a different system of values
brought in by the migrants).
But the development of many poor parts of the globe, even if it were more successful
than during the last 60 years, would still take a long time to bring their average incomes to
anything close to the incomes in the rich world. Thus migration will remain, by its sheer human
importance, the key mechanism whereby incomes of the poor people in the world are to be
raised. Faster development of poor countries and migration are two complementary, and in
many ways very similar, mechanisms. In both cases, the end result is increase in income of the
poor. In one case, the increase is achieved in the place where they live now; in the other case, it
necessitates their movement to a different locale.
The process moreover is hardly new: in
terms of relative importance, migration was much more significant at the end of the 19
The same point is argued by Pritchett (2006, Chapter III in particular).
and early 20
century than it is today.
Not because the demand for migration is less today but
because the impediments are greater.
Impediments to migration pose one methodological issue for economic theory, a point
raised by Edwin Cannan almost a century ago, in his discussion of Adam Smith’s ―invisible
hand‖. Cannan asked ―if…indeed, it [ìs] true that there is a natural coincidence between self-
interest and the general good, why…does not this coincidence extend, as economic processes do,
across national borders [?]‖20 Smith´s argument of ―invisible hand‖ is general and cannot depend
on arbitrary lines such as national borders. The relocation of people ought to be beneficial to
total world output and therefore to reduction of global poverty and (very likely) global
inequality. Hanson (2010, p. 22) calculates that the annual flow of Mexican migrants into the
United States raises global income by an amount equal to about 1% of American GDP.
Walmsley and Walters (2006) find, using CGE simulations, that an increase in both skilled and
unskilled migration equal to 3 percent of their respective labor forces in developed countries,
would yield severalfold more, in net welfare gains terms, than the current development aid.
Migration restrictions are, they argue, more costly than the existing restrictions on trade. In a
recent study of the effects of migration in Spain, Bruquetas Callejo and Moreno Fuentes (2011)
find that the immigration surge that has, in less than two decades, transformed the country from
an emigrant nation to one where foreigners account for 12% of the resident population, has
Pritchett (2006, p.95) criticizes economists who, while acknowledging that both trade and
migration are welfare-enhancing, tend to treat them very differently. They argue for trade
liberalization on general welfare grounds, and then, in a second step, address income distribution
concerns of those who are affected through redistribution. But they never use the same approach
Pritchett (2006, p.69) estimates that the net flows (relative to population) are today between ½ and 1/5 of what
they were in the period 1870-1910.
Cannan´s question is quoted from Frenkel (1942, p. 177), I am grateful to Tony Atkinson for bringing to my
attention this undeservedly obscure reference.
See also the article ―Inmigracion: Espaa sale gañando‖, El Pais, 23 May 2011, available at
when it comes to migration: the equivalence would imply being in favor of migration as a default
position, and mitigating its negative effects through specific additional policies.
That from the global perspective migration should be desirable leaves very little doubt,
even when we think of it using the simplest economic principles: if people are allowed to move
where they think they would do better, it is very likely that total output would increase compared
to the situation when people are not allowed to move. For if the reverse were the case,
impediments to migration similar to the ones that currently exist at the international level would
be found useful and imposed at the national level as well. As Frenkel (1942, p. 183) writes, ―The
movement of men and women from areas where they are poverty-stricken to areas where they
can make their full contribution to the world’s income stream is of advantage to all‖.
While the desirability of both increased aid and greater migration may be established in
principle and for the world as a whole, it does not mean that it would be to the advantage of each
particular country, or particular sections of population in each country. Greater migration may
be associated with reduced wages or increased unemployment for the groups of people whose
skills are most similar to the skills of migrants. Thus even if for the recipient country as a whole
migration is advantageous, sufficiently powerful economic and political groups may be able to
block it or impose tough limits on it. Yet, as nicely put by Hanson (2010, p. 22), ―in considering the
migration rights that maximize global welfare, one could not argue that US workers be weighted
more heavily than both richer [US] employers and poorer migrants‖. Again, going back to the first
principles, we can either treat every person in the world as equally important, or we can put a greater
weight on the welfare of poorer individuals because of diminishing marginal utility of income but we
cannot (logically) pick and choose one group as such in preference to others.
Perhaps more importantly, greater migration faces non-economic obstacles that can be
vaguely described as belonging to the area of social acceptance of migrants and their effects on
domestic culture. These issues have been exacerbated by the current economic crisis and the
unexpectedly great difficulties that many European countries have had in ―absorbing‖ migrants,
particularly those with Islamic background. Thus, in a close succession, both British and German
Prime Ministers have declared the ―multiculaturalist‖ model, which was supposed to be Europe’s
answer to migration, to have failed. Angela Merkel, moreover, pronounced such a model
unambiguously ―dead‖. These problems hold particular poignancy for Europe because it is, due
to its low fertility rates, much more in need of migrants than the United States. Moreover, in its
southern fringes, Europe is surrounded by countries whose demographic profiles are exactly the
opposite of the European: these are countries with a very large share of young population and
relatively few older people. Were it not mostly for non-economic reasons, the demographic
match between the two shores of the Mediterranean would seem almost perfect. However,
Europe which has, at least since the end of the 15
century, ―exported‖ its people elsewhere has
a serious difficulty viewing itself as an immigrant continent and accepting migrants who mostly
come from the areas that were formerly colonized by the Europeans. This reversal of migration
flows seems politically difficult to manage.
From the facts that (i) most of today’s global income inequality is due to differences in
mean incomes between the countries, (ii) in an era of globalization such differences are well-
known to people in poor countries, and (iii) the costs of moving from one place to another are
not prohibitive, it follows that migration, in the absence of significant acceleration of growth in
poor countries, will be a great 21
century mechanism of ―adjustment‖. It will be driven by the
self-interest of individuals but its ultimate result would be a reduction in global inequality and
global poverty. Aid and migration ought to be regarded as two complementary means for
achieving these goals. Policy makers in developed countries shall come to realize that either
poor people will become richer in their own countries or they will migrate to the rich countries.
Bairoch, Paul (1997), Victoires et déboires, volume 1, Paris: Gallimard.
Bourguignon, François and Christian Morrisson (2002), ―The size distribution of income among
world citizens, 1820-1990‖, American Economic Review, September, pp. 727-744.
Broadberry, Stephen and Bishnupriya Gupta (2006), ―The early modern great divergence:
wages, prices and economic development in Europe and Asia, 1500–1800‖. Economic History
Review, vol. LIX, 1, pp. 2–31.
Bruquetas Callejo, Maria and Francisco Javier Moreno Fuentes (2011), Inmigración y Estado de
bienestar en España, Madrid: Colección de Estudios Sociales de la Obra Social La Caixa.
Frenkel, Herbert (1942), ―Presidential Address: World Economic Solidarity‖, The South African
Journal of Economics, vol. 10, No. 3, pp. 169-192.
Hanson, Gordon H. (2010), ―International migration and human rights‖, National Bureau of
Economic Research Working Paper No. 16472, October.
Hobsbawm, Eric (1996), The Age of Capital, 1848-1875, Vintage books. First published in
Li, Bozhong and Jan Luiten van Zanden (2010), ―Before the Great divergence: Comparing the
Yangtzi Delta and the Netherlands at the beginning of the nineteenth century‖, Centre for
economic policy research, London, Working Paper No. 8023, October 2010.
Maddison, Angus (2004), ―World population, GDP and GDP per capita, 1-2000 AD‖. Available
Maddison, Angus (2007), Contours of the world economy, 1-2030: Essays in macro-economic
history, Oxford: Oxford University Press.
Marx, Karl and Friedrich Engels (1978), ―Manifesto of the Communist Party‖ in Robert C.
Tucker (ed.), The Marx-Engels Reader, Second edition, W.W. Norton and Co. First published in
Milanovic, Branko (2011), ―Global inequality recalculated and updated: The effect of new PPP
estimates on global inequality and 2005 estimates‖, Journal of Economic Inequality,
Milanovic, Branko (2011), ―A Short History of Global Inequality: The Past Two Centuries‖,
Explorations in Economic History, forthcoming.
Moellendorf, Darrel (2009), Global inequality matters, Palgrave McMillan.
Kenneth Pomeranz (2001), The Great Divergence: China, Europe, and the Making of the
Modern World Economy, Princeton University Press.
Pritchett, Lant (2006), Let Their People Come, Breaking the Gridlock on Global Labor Mobility,
Washington, D.C.: Center For Global Development.
Union of Banques Suisses (2009), Prices and Earnings: A Comparison of Purchasing Power
Around the Globe, 2009 edition.
Warner, Andrew M. (2002), ―International wage determination and globalization‖, typescript.
Available at http://www.nes.ru/NES10/cd/materials/warner.pdf.
Walmsley, Terrie L. and Alan L. Winters (2006), ―Relaxing the Restrictions on the Temporary
Movement of Natural Persons: A Simulation Analysis¨, Journal of Economic Integration,
Volume 20, Number 4, December 2005, pp. 688 – 726.