GLOBALIZATION AND CONVERGENCE (pagg.

77 90)
INTRODUCTION
The assignment is to analyze and understand how forces pushing for increased
global economic integration have helped or opposed the process by which
economies are supposed to converge in capital intensity, techniques of
production, productivity levels and living standards.
We see globalization everywhere.
During the XIX century we saw falls in the costs of transporting goods all around
the world that made large-scale international trade in staples rather than
luxuries possible for the first time, mass flows of capital, mass migrations.
From 1950 we saw a further advance in international economic integration.
By contrast we dont see convergence everywhere. We certainly see
convergence at some times and in some places, but it has been limited in
geography and in time.
Some decades ago, geo-politicians spoke about countries of the world being
divided into three worlds: first, second (the Communist one that is gone) and
the third (still useful word because it underscores the differences living
standards and productivity levels- in the world today).
Economists expected a pattern of economic growth all around the world for XX
century, instead we notice a striking divergence. World trade, flows of capital,
migration should, together with transportation and communication costs fall,
erode the differences in productivity and living standards between continents
and between national economies. To consider also the fact that every poor
economy has the opportunity to catch up with the rich ones by adopting and
adapting the modern technologies. Yet economists expectations have been
disappointed throughout the past century.
Why economists have been disappointed in their expectation that economic
forces international trade, international migration, international investment and
technology transfer- will narrow the big gap in productivity levels, real incomes
and living standards around the world?
In 1988, Baumol and Wolff set out the idea that it would be useful to analyze the
pattern of economic growth in terms of membership in a convergence club.
They proposed that we should think about this problem by examining the
membership over time of the convergence club (set of economies where the
forces of technology transfer, increased international trade and investment,
spread of education, were powerful enough to drive productivity levels and
industrial structures to or at least toward those of the industrial core). Baumol
and Wolff believed that examining how that economies enter and fall out of this
convergence club should reveal clues to what are the particular economic,
political, and institutional blockages that keep convergence the exception in the
world and not the rule.
The first task to accomplish is to define convergence; convergence is as much a
structural and organizational target as a target indicated by levels of GDP per
worker. In fact if we analyze Saudi Arabia and the Persian Gulf Emirates, we

notice that levels of GDP per worker and standards of living are equivalent to
those of Western Europe, but we wouldnt claim that they have converged to
the industrial core. If we analyze Argentine or Australia during a generation
before WWII, their productivity levels were falling relative to those of the
industrial core because of declining trade and drought; these countries in this
period built their industrial sectors and raised their educational levels. We must
say that they converged to the industrial core in economic structure even if they
were losing ground in terms of standards of living and value of output per
worker.
So our definition of which economies are in the convergence club over a time
period is not merely those countries in which GDP per capita as a proportion of
the North Atlantic level rose over the time period in question; it looks at the
extent of industrial development and structural change as well.
THE CONVERGENCE CLUB
We examine the evolution of the worlds convergence club over time by taking
snapshots of its membership during four different eras over the past two
centuries:
-

1820-1870: because industrialization is considered the key measure of

modernity or development in the middle and late nineteenth century, the
convergence club of this period was very small indeed.
In 1820 the Industrial Revolution was in full swing in Great Britain and as the
pace of structural change and industrial development accelerated, its
technologies began to diffuse to the continent of Europe and overseas to
North America.
In this interval only Great Britain, Belgium and the northeastern United
States are part of the convergence club. Despite other countries began to
industrialize (Canada, the rest of United States, Netherlands, Germany,
Switzerland, Austria, Czech Republic, France), all of these economies were
further from Great Britain in industrial structure in 1870.
Lets consider that for this period we focus on industrial structure rather than
economy productivity or labor productivity; if we consider Netherlands, its
beyond a doubt that it was more prosperous in overall terms than Great
Britain in 1820 and even in 1870 the gap was relatively small. But it was still
a rich merchant economy, not yet an industrializing one.
Thats why in this interval the convergence club is very small.

1870-1913: we call this interval First Era of globalization.

The convergence club expands considerably: the countries defined of
temperate European settlement Canada, western United States, Australia,
New Zealand- plus Argentina, Chile, Uruguay, perhaps South Africa and
obviously all the countries of inner Europe- Belgium, Netherlands, France,
Germany, Switzerland, Spain, Austria, current Hungary, current Czech
Republic, Denmark, Norway, Sweden, Finland and Ireland-.
Because of there are a lot of countries and regions that fail to join the
convergence club, we state that the scale of this first era of globalization is
limited.

Lets think about the consequences of international trade, international

investment, international migration and international conquest. They
profoundly affected economic, social and political structures throughout the
world. Lets think about British Empire that brought for example the tea plant
from China to Ceylon; but this did not trigger any rapid growth in real wages,
acceleration in productivity growth or industrialization, rapid growth in
factory employment, convergence to the worlds economic core.
-

1913-1950: Its an extended interwar period defined by an enormous

destruction caused by two world wars and an enormous destruction brought
by the Great Depression.
The convergence club increase its members: Japan, Korea and Taiwan (part
of Japanese Empire), southern United States, the Soviet Union (Stalinist
industrialization was a disaster for human life, social welfare and economic
efficiency, but it was a powerful motor of industrialization), Venezuela, Peru,
Brazil (in Latin America), Morocco, Algeria and Tunisia (French North Africa).
These countries appears to grow as rapidly as in the industrial core.

1950-2000: in this period the convergence club both expanded and

contracted massively because for the first time many economies joined and
many dropped out. The implications of globalization for the size of the
convergence club in this second era of globalization are less clear because
its difficult to understand why it has been favourable to some countries such
as East Asia and unfavourable to others like Latin America.
Countries like Venezuela, Peru, Argentina, Chile and Uruguay declined.
Costal West Africa and Eastern Africa as well fell out of the convergence club
(if they ever joined); the countries of French North Africa fell out of the
convergence club because Morocco, Tunisia and Algeria fell behind France in
productivity and industrial structure; former Soviet Union dropped its
membership because of first the stagnation and later the economic collapse
of the communism. We can argue that the failures in economic development
after WWII were largely political failures.
As these economies fell out of the convergence club, other economies
joined: East Asian miracle (Japan, South Korea, Taiwan, Hong Kong,
Singapore, Thailand, Malaysia, Indonesia (after 1965) and China (after
1978)), the Balkans (Yugoslavia, Romania, Bulgaria), in the Eastern
Mediterranean (Greece, Turkey, Israel, Egypt), in Latin America (Colombia
and Mexico) and after 1980 India begins to grow and to narrow the gap in
productivity and industrial structure.

DEBATING CONVERGENCE WHILE INCOMES DIVERGE

Recent debates on growth theory have contrasted the convergence predictions
of the Neoclassical Growth Model of Swan and Solow with new models
predicting non-convergence.
The neoclassical representatives state that in the short run growth is
determined by moving to the new steady state which is created only from a
change in capital investment (from the change in the savings rate), labor force
growth and depreciation rate (population increase doesnt actually affect the
amount of capital in the economy, but it decreases the amount of capital per

worker; consequently depreciation, that is the rate at which capital wears out,
increases. To retain an unchanging level of capital per worker over time, we
have to invest enough to create new capital to counterbalance the loss over
time.)
In the short run an increase in the saving rate raises the growth rate of capital
per worker; in the long run the growth rate of capital per worker is the same
zero for any saving rate. In this long-run or steady-state situation a higher
saving rate leads to higher steady-state capital per worker (denoted with k*),
not to a change in the growth rate which remains zero.
In the long run the model predicts that growth is achievable only through
technological progress because in the short run an increase in the technology
level raises the growth rates of capital per worker and GDP per worker and
these growth rates remain higher during the transition to the steady state, so in
the long run, the growth rates of capital and real GDP per worker are the same
zero- for any technology level, but in this long-run or steady-state situation, a
higher level of technology leads to a higher steady-state capital k* and real GDP
y* per worker, not to changes in their growth rates which remain at zero.
So we found that the Solow model predicted convergence across economies in
capital per worker. But the key assumption that the determinants of the steadystate capital per worker k* are the same for all economies is reasonable only for
similar economies, not for countries having different economical, political and
social characteristics.
When we consider convergence, we observe a lack of absolute convergence for
a broad group of countries, making us thinking that the Solow model is failing.
An extension of the model is to consider Conditional Convergence in order to
explain this apparent failure. Conditional Convergence allows us to understand
many other features of economic growth in the world.
The idea is to measure an array of variables, each of which influences a
countrys steady-state capital per worker k*.
Perhaps the most important advocate of the position that there is convergence
in the world is economist Robert Barro, which focus on the notion of conditional
convergence. In his view, strong and powerful forces are pushing countries
together; he create a regression that shows that countries can close between
2.5 and 4.5% of the gap between them and the industrial leaders countries each
year BUT with the assumption that other things (defined right-hand-side
variables of Barros regression) could be made equal.
Which are these other things?
They are an index of democracy, an index of the rule of law, government
speculative spending as a share of GDP, life expectancy, male secondary-school
attendance rate and fertility rate.
Its not difficult to admit that these values cannot be brought to the mean
values of the core industrial countries if the country we analyze hasnt already
achieved the productivity level and the socioeconomic structure of the industrial
core leaders.

If we consider a country like Mozambique for which the book gave us the idea
that, other variables being equal, it could close the gap in productivity with the
most industrialized countries in 16 to 29 years, we can state that it isnt
possible because for example Mozambique couldnt have a life expectancy and
a secondary-school enrolment as high as those of the industrial core.
We can see no way of bringing the poor-countries values of Barros variables to
their rich-countries means because it means that full economic development
hasnt already been accomplished.
DeLong and Summers include in the list of right-hand-side variables some
measures of investment.
Poor countries face a high relative price to acquire the capital equipment they
need to turn savings into productive increase in capital stock. It means that
poor countries get less investment out of saving some fixed share of their
income.
The coexistence of actual divergence and conditional convergence reflects the
observation that some of the conditioning variables in the convergence
regression are distributed in such a way as to promote divergence.
Lets consider an autarkic model; it implies the existence of a development or
poverty trap: population living close to subsistence is unable to mobilize the
surplus required for substantial domestic investment because they typically
face high prices for imported capital goods.
An explanation for the limited convergence over the past century and a half is
promoted by Easterlin who attributes limited convergence to a lack of formal
education throughout much of the world; he says that the diffusion of modern
economic growth has depended principally on the diffusion of knowledge about
the productive technologies developed since the Industrial Revolution; this
knowledge cannot diffuse to populations that havent acquired the
characteristics produced by a formal schooling. Obviously political conditions
and ideology play the biggest role in restricting the spread of education, but
Easterlin hopes for a world in which formal education spreads and so blockages
to convergence disappear.
The most interesting contribution comes from Gregory Clark who suggests that
the chief obstacle to convergence was the relative inefficiency of labor.
In order to demonstrate his idea he studied the state of cotton mills worldwide
around 1900; he found out that technologies of automated cotton-spinning had
been successfully transferred all around the world (Manchester, Shanghai,
Tokyo, Bombay, all used the same technology and the same machinery). Labor
productivity varied by a factor of ten-to-one worldwide offsetting the variation in
real wages worldwide and so making the profitability of the cotton-spinning mills
equal everywhere.
Clark points out that the biggest part of the total cost was labor cost, but, if
labor productivity havent varied directly and proportionately with the real
wage, all the cotton-spinning industry would have been located in low-wage
economies. Because it wasnt, Clark understood that the obstacle to
convergence was the relative inefficiency of labor.

We can contest Clarks idea making clear that, in order to do his comparisons,
Clark needed to find an industry in which its profitable to locate in a country no
matter which level of real wages it presents. When is it profitable to locate in
whichever place? When labor productivity is proportional to the local real wage.
So Clark could arrive at his conclusions without any of his calculations, simply
by observing the reality.
Lets conclude asking how common are industries like the cotton-spinning one;
its difficult to find industries in which labor productivity worldwide is
proportional to the local real wage. When labor productivity varies less than the
real wage the industry tends to be located in relatively poor countries; when
labor productivity varies more than real wage the industry is located in the
richest and most technologically capable economies.
We are in a world in which some forces push towards convergence while others
towards divergence. We presume that globalization should reinforce the
convergent trend, so why it failed to produce convergence over the past fifty
years?
Lucas suggests that human capital complementarity may block the capital
channel. When physical and human capital are complementary, the problems of
moral hazard in human capital investment explain the failure of international
capital markets to invest in the capital-poor economies.
Similarly for the international transfer of technology that has the lack of social
capability as the major obstacle preventing the backward countries from
acquiring the technologies of the developed countries. This hypothesis is
supported by scholars who find evidence that the growth rate of total factor
productivity depends on the national stock of human capital.
An alternative explanation for the failure of globalization to bring about
convergence comes from Sachs and Warner that constructed an index of
openness for the period 1970-1989 (1 for open economy while 0 for close one
according to at least one of the following criteria: tariff rates over 40%, nontariff barriers covering at least 40% of imports, socialist economic system, state
monopoly of major exports, black market premium of >= 20% on foreign
currency).
They found evidence of strong convergence in per capita GDP among the
countries classified as open, but no convergence among the closed ones. Sachs
and Warner states that globalization has indeed promoted both growth and
convergence but only to those countries that allowed free movement of goods
and capital.
The Sachs and Warner evidence has been criticized by Rodriguez and Rodrik
who said that S.&W. measure is so correlated with plausible alternative
explanatory variables that it is risky to draw strong inferences about the effect
of openness on growth.
Lets examine the robustness of the Sachs-Warner result.
Descriptive statistics are given for 109 countries in three intervals 1960, 1980
and 1998 and we are considering the real GDP per capita.

Proceeding with a first analytical examination for the period, we notice that
dispersion is measured by variance and it is connected to divergence. Lets see
our results: we have divided the sample into 3 groups depending on whether
1960 real GDP was above I$ 5,000 for rich countries, below I$ 1,500 for poor
countries and between the values of the range for the middle ones (I$ are used
to use international prices and normalize to the purchasing power of the US
dollar in 1985).
All three measures show increasing dispersion: divergence has occurred within
each group except for the richest 19 countries (between 1960 and 1980); the
richest 19 countries converge during the interval 1960-1980 and we understand
that because of variance that declines from 1960 to 1980. The principal cause
of divergence has been the failure of the poorest to match the growth of the
more developed countries; between 1960 and 1980, the middle-income
countries grew fastest followed by the rich ones and the poorest. Pay attention
to the fact that over the subsequent two decades growth rates slowed for all
groups!
A second analytical examination takes place for the interval 1960-1998 on a
sample of 96 countries because we introduce also the S&W distinction between
open and closed economies and reclassifying countries as open if S&W report
that they have been open for a significant number of years since 1980. We are
testing the impact of openness on convergence with regressions.
The results for the interval 1960-1980 are that openness lifts per capita GDP (so
open economies grow faster than closed ones), openness tends to be more
important for poor countries because growth rates are higher for low-income
economies that open, investment rates and demography matter because lower
investment rates and high fertility slow down economic growth of poor
countries. By adding a variable measuring the level of schooling in the adult
population we see that it has no significant explanatory power.
Performing a similar analysis to explain growth between 1980 and 1998 we
verify that openness is less important (it isnt clear the advantage of being
open), openness is less beneficial for poor countries than before (so it questions
Sachs and Warners results), investment rates and demography matter more
than before (these bottlenecks slow down poor countries growth than the
previous period), education begins to matter because human capital and
technology become relevant because there is support for the hypothesis that a
high level of initial human capital does promote growth especially when initial
income is low and that successful technology transfer requires a certain level of
social capability).
So for the overall period from 1960 to 1998 we can say that we observe
divergence rather than convergence especially for low income economies, and
that openness is correlated to economic growth but not always to convergence
(a large number of the poorer countries have opened since 1980, but precisely
during this period the benefits of openness have diminished) and it stresses the
main factors of divergence (low income, low investments, lack of education and
population increase).
What do we conclude?

There isnt evidence of a direct relation between globalization and

convergence
Openness doesnt seem to be the solution for the great divide; after WWII
it seemed the solution making for convergence, but from 1980 it isnt like
that anymore
Pre-industrial traps (conditions of subsistence income, low saving and
investment, low level of education and high fertility), heritage of old
biological regime, seem to be the main blockades to convergence for poorest
countries
It remains an open question whether the growth benefits of openness have
really declined in recent decades or whether an early turn to openness is
correlated with other growth-promoting factors omitted from standard crosscountry studies.