Waves of Structural Globalization since
1800: New Results on Investment Globalization
Christopher Chase-Dunn, Andrew Jorgenson, Rebecca Giem,
Shoon Lio, Thomas E. Reifer, and John Rogers
Institute for Research on World-Systems
https://irows.ucr.edu
College Building South
University
of California-Riverside
Riverside,
CA. 92521-0419
chriscd@mail.ucr.edu
To be
presented at the annual meetings of the American Sociological Association,
August 16-19, 2002, Chicago. This
research is supported by the National Science Foundation, Sociology Program. Draft
v. 8-14-02. words=8153. Thanks to Carol Bank and Siobhan Tuthill for their help
with this research. This paper is available at https://irows.ucr.edu/papers/irows10/irows10.htm
This paper discusses research that is designed to examine
the historical trajectory of structural globalization as an attribute of the
whole world-system. Did the globalized world economy arrive all at once in a
rapid and recent transition from national to global economic networks? Or is
the process of international integration a long-term trend that has been going
up for centuries only to be noticed recently because it has reached such a high
peak? Or, alternatively, is globalization
a cyclical phenomenon in which the world-system alternates between periods of
national autarchy followed by periods of international economic and political
integration?
We propose a conceptualization
of structural globalization as several inter-related dimensions of the
expansion and intensification of interaction networks. The real trajectories of
different kinds of globalization over the last two centuries are knowable only
if we gather comparable data over time.
Studies of recent decades do not answer the question of the shape of
long-term trajectories. Our project
improves upon data for the nineteenth century and splices earlier, cruder
measures with later, more refined and more complete data series.
The two main objectives of our
research are:
ｷ to
determine the trajectories of trade and investment globalization; and
ｷ to
empirically examine the relationship between these and several other
world-system variables that have been hypothesized to cause international
economic integration.
The trajectories of different types of globalization have
important implications for our understanding of the processes of development in
the modern world-system. In this paper we present new results on the trajectory
of investment globalization.
Both the popular and the academic
discourses about globalization contain great confusion and disagreement
regarding the meaning and connotations of this contested term. We contend that
the scientific study of globalization can move forward by making a clear
distinction between globalization as greater integration and interdependence of
the world-system, on the one hand, and the political discourses that employ
ideas about global integration and competition to justify actions and policies
on the other. Our research will distinguish between:
ｷ
globalization as ideology, and
ｷ
globalization as objective structural trends of spatial
integration.
Our main focus is on different types of structural
globalization, but we are also interested in understanding changes in the
ideologies that are used to legitimate the actions of the powerful. Giovanni
Arrighi is researching the transition from Keynesian theories of national
development to the neo-liberal 展ashington Consensus. Phillip McMichael (1996)
describes the emergence of what he calls the 堵lobalization project a
revitalized glorification of market mechanisms as allegedly efficient antidotes
to rent seeking and the 砺ampire state.
The 堵lobalization project emerged with Thatcherism and Reaganism in
the 1980s, and has swept around the world as a justification for attacking and
dismantling welfare states and labor unions following the demise of the Soviet
Union. While this is an interesting and consequential phenomenon, it is not to
be the main focus of the research here proposed.
Rather we intend to determine
the real temporal trajectories of structural dimensions of global integration
over the past two hundred years. We understand structural globalization as
composed of different inter-related dimensions of expanding and intensifying
interaction networks especially political, economic and cultural
globalization (Chase-Dunn 1999). We
specifically reject the notion that these dimensions constitute completely
different aspects of social reality that should be studied separately by
different academic disciplines, but we contend that it is useful to distinguish
between them in order to understand how they have affected one another.
Our research has already shown
that trade globalization is primarily a cyclical phenomenon, though the most
recent upsurge has reached a level that is significantly higher than the level
reached at earlier peaks (Chase-Dunn, Kawano and Brewer 2000). In the future we propose to determine the
trajectories of:
ｷinvestment
globalization the extent to which international capital flows and
investments increase (or decrease) in
relationship to the size of the world economy; and
ｷ
political globalization -- the degree
to which the multicentric international system has moves toward centralization,
integration and hierarchy.
The quantification of political
globalization will require measures of the relationship between the power and
sizes of large and small political and military organizations in the
world-system. We propose to operationalize these characteristics of the whole
world-system over the past 200 years in order to compare their temporal
trajectories with that of trade globalization, and to examine their
hypothesized causes.
We
define structural globalization as the
increasing spatial scale and intensity of interaction networks. Charles
Tilly (1995:1-2) proposes a similar definition of globalization as "an
increase in the geographic range of locally consequential social interactions,
especially when that increase stretches a significant proportion of all
interactions across international or intercontinental limits." If both
national level and global networks increase in intensity at the same rate, this
approach would not see an increase in the globalization of interaction. Globalization in the structural sense is
increasing integration and interdependence. As with other efforts to
measure globalization (e.g. Chase-Dunn, Kawano, and Brewer 2000), the
estimation of a global characteristic needs to take account of the changing
size of the system as a whole. Of
course there are more transnational interactions now than there were in the
nineteenth century. There are also more
within-nation interactions because the world population and the world economy
have become larger. It is the ratio of
these that must be studied.
Human interaction networks have
been increasing in scale and intensity for millennia as transportation and
communications technologies have made regular trade and interaction over
greater distances possible. It is obvious
that the railroad and the steam ship facilitated a massive increase in the
spatial scale of interaction networks.
Ideally we would like to trace the relative degrees of interaction
integration at several levels.
Households exchange goods and ideas with other households. Neighborhoods
and towns exchange with other neighborhoods and towns, cities with cities and
etc. But such a study is not feasible
at the present for two reasons: our unit of analysis is the world-system as a
whole meaning all the countries of the world; and we want to examine trends
over the past 200 years. It might be
possible to examine local and regional interaction networks for a particular
country or for a few core countries in recent decades. But in order to study
the whole system over two centuries we must necessarily use data on the units
that have been the main data-gatherers in this period of human history the
national states.
By this decision we do not mean to imply that national states are
the only, or even the most important, actors in the world-system. We recognize
the importance of transnational relations emphasized by political scientists
thirty years ago (Keohane and Nye 1970; and more recently by Sklair (1995,
2001). The world-systems perspective
has long pointed out that the interstate system the system of sovereign
national states -- is only one institutional structure of the global political
economy. The world-system is composed
of individuals, households, towns, cities, regions, firms, classes, states, and
other non-governmental and international organizations. It is not simply a matter of 妬nternational
relations. The world-system is the whole
system, not just relations among states.
Transnational relations occurring across state boundaries between all
these social actors are not a new, or a recent, phenomenon. Intersocietal migrations and trade among
individuals, families and firms have been important aspects of small, medium
and large world-systems for thousands of years. There was never a time in which
the members of different societies did not importantly interact with one
another (Chase-Dunn and Hall 1997). But
the spatial scale of both societies and intersocietal interactions has grown.
And intersocietal integration only became global in the sense of linking every
region of the Earth into a single network in the nineteenth century. The unit of analysis we will study in this
research is the modern Europe-centered system as bounded by the system of
allying and conflicting states. It was during the nineteenth century that the
states systems of East Asia and Europe became linked by political/military
interaction, though they had long been linked by the exchange of prestige
goods..
If we think of the world
economy as a system the phenomenon of globalization should represent increases
in the intensity of global interaction networks relative to the intensity of
local interaction networks. If both
local and global interactions increase at the same rate it would be mistaken to
say that the system is becoming more globalized. It is when global interactions increase at a greater rate than
local interactions that the system qua
system is more integrated at the global level. In order to study globalization
in this sense we need to measure the intensity of both global and local
interactions.
This study will focus on variable characteristics of the
world-system as a whole. The questions
we are asking here are about the continuities and changes at the level of the whole
system, and so our empirical strategy will be to construct measures of how this
single larger system changes over time. For this reason we have only one
田ase, though we can utilize the method of time series analysis to test
propositions about the relationships among variables in this single case
(Chase-Dunn 1998:Chapter 15).
Internationalization
of finance and investment, the growth of international trade as a proportion of
all economic interaction, and the organization of production on a global scale
by transnational corporations have undoubtedly increased in the last two
decades. There are potentially a large number of different indicators of
economic globalization and they may or may not exhibit similar patterns with
respect to change over time.
Trajectories of Trade and Investment Globalization
We have constructed an improved
measure that shows that there have been three waves of trade globalization
since the early nineteenth century. Our
new measure of trade globalization extends yearly data further back in time and
greatly improves the time resolution relative to the widely spaced estimates
that had been previously available. Our new measure of 殿verage openness trade
globalization estimates the world level based on averages of country ratios of
GDP to imports. Because both GDP and
imports are available in country currencies (e.g. francs, pesos, etc.) we are
able to estimate trade globalization without resorting to the problematic
assumptions involved in converting country currencies into a single currency
(e.g. U.S. dollars), and we did not have to convert the current values into
constant values using estimates of inflation and deflation. The results of our study was published in
the American Sociological Review
(Chase-Dunn, Kawano and Brewer 2000).[1] Our
study of trade globalization shows that it is a cyclical phenomenon, as well as
containing a long-term upward trend based on the comparison of the peaks of the
cycles (see Figure 1).
Figure
1: Average Openness Trade Globalization, 1830-1992 (Weighted)
It is possible that investment globalization behaves in a similar
way, but we do not know for sure. Existing estimates of investment
globalization (e.g. Bairoch 1996) are even more intermittent than estimates of
trade globalization were before we undertook our ASR study. It would be
desirable to have a better understanding of the relationship between investment
and trade globalization and to be able to study the causes of both.
The Trajectory of Political Globalization
We
conceptualize political globalization analogously to our understanding of
economic globalization as the relative strength and density of larger versus
smaller interaction networks and organizational structures. Much has been written about the emergence
and development of global governance and many see an uneven and halting upward
trend in the transitions from the Concert of Europe to the League of Nations
and the United Nations toward the formation of a proto-world state. The
emergence of the Bretton Woods institutions (the International Monetary Fund
and the World Bank) and the more recent restructuring of the General Agreement
of Tariffs and Trade as the World Trade Organization, and the visibility of
other international fora (the Trilateral Commission, The Group of Seven
[Eight]; the Davos meetings, etc.) support the idea of emerging global
governance. The geometric growth of
international non-governmental organizations (INGOs) is also an important
phenomenon of global governance and the emergence of global civil society
(Murphy 1994; Boli and Thomas 1999).
All
world-systems go through cycles of political centralization and
decentralization with occasional leaps toward new and higher levels of
political integration (Chase-Dunn and Hall 1997). In the modern world-system
the cycle for the last 400 years has taken the form of the rise and fall of
hegemonic core states. Some claim that this hegemonic sequence is now morphing into a new structure of core
condominium (Goldfrank 1999). We intend
to study both the hegemonic sequence and emerging global governance. While
these might be combined into a more general concept of political globalization,
we contend that it is important to keep them separate because hegemonic rise
and fall is an old feature of the world-system, whereas political globalization
is arguably much more recent. Political
globalization can be analytically reduced to the question of the relative
strength of larger vs. smaller political and military organizations (including
also the functionally 兎conomic ones (IMF, World Bank, WTO) mentioned above.[2].
There is a single size
distribution of political/military organizations in the world-system. We plan
to operationalize three different parts of this size distribution, as well as
the whole thing. Our conceptualization
of political globalization will be analogous to our understanding of economic
globalization a ratio of the size and importance of global and international
organizations vs. the sum of size and importance of national (and
multinational) states. But we will also
operationalize the hegemonic sequence by examining changes in the distribution
of economic and military power among the core states using the research of
Modelski and Thompson (1996). And we
will study the changing shape of the whole system of states as well, taking
into account the processes of colonization and decolonization (Bergesen and
Schoenberg 1980), the incorporation of the peripheral and semiperipheral
regions into the interstate system, and changes in the distribution of economic
and political/military power in the whole system of states. We will also combine political
globalization, hegemonic rise and fall, and state power stratification into a
single overall measure of the distribution of power among state and proto-state
institutions. This latter we will call 登verall global political/military
inequality.
Measuring Investment Globalization
We have assembled an annotated
bibliography on sources of information about international investments since
1800.[3] In
principle, investment globalization is the proportion of all invested capital
in the world that is owned by non-nationals (i.e. 吐oreigners). In practice we cannot easily measure the sum
total of all invested or loaned capital (or the amount of domestically owned
capital) over the desired time period, so we use the total of all the national
GDPs to estimate the economic size of the world economy. World GDP will serve
as the denominator of our 努orld totals estimate of investment globalization.
The numerator will include
most, but not all, international capital flows, ownership claims and debts. We
will not include transfer payments made by individuals to their families in
other countries because these are not economic investments of the kind we want
to study. We do not include payments for imported or exported goods -- these
are the basis of our measure of trade globalization. Nor do we include foreign
reserves held by central banks in order to support their currencies in the
world money market. But we do include
loans and direct equity investment, profits (repatriated or not) and intrafirm
transfers that cross state boundaries regardless
of whom the parties to the
transactions are. The transacting
parties may be individuals, firms, banks or governments. In principle we want to measure all of the
international financial transactions that involve claims of ownership,
control or debt irrespective of who the parties are. And ideally we would like to systematically
distinguish among these different kinds of international capital flows and
obligations to see how they are similar or different in their geographical and
temporal distributions.
This latter desideratum will
only be possible for the period after World War II. Before that we will find
different combinations of the several types of international capital flows and
obligations in the available data and we will need to be careful about how we
combine and splice data series that contain different components. For example, for early periods it will be
easier to get data on loans made to governments, than to find information on
loans made to private firms or individuals within a country. Whenever possible we will continue the less
inclusive measures into periods in which more comprehensive and decomposable
information is available and we will overlap less complete indicators with more
complete ones. This will enable us to splice different data series in a more
sophisticated way than simply switching from one to another as more complete
data become available.[4]
We also
need to pay close attention to the important distinction in international
capital data between stocks and flows. Stocks are the total accumulation
of debts or the book value of foreign investments at a particular point in
time, while flows are the amount of moneys that flowed in (or out) over a short
period, usually one year.
We are
pursuing two different strategies for constructing long-term measures of
investment globalization. These are
loosely analogous to the 努orld totals and 殿verage openness strategies that
we describe in our ASR study of trade globalization. The first will
involve gathering data on the main investing countries, (e.g. Britain, France,
Germany, the United States, the Netherlands, Belgium and Switzerland) on both
the outflows and the accumulated values of foreign loans and investments. This
is the strategy that has been employed in earlier studies. It assumes that the
great bulk of foreign capital comes from these countries and so efforts are
concentrated where they reap the greatest informational returns. The disadvantage is that the number of
countries with significant capital outflows increases over time and it is
difficult to know how the missing cases might be affecting the estimate of the
value of international capital. This method also requires the problematic
assumptions involved in converting values into a single currency unit for
purposes of comparison of different countries, and converting current into
constant units for comparing over time.
Nevertheless we propose to
upgrade the currently available estimates that use this approach by adding data
from more investing countries. Suter
[1992:Appendix (f)] has compiled the most complete long-run data series on the
value of international capital holdings. We propose to improve upon Suter痴
compilation by adding data from additional investing countries and splicing the
early series to a series compiled from more complete data after 1950. We will
also disaggregate the 渡et figures used by Suter whenever possible for the
countries that he did cover. Net figures
are the balance of credits and debits. In most previous studies of changes in
the level of foreign investment net values have been used (e.g. Suter 1992).
The problem here is that a country may have large amounts of capital invested
abroad and large investments from abroad and these will cancel each other when
net values are used. [5]
Our
second strategy is similar to the 殿verage openness approach we developed for
studying trade globalization. This will
involve estimating 妬nvestment dependence for each country -- the ratio of the
foreign debt to the national income (GDP), and then taking the weighted average
of these as our indicator of world investment globalization. The advantage of this approach is that it
does not require converting into a single currency and computing constant from
current values. We already have the
country currency GDPs (national income estimates) from Mitchell
(1992,1993,1995) that we used for our measure of average openness trade
globalization. For our new 殿verage
investment dependence measure of investment globalization we will need to
collect estimates of inflows and accumulated stocks and debts of foreign
capital in country currencies for each country.[6] With complete data these indicators would be
equivalent to the total sum of international capital flows and obligations
divided by the world GDP. But as with average openness, we will not have
complete data for the years before 1950. This will be a 都ampling problem in
which the sample is biased because we will have more core countries than
peripheral countries. This indicator
will be compared with the results of our first strategy discussed above.
Analyses
The first task of analysis will
be to use the new data we have coded on international capital flows and
obligations to construct several new indicators of investment globalization.
Then we will see how these relate to one another and study their temporal
trajectories in comparison to what we have found for trade globalization. We expect that investment globalization will
show a similar cyclical pattern, but it may not. We also will consider the
question of a long-term trend in investment globalization, much as we did in
our study of trade.
At the time of this writing, August 2002, we have coded and analyzed data on non-net British foreign investment flows from 1865 to 1914 that were not available to Suter (Stone 1999), and we have coded several measures of investment globalization from 1938 to 1999. Some preliminary results and discussion of British foreign investment flows in the nineteenth century are included in Appendix below.
Investment Globalization, 1938-1999
These measures are based on credits and debits on investment income coded from the International Monetary Fund Balance of Payments Yearbooks. Investment income includes the repatriated profits on direct investment and dividends on portfolio investment. Credits means that such was earned from abroad. Debits means that investors abroad were paid profits and dividends by their holdings in the country.[7] In principle, if we had data on the whole world, the sum of credits should equal the sum of debits. But of course we do not have complete data until very recent years, so it is useful to compare credits with debits to see how our 都ample of countries may be biasing our estimates of this variable characteristic of the whole world economy (investment globalization).
Investment income has been used as a proxy for estimating the total book value of foreign investment, because there is a general profit rate that averages around 10% and so the profit can be used as a proxy.[8] And the amount of profit and dividends produced by foreign investment are also of interest in their own right.
We coded investment income debits and credits for all available countries from the IMF Balance of Payments Yearbooks from 1938 to 1999. We found that the CD ROM datasets made available from the IMF do not always include all the information that was published in the original Balance of Payments Yearbooks and so it is important to use both the printed volumes and the CD ROM.
Our new measures of 殿verage
investment dependence/dominance investment globalization estimate the world
level based on averages of country ratios of GDP to credits and debits of
investment income. Because both GDP and
investment income are available in country currencies (e.g. francs, pesos,
etc.) for many countries we are able to estimate investment globalization
without resorting to the problematic assumptions involved in converting country
currencies into a single currency (e.g. U.S. dollars). Exchange rates vary
because of monetary regulation regimes such as the Bretton Woods agreements,
and because of speculative activities of money traders. And we do not need to
convert the current values into constant values using estimates of inflation
and deflation. Removal of the error
introduced by exchange and inflation rates provides superior estimation of
investment globalization.[9]
Another advantage of using the IMF Balance of Payments Yearbook data is
the avoidance of net values. In our analysis of Balance of Payments data we
examine both credits (the returns received by a country for investments
abroad), and debits (the amounts paid out to foreign investors).
In order to estimate the world level of investment globalization we need to weight the country values. Treating large and small countries equally produces an average that overvalues the information from small countries. We weighted the investment ratios for each country by the ratio of the country痴 population size to the world population. The weighted and unweighted averages were compared to make sure that our weighting did not produce strange results. The problem is that the 都ample of countries upon which we are estimating the world level of investment globalization changes over time. Most usually we have more complete data on core countries than on non-core countries in earlier time periods,[10] but the number of countries does not rise in a nice even trend. Figure 2 shows the changes in the number of countries for which we have estimates for debits of investment income since 1938.
Figure 2: Number of Countries with Debits on Investment Income Data, 1938-1999
Notice that in Figure 2 the number of countries for which we have information on debits on investment income decreases during the 1970s. Researchers at the IMF have been unable to explain to us why the availability of data decreases in that period. Notice also that the data availability decreases after 1995. This is because some countries are tardy in reporting their international financial statistics. The pattern of availability is almost exactly the same for Investment Income Credits.
We present the results of our
study using three different indicators from the IMF Balance of Payments
Yearbooks total investment income; direct investment income and portfolio
investment income. Total investment income is the sum of direct investment
income, portfolio investment income, and 登ther investment income. The
definitions of these current account items are as follows:
Investment income covers
receipts and payments of income associated, respectively, with holdings of
external financial assets by residents and with liabilities to nonresidents. Investment
income consists of direct investment income, portfolio investment income,
and other investment income. The direct
investment component is broken down into income on equity (dividends, branch
profits, and reinvested earnings) and income on debt (interest); portfolio
investment income is broken down into income on equity (dividends) and income
on debt (interest); other investment income covers interest earned on other
capital (loans, etc) and, in principle, imputed income to households from net
equity in life insurance reserves and in pension funds (Balance of Payments
Statistics Yearbook, 2001).[11]
We have coded the three subcategories of Investment Income because we are interested in the differences between foreign direct investment and portfolio investment. Direct investment involves organizational control, as when a transnational corporation invests in its subsidiary. Portfolio investment does not involve direct organizational control. The purchase of stocks of firms or of bonds by foreigners is considered portfolio investment. The subcategories are available for many fewer cases and for only recent years. Nevertheless we shall be interested in comparing direct and portfolio investment globalization.
Figure 3 shows the weighted average investment income debit ratios for 1938-1999. Recall that investment income debits include the profits, interest, and dividends earned by foreigners within a national economy. Not surprisingly this estimate of worldwide investment globalization goes up over this forty-year period. But the yearly data enable us to examine the exact temporality of the rise.
Figure 3: Weighted Total Investment Income Debits, 1938-1999
The results in Figure 3 show the averages for all the countries for which we have data (the number of which increases greatly over time as shown in Figure 2), for just the core countries and for the non-core countries. The changes in Figure 3 are partly due to changes in the level of worldwide investment globalization and partly due to changes in the availability of data over time. In order to take out the part due to changing N we use the method of constant groups, looking at a set of countries for which we have data over the whole span of time. In Figure 3 the constant group includes eleven countries ( Australia, Canada, Denmark, Finland, Ireland, Italy, Netherlands, Norway, Sweden, United Kingdom and the United States). The close similarity between the results for the constant group and for the core is due to the fact that the core countries are the ones for which we have the most data further back in time.
The trajectory of investment globalization indicated by including all the countries for which we have data on debits on investment income shows that there was a decrease between 1938 and 1946, and that was followed by a slowly accelerating upward trend. The 1938 level was not reached again until 1979, but the level reached by 1999 was almost three times higher than the 1938 level. Thus the perception that the world economy experienced a wave of investment globalization in recent decades is confirmed by our results using our new 殿verage investment dependence estimator based on debits on investment income.
Figure 3 also shows that the trajectory has been rather different for the core and the non-core. In 1938 the non-core countries were more than twice as dependent on foreign investment as were the core countries and the non-core declined from this high level until the late 1950s, while the core began a new ascent from 1946 on. In 1975 the core began a sharp upturn in payments out on investment income, whereas the non-core was experiencing a much slower increase. This indicates that investment globalization occurred much earlier and to a much greater extent for the core countries than for the non-core.
Now let us examine the credit component of investment income. This is composed of the profits, interest and dividends taken in by countries as a result of their investments abroad, a feature we call 殿verage investment dominance. Figure 4 shows the weighted trends since 1938 for the same groups studied Figure 3.
Figure 4: Credits on Total Investment Income, 1938-1999
The results in Figure 4 are similar in most respects to Figure 3, except that the upward trend for the whole world and for the non-core countries is much weaker.[12] Credits on investment income have gone up mainly for the core, and this acceleration is temporally similar to the results for core debits shown in Figure 3. The other difference is that core credits were already higher than non-core credits in 1938. This is just the opposite of the pattern for debits. Core countries are the main global investors and this difference between core and non-core has increased over time despite all the talk of emerging markets and transnational corporations based in the semiperiphery.
Figure 5 compares two of the components of investment income debits with the total investment income debits for the core countries.
Figure 5: Total, Direct and Portfolio Investment Income Debits, Core Countries
Figure 5 shows some interesting differences for the core countries across different kinds of investment income debits. The total investment income trajectory is the same as that shown in Figure 3 above. Average weighted direct investment income debits for the core countries do not show an upward trend. The upward trajectory shown by total investment income debits is entirely due to the increase of payments out on portfolio investment and on 登ther investment income (not shown). [13] 徹ther investment income primarily includes items that can be considered as parts of international investment. [14] Thus the increase in core investment globalization as indicated by debits on investment income appears to be largely due to the rise of portfolio investments corresponding to the financialization of the world economy that has accompanied the ideological hegemony of neoliberalism since the 1970s. Debits on direct investment income in the core appear to have remained flat relative to the size of the world economy in this period, meaning that the amount of profit paid out by core countries resulting from the activities of subsidiaries of transnational corporations has apparently not increased.
Let us now examine the trends for the subtypes of credits on investment income. Figure 6 shows the corresponding credits categories to those shown in Figure 5.
Figure 6: Total, Direct and Portfolio Investment Income Credits, Core Countries
Figure 6 shows a very similar pattern as Figure 5. The trend in total investment income credits for core countries is the same as in Figure 4 above. As in Figure 5, direct investment globalization is flat, while portfolio investment globalization shows a strong rise since the late 1970s. But some of the trend in Figure 6 is due to changes in the cases. Figure 7 shows the results when we hold the cases constant for each of the subtypes.
Figure 7: Credits on Total, Direct and Portfolio Investment Income for Constant Cases
Figure 7 indicates that the flatness of the trend in direct investment credits shown in Figure 6 may be due to the addition of cases as we move through time. When we hold the cases constant, direct investment credits show a modest upward trend.[15]
The conclusions suggested by our study of investment globalization from 1938 to 1999 are as follows: there was indeed an upward trend of investment globalization during this period but it began after a decline during World War II. The big rise began in the late 1970s corresponding with the abandonment of the Bretton Woods regulations over international investment and the deregulation of international monetary arrangements. There were important differences between the core and the non-core with regard to investment globalization. The core did it earlier and rose to a much higher peak. Portfolio investment was a major contributor to the big wave of investment globalization that occurred after the late 1970s. This corresponded to the shift of capital accumulation away from investment in production and trade and into finance capital.
What remains to be done for this part of our study of trends in investment globalization is to compare our 殿verage investment dependence/dominance measures with world totals measures, and to compare our results for investment globalization with yearly trade globalization figures and with other types of globalization. We also need to finish linking our study of nineteenth century investment globalization with the post-1938 measures, splicing the series in a way that permits us to address the questions of cycles and long-term trends that motivate our comparison of the nineteenth century wave of globalization with the twentieth century wave. How big is the upward trend? What is the precise temporality of investment globalization and how does it relate temporally with other types of global integration? Which kinds of integration lead and which ones lag? What causes global integration and what does it cause? And how can our study of global integration help us lower the probability of future conflicts and increase the probability of global cooperation, stability, democracy and equality.
Appendix A: Yearly Non-net British Foreign Investment Flows,
1862-1914
Our best yearly estimate so far of investment globalization in the nineteenth century comes from Irving Stone痴 (1999) new study. This is a much-improved basis for estimating the trajectory of investment globalization in the nineteenth century because it does not rely on net figures. We have used Stone痴 estimates of total private capital exports (flows) from the United Kingdom to thirteen countries (Argentina, Australia, Canada, Chile, Germany, Spain, France, India, Italy, Japan, Mexico, United States and South Africa) in pounds stirling and national income estimates in country currencies (converted into British pounds using exchange rates). These have then been weighted by the ratio of the country痴 population to the world population and then the average for all the countries available for each year have been calculated with Figure 8 as the result.
Figure 8: Investment Globalization Based on British Private Capital Exports, 1865-1914
The method used in Figure 8 to produce yearly estimates of investment globalization is based on flows, the average amount of capital exported from Britain each year rather than on estimates of the total book value of British capital within each country. This may contribute to the volatility of the values, but the upward and downward trends seem to track the ten-year business cycle fairly closely. Interestingly the Great Depression that began in 1873 initiated a fall of capital exports that was no deeper than the previous decadal downturn, and in 1885 a steep ascent began that reached the highest peak of the whole period. From these data it appears that investment globalization has a similar trajectory to that of trade globalization in the nineteenth century. Recall that the trajectory of trade globalization in Figure 1 above has been transformed by a five-year moving average that smooths out short-term variations.
We are not satisfied with this measure of nineteenth century investment globalization for several reasons. First, the United Kingdom was perhaps the most important investing country, but rather important foreign investments were also flowing from France, Belgium, Germany, and after 1900 from the United States. The other big problem is that these yearly export flows do not provide good estimates of the total value of foreign-held assets in the countries to which they are flowing. The way to estimate the total value of capital is to sum current investments over a period of years depreciating the earlier years with an amortization schedule. This is our next step for the British flow figures, and these will need to be weighted by estimates of world GDP. To accomplish this latter we will need to transform Maddison痴 (1995, 2001) estimates from 1998 Geary-Khamis U.S. dollars into current British pounds for the nineteenth century (or transform current pounds into 1998 Geary-Khamis U.S. dollars). This will provide a better estimate of nineteenth century investment globalization that the flows numbers shown in Figure 8.
Appendix B: List of Core and Non-core Countries Studied
with IMF Balance of Payments Data.