Economic Globalization since 1795: structures and cycles in the modern world-system
Christopher Chase-Dunn
Yukio Kawano
Benjamin Brewer
Department of Sociology
Johns Hopkins University
Baltimore, MD. 21218 USA
 

To be presented at the International Studies Association session on "Globalization in Contemporary and Historical Perspective," Section on Comparative Interdisciplinary Studies, Washington, DC, February 29, 1999. The graphic above is a network visualization of the global automobile trade in 1994 from Krempel and Pluemper.
  1. 2-16-99. Draft. Please do not quote without permission. Comments and criticisms to chriscd@jhu.edu
This paper is available at /archive/papers/c-d&hall/toc.htm

ABSTRACT: The term globalization is used in social science and in popular discourse to mean quite a number of different things. We contend that it is important to distinguish between globalization as a contemporary political ideology and globalization as the increasing density of globe-wide interaction networks relative to the density of national-level networks, which we call structural globalization. This paper conceptualizes and begins to operationalize several important types of structural globalization. And we present the results of our study of one type of economic globalization: the trajectory of international trade as a proportion of global production over the past 200 years.

One big disagreement among social science approaches to globalization is the temporal form that is assumed regarding changes in the structure of the world economy. Many social scientists believe that there was a period in the recent past in which national economies were independent entities, and that a new global economy has emerged in the last decades in which national societies have become integrated into a global network of trade and an interdependent division of labor. A rather different approach imagines a long-term trend toward greater and greater globalization as transportation and communications costs have declined. And yet another approach asserts a cyclical process of phases of increased integration followed by phases in which national economies or subregions return to more autarchic economic interactions. Our study is relevant to these divergent temporal images and to the important conceptual and theoretical issues that are linked with them.
 

Table of Contents

Introduction

Structural Globalization

Unit of Analysis: The Whole System

Trade Globalization

Average Openness

Weighting

Checking

Three Waves and a Trend

Explanations

Integration and World Order

References

Appendix

We are sociologists who study the world-system as a whole, as well as its parts. Though much of the research presented in this paper focuses on economic networks, our theoretical approach is not economics nor even economic sociology. Rather we seek to understand continuities and changes in institutional structures of the modern world-system over the past 200 years. Institutional structures are fundamentally cultural inventions. Market exchange, firms, states, global governance organizations and the civilizational ideologies that naturalize them are all grist for the analysis of institutional structures that is the framework for our consideration of globalization.

Since the 1980s the term 'globalization' has been used to describe allegedly recent and important changes in the world economy. It generally refers to changes in technologies of communication and transportation, increasingly internationalized financial flows and commodity trade, and the transition from national to world markets as the main arena for economic competition. These ostensible changes have been used to justify economic and political decisions such as deregulation and privatization of industries, downsizing and streamlining of work forces, and dismemberment of the welfare services provided by governments. The expansion of the global economy has also been painted as the victory of progressive and rational capitalism over the anachronistic ideologies of socialism and communism. People naturally want more and bigger and faster commodities, and global capitalism is the most efficient feasible system for providing these.

This discourse about globalization is itself a phenomenon worthy of social science research. The emergence of neo-liberal political ideology is the topic of Phillip McMichael's (1996) analysis of the "globalization project." This phenomenon emerged with Reaganism and Thatcherism 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. How did this somewhat revised and expanded rendition of the private property version of the European Enlightenment become the global hegemonic ideology at the end of the twentieth century? This fascinating question is not the focus of our research reported here. Rather we will distinguish between globalization as an ideology and globalization as objective structural trends of spatial integration, and focus on the latter.

Structural Globalization

Human societies are composed of interaction networks and the institutions and forms of consciousness that make various kinds of interaction possible. The world-systems perspective 1 asserts that interaction networks have been importantly intersocietal since at least the emergence of cities and states, but comparative studies reveal important intergroup interaction networks even in systems composed entirely of nomadic hunters (Chase-Dunn and Hall 1999).

While the institutional nature of interaction networks has undergone major transformations with the evolution of social complexity and hierarchy, one important aspect of interaction networks has always been their spatial scale and the relative intensity of smaller and larger nets. And comparative research reveals that all world-systems small and large have exhibited the phenomenon of "pulsation" in which exchange networks alternately expand and contract. For the modern world-system we will conceptualize globalization as, in part, changes in the intensity of international and global interactions relative to the local or national networks. 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 both integration and interdependence.2

Different kinds of interaction have long had different spatial attributes. Most world-systems are multicultural in the sense that important political/military and trade interactions link groups with very different languages and cultures. The modern world-system is mainly composed of national cultures, though the most powerful countries have long been able to impose, sell or diffuse their cultural characteristics widely, and their may now be emerging a truly global culture that is more than just the cultural reach of the most powerful national states (Meyer 1996; Boli and Thomas 1997).

Structural economic and political globalization are conceptualized here as the differential density and power of larger vs. small interaction networks and organizations. Though we do not contend that politics and economics are separate realms that can be independent objects of scientific inquiry, we do find it convenient to distinguish between political and economic forms of globalization.

Economic globalization means greater integration in the organization of production, distribution and consumption of commodities in the world-economy. We are all aware that our breakfasts have been increasingly coming from distant locations. Sugar has importantly been a global commodity for centuries, in the sense that both its conditions of production and consumption have been massively affected by intercontinental market forces and the policies competing states. But fresh grapes have only become global commodity since jets started transporting them between the southern and northern hemispheres. If you do not eat sugar or grapes for breakfast, no matter. The energy that was used to produce whatever you eat has long been a global commodity as well, though there have been important changes in the nature of energy production, the organizational structures and ownership of energy-producing firms and the impact of state policies on energy production and consumption (Podobnik 1999).

Political globalization is here conceptualized as the institutional form of global and inter-regional political/military organizations (including "economic" ones such as the World Bank and the International Monetary Fund), and their strengths relative to the strengths of national states and other smaller political actors in the world-system. This is analogous to our conceptualization of economic globalization as the relative density and importance of larger versus smaller interaction networks.

This paper is part of a larger project in which we study trajectories at the global level of different kinds of political and economic globalization. Here we report the results of our study of one kind of economic globalization - the globalization of trade over the past 200 years. We also plan to analyze the globalization of ownership of capital, but that part of our project has not yet proceeded to the point where we can report results. 3

Unit of Analysis: The Whole System

There are two matters that we need to clarify regarding our study. First, our research is about changes in the organization of the world-system as a whole. Thus we will operationalize trade globalization as a variable characteristic of the whole system. We conceptualize the world-system as a complex network of nested and overlapping subnetworks. It is not a matter of "international relations," but rather the focus is on all the interactions of the people of the Earth, local as well as global. This includes individuals, households, neighborhoods, communities, villages, towns, cities, local polities, substates, national states, firms, political parties, classes, zones (core, periphery and semiperiphery), transnational and global level organizations and networks of all kinds - all the local, regional, inter-regional and global networks - the whole system. This is our unit of analysis. The question we are asking is how has the organization of this unit changed over the last 200 years? Was there a recent leap from national-level economic networks that were largely independent of one another to a global interdependent network? Was there a long-standing upward trend from local to regional to national to international regional to global-level interactions? Or was there a cycle of changing intensity of global-level interactions relative to the intensity of local or national-level interactions?

Whereas ideally we would like to have data on all levels of interaction, this is not possible for the whole time period that we want to study. To answer the questions we are asking it is necessary to compare recent decades with earlier periods, and with the nineteenth century. In order to do that we must utilize data on national societies because it was these entities, their states, that developed "statistics." This does not mean that our analysis is "state-centric." We have already clarified that we want to focus on interaction networks at all levels, including global and transnational ones. But data for long-term comparisons are only available for national states. We can use these data if we are chary about what sorts of distortion or false inferences might be introduced by having information only on national states. 4

Of what theoretical import is the question we have posed about the long-term trajectory of globalization? At the level of general discourse even asking the question challenges the "presentism" of an age in which experiential change has become so rapid that people dismiss last year's TV programs as "history." Within social science there is the repeated discovery of new organizational features or technologies that are alleged to have ushered in (or to be ushering) the next stage of development. Global capitalism, the information age, flexible specialization, the new world order and/or post-modern philosophy and consciousness have changed everything. The idea that there may be important institutional and structural continuities with the past is lost in the celebration or critique of these new departures. In this mentality the recent jump from national economies to a global economy has the status of a known fact.

Among those who admit the possibility of important historical continuities and long-run processes there are great differences regarding the conceptualization of these. Some see political/military cycles and economic cycles at the system level. Both the world-systems perspective (Wallerstein 1974) and the power cycles approach (Modelski and Thompson 1994) focus on the sequence of the rise and fall of hegemonic core powers as well as the forty to sixty year business cycle called the Kondratieff Wave. Others see these and other cycles, as well as long-term trends (Chase-Dunn 1998). And others see cycles, trends and occasional systemic transformations in which major changes have occured in the organizational nature of accumulation (Arrighi 1994). We will discuss these models in more detail after we have presented the results of our study of trade globalization.

Trade Globalization

By trade we mean the buying and selling of commodities. The question we are posing is about the spatial nature of trade networks. All would agree that some commodities have been traded on an intercontinental basis for centuries. But the amount of goods and services that are bought and sold within communities and within national societies has also increased during these same centuries. We follow economic historians in conceptualizing the globalization of trade in terms of changes in the relative intensity of international vs. within-nation trade (e.g. Bairoch 1996). Thus it is possible for the amount of international trade to increase, but if the amount of within-nation trade increases faster, then the world economy would be experiencing decreasing globalization.

The standard operationalization of trade globalization is the sum of all international exports as a percentage of the global product, which is the sum of all the national Gross Domestic Products (GDPs). The Gross Domestic Product of a country is the sum of all the economic transactions within that country minus the value of imports.5 The sum of all the national GDPs is a good indicator of the total amount of economic trade in the world economy because GDP includes exports. The sum of all the national imports is a good indicator of the total sum of all international trade. And the ratio of total imports to total GDPs is a good measure of how globalized the world economy is.

We could use either imports or exports to estimate the total international trade. Using both would be double-counting. Imports are preferred because statistics on imports have been much more accurate because governments have long taxed imports and so they have paid close attention to them. 6

Earlier long-term studies of trade globalization have found interesting results. Using data from Angus Maddison (1995) we can combine estimated world totals of exports and GDPs to get a rough idea about the trajectory of trade globalization (See Figure 1).

Though estimates of the world totals of international trade and GDPs might seem to be the most direct route, this approach has serious problems. The studies of trade globalization that take this approach usually include only widely spaced estimates of the degree of trade globalization for the nineteenth and early twentieth centuries (e.g. Bairoch 1996). Maddison's (1995: 227) data for total world GDP jumps from 1820 to 1870, and then to 1900, 1913, 1929 and then to 1950 ( See Figure 1). This makes it difficult to see the finer temporal aspects of changes in the level of trade globalization. It would be desirable to have a yearly measure in order to see whether or not there have been important short-term changes and if the time points in other studies are representative or are deviations from the years surrounding them.

Another problem involves the need to transform country currencies into a comparable standard in order to sum imports and GDPs to compute the global totals. Typically these have been transformed into United States dollars. This is done using the exchange rates between U.S. dollars and the individual country currencies for each year that the transformation is needed. The assumption here is that the market exchange rate of the currencies accurately reflects the relative values of goods and services in different countries.7 After this transformation is made, the U.S. dollars are then adjusted for the U.S. rate of inflation to transform the values from current into constant dollars for a comparison year. This is important for the purposes of cross-temporal comparisons.

All these moves involve assumptions that may introduce large errors into the estimates. And these could be influencing what we see in Figure 1. Figure 1 implies that there is both a cycle and a trend in trade globalization.

Here are the trade globalization ratios calculated from Maddison's world totals.
Year  1820  1870  1900  1913  1929  1950  1955  1960  1965  1970  1975  1980  1985  1990 
Exports/ 

GDP 

.010  .050  .070  .087  .090  .070  .074  .083  .091  .105  .108  .115  .113  .125 
Table 1: Trade globalization ratios from Maddison's (1995, 227,239) world totals

It would appear that, from a low degree of trade integration in 1820, the world economy increased to a peak in 1929 and then dropped during the depression and World War II to a low in 1950, from whence it began an upward movement that slowed slightly between 1980 and 1985, and then increased again until 1990, when it reached a level of integration greater than ever before.

The results from Maddison's data are intriguing because they imply a number of conclusions. First, the idea that globalization is only a recent and unique phenomenon is dispelled. There would appear to have been a huge wave of trade globalization during the nineteenth century that extended well into the twentieth century. Estimates of changes in the level of foreign investment relative to the size of the world economy confirm this general result (Bairoch 1996) and several recent discussions of economic globalization have compared the nineteenth century wave with the post-World War II period (e.g. Gordon 1988; Sachs and Warner 1995; Bairoch and Kozul-Wright 1998). Secondly, the numbers in Figure 1 and Table 1 imply that there is a trend in the sense that the recent wave of globalization has been larger than the nineteenth century wave. If this is true the world may be experiencing an upward spiral of global trade integration that is interrupted by periods of backsliding. This is the general thrust of many of the discussions that seek to infer policy implications from an understanding of changes in the world economy (e.g. Sachs and Warner 1995).

While the single step function of a recent jump from national economic autarchy to globalization can be ruled out, there are many remaining questions? Is it true that there is a trend as well as a cycle of trade globalization? Is the cyclical pattern more fine-grained than the spotty data prior to 1950 might suggest? To answer these questions we have devised a different approach to measuring trade globalization. 8

Average Openness Measures of Trade Globalization

The problematic assumptions about currency equivalence, exchange rates and corrections for inflation can be cut out by estimating trade globalization using average national levels of "openness." Openness is traditionally operationalized at the country level by determining the ratio of external trade (either imports or exports) to the GDP. The (weighted) average of all the national degrees of openness will equal the world level of trade globalization (see below), and the advantage of this is that we have both GDP and the value of imports in country currencies (Mitchell 1992,1993,1995). When we compute the ratio using country currency units in both the numerator and the denominator, the currency itself drops out. This makes levels of openness comparable without having to convert everything into U.S. dollars, and so we can dispense with all the problematic assumptions discussed above.

Recall that the measure of trade globalization using world totals is the sum of world exports (or imports) divided by the sum of all the country GDPs. If all countries were the same size, the sum of the country openness levels divided by the number of countries (the mean openness) would exactly equal the World Totals measure of trade globalization.

So,

                                                                        Imports
(WORLD TOTALS)  Sum(Imports)  =    Sum (GDP)        (AVERAGE OPENNESS)
                                      Sum(GDPs)                   N

Of course countries are not all the same size. But we can weight the openness scores by the population sizes of the countries, and when we do that our measure of average openness estimates trade globalization without having to convert country currencies into U.S. dollars and without having to adjust for inflation.

There is more good news. We have yearly data on openness, so we can see the finer changes in the level of trade globalization. The bad news is that our data are less complete as we go back in time.9 This means that we are using a "sample" to estimate trade globalization, but the sample is not randomly chosen. Generally the core countries have more complete data earlier on, whereas for peripheral countries we have only recent data. Since we are interested in both the temporal ups and downs and the comparative levels in different periods, the biased pattern of missing data is a concern. It is well known that countries differ in terms of their degree of openness. Generally, smaller countries, and especially peripheral ones, tend to have higher levels of openness (or "trade dependence" as it has also been called).

The number of countries for which we have data on trade openness that were used to estimate world trade globalization is graphed in Figure 2.

Figure 2: Number of Countries with Data on Openness, 1795-1995

We excluded country/years in which the country: 

The data were mainly found in Mitchell (1992,1993,1995) but from 1965 we use numbers from the World Bank (1980, 1998).

Weighting the Country Openness Scores to Estimate Trade Globalization

It was mentioned above that we need to weight the country openness levels so that our average reflects the differential sizes of countries. The unweighted mean erroneously assumes that e.g. El Salvador and Mexico should have the same importance in determining the world level of trade globalization. We weight the country openness scores by multiplying them times the ratio of the country population size to the average population size of all the countries for which we have openness scores at each time point.
 

                                                                      Sum(i-N) _____POPi __    x   IMPORTSi
WEIGHTED AVERAGE OPENNESS  =                     Sum(i-N)  POP            GDPi
                                                                                                  N___________________
                                                                                                                 N

This is mathematically equivalent to weighting by the total population, but it produces a metric which is more comparable over time because the scale does not change radically with population growth-an important consideration when we seek to compare the nineteenth and twentieth centuries. Figure 3 shows the weighted and unweighted estimates of trade globalization using all the countries for which we have data.

Figure 3: Average Openness Trade Globalization estimated from all countries for which we have openness data (5 year moving average). Pearson's r correlation coefficient for weighted and unweighted is .73.

The correlation between the weighted and the unweighted series is .73, so weighting alters our estimate substantially and is an important correction. Weighting is especially important in the decades since 1950 when many small countries enter the calculations. Most of these are peripheral countries that have relatively higher scores on openness. The unweighted series over-estimates the world level of trade globalization because these countries are weighted equally with larger countries. 10

Checking the Results

There are two techniques we can use to check the possibility that our "sample" of countries with available data is not providing a reliable estimate of world-wide trade globalization. The first is to compare our estimated level of trade globalization based on Average Openness with the World Total approach based on Maddison's (1995) estimates. For the period after 1950 Maddison has yearly data on GDP and the problems of the World Totals approach are probably reduced because exchange rates are more reliably known, and so are inflation rates. Thus we can use Maddison's (1995: 227,239) data to see how our Average Openness measures compare after 1950.

Figure 4: World Totals and Average Openness Measures, 1950-1992

Figure 4 indicates that the Average Openness measures may overestimate the levels of trade globalization. Especially the unweighted Average Openness does this, probably because the small countries with high levels of openness are over-weighted and raise the level of the estimates. The weighted Average Openness is a much closer estimate of the real levels of trade globalization in this period. Both the weighted and the unweighted Average Openness estimates show a similar upward trend with the World Total approach, but the unweighted series shows a much steeper ascent. This may be due to the adding of more peripheral countries as we approach the present.

Another technique we can use to examine the errors due to missing data is to select different subgroups of countries and hold these groups constant over time and then to compare the groups to see if they are revealing similar temporal sequences and similar levels of openness. If we find that constant subgroups exhibit patterns similar to those found for the whole data set we may be comforted regarding the proposition that our restricted sample earlier in time is not a bad estimate of the true world level of trade globalization, though this will not at all be certain.

Figure 5: Average Openness of Constant Groups of Countries (5 year moving averages)

Figure 5 graphs the weighted Average Openness values for six groups of countries, with the groups held constant over time so that changing country composition does not affect the averages. The yearly values are smoothed a bit by means of a five year moving average so that we can better see the waves. The trajectories get shorter as we add countries. The first "group" is not a group at all. It is the United States, the only country for which we have data for the whole period from 1795 to 1995. We will discuss the trajectory of U.S. trade openness below after further consideration of the validity of our measures of trade globalization. The second group is composed of the U.S., the United Kingdom and France with average scores beginning in 1830. The third group, beginning in 1861, adds to these Australia, Denmark, Italy and Sweden (seven countries). The fourth group (fourteen countries), beginning in 1905, adds Cuba, Spain, India, Japan, Mexico, the Netherlands and Taiwan. The fifth group (24 countries) begins in 1927 and adds Austria, Canada, Colombia, Greece, Guatemala, Honduras, Hungary, Indonesia, South Africa and Zimbabwe. The sixth group (50 countries), begins in 1950 and the seventh group with 89 countries begins in 1965.

Inspection of Figure 5 clarifies some aspects of Figure 3 and generally supports the idea that average levels of openness of a subgroup of countries can be used as a reasonable proxy for both the level of world trade globalization and for periods of rise and fall in that level. The jumping around between 1795 and 1830 is the fledgling United States of America feeling its way through a world war in which it was allied with the losing side. Both vast blockades breeched and poor statistics probably account for the wildness of U.S. trade openness in this period. We include these data mainly to display the slim reed that is our window on world trade globalization before 1830. And indeed, the rest of the U.S. performance until about 1960 shows that the United States by itself is a poor reflector of world trade globalization. It was not until the 1960s that the United States experienced increased openness of its trade to the world division of labor. Perhaps the historical uniqueness of this aspect of U.S. history partly explains the amazement and presentism that seems to dominate considerations of globalization in the U.S.

The trajectory of U.S. trade openness may not be a good window on the world economy as a whole, but it is an important window on what happened within the U.S. and its relationship with the larger world-system. While most of the other countries were experiencing trade globalization, the U.S. was expanding its domestic economy faster than its international trade, and it was moving up the value-added hierarchy of the international division of labor.

While the rest of the world was going through at least one, and possibly two waves of trade globalization between 1830 and 1929, the United States enjoyed a low, and mainly declining, level of trade dependence. This was probably due mainly to the relatively fast rate of growth of U.S. GDP during this period of territorial expansion, rapid population growth, industrialization and upward mobility into the core of the world-system. U.S. imports did grow mightily, but the domestic economy grew even faster. This amazing performance was the outcome of internal and international struggles among classes, different sectors within the same classes, and national states. Indeed, it has been argued elsewhere that the U.S. Civil War was mainly a struggle over how the U.S. would be inserted into the larger core/periphery hierarchy (Chase-Dunn 1980). The struggle over tariff policy between 1816 and the Civil War showed how the southern exporters of peripheral agricultural products had interests that were quite different from northern manufacturers. The victory of the north in the Civil War meant a consistent policy of trade protectionism to promote import substitution industrialization, a policy that lasted until after World War II. Thus the U.S. success is a poor example for those who want to argue that free trade is a central pillar of economic development.

The main point we need to emphasize here is that we are not really studying the trade history of all the countries and neither do we need to assume that all the countries have the same trajectories of trade openness. Rather we are studying the whole system, which is composed of diverse parts with different histories. The notion that average country openness levels can be used to estimate the world level of trade globalization becomes less and less problematic as we add more and more countries.

We also need to notice in Figure 5 that, except for the U.S. the other groups display generally similar levels of average trade openness and these levels go up and down rather synchronously. Additional support for our contention that the estimated scores for the all the combined cases, which are added more gradually than in Figure 5, is the table of correlation coefficients between the group scores and the values estimated using all the cases. These are shown in Table 2.
Country 

Groups 

USA  US/GBFrance  Seven  Fourteen  Twenty-four  Fifty  Eighty-nine 
All 148  .51  .84  .85  .73  .92  .94  .74 
Table 2: Pearson's r correlation coefficients of subgroup scores with our estimate of world trade globalization that uses all data.

While the subgroups vary as to how well they are correlated with our overall measure, they are all fairly well correlated with it, except for the United States. We have just discussed that deviant case.

Another approach for evaluating our Average Openness measure is to examine systematic differences among countries that may be affecting our estimates for earlier years. We have already mentioned that core and peripheral countries often differ in terms of their levels of openness or trade dependence. Peripheral countries tend to be smaller and more dependent on imports and exports. But there are also small core countries that have high levels of openness (e.g. Switzerland, the Netherlands).11 In any case one big problem with our measure of trade globalization is that we have data on few non-core countries early on, so an important piece of the world-system is missing, and this could be biasing our estimation of the level of world trade globalization downward. This is of concern because one of questions we want to answer is whether or not there is a real upward trend, in addition to the obvious cycles. It is possible that the high level indicated for recent decades might be due to the addition of more data on peripheral and semiperipheral countries rather than a real increase in the level of world-wide trade globalization.

We have divided our list of countries into core and non-core groups (see Table A1 in the Appendix). Figures 6 plots the weighted and unweighted values of these groups since 1950.

Figure 6: Average Openness Scores for Core and Non-core Groups, 1950-1995

The core/non-core correlations for these are .xx (weighted) and .xx (unweighted). The differences between the weighted and unweighted series due to small countries being over-represented in the unweighted series we have discussed above. The unweighted series are higher and they go up more, but they are not affected by changes in the populations of countries. Inspection of Figure 6 shows that there is a rather close tracking of core and non-core groups regarding the ups and downs of openness until 1975. Both groups wiggle a bit and then rise precipitously between 1955 and 1975. After 1975 the core countries reach a plateau that they cycle around until the end of the series, whereas the non-core countries rise to a height greater than ever before and then fall back in the mid '80s and then rise again to their highest height in the early 90s when our series ends. None of these changes or comparisons are due to adding cases nor, in the case of the unweighted series, to changes in relative populations.

These results support the notion that both core and non-core countries are experiencing changes in trade globalization synchronously, and up until 1975 these groups have very similar levels of openness. After 1975 we see a divergence. The core countries plateau at level that is higher than the level reached in the earlier waves of globalization, but the peripheral countries continue to rise to an even higher level.

Three Cycles and A Trend

The conclusion we can reach here is that there is indeed a trend as well as a cycle, and this is not due solely to the addition of peripheral countries to the data. When we look only at the core countries the level of openness reached by 1975 is already higher than was reached in the earlier waves of globalization. The apparent continuing upward trend in the combined measure (including both core and non-core countries) after 1975 is due mainly to the peripheral countries.

This is an important finding if we think about how our estimation of the highest level of the earlier peaks might have been affected if we had had data on non-core countries then. It is quite possible that a similar divergence between core and non-core occurred near the end of earlier waves. If that were so the average level would have been higher than our estimated level indicates. But we doubt that it would have been as high as the level reached by 1995. And so we think that the notion of a trend as well as a cycle can stand. It should be noted that the magnitude of the increase due to the trend is significant, but not a qualitative leap to a vastly different degree of global integration. There is simply no support for the idea that a completely new stage of global integration has been entered in recent years. What we have instead is an unprecedented high level of integration, but not one that is of an entirely different magnitude than ever before.

We have already concluded that there are cycles. But how does our more temporally fine indicator improve upon what we saw from the World Totals data from Maddison in Figure 1 above? Recall that Maddison's data before 1950 were quite spotty, though an earlier wave of globalization in the late nineteenth and early twentieth centuries could be seen. Our new measures show that there were really two earlier waves, not one. Figure 3 (above) shows both our weighted and unweighted measures of Average Openness Trade Globalization.

After 1830 we have better data from the U.S. and we have added the United Kingdom and France -- a wider and more reliable window on trade globalization at the world-system level. Figure 7 begins in 1830 so that we can focus on the ups and downs since then. Looking at these weighted scores we see that from the 1830s there was a rise to a high mound between 1850 and the late 1880s, then a decline until 1905 and then another wave that rose before World War I, declined a bit during it, and then rose again for the roaring twenties. A big downturn corresponded with the stock market crash of 1929, and with some wiggles it descended to a very low level reached in 1949, and hence began the most recent great wave of trade globalization.

Figure 7: Average Openness Trade Globalization, 1830-1995 (Weighted)

So we have three waves and a trend. The trend looks more pronounced when we look at the unweighted scores (as we have discussed above), but this could be due to overly weighting smaller (and more peripheral) countries and their increasing presence in our "sample" as we move toward the present. But the weighted scores also indicate a trend. Unfortunately this might be a consequence of how the weights are affected by generally increasing populations. The addition of smaller countries lowers the average population size of countries in the "sample," which increases the scores of the largest countries. This could lower the level of the weighted measure because their is a small negative relationship between population size of countries and openness. 12

The most surprising finding of our research on trade globalization is the existence of three waves instead of two - the middle wave from about 1905 to 1929. Most long-run considerations of the history of the world-system have recognized the late nineteenth century wave and compared it with the contemporary period. The patchy World Total data from Maddison makes it appear that there is a long earlier wave that extends from the nineteenth century until 1929 (see Figure 1 above). But we see in Figure 7 what appears to be a middle wave. Let us look more closely at this middle wave for clues about its nature. Figure 8 shows the trajectories of the constant country groups for which we have data over the relevant time period so that we can see which countries do the middle wave.

Figure 8: Weighted Average Openness of Constant Groups 13

There are somewhat different things happening with the different groups. The big three (U.S. Great Britain and France) declined from 1883 to 1902 and then made a small rise from about 1905 until 1913, declined during the war, and then recovered a bit during the 20s to plunge at the end of the decade. This must have been due to the changes of openness in Great Britain and France because, as we have seen above, the U.S. experienced slowly declining openness throughout the period (see Figure A2 in Appendix). The middle wave for the big three was not really a wave, but was rather an oscillation around a relatively high level of openness.

For the group of seven (the above plus Australia, Denmark, Italy and Sweden ) the first wave peaked in 1887, then they declined somewhat and stayed down from 1897 to 1905 and then we see a definite middle wave that rose from 1905, wiggled and soared to a single peak in 1921 that was quite higher than their earlier peak in 1887. Then they dropped, recovered a bit in 1927-28, and then the plunged with the rest.

The group of fourteen adds Cuba, Spain, India, Japan, Mexico, the Netherlands and Taiwan. The data on this group do not start until 1905 so we cannot see the great wave of the late nineteenth century. But from 1905 we see a flat wiggle with a very small drop during World War I, and a rise that began in 1918 to a peak in 1924, and then a sharp decline that slowed a bit from 1937 to 1939 and then went down until 1948. The fourteen had a definite middle wave, but it was temporally later and shorter than that of the group of seven.

The group of 24 has data only from 1927 on. Tantalizing though, are the first three years. They show an increase in openness from 1927 to 1929. Then the plunge. Is this the tail of the middle wave? We do not know.

What we can say of the middle wave is that it was mixed in its composition and its temporality.14 Of course so were Waves 1 and 3, but to a lessor extent.

Explanations

How can we explain the trajectory of trade globalization? There are two things that need to be explained: the trend and the cycles. For the trend we can acknowledge common wisdom in seeing the falling costs of transportation and communications in these centuries as a main driving force of upsurges of globalization that reach higher peaks (or mounds). But these declining costs of long-distance transport and communications are facilitating background factors that cannot explain the periodic collapses of globalization because costs did not radically increase when globalization declined. Why did transportation and communications costs decrease? This was a complicated outcome produced by the effects on technological change of industrial capitalism, competition among states and firms and political struggles for power and against oppression and domination in the world-system.

Let us look more closely at the cycles. The first thing we can ask is how do they correspond temporally with other known cycles? Causality should be revealed in the temporal relationships among variables. The contenders here are business cycles (the Kuznets cycle and the Kondratieff Wave), the rise and fall of hegmonic core powers (the hegemonic sequence), the war wave, and the incidence of world wars, the debt cycle and changes in the level of trade protection.

The hegemonic sequence has been quantitatively measured in terms of military power (or rather naval and air power) by Modelski and Thompson (1988). They examine the proportion of intercontinental power capability that is controlled by the most powerful country. In the period we are studying they find the rise and decline of Britain in the nineteenth century and the rise of the United States in the twentieth century. The world-systems perspective has emphasized the importance of economic power in the hegemonic sequence. These two approaches have influenced each other and Modelski and Thompson (1994) now include economic power as an important part of their conceptualization and measurement of "global leadership." Arrighi (1994) recognizes the importance of ideology and legitimacy in the successful performance of the hegemonic role

Regarding the problem at hand, both the first and the third waves of globalization correspond to the rise and consolidation of hegemonies, the British in the nineteenth century and the U.S. after World War II. But the middle wave, that rose from about 1900 through the 1920s, occurred in a period in which hegemony was being radically contested. This middle wave cannot be a function of hegemonic rise and fall, because there was no rise or fall in this period.

The Kuznets business cycle is a twenty year cycle in which economic growth increases for about ten years and then stagnates for about ten years. This is too short a period to account for the waves of globalization. The Kondratieff Wave is a longer business cycle that varies from 40 to 60 years. This is a closer match to the waves of globalization. The 1929 crash fits with the decline of the second wave. But there are some non-fits as well. The period of the Great Depression of the 1870s was during the latter part of the rise of the first wave of trade globalization, which did not begin its decline until the 1880s. Most K-wave studies find a K-wave decline (B-phase) beginning in about 1970. This one is not associated with a drop in globalization. Indeed after 1975 our measure indicates a rise to the highest level of globalization known, though this is due mostly to the increasing openness of non-core countries.

World Wars do not fit well with the three waves of globalization. The only data we have prior to and during the Napoleonic War is from the U.S., and they are unreliable as a basis for studying trade globalization. After the Napoleonic Wars there were no world wars in the nineteenth century in the sense of major wars among core powers. There were three "Great Power Wars" between 1815 and 1914, but none of them were very big (Levy 1983:72-3). World War I was during the first rise of the middle wave. World War II began before the trough of the middle wave and ended before the beginning of the rise the third wave of globalization. There is no regular relationship between world wars and globalization cycles.

Sachs and Warner (1995) have argued that trade liberalization policies in the form of tariff reductions and free trade agreements were a major factor producing the high rates of growth and the economic globalization waves of the late nineteenth century and the post-World War II period. But Paul Bairoch (1993) contends that trade liberalization and trade protectionism cannot be main causes of either economic growth or globalization because the pattern of trade liberalization does not fit temporally or geographically. We will focus here on the relationship between trade libaralization policies and the waves of trade globalization, though the question of changing rates of economic growth is relevant also.

Bairoch notes that the period between 1815 and 1860 was one in which the British opened their home market to foreign goods and advocated that other countries should do the same. This was the heyday of Cobden and Bright and their Anti-Corn Law League.15 But it was only between 1860 and 1879 that other countries on the European continent decreased their tariff barriers to imports. The United States, as we have mentioned above, adopted greater tariff protection following the northern victory in the U.S. Civil War.

After 1879 the European states gradually slipped back toward protectionism, while the British maintained low tariffs until 1914 despite huge political arguments over this policy (Taylor 1996). Bairoch (1993:51) shows that the reintroduction of protectionism did not have a long-term negative affect on the growth of exports for those countries that went protectionist. In the second decade following their reintroduction of protectionism France, Germany, Italy, Denmark and Switzerland all had higher rates of export growth than in the decade before they went protectionist, and this was true for Europe as a whole. And the United Kingdom, where a liberal trade policy was maintained, had a declining rate of export growth over this same general period. Bairoch does not go so far as to claim that protectionism causes globalization, but he does assert and support the contention that trade liberalization did not cause globalization in the late nineteenth century.

Our average openness measure of world-level trade globalization is similarly contradictory with the hypothesis that trade liberalization causes globalization. The first wave of trade globalization began well before the European shift toward free trade. And the downturn in the early 1880s preceded by several years the readoption of protectionism policies by the European states.

Despite the common belief that the economic collapse of the 1930s was caused by protectionism, Bairoch (1993) shows that protectionism was not particularly high in the 1920s and the Smoot-Hawley tariff was adopted in the U.S. after the stock market crash of 1929 and after the decline in trade globalization had already begun. The rise of the middle wave occurred during a period in which tariffs were high but not rising or falling, and the rising tariffs of the 1930s occurred after trade globalization had already begun to fall.

The third wave began well before the trade liberalization advocated by the now-hegemonic U.S. had been adopted by many countries. The greater trade openness of the peripheral countries subject to International Monetary Fund restructuring has generally occurred in a period of very slow GDP growth. The countries that grew during the "Asian miracle" period on the basis of export promotion did contribute to the rise of trade globalization and their successes were largely due to their access to the U.S. market, so trade liberalization in recent decades has probably had a positive impact on the level of trade globalization. But the cycles of trade globalization over the long run do not correspond very closely with changes in the degree of international trade liberalization.

Integration and World Order

So what does cause the cycles of trade globalization? The structural feature that has the closest temporal correspondence with the cycles of trade globalization is the hegemonic sequence. The British rise and fall of the nineteenth century corresponds nicely with the first wave. The second and third waves are somewhat more complicated. The third wave corresponds closely with "America's half century" (McCormick 1989), except that its apparent continuation after 1975 does not seem to fit. As we have already seen however, the trade globalization of most core countries leveled off in 1975, and the continuing upsurge is due primarily to increases in the exports of non-core countries in an era of very slow GDP growth. We contend that the U.S. economic hegemony has declined since 1975, and that the relative openness of core countries has not increased since then in part because of the U.S. hegemonic decline.

The U.S. itself, however, unlike most of the other core countries, has continued to increase its trade openness since 1975 (See Figure A2 in Appendix). Other indicators of economic globalization, e.g. global sourcing in the automobile industry, also show that it is only the U.S. among major manufacturers of cars, that has increased its reliance on the production of automobile parts at diverse international locations (Kremple and Pluemper 1999). The most recent period of the third wave of trade globalization is similar to the late nineteenth century, when the declining hegemon continued to advocate globalization and to globalize its own economy, while competitors began to pursue a more local or regional approach. In this view the process of European integration is seen mainly as the formation of a new larger integrated competitor for hegemony rather than increased openness among states.

The middle wave would not seem to fit the interpretation that ties waves of globalization with the rise and fall of hegemonic core states. As we have said above, it occurred during a period of hegemonic rivalry when the question of which country was going to move into the position of hegemon had not been settled. The British no longer had the economic power, nor the leadership prestige, to set a standard for the globalized world economy to revolve around. The German challenge in World War I was unsuccessful and did not lead, right away, to the emergence of a new hegemon because the United States refused to take upon itself the burden of leadership. Though it was by far the most economically powerful country in the world after World War I, and though Wilson and his supporters had a vision of international liberalism that could have provided the legitimation for hegemony, the Senate rejected U.S. membership in Wilson's creation, the League of Nations, and the U.S. went back to following George Washington's admonitions about not getting involved with the decadent and quarrelsome Europeans.

This led to a huge vacuum at the center, and the economic crisis and powerful political challenges to liberal capitalism in the guise of Bolshevism and Fascism broke the world order into another violent global struggle for power. After World War II the U.S. did take up the mantle of hegemony, which then did provide the institutional basis for a new wave of economic globalization.

In this interpretation, the forces of falling costs of transportation and communication led to the middle wave, but the lack of an institutional guarantor of world order in the form of a hegemonic core state did not allow that wave to be consolidated and sustained. The middle wave was uneven in time and space, starting and ending in different years for different countries. In that respect it is similar to the recent continuation of the third wave, and there is a similar problem in both periods. In the middle wave there was no hegemon, and in the decades since 1975 there has been a declining one.

Economic globalization creates a great demand for political globalization because markets are not themselves able to resolve the problems of distributional fairness and uneven development that they create. Of course, political globalization does not have to take the form of the hegemony of a single core state. It is possible that new or renewed international political organizations could be an effective global proto-state with the power to adjudicate disputes and to balance and sustain the processes of economic growth. Something like the latter is likely to eventually emerge if the world-system survives the next window of vulnerability to war among core states (Chase-Dunn and Podobnik (1999).

The question now is whether or not sufficient legitimate consensus will be constituted to prevent the world-system from entering another period of violent hegemonic rivalry of the sort that was seen twice in the twentieth century. Such a new period of conflict would undoubtedly once again lower the level of trade globalization. But this would be more the consequence rather than the cause of rising conflicts.

Our study of trade globalization needs to be supplemented with a study of the long-run trajectory of investment globalization. Without this any understanding of the modern world-system is incomplete. Existing evidence indicates that, as with trade, there have been waves of the globalization of capital. We would expect to find a somewhat different trajectory of capital globalization if we were able to study it with more temporal resolution than currently available data make possible. This would certainly be a valuable addition to the study of structural globalization, and it might alter the conclusions we have reached in this study. We also propose to study political globalization over the long run, though this is more difficult both conceptually and empirically. In the mean time our study of trade globalization provides a new window on the continuities and changes in the structure of the modern world-system over the past 200 years.

Notes

1. Shannon (1996) provides a helpful overview of the world-systems perspective as applied to the modern system.

2. While integration and interdependence overlap, they are not exactly the same. Two entities may be linked by some kind of frequent interaction, but they may not be interdependent if the frequent interactions are not important. The exchange of food or other strategic raw materials is always more important than the exchange of luxury goods. Thus it is desirable to consider both the amount of exchange and what is exchanged. The research we report in this paper considers only the value of trade, but we intend to study the changing nature of the world division of labor as well.

3. Ideally we would like to be able to use network analysis to study long-term change in the world-system but data constraints do not make it possible to compare the nineteenth century with the twentieth century. Tieting Su's (1995) fine network study of world trade examines 1938 and compares it with more recent decades. Bergesen and Fernandez (1999) use network techniques to examine the Fortune 500 largest world companies. The now classic network study of the world division of labor by Smith and White (1992) examined change over recent decades. And new work by Kremple and Pluemper (1999) on the global automobile industry also examines change over time.

4. This problem has been generally considered in Chase-Dunn (1998: Chapter 15).

5. In national accounts statistics Gross National Product (GNP) is GDP less Net Factor Income from abroad. Net Factor Income includes credits and debits due to labor and capital and some other remittances (transfer payments). Repatriated wages earned by nationals abroad and wages paid to the nationals of other countries that are sent abroad are deducted, as well as profits earned by foreign-owned corporations that are sent out, and profits returned from domestic corporations having subsidiaries in other countries all constitute Net Factor Income. We use GDP instead of GNP because it includes transactions that are important parts of world product.

6. Of course tariffs have also caused smuggling, a practice that introduces error into trade statistics.

7. This assumption is unrealistic because exchange rates among currencies reflect many other things besides the relative value of goods and services in different countries. Kravis, Heston and Summers (1982) sought to correct this problem and to produce comparable estimates of real gross product by weighting GDP figures using a correction for the prices of a basket of typical consumer goods in each country, so-called purchasing power parity or PPP. These corrections were made for 1980. Korzeniewicz and Moran (1998) have pointed out that PPP weights are unrealistic for studies over long periods of time unless the weights are recalculated for the earlier time periods. Many studies have used the 1980 weights for earlier or later years, but this is certainly inappropriate if we want to do long-term cross-temporal comparisons.

8. Another way to consider the growth of world trade is to examine the value of total world exports per capita (of world population). This involves the problematic assumptions about exchange and inflation rates discussed above, but at least Maddison has more data on exports than on GDP so that we can see what happened since the 1880s. The exports per capita in constant 1990 U.S. dollars is geometric trend with a few suggestive bumps (see Figure A1 in the Appendix). The population data in the denominator are mainly interpolations, which smooths things out. But the size of this trend is impressive, reminding us of the huge increases in the economic value of goods that are traded internationally. This is not a good indicator of economic globalization because the amount and value of interactions within national economies also increase hugely over this same period, and exports per capita misses this entirely.

9. Of course the estimates of total world exports and GDP also suffer from incomplete data as we go back in time.

10. We are not entirely pleased with our method of weighting because changes in the composition of the "sample" can change the weights, and so the estimated values of trade globalization. Ideally we would like to have a method of weighting that allows completely meaningful comparisons between the nineteenth and twentieth centuries. This is important for whatever conclusions we will draw about a long-term secular trend in trade globalization, and the slope of such a trend if there is one. We have experimented with "time constant" weighting that uses a single grand mean of the world population for the whole period of study as the weighting denominator. This eliminates one source of variation over time, but it imposes a scale that is probably too small for the early nineteenth century and too large for the later twentieth.

11. Openness is not itself a good measure of dependency. What matters in the hierarchical division of labor in the world economy is whether the national exports are high or low in the value-added hierarchy. Little Switzerland, classically exporting fine watches, has a very different sort of openness from Honduras, which exports bananas. "Trade composition" is the concept that captures the nature of exports and imports and this notion has been studied for the whole world-system using network analysis by Smith and White (1993).

12. The cross-national correlation in 1991 between country openness and population size is -.21.

13. Figure 8 is a bit jumbled. A clearer comprehension is provided by examining the data series and the separate graphs for the different groups. The graphs are A2-A6 in the Appendix.

14. The standard deviation of the distribution of country openness is very high from 1900 to 1914, indicating an unusual level of diversity in this period.

15. One of their posters said, "Down, down with the infamous Corn Law." The Corn Law was a tariff on the import of foreign cereals that protected domestic farmers from competition from abroad. Cobden and Bright argued that this was a tax on the working class that augmented the incomes of landlords. 

References

Arrighi, Giovanni 1994 The Long Twentieth Century. London: Verso.

_______________ 1999 "The world market." Journal of World-Systems Research.

Bairoch, Paul 1993 Economics and World History: Myths and Paradoxes. Chicago: University of

Chicago Press.

Bairoch, Paul 1996 "Globalization Myths and Realities: One Century of External Trade and

Foreign Investment", in Robert Boyer and Daniel Drache (eds.) , States Against Markets: The Limits of Globalization, London and New York: Routledge.

Bairoch, Paul and Richard Kozul-Wright 1998"Globalization myths: some historical

reflections on integration, industrialization and growth in the world economy," pp.37-

68 in Richard Kozul-Wright and Robert Rowthorn (eds.) Transnational Corporations

and the Global Economy. London: MacMillan

Banks, Arthur S. 1975 Cross-national Time Series Data Archive. Binghamton, NY: Center for

Comparative Political Research, Binghamton University.

Barnett, George A. and Joseph G.T. Salisbury 1996"Communication and Globalization: A

Longitudinal Analysis of the International Telecommunication Network" Journal of World-Systems Research Vol. 2, No. 16. http://jwsr.ucr.edu/

Bauman, Zygmunt 1998 Globalization: The Human Consequences. New York: Columbia

University Press.

Bergesen, Albert and Roberto Fernandez 1999 "Who has the most Fortune 500 firms?: a

network analysis of global economic competition, 1956-1989" in Volker Bornschier and Christopher Chase-Dunn(eds.) The Future of Global Conflict. London: Sage

Boli, John and George M. Thomas 1997 "World culture in the world polity," American

Sociological Review 62,2:171-190 (April).

Borrego, John 1999 "Twenty-fifty: the hegemonic moment of global capitalism" in

Volker Bornschier and Christopher Chase-Dunn (eds.) The Future of Global Conflict, London: Sage.

Boswell, Terry and Mike Sweat 1991 "Hegemony, long waves and major wars,"

International Studies Quarterly 35,2:123-150 (June).

Boyer, Robert and Daniel Drache 1996 States Against Markets: The Limits of

Globalization. London: Routledge.

Dunning, John H. 1983 "Changes in the level and structure of international

production: the last one hundred years," Pp. 84-139 in Mark Casson (ed.) The

Growth of International Business. London: Allen and Unwin.

Chase-Dunn, Christopher 1980 "The development of core capitalism in the

antebellum United States: tariff politics and class struggle in an upwardly

mobile semiperiphery" in Albert J. Bergesen (ed.) Studies of the Modern World-

System. New York: Academic Press.

______________________1998 Global Formation: Structures of the World-Economy.

(2nd edition) Lanham, MD. : Rowman and Littlefield

Chase-Dunn, Christopher and Peter Grimes 1995 "World-systems analysis" Annual

Review of Sociology 21:387-417.

Chase-Dunn, Christopher and Thomas D. Hall 1999 "The Chesapeake world-system:

complexity, hierarchy and pulsations of long range interaction in prehistory,"

Presented at the International Studies Association, Washington, DC. February

19, 1999.

/archive/papers/c-d&hall/isa99a.htm

Chase-Dunn, Christopher and Bruce Podobnik 1999 "The next world war: world-system cycles and trends" in Volker Bornschier and Christopher Chase-Dunn (eds.) The

Future of Global Conflict. London: Sage.

Freeman, J.R. 1983 "Granger causality and the time series analysis of political

relationships" American Journal of Political Science 27:327-58.

Goldstein, Joshua 1988 Long Cycles: Prosperity and War in the Modern Age. New Haven:

Yale University Press

Gordon, David 1988 "The global economy: new edifice or crumbling foundation?"

New Left Review 168: 24-64.

Grimes, Peter 1996 "Economic cycles and mobility." Ph.D. dissertation, Johns Hopkins University, Department of Sociology.

Harvey, David 1989. The Condition of Postmodernity. Cambridge, MA.: Blackwell.

_____________ 1995 "Globalization in question." Rethinking Marxism 8,4:1-17

Hirst, Paul and Grahame Thompson 1996 Globalization in Question: The International

Economy and the Possibilities of Governance. Cambridge: Polity Press.

Hugill, P. J. 1993 World Trade Since 1431. Baltimore: Johns Hopkins University Press

Junne, Gerd 1999 "Global cooperation or rival trade blocs?" in Volker Bornschier and

Christopher Chase-Dunn (eds.) The Future of Global Conflict. London: Sage.

Keohane, Robert and Joseph Nye (eds.) 1970 Transnational Relations and World Politics.

Cambridge, MA. Harvard University Press.

Korzeniewicz, Roberto Patricio, Timothy Patrick Moran and David Consiglio 1998

"Global divergence since 1820: challenging the consensus,"

Kravis, Irving B., Alan Heston and Robert Summers 1982 World Product and Income:

International Comparisons of Real Gross Product. Baltimore: Johns Hopkins

University Press.

Kremple, Lothar and Thomas Pluemper 1999 "International division of labor and global

economic processes: an analysis of the international trade of cars with world trade

data," http://www.mpi-fg-koeln.mpg.de/~lk/netvis/globale/

Levy, Jack S. 1983 War in the Modern Great Power System, 1495-1975. Lexington:

University Press of Kentucky.

Lewis, Cleona 1945 Debtor and Creditor countries: 1938,1944. Washington, DC: The

Brookings Institution.

Lipietz, Alain 1987 Mirages and Miracles: The Crises of Global Fordism. London: Verso.

Maddison, Angus 1995 Monitoring the World Economy, 1820-1992. Paris: OECD

Markoff, John 1996 Waves of Democracy: Social Movements and Political Change. Thousand Oaks, CA.: Pine Forge Press.

McCormick, Thomas J. 1989 America's Half Century: United States Foreign Policy in the

Cold War. Baltimore: Johns Hopkins University Press.

McMichael, Philip 1996 Development and Social Change: A Global Perspective.

Thousand Oaks, CA.: Pine Forge.

Meyer, John W. 1989 "Conceptions of Christendom: notes on the distinctiveness of

the West." Pp. 395-413 in Melvin L. Kohn (ed.) Cross-national Research in Sociology. Newbury Park, CA.: Sage.

___________1996 "The changing cultural content of the nation-state: a world society

perspective." In George Steinmetz (ed.) New Approaches to the State in the Social Sciences. Ithaca: Cornell University Press.

Mitchell, B.R. 1992 International Historical Statistics: Europe 1750-1988. 3rd ed. New

York: Stockton

____________ 1993 International Historical Statistics: The Americas 1750-1988. 2nd ed.

NY:Stockton

____________ 1995 International Historical Statistics: Africa, Asia, & Oceania 1750-1988.

2nd ed. NY:Stockton

Mittelman, James H. (ed.) 1996 Globalization: Critical Reflections. Boulder, CO.:

Lynne Rienner

Modelski, George and William R. Thompson 1988 Seapower in Global Politics, 1494-

1943. Seattle: University of Washington Press.

______________________________________1994 Leading Sectors and World Powers:

the Coevolution of Global Economics and Politics. Columbia: University of South

Carolina Press.

Murphy, Craig 1994 International Organization and Industrial Change: Global Governance since 1850. New York: Oxford.

Pfister, Ulrich and Christian Suter 1987 "International financial relations as part of the

world-system" International Studies Quarterly 31,3: 239-272.

Podobnik, Bruce 1999 Global Energy Shifts: States, Firms and Social Unrest in the

Evolution of Commercial Energy Industries. Ph.D dissertation, Sociology, Johns

Hopkins University.

Rasler, Karen and William R. Thompson 1994 The Great Powers and Global Struggle,

1490-1990 Lexington: University Press of Kentucky.

Robertson, Roland 1992 Globalization: Social Theory and Global Culture. London: Sage.

Ross, Robert and Kent Trachte 1990 Global Capitalism: The New Leviathan. Albany:

State University of New York Press.

Sachs, Jeffrey and A. Warner 1995 "Economic reform and the process of global

integration," Brookings Papers on Economic Activity, 1, 1-118.

Schaeffer, Robert 1997 Understanding Globalization: The Social Consequences of

Political, Economic and Environmental Change. Lanham,Md.: Rowman and Littlefield

Shimizu, Kosuke 1996 "The dialectics of globalization: whose globalization is it anyway?" Review of Sekai Hendo no Nikata by Takashi Inoguchi. Mershon International Studies Review 40:266-271.

Sklair, Leslie 1995 Sociology of the Global System. Baltimore: Johns Hopkins University Press.

Smith, David A. and Douglas White 1992 "Structure and dynamics of the global

economy: network analysis of international trade, 1965-1980." Social Forces

70,4:857-893.

Soros, George 1998 The Crisis of Global Capitalism. New York: Pantheon.

Su, Tieting. 1995. "Changes in World Trade Networks: 1938, 1960, 1990" Review XVIII, 3 :431-459 (Summer)

Suter, Christian 1992 Debt Cycles and the World-Economy: Foreign Loans, Financial Crises and Debt Settlements, 1820-1990. Boulder, CO. : Westview.

Taylor, Charles L. and M.C. Hudson 1972 World Handbook of Political and Social Indicators. New Haven, CT.: Yale University Press.

Taylor, Peter 1996 The Way the Modern World Works: World Hegemony to World Impasse New York: Wiley.

United Nations 1994 World Investment Report 1994: Transnational Corporations, Employment

and the Workplace, New York

Wallerstein, Immanuel 1974 The Modern World-System, Vol. 1. New York: Academic

Press.

The World Bank (I.B.R.D.) 1980 World Tables. 2nd Ed. Baltimore: Johns Hopkins

University Press.

_______________________1998 World Development Indicators,

http://www.worldbank.org/html/extpb/wdi98.htm

Woytinski, Wladimir S. and E.S. Woytinski 1955 World Commerce and Governments.

New York: Twentieth Century Fund.