The Growth of Transnational Corporate Networks : 1962 – 1998 *

This is a study of the growth of organizational power in the world-economy over the past forty years. It takes the position that transnational corporations (TNCs) are increasingly significant actors in the world-economy, independent of the nation-states within which they are located. The goal of this work is to identify the expansion, spatial distribution, and concentration of this global power over time, and to consider its impact on the global economy. The TNC networks are identified by locating the headquarters and foreign subsidiaries of the worlds 100 largest manufacturing corporations in 1962, 1971, 1983, 1991 and 1998. The distribution of ownership and location of these foreign subsidiaries are examined, both globally and bilaterally. I find high levels of concentration in ownership of these global networks that decrease over time, in contrast to a high degree of dispersion in the location of these linkages. U.S. corporations are clearly the dominant actors from 1962 to 1971 but decline dramatically through 1998, while Japanese and Western European TNC control over transnational networks grows significantly over this period. An empirical measure of economic dominance in the global economy is also presented.


the shifting locus of economic power
Th e central tenet of this work is that transnational corporations, and their global networks, represent a distinct locus of power that have a signifi cant impact on an increasingly global economy.Th e theoretical basis for this assertion begins with Charles Tilly' s (1994) work on the emergence of the modern nation-state system.Tilly argues that the current inter-state system is the result of a merging of coercive and economic power between a.d.1000 and 1800.Prior to this time, economic and coercive, or military, power were separate.Political units, such as states, feudal areas and empires, were essentially containers of coercive power, used to acquire the necessary goods, and people, to maintain their sys-tems.Economic power resided within cities, the centers of economic activities in these times, and where capital was accumulated by the emerging burgher class.As military technology progressed and warfare became more expensive, these political organizations were forced to look to cities for the fi nancing of their military activities.Th e resulting relationship between state and city, of coercive and economic power, solidifi ed over this 800-year period, giving rise to the modern nation-states of today.Th ese modern nation-states controlled both military and economic power.
During the last few decades of the twentieth century, however, this coalescence of economic and coercive power began to fracture, due primarily to the emergence of the transnational corporation.Saskia Sassen (1991) describes the global dispersion of production that began in the 1970s, as corporations searched for lower wages, closer proximity to markets and raw materials, and a way to diffuse the power of labor.Th e corporate headquarter-foreign subsidiary linkages that emerged as a result of this process of production dispersion have formed the basis for a new dimension of economic power.It has allowed the transnational corporation to circumvent, to a signifi cant extent, the regulation of their activities by the nation-states within whose boundaries they are located.¹Christopher Ross (1994) provides an additional perspective on these organizational networks.Ross, a human ecologist, studied the structure of city-systems within the United States.He argued that cities were essentially containers of organizations and city-systems were, therefore, refl ections of organizational networks, primarily corporate.Th e hierarchy of these city-systems was determined by the relative power of the corporations residing within these cities in terms of their control over the economic activity in other cities.Ross operationalized these power relationships in terms of corporate headquarter and subsidiary locations.To the extent that a corporation headquartered in New York, for example, has a subsidiary in Pittsburgh, some amount of control or power is acquired by New York over Pittsburgh.Th e New York based corporation has an impact on employment and capital activities in Pittsburgh, which reduces the control that Pittsburgh has over its own economic activity.In other words, it refl ects a loss of autonomy for the host location.
¹.A key point for Sassen is that this diff usion of production was accompanied by the expansion and concentration of certain service functions, primarily producer services, which facilitated the coordination and expansion of these manufacturing activities, giving rise to a new network of "global cities." Jeff rey Kentor 266 The Growth of Transnational Corporate Networks  immense in scope.Philip Morris, for example, operates in 170 countries.In 1983, corporate revenues of the world' s 500 largest corporations equaled 15 of world GDP.By 1998, this ratio had grown to 28 (Kentor 2002).Revenues of the largest 200 corporations now exceed the combined economic activity of 182 nations (Anderson and Cavanagh 2000).³Clearly, these TNCs have become signifi cant actors in the global economy, apart from the nation-state system within which they are geographically located.

the analyses
As stated earlier, the goal of this work is to describe the economic and spatial expansion of transnational corporate networks over time, in terms of both individual countries and the global network as a whole.I employ the methodology used by Ross (1996) discussed above, by charting the shifting patterns of TNC headquarter-subsidiary linkages for the world' s 100 largest manufacturing corporations at fi ve time points : 1962, 1971, 1983, 1991 and 1998.Th ese dates were chosen for data availability.Th is study is limited to manufacturing fi rms because, as Ross points out, these organizations have the greatest impact on the host economy relative to other sectors.Generally, they have the largest fi xed investments, employ the most workers and have the greatest impact on the environment.Th e location of each corporate headquarters is identifi ed, along with the number and locations of all foreign subsidiaries, for each time period.Hence, the specifi c corporations are not constant across periods.
Th is methodology has two possible drawbacks.First, it is conceivable that limiting the data set to the largest 100 industrial corporations could bias the sample.Th is methodology would exclude countries with only mid-sized corporations or those outside the industrial sector of the economy.To address this issue, another data set was constructed that includes the largest 250 fi rms in 1998, irrespective of the sector within which it is located.Th is list is taken from the Fortune Global 500.Th is allows a comparison of network structures, which will provide a reasonable indication of the possible bias in the methodology.Second, it is assumed that all foreign subsidiaries have the same impact on the host country.Th ere is no explicit control for subsidiary size, in terms of assets, sales, or number of employees.Th ese would certainly be potentially useful data to include in the analyses, but are not available.However, there is an implicit control on Ross created an organizational matrix to describe the city-system hierarchy in the U.S., by identifying the locations of corporate headquarters and their subsidiaries in major metropolitan areas in 1950 and 1980.Ross referred to these headquarter-subsidiary networks as "control linkages."He selected only manufacturing corporations for his study, arguing that industrial activity has a greater impact on a city' s overall economic activity than primary or tertiary activities.²Ross concluded form these analyses that (a) urban systems are pyramidal in nature, with a few dominant cities at the top (New York and Chicago) and an increasing number of cities at lower levels of the hierarchy, and (b) dominance, or power, decreases at lower levels of these networks.
Th is global diff usion of production was facilitated by the emergence of what John Meyer et al. (1997) refer to as the "world society," an ideology that legitimates and facilitates the penetration of foreign interests into less developed countries.Th is is refl ected in an isomorphism of laws and conventions throughout the world-economy concerning foreign ownership of private property, repatriation of capital, employee/employer rights and accounting practices that permit transnational corporations to locate and operate with consistency and predictability within most countries around the globe.
I extend the above arguments to formulate the following theory of international dominance.Th e coalescence of economic and coercive power, which generated the modern interstate system, has begun to unravel with the global diff usion of manufacturing that has occurred over the past forty years.Th e transnational corporate networks that have emerged over this period refl ect a distinct locus of economic power.I do not suggest that these TNC networks are the sole conduits of power in the world-economy.Th ey are only one network structure embedded within a series of hierarchical, overlapping networks that includes fl ows of information, migration, transportation, culture, and coercion.But they certainly play a signifi cant, if not primary, role in the evolution of the world-economy.

empirical evidence
A brief examination of available data strongly suggests that transnational corporations represent a signifi cant, and growing, dimension of economic power in the global economy.In 1962, the world' s 100 largest industrial corporations owned 1288 foreign subsidiaries.By 1998, the 100 largest industrial fi rms owned nearly 10,000 foreign subsidiaries.Some of these transnational corporate networks are ².Meyer (, ) conducts similar analyses of fi nancial networks in the Southern U.S., and South America.
³.It is interesting to note that while TNC economic activity has grown dramatically over the past forty years, its share of workers has steadily declined.Th e world' s  largest corporations employ less than one percent of the global work force.

Jeff rey Kentor 268
The Growth of Transnational Corporate Networks  the size of these subsidiaries, because they are all owned by the world' s largest corporations.Th e TNC headquarter and subsidiary data used in these analyses have been obtained from various years of Moody' s Directories, Dun and Bradstreet, Directory of Inter-Corporate Ownership, Th e National Register, Standard and Poor' s Register of Corporations, Th e Directory of Multinationals, Th e World Directory of Multinational Enterprises and the Directory of American Firms Operating in Foreign Countries.Not all of the corporations listed in the top 100 industrials (or Global 500) had foreign subsidiaries, according to the sources listed above.It is unclear, however, whether this indicates there are no subsidiaries, or that these data are missing.In some cases it was possible to contact a specifi c corporation to confi rm this information, in others it was not.Th e impact of these possible missing data would be to underestimate the extent of these TNC networks.

results
Th e expansion of these TNC networks during the 20 th century has been dramatic.Table 1 presents the ratio of revenues of the 100 largest industrial fi rms to world GDP.
Th is ratio grew from .04 in 1912 to .11 in 1991, before declining to .09 in 1998.Th is decline refl ects the growth of the service sector.Th e Global 100, which include all sectors of the economy, grew from .09 of world GDP in 1983 to .13 in 1998.Th e revenues of the Global 500 grew from .15 to .28 of world GDP between 1983 and 1998.Th ese fi gures refl ect the expansion of producer services that arose in response to the diff usion of production that began in the 1980s (Sassen 1991).Another indicator of the growth of producer services is the ratio of revenues of the 100 largest industrial fi rms to the revenues of the Global 500, which declined from 60 in 1971 to 46 in 1998.
I now turn to an examination of the distribution of TNC headquarter-foreign subsidiary linkages for the 100 largest industrial TNCs from 1962 and 1998, as shown in Table 2.⁴ Th ere are two ways to think about these linkages.
We can examine both the total number of TNC headquarter-subsidiary connections, and the number of dichotomous (country to country) linkages.Th ese two aspects of TNC networks have diff erent meanings and diff erent impacts.Dichotomous linkages represent relationships between countries.From a dichotomous perspective, it is the existence or non-existence of a linkage that is important rather than the absolute number of linkages.Whether there are one or several linkages between two countries, a whole host of political, economic, social (and possibly even military) laws, regulations and norms are required for a TNC to be able to locate within a given host country.Th e total number of linkages is more clearly a measure of penetration of a host country by transnational corporations.Total TNC headquarter-foreign subsidiary linkages for the 100 industrial TNCs grew from 1,260 in 1962 to nearly 10,000 in 1998, with the sharpest increase occurring between 1991 and 1998.Th e number of dichotomous TNC linkages, or country to country connections, grew from 220 in 1962 to 780 in 1998, with the largest growth occurring between 1971 and 1983.
Th e distribution of TNC headquarter-subsidiary linkages for the top 250 fi rms in 1998 is, as expected, somewhat larger.Th ere are nearly twice as many total linkages (19,481 / 9988) and approximately 50 more dichotomized connections (1241 / 780).However, the correlation between the two data sets is extremely high, both for out and in-degrees (.97 and .91, respectively), suggesting that the 100 largest manufacturing TNCs are a reasonable refl ection of the organizational network as a whole.
An examination of the distribution of these linkages by country, given in Table 2, indicates a high degree of concentration in the ownership of these networks, and a dispersion of the location of these subsidiaries.
In 1962, the 100 industrial TNCs with foreign subsidiaries were concentrated in only fi ve countries: the U.S., the Netherlands, the U.K., Germany and Italy (see Figure 1).U.S. corporations were clearly the dominant fi rms, controlling 1040 of the total 1260 subsidiaries.
In other words, U.S. based corporations owned 82 of all foreign subsidiaries of the 100 largest industrial TNCs in 1962.Th e position of U.S. TNCs strengthened in 1971, controlling 1337 of the 1566, or 86, of all foreign subsidiaries.Dutch TNCs, the next largest group in 1971, owned only 115 (9) subsidiar-⁴.Th ese analyses do not directly examine diff erences in revenues among these subsidiaries.However, the parent TNCs (world' s  largest industrial TNCs) are from a uniformly high revenue group, which indirectly controls for economic size.The Growth of Transnational Corporate Networks  4 th , owning 18 and 14 of all foreign subsidiaries, respectively.A useful measure of transnational organizational dominance can be constructed from the above data by examining the ratio of the greatest number of foreign subsidiaries owned by TNCs from a single country to the number of foreign subsidiaries owned by TNCs of all other countries, somewhat analogous to the notion of urban primacy (Walters 1985).
Th is measure quantifi es the extent to which TNCs from a single country dominate the organizational networks in the global economy; much like urban primacy refl ects the dominance of a single largest city over the entire city system of a given country.For example, U.S. TNCs owned the largest number of foreign subsidiaries in 1971, 1040 out of 1260, resulting in a ratio of 1040/220, or 4.7.Th is means that U.S. TNCs owned 4.7 times as many foreign subsidiaries than TNCs from the balance of the world combined.According to this measure, U.S. TNC dominance peaked in 1971, with a score of 6.1.Th ere was a dramatic decline ies.TNCs in seven other countries owned foreign subsidiaries: Germany, France, U.K., Italy, Japan, Australia and Switzerland.
Th ere is a signifi cant decline in the concentration of subsidiary ownership after 1971.Th e number of countries with TNC headquarters in 1983 grew 30 (from 9 to 12), refl ecting a continued diversifi cation of control.TNCs from Belgium, Sweden and Canada now own foreign subsidiaries, while Australia has been dropped.Th is trend continued in 1991, with the U.S. TNC share of ownership dropping to 28.4 (982 out of 3454 total subsidiaries).TNCs from 14 countries now control foreign subsidiaries, with signifi cant country movement.Finland, Spain and Venezuela have been added, while Canada, Brazil and Australia have been dropped.Th e U.S. position stabilized in 1998, now accounting for 2901 of 9977, or 29, of total out-degrees (see Figure 2).Japanese TNCs, now the second largest group, had the most dramatic growth, controlling 2296 foreign subsidiaries, or 23 of the total.German and Swiss TNCs ranked 3 rd and It is important, however, to distinguish between the concentration of control of the total number of foreign subsidiaries on a global level and the concentration of ownership within a given country.If we look at the concentration of ownership within a given country, a diff erent picture emerges.In 1971, the peak year for U.S. TNC dominance, U.S. industrial corporations owned the majority of foreign subsidiaries in 86 countries, with an average concentration of 88.Th ese countries, listed in Table 3, include nearly all of the developed world; the U.K. (96), Canada (95), Japan (91), Germany, Italy, Sweden and Switzerland (90) Netherlands (89) and Australia (84), among others.U.S. industrial TNCs are also dominant in the developing countries of that period, including Argentina (89), Brazil (88), Mexico and Taiwan (87) and India (58).By 1998, this situation had changed signifi cantly.Th e U.S. industrial  fi rms now dominated the foreign subsidiary structure of only 50 countries, with an average concentration of 49 percent, as listed in Table 4.U.S. industrial TNCs still own the largest number of foreign subsidiaries in many of the same countries.However, the concentration is substantially reduced: Japan (64), the U.K. and the Netherlands (47), Germany (42), Australia and Canada (38) and Italy (32).Japanese TNCs, with 22 of total foreign subsidiaries in 1998, dominated the ownership of foreign subsidiaries in only 18 countries, as listed in Table 5. Th ese include only one developed country, the U.S. (43), along with several regional economies; Th ailand and Indonesia (60), Malaysia (50), Taiwan and the Philippines (40), China (38) and Singapore (37).⁵Th e location of foreign subsidiaries has also shifted signifi cantly between 1962 and 1998.Foreign subsidiaries of the 100 largest industrial fi rms were located in 99 countries in 1962 (see Figure 4).Canada was host to the largest number of these (169, 13), followed by the U.K. (132, 11), Germany and France.
Only 22 subsidiaries (2) were located in the U.S., and only 18 in Japan.By 1998 this picture had reversed.Th e U.S. was now the single largest host, accounting for 15 (1479 out of 9977) of all foreign subsidiaries.Th e next largest, the U.K., accounted for 8 (827) of the total.Japan, however, maintained its lim-   ited role as host to the world-economy, with only 302 foreign subsidiaries being located there in 1998 (see Figure 5).Th e above analysis has focused on the relationships between developed and less developed countries.A more complex set of power relationships emerges when we examine the TNC networks among developed countries where there is a more equal distribution of control over these linkages, especially for the more recent period.
Tables 6 and 7 are ownership/location matrices of industrial TNC linkages The Growth of Transnational Corporate Networks  among developed countries for 1971 and 1998.China is included in these matrices for discussion purposes only, to highlight the fact that they do not participate in these bilateral relationships during these time periods.Th e rows of these matrices represent the country of ownership and the columns indicate the location of the foreign subsidiaries.Th is format allows us to examine the distribution of power among developed countries as bilateral relationships, rather than aggregate characteristics for a given nation.Th e U.S. maintained dominant bilateral relations over all other developed countries in 1971.For example, U.S. TNCs owned 109 subsidiaries in the U.K., while U.K. TNCs owned only 9 subsidiaries in the U.S. A similar imbalance existed with Germany (69 to 0), France (53 to 3), Switzerland (45 to 0), and Japan (31 to 5).By 1998, however, many of these relationships had reversed.Th e most dramatic change occurred between the U.S. and Japan.Japanese TNCs now own 619 subsidiaries in the U.S., compared to 182 U.S. subsidiaries in Japan.German and Swiss TNCs also had more subsidiaries in the U.S. than U.S. TNCs had in these countries.
Th e interpretations of these bilateral relationships require two additional considerations.First, we need to control for the overall sizes of the two economies when considering the relative impacts of foreign subsidiaries.Switzerland may Jeff rey Kentor 280 The Growth of Transnational Corporate Networks  own more subsidiaries in the U.S. than vice-versa, but the potential loss of U.S. subsidiaries in Switzerland would likely have a greater impact on Switzerland than the removal of Swiss subsidiaries from the U.S. Second, we need to consider the potential impact of the European Union.To the extent that the EU becomes a single integrated economy, it may be more appropriate to examine these bilateral headquarter-subsidiary relationships between the U.S. and Europe as a whole, rather than individual European countries.
discussion Th e expanding and shifting organizational networks described above provide a productive level of analysis for understanding the way in which power is concentrated and distributed in the world-economy.Th ese descriptive analyses clearly indicate a signifi cant expansion of economic power of transnational corporations, in terms of both dollars and global linkages.We should be careful, though, not to overestimate the strength of these transnational corporations visà-vis nation-states (Wolf 2001).It could be argued, for example, that corporate revenue is not equivalent to a national gross domestic product.Further, the state still holds a monopoly on coercive power.Th ere can be little doubt, however, that these giant fi rms wield a signifi cant and increasing amount of power that is, to some extent, beyond the control of nation-states.
One question that arises from this work concerns the value of using the number of foreign subsidiaries as a measure of TNC penetration in place of the traditional aggregate indicator of foreign capital penetration: foreign direct investment as a percentage of GDP.Th ere are several theoretical reasons for exploring the impact of globalization from an organizational perspective.It is, fi rst, an eff ort to examine these global processes at the level of the operant actors.Foreign investment is typically controlled by individual corporations.Th ese corporations determine the place and amount of foreign investment, the transfer of technology, access to their international markets, repatriation of profi ts, number of employees, etc. Th ese TNCs also control a variety of non-equity items, such as licensing agreements.And in less developed countries, these corporations may carry signifi cant infl uence on the host country' s political processes.Th e notion that "foreign investment" is a homogeneous variable with singular eff ects seems increasingly less likely (Kentor and Boswell 2003).Th e more productive avenues of research appear to be in examining the components or structures of foreign investment.Decomposing foreign investment allows us to understand the multiplicity of eff ects, be they positive, negative, or benign.
Th ere are empirical reasons to suggest that foreign subsidiaries may have signifi cant, independent eff ects.Preliminary empirical analyses indicate that relatively high concentrations of foreign subsidiaries retards economic growth in less developed countries (Kentor 2002).
Th ese TNC networks also appear to provide a useful level of analysis for examining changes in the international nation-state hierarchy, which we commonly refer to as "hegemony."Th e headquarter-subsidiary data presented above seems to refl ect well the apex and decline of U.S. hegemony from 1970 to 2000, and the dramatic rise of Japanese and European TNC networks during this  The Growth of Transnational Corporate Networks  summary Th is descriptive study of transnational organizational networks provides a new vantage from which to examine the process commonly referred to as "globalization."It adds to our theoretical understanding of the ways in which power is organized and transmitted around the world.It also suggests new ways of thinking about the future distribution of power in the world-economy.Peter Taylor (1996) has written of the end of national hegemony.He argues that the world-economy is so large that a single country can no longer exert dominance over it and questions whether future hegemons may be coalitions among several countries.Th is may be the wrong question entirely.Global dominance may no longer be within the domain of nation-states at all.Th e locus of power in the world-economy may be shifting to transnational organizations.Th e implications of such a shift are profound.John Markoff (1996) examines the impact of this redistribution of power on democracy.He argues that, to the extent transnational corporations increasingly "make the rules," democracy is threatened.When nations make laws, there are political mechanisms by which these laws can be challenged.Th ere are no such mechanisms for challenging the "laws" made by transnational corporations.
As noted above, this is a descriptive study of the growth of transnational corporate power, a necessary fi rst step in exploring the expansion of transnational corporate power over time.An obvious extension would be to expand this study to the Global 500.Th is would provide a more complete picture of foreign subsidiary penetration as well as changes in the sectoral distribution of these subsidiaries.Th ese data would also provide a more comprehensive measure of the international nation-state hierarchy.A variety of causal analyses can be generated from these data as well, such as the impact of foreign subsidiaries on economic growth and inequality.Network analyses of these data could also tell us much about changes in the global economy over time, in terms of centrality, density, cliques, and other network measures.Finally, these studies could profi tably be merged with other aspects of transnational corporate networks, such as interlocking directorates.period.It is interesting to note, though, that the expansion of Japanese TNC networks in the 1990s was not accompanied by expansion of the Japanese national economy, which has been in recession for the entire decade.Th is may refl ect the growing independence of transnational corporations from their territorial bases or, in other words, the growing chasm between economic and coercive power.
Alternatively, the answer may be found in the relationship between ownership and penetration of TNC subsidiaries.Following Ross (1996), it seems reasonable to assume that ownership of a foreign subsidiary transfers some amount of power from the host country to the TNC and, in some cases, to the country in which the TNC is headquartered.Th e economic and political implications of having a foreign subsidiary located within a country are less clear.Is it more benefi cial for a country to have few or many foreign subsidiaries located within its borders?
In political economy terms foreign subsidiaries may result in a loss of power, or autonomy, for the host country.To the extent that an outside actor has control over the internal dynamics of the host country, it reduces that country' s ability to control or direct its own economy.Th is loss of control may be exacerbated by the political and social ties that emerge from these economic linkages.But the impact of foreign subsidiaries may not be uniform.Th e impact may vary as a function of the economic and political strength of the host economy.In developed countries like the U.S., having foreign subsidiaries within its borders may be benefi cial, both from economic and political vantages.Th ese foreign subsidiaries become, to some extent, captured resources of the host country.Laws of the host country may preclude TNCs from unilateral actions that might be harmful to its economy or its work force.In less developed countries, however, the host country may not have the economic or political strength to confront TNC penetration.A related way to think about this is in terms of a balance between ownership and location.Th e impact of foreign subsidiaries in a host country will be most pronounced if the host country is bereft of TNCs with subsidiaries outside the host country.In countries with more of a balance between ownership and penetration, the eff ect may be lessened or even reversed.
Th e role of China also deserves mention.Many would argue that China is upwardly mobile in the world-economy, and may be an ascending hegemonic power (Frank 1998), citing China' s rapid growth in market size, national income and military strength.I would suggest otherwise.Th ese results clearly illustrate that Chinese TNC networks have little impact in the global economy.It is possible that China' s mobility in the world economy may be hindered by its inability to project its national interests via these networks.Th e recent eff orts of Chinese fi rms to purchase large U.S. fi rms such as UNOCAL may be attempts to "buy their way" into these networks.

Figure
Figure 1 -Distribution of Subsidiary Ownership in 1962

Table 3 (Cont.) -Countries Dominated by US Subsidiaries 1971 (IND 100)
⁵.It is interesting to note that Dutch industrial TNCs, which controlled fewer than  of total foreign subsidiaries in , dominated ownership of foreign subsidiaries in  countries, with an average concentration of .

references
Anderson, Sarah, and John Cavanagh..Top : The Rise of Global Corporate Power.New York: Global Policy Forum.Angel, Juvenal.Various Years.Directory of Inter-Corporate Ownership.New York: Simon and Schuster.Bergesen, Albert, and John Sonnett.."The Global : Mapping the World Economy at Century's End."American Behavioral Scientist : -.