GaWC Research Bulletin 372

GaWC logo
  Gateways into GaWC

This Research Bulletin has been published in Tijdschrift voor Sociale en Economische Geografie, 102 (5), (2011), 499-514.


Please refer to the published version when quoting the paper.


The Map of Multilateral Treaty-Making 1600-2000: A Contribution to the Historical Geography of Diplomacy

H. van der Wusten*, R.A. Denemark**, M. Hoffmann*** and H. Yonten****


The paper describes and analyzes the successive geographical distribution of places where multilateral treaties have been signed over the life course of the state system. A large proportion of all negotiations occurred in just a few places and the collection of most frequently selected places shows considerable continuity over time. Treaty-making emerges as more of a secular trend than a cyclical pulse, being insignificantly impacted by economic cycles, and inconsistently impacted by hegemonic cycles. The work presents a measure of specialization that helps to identify types of central venues in the multilateral treaty-making system. The actual selection of specific venues suggests functional and political considerations to have been most important. The sustained preference for national political centers expresses the importance of such considerations, while the actual choice of a venue in a specific case can be highly contingent.

Key words: diplomacy, hegemony, multilateral treaties, political centers, world-systems analysis


A state cannot be single. States evolved and exist in groups. To avoid or resolve disputes, to prevent future differences from upsetting relationships, and to coordinate or commonly execute tasks, states occasionally engage in treaty-making. An interesting, rarely studied feature of international negotiations and treaty-making concerns the set of places where those activities have occurred (examples include Henrikson 2005 and Van der Wusten 2007). In this paper we analyse the pattern of places where multilateral treaties have been produced and signed over the life course of the state system. This initial section marks the specificity of the multilateral track, indicates the variation in the types of treaties with which we have to deal under that label and suggests the possible importance of studying the geographical distribution of the venues where delegates gathered, the ‘production sites’ of those treaties. The following sections describe the maps of these sites over the last 400 years and interpret the regional distributions and the repeated selection of specific places.

The Multilateral Track

Multilateral treaties connect more than two states or other actors. This numerical/literal definition is common in the literature and indicates a ‘natural’ distinction among treaties. There are treaties with two signatories (bilateral treaties) that reflect the traditional view of activity in the state system. Alternatively, and in exceptional circumstances, a large number of states come together to conclude much broader agreements. Such multilateral fora are considered a separate way of operating (e.g. Berridge 2002 chapter 7 versus chapter 9; Muldoon et al. 2010 to be published on new developments).

In two-actor games the number of possible moves is small given the limited range of positions that only two actors might possess. Resulting treaties tend to reflect straightforward power differences between the parties. In multiple actor games there are both more actors whose intentions have to be taken into account, and different coalition strategies that may be employed. Such treaties, when successfully concluded, more often help to establish and reinforce generalized rules of behaviour for all the actors involved. These can only partially reflect the power differences between the actors concerned. The resulting multilateral frameworks may introduce expectations and norms that can subsequently be used for further cooperative ventures. It is Ruggie (1992) who most forcefully introduced this perspective in the modern literature. For Ruggie, multilateralism is a mode of organizing the state system that speaks to the creation of norms and rules for relations across the widest number of variables. Multilateralism as a preferred policy has been disputed and vehemently discussed in the recent past (e.g. Ikenberry 2003, 533-550, Van Oudenaren 2003, 33-47).

The Map of Venues

This paper will focus on the contribution of the venues of multilateral treaty-making to the spatial distribution of political centers generally, and consider the selection process of those places by the initiating parties. Sites of multilateral treaty-making add to the array of political centers within the state system. Multilateral treaty sites constitute one type of political center. They may coincide with other types of centers, they may specialize in agreements in specific policy domains (e.g. war and peace, or the environment), and they may prove to be long or short-lived in their roles.

In this paper we describe the successive maps of places where multilateral agreements have been signed over four centuries. In section 3 we use a conventional zonal partition of the world and also world-system analysis to account for the successive regional distributions of signature places. Section 4 is about the selection of specific places based on hegemonic position, functional conditions and political interactions.


A new dataset that covers over 7000 multilateral treaties signed in the last 400 years forms the basis for the present analysis. The Multilateral Agreement and Treaty Record Set (MATRS) was constructed by consolidating information from several compendia of multilateral agreements (among others Mostecky 1965, Bowman and Harris 1984, Wiktor 1998), and augmented by a variety of national sources. When available, the title, signators, date and place of signature, and two substantive categorizations (one broad and one more specific) were recorded (Denemark and Hoffmann 2008).

Figure 1: Multilateral treaties signed per year 1596-1995

Figure 1

Figure 1 shows the temporal distribution of multilateral treaties concluded between 1596 and 1995. The data suggest about 200 years of minimal activity. There is a flash of treaty-making at the end of the period, with the wars that concluded the Napoleonic era. This is the highest peak of the 19th century, suggesting an urgent need by victorious powers for a thorough re-arrangement. The Concert of Europe was the result. From about 1850 to the 1960s multilateral treaty-making increased exponentially with two deep declines during the two World Wars. At the end of these brief war-time periods the growth curve quickly returned to its original trajectory. The underlying process gives more the impression of being crisis interrupted than driven by the victorious architects of successive new world orders. Since the 1960s the frequency of multilateral treaty-making has not grown. Volatility has increased and the trend is perhaps somewhat declining. Unfortunately we cannot see the consequences of the aftermath of the Cold War in these data.

One major question raised by the data concern its apparent periodization in three parts: before 1815, 1815-1960s, after 1960s. We look at the first two parts as essentially determined by the history of the European part of the state system. The Europe of the ancien régime was built upon a number of states with fairly temporary demarcations, and societies with a relatively small number of transnational links. A flat, low level curve resulted. The volume of these links increased immensely during the 19th and 20th centuries, encouraging a host of multilateral treaty-making. Treaties provoked additional treaties, while additional parts of Europe and the rest of the world became involved in these processes (Denemark and Hoffmann 2008). The third period, starting about 1960, shows a flattening trend in treaty-making. We suggest that the additional regulatory activity surrounding new cooperation was taken up by existing organizations, particularly what eventually became the European Union (van Middelaar 2009, 31-60) but also the International Labour Organization, the Council of Europe and the United Nations. New cooperative interaction needed less often to appear in the form of separate treaties. For 6769 of all these treaties we have data on their subject matter. They have been categorized in six classes. The shifts in the subject matter of the agreements over time can be read from table 1

Table 1: Distribution of all treaties in six types for three periods


War &


Comm. & Transport.


Trade &



































The table shows the dominance of war and peace issues on the multilateral agenda during the ancien régime. Survival and demarcation of states were the problems that mattered most in this period. Since Napoleon, the etiquette of state relations became and remained an important issue. Additional cooperation focused on transboundary problems: e.g. rivers, railroads, labour migration, modes of communication and transportation and contagious diseases. International trade and economic relations became much more important. The most recent period shows an increase in the number of treaties that cover environmental issues. Multilateral cooperation is not tied to any particular sector of public policy, it is a potentially fruitful diplomatic method across the international agenda of state concerns.

For 5958 treaties (85% of the total number) we know the places, usually cities, of signature. Negotiations have been hosted in 416 places, but 286 (69%) were used less than 5 times for this purpose. Twenty-one places occupy the top positions of the successive periods during these four centuries, accommodating half of all treaty negotiations in their respective periods. This is a highly skewed distribution with considerable stability in the top positions over time.

Figure 2a-c: Prominent signature places during three periods

Figure 2

Figure 2 portrays the most prominent venues for multilateral treaty-making in our three periods. The lists have been constructed in the following manner. Prominent venues in each period or stage are those with the highest frequencies that collectively make up 50% of all treaty-signing locations in that period or stage. The portrait of the second period has been subdivided in 6 parts to take into consideration its different stages. In the first period (1596-1814) 5 places make up 50% of venues. In the last period (1967-1989) when the overall number of treaties had become much larger, there were 8 prominent places. Since the number of treaties strongly climbs between 1815 and the 1960s it hardly makes sense to produce a single list of top ranks as it would be strongly biased towards the most frequently selected places in the later years of this period. Instead, we divided this period into 6 roughly equal segments. Using the same approach for each subfield (listing cities that collectively make up 50% of all treaty-signing venues) we found 3, 5, 6, 9, 7, and 7 prominent venues. This again shows the number of front ranks going up somewhat with an early peak in 1897-1919. Over the course of four centuries there are 50 high ranking positions. The overlap of selected places between periods results in 21 different places (in Figure 2) occupying these 50 positions. Maximum overlap would have resulted in 9 different places, no overlap in 50, so the continuity is considerable. One venue turns up amongst the most frequently selected places in all the different time segments, another 7 in more than one, 13 appear only once. While continuity is considerable, there is also some non-trivial change.

In the next two sections we consider first the zonal component in the successive distributions of selected places, particularly of those most frequently selected, and then the possible arguments for selecting specific places within those zones, particularly those that were repeatedly selected.


A first impression of the zonal distribution of venues can be derived from the data on the top ranks presented in Figure 2. The preponderance of Europe is obvious. During the first period, prominent places are exclusively European. In the second period first Constantinople (already long involved in the European diplomatic network), emerges. As a more or less separate Latin American diplomatic network is constituted in the later part of the 19th century, their respective capitals briefly act around 1900 as prominent places in the production of region-wide multilateral agreements. This build-up of a new regional extension of the state system and the still unstable nature of such arrangements account for the more dispersed character of prominent places from 1897 to 1919. At about the turn of the 20th century US power increases and Washington DC becomes a prominent venue which has endured since. Even in the last period, as decolonization is complete and nearly the entire inhabited world has been incorporated in the state system, prominent places for multilateral treaty-making remain heavily concentrated in Europe. Geneva enters the list in 1920 and remains throughout, while only one new place (Strasbourg) emerges in the last period. The only non-European venue remains Washington DC. It remains to be seen why signature places elsewhere have not become relatively more prominent as the state system has geographically been extended. Is it because multilateral cooperation has more quickly grown in the older parts of the system or have agreements, global in scope and primarily agreed in well established locations, become more preponderant?

The other way to approach the question of the successive zonal patterns is to avoid conventional zones like continents and look for theoretically relevant zones that may assist in explaining the distribution of multilateral venues. World-systems analysis provides a theory- driven zonal differentiation with a long term perspective on the evolution of the global political economy. Its spatial elements were an integral part of the earliest treatments (Wallerstein 1974), and have been a prominent part of the literature for some time (see especially Chase-Dunn 1981, Arrighi and Drangel 1986, Terlouw 1985; 1992, and Taylor & Flint 2000).

From the world-systems perspective, the capitalist world-economy expands from its emergence in the long 16th century through the incorporation of nearly the entire planet by about 1900 (the shrinking external arena has no place in the following analysis). The world-system includes a multicentric system of states that evolves in tandem with the capitalist world-economy. Areas that initially combine superior strategies of accumulation with the elements of politico-military success take their place as the core of the system. All areas (including states, or sometimes parts of states) are stratified into core, semiperipheral and peripheral zones. The world-system is subject to both secular trends and cyclical processes. Among the secular trends we find geographic expansion, commodification and mechanization. Among the cyclical processes we find periods of economic rise and decline (often identified as Kondratieff waves), periods of hegemony and rivalry, and periods of varying levels of social upheaval (usually among the holders of capital and those who do not own the means of production). In this section we concentrate on the patterns and possible significance of the successive zonal configurations that world-systems analysis indicates. We find evidence that multilateral treaty-making is a secular, not a cyclical activity, by looking first at the location of multilateral treaty-signing behavior, and then at its vulnerability to disruption by waves of economic rise and decline.

Table 2: Signature places across zones for different phases of the world-economy


























The demarcations for core, semiperipheral and peripheral zones are derived from three maps in Terlouw (1985) for circa 1700, 1900 and 1980. Terlouw culled these from the world-systems literature. Core states are those with the greatest levels of social, economic and political power. One might expect multilateral treaty-signing behavior to be focused in such areas. The data in table 2 suggest that this is only the case after 1850. Prior to the mid-19th century, most treaties were signed in semi-peripheral venues. The semiperiphery is the zone where core and peripheral processes intermingle. Semi-peripheral areas may emerge from different processes. Some semi-peripheral regions result from the decline of core areas in the wake of shifts in resource or trade patterns, or political upheaval. Older, former core regions might provide a relatively safe and neutral haven for negotiation in larger semiperipheral zones with many conflicts to resolve. Semi-peripheral areas may also emerge from peripheral areas that rise on the foundation of new economic activities (often the discovery of new resources) and/or in the wake of politico-military innovations. The rulers of these new areas might angle for a fuller role in the processes of global accumulation. Perhaps it is no surprise then that early multilateral treaty-making venues were frequently situated in the semi-periphery. In the later 17th and 18th centuries the ‘congress system’ of diplomacy prevailed in Europe. Sovereigns, with their relevant advisors and experts in tow, would meet for upwards of one month in a neutral town to work through their most important problems.

With the growth of revolutionary movements in Europe, semi-peripheral areas may have lost their allure in this regard. Enhancements in communication, transportation, democratization and bureaucratization made the ‘congress system’ less necessary. It enjoyed a brief renaissance in the second half of the 19th century, but disappeared after about 1920. The semi-periphery probably became less prominent as a multilateral venue for still another reason. This zone has often been considered the most volatile part of the world-system (Chase-Dunn 1981). It is here that the sharpest polarization is to be found. This gives rise to the most serious conflicts, upsetting the calm neutrality that would be identified as necessary for the hosting of large international meetings. As the expansion of transportation opportunities provided more freedom of choice for the selection of venues, the semiperiphery became less attractive even if the issues to be treated derived from there.

Among the most important of the cyclical processes, waves of economic rise and decline (Kondratieff waves) appear to drive, or at a minimum fundamentally impact, many world-system processes (Hopkins et al. 1979). It is reasonable to expect that multilateral treaty-making might be one of those processes. In periods of economic rise, there may be more interaction among states and regions, and greater levels of wealth to be acquired if states can organize their relations in a peaceful manner. In periods of economic decline, there may be less interaction and declining opportunities might engender greater levels of military, as opposed to diplomatic competition. There is, however, no evidence (the details are not shown here) of shifts in the distribution of venues across zones related to the successive stages of the Kondratieff waves, nor does the intensity of diplomacy as expressed by the frequency of treaty conclusions differ in any systematic way.

Multilateral diplomacy appears unaffected by Kondratieff waves. During its long-term upward trend it evidences more of a secular increase than a cyclical pattern. The upward trend in the number of treaties is accompanied by a shift in the geographical distribution of signature places to the core.


Certain venues are frequently selected for treaty-making. We offer three possible explanations for this. First, we deal with possible consequences of hegemony. Does the (temporary) capacity to unilaterally establish the relevant rules of the state system also imply a frequent role as a venue for the negotiation of multilateral agreements? We also consider the effect of favourable functional conditions, in particular the local availability of the necessary prerequisites for shaping a successful negotiating environment for multilateral agreements in general, or for those in a given issue area. Finally we look at the political arguments involved in selection: the kinds of places that politicians and diplomats generally tend to consider, questions of status and power in the acquisition of these roles, qualitative considerations in the context of specific negotiations.


In Wallerstein’s view three core countries have temporarily played an extraordinary role in the development of the world-system. These states acquired the full benefits of a concentration of innovative technologies on their territory (rise to hegemony), translated this into commercial advantages (as a victorious hegemon) and moved into a central position in the networks of international finance (as a mature hegemon). These phases are inevitably followed by a decline (Wallerstein 1983). Three temporary hegemonies of the United Provinces, the United Kingdom and the United States are most often mentioned in the world-systems literature and have been comparatively portrayed by Taylor (1996). The sequencing of the last two is supposedly determined by the Kondratieff waves, with each hegemon rising in an expansionary phase of the world economy and extending its power across two cycles each.

In terms of diplomacy, we might expect that the unilateral ability to make the rules would translate into a miniature golden age of coordinated interaction and treaty-making. The paucity of formal, treaty-based multilateral activity during the Dutch period makes it difficult to assess the diplomacy of that era. While many of the sparse efforts at multilateral treaty-making during the period generally associated with Dutch dominance were concentrated in the United Provinces, data are rare. We therefore move directly to a comparison of the UK and US as outlined in Table 3 below. An analysis of these actors suggests a complex role for both.

Table 3: The phases of the UK and US hegemonic cycles

Hegemonic phases















Source: Hopkins, Wallerstein, and Associates (1979)

We use this periodization to explore the relationship between hegemony and centrality in multilateral treaty-making. First, we look at the absolute numbers and proportions of multilateral agreements in which either hegemon took part during these two successive cycles. Then, we consider the positions of the two capitals as venues for multilateral agreements over the course of those cycles.

Figure 3a-c: UK and US participation in multilateral treaties in hegemonic cycles

Figure 3

Figure 3

Figure 3

The absolute number of multilateral treaty-signings by hegemonic powers increases from the rise of hegemony to the stage of victory and on to the stage of maturity. Treaty-signings decrease as decline sets in (Figure 3a). The growth is much steeper for the US than for the UK in its time. This is partly due to the steeper general growth during part of the US hegemonial period. Another reason is the different position of both hegemons at the start of their respective periods. While the UK had long been a well established member of the society of states and had acted as a major power, the US had been a virtual bystander immediately prior to its rise. It is important to stress the decrease in absolute and relative terms for both hegemons from their periods of maturity to decline. Such decreases cannot be an artefact of the overall (upward) trend of multilateral treaty signings. This suggests a specific feature of the life course of hegemonial powers.

In relative terms (Figures 3b and 3c), controlling for the effects of the general trend, we find a proportionally larger involvement of the UK in multilateral treaty signings. Larger (overall) volumes of treaties in the US dominated period (Figure 3a) often do not include the hegemonic power. In terms of relative proportions, multilateral activity for the UK declines from the initial stage of hegemony to victory. This is consistent with the different pre-hegemonic positions of the two powers. In addition, it undermines the idea that hegemonial powers ipso facto should be increasingly engaged in multilateral action during all the most active stages of their hegemonial tenure. In the periods from victory to maturity hegemonic involvement increases and then falls off, suggesting a specific hegemonial trajectory. It looks as if a hegemonic position asserts itself at the stage of maturity as regards involvement in multilateral treaty-making and declines as hegemony declines, contrary to the general trend in the system as a whole.

It is interesting to note that the hegemonic power does not tend to turn to multilateralism as hegemony dissipates. Keohane (1984) suggests that multilateral activity should increase in an attempt to find a means to support rules the hegemon was able to establish and support unilaterally. We find that as the hegemon’s position declines so does its aggregate participation in both the absolute and relative level of multilateral treaty-making. (There is some evidence that a burst of activity emerges late in the decline phase, b ut that is the subject of ongoing analysis.) In general, the data suggest that hegemonic cycles play only a modest role in accounting for the numbers of treaties agreed upon across time. Once again, multilateral treaty-making appears to be a more secular than a cyclical phenomena.

As regards national capitals as venues for multilateral treaty-making, the significance of a hegemonic position is not clearcut. London remains a prominent place for treaty signings through both of these hegemonic cycles, not only the UK cycle. Washington becomes a prominent place as the US hegemonic cycle begins, and it retains its place during all its phases. Washington’s rank among the prominent places does not improve during the US cycle. If anything, there may have been some decline over time. There is no strong general relation that connects the position of the hegemon’s capital to the processes of multilateral treaty-signing.

Functional Arguments

While hegemony plays a relatively minor role in the locus of treaty-making activity, more functional considerations possibly guided the selection of venues for the making of multilateral agreements. These gatherings involved relatively large groups of people that had to be accommodated for sustained periods of time. They require amusement and leisure facilities, the security and the support of specialized personnel for creating and circulating documents, as well as expert advice on substantive points in the negotiations. Such items are part of what Jean Gottmann (1990) has baptised ‘the hosting environment’ that is typical for capital cities. As the home of political centers these cities have through the ages also been home to these very relevant support functions. Indeed, on average in all periods that we have already distinguished and in all the different zones with only minor variations, 76% of all these multilateral conferences have been held in capital cities.

The availability of a ‘hosting environment’ has not been the only cause of this pointed preference (see further below). ‘Hosting environements’ also existed elsewhere e.g. in resorts. During the ancien régime Aachen was selected in the 1740s in part for its construction of leisure facilities around the waters (Duchhardt 1999). More recently, Cancun (Mexico) has been chosen as the venue for multilateral conferences for similar reasons. Finally some places are suitable as they are close to the subject matter to be discussed: Mainz as a regular venue to discuss matters concerning the Rhine in the early 19th century, Algeciras in southern Spain, hosted a 1906 meeting to discuss the Moroccan question in a location with a view of the Moroccan coastline.

As the range of policy sectors involved in multilateral treaty-making widened and the number of agreements rose, there was increased need for more specialized expertise and one could imagine the emergence of more specialized places for treaty-making in specific sectors. This was undergirded by the first generation of inter-governmental organizations, the Public International Unions that became operational in the later part of the 19th century. They established permanent secretariats at certain places where on a more or less regular basis member states were invited to discuss new regulations (Murphy 1994, 111-113). This is well illustrated with the creation of the League of Nations, and the emergence of Geneva as a primary city in every period after 1920. Table 4 captures the amount of specialization amongst frequently selected places (a category earlier delineated in Figure 2) for multilateral treaty-making from 1815.

Table 4: Specialization of prominent venues in six treaty types in five periods

War & Peace

State Relations

Comm. & Transport

Social Affairs

Trade & Economy
























Specializations have been calculated as location quotients. They are ratios of numbers of treaties in an issue area signed in one city relative to all treaties signed there compared to numbers of all treaties in that issue area signed everywhere relative to the numbers of all treaties signed everywhere. In this way we show the specific attraction of a particular city for treaty-makers in a particular issue area ( LQ= e(i):e/E(i):E). If the attraction in that issue area equals the general attraction of that city, the score is 1.

Table 4 shows places with location quotients of at least 2 for the different issue areas for five periods. Most of these periods recapture our earlier classification, but the initial period is lacking and most of the 19th century has been taken as one single period. This is because during the ancien régime there is hardly any variation in the issue areas that treaties cover: they are nearly all on war and peace (see table 1). Consequently, no inter-city specialization is possible, or in other words, all signature places ‘specialize’ in war and peace treaties. Particularly in the earlier part of the 19th century numbers of treaties are rather small. The calculation of location quotients does not provide robust results under those circumstances. Therefore, some time periods have been aggregated.

All in all 11 out of 19 places mentioned as frequently used venues in Figure 2 for the period from 1815 show up in Table 4 with mostly one, sometimes two specializations in any one period. There are four examples of cities that sustain a specialization over two periods (The Hague, London twice in different treaty types and Brussels). All this does not point to specialization as a very important condition or part of a trajectory towards front rank position among multilateral treaty production sites. This is further underlined by other specialized cases in our collection ( 21 in all with location quotients of at least 2 in at least one treaty type) that are not among the most frequently selected ones that we consider so far and consequently are not shown in table 4 (e.g. New York). Their specialization was neither derived from, nor did it so far provoke a very large number of treaty-producing gatherings. On the other hand, some of the top ranking venues in terms of treaty signings host conferences concerned with a variety of issues. The best example is Washington DC (1897 to 1989).

Specialization is pretty common among frequently selected venues, but it is by no means a necessary or sufficient condition for continued choice as a preferred location. In some cases specializing may facilitate growing into a frequently selected venue. Only rarely do specializations endure over sustained periods. In just as many or more cases a city will cease to serve as a specialized venue as it attracts a wider variety of conferences, or because it is less often selected. Specialization does not assure more frequent selection.

Political Arguments

We have already mentioned the strong preference for national political centers. National political centers were the places primarily considered because they were most familiar. National political centers were also ranked as the centers of power that had to be defended or underlined by hosting a well attended conference. Henrikson (2005) discusses this as the ‘my place’ argument, but this is insufficient in a competitive multilateral context. Part of that preference may be due to the availability of a suitable hosting environment, a practical functional argument that we already mentioned.

Henrikson (2005) has suggested a number of arguments that have been used in the selection of individual venues for international meetings in recent decades. Many of them are also relevant in the selection of places for multilateral treatymaking: neutral places, halfway places (optimal overall accessibility), metropolitan places, everyone’s places (felt to be equally inviting and receptive to all), safe places, exotic places, demonstration places (where the issues of the conference are unavoidably present). Which arguments end up victorious in individual cases can best be considered through a detailed knowledge of the relevant actors and decision processes. The data presented in this work provide a good start for such a study in four ways.

First, they provide long-term information on the most frequently selected places. Second, the data suggest that the traditional locations of bilateral diplomacy, like Paris, London and Vienna, played a vital role as meeting places for multilateral agreements as well. Paris in particular, but also London and Vienna (as the seat of the predominant Habsburg overlord of the Holy Roman Empire) played pivotal roles in this regard twice: after having been the major actors in the European diplomatic web (Anderson 1993, in particular 7-11, 27-29, 60), then again as they played vital roles as evidenced by the 400 years of data reviewed here. They were in Henrikson’s terms, the symbol of ‘our places’ where the many diplomatic posts linking the members of the society of states most ‘naturally’ came together. The Hague was to a large extent also a member of this group. Lacking a prince, it could initially not follow all the diplomatic mores, but its hegemonic status in the mid-17th century allowed it a privileged position among the powers of the day.

Third, Europe’s urban spine, the densely urbanized corridor from Southeast England through the Low Countries, the Rhine Valley and Switzerland into Northern Italy, has been home to a number of frequently selected multilateral venues in the course of time. Starting from Aachen during the ancien régime and later with Mainz, Berne, Brussels, Geneva and Strasbourg, this region has accommodated a remarkable fraction of the multilateral conferences (obviously The Hague and London could also be put in this category). From the world-system perspective, this was among the earliest and most important European corridors for long-distance commerce and economic innovation. As Henrikson suggests, places in this area could well have gained a reputation of ‘halfway place’ or ‘everyone’s place’. This area is also known for late state formation due to its dense pattern of urban life that was supposedly resistant to rigid territorial rule. Most of the cities mentioned as frequently selected places in this corridor were incorporated into states, as traditionally defined, once state formation was well along and began to offer important advantages.

Fourth, newly established or aspiring major powers (in world-system terms, semi-peripheral marcher states), may attract or organize multilateral treaty negotiations as a way of signalling their emergence and getting acquainted. (Such a category could be an extension of Henrikson’s list.) These include Berlin in the late 19th century, Washington DC in the early 20th century, and perhaps Rome in the interwar years. In the cases of Berlin and Rome their new status proved temporary, but the signals they sent as host states were clearly understood. Washington DC remained a major venue over the entire 20th century. From a new place, it became a hybrid, ‘our place’ for Americans, ‘everyone’s place’ for some others, ‘exotic place’ for the diplomats of old Europe and new entrants to the state system in more recent times. The Latin American capitals that for a short while acted as frequently selected venues mainly in the late 19th century, were part of a new extension of the international state system. But their governments were also sorting out a rivalry amongst national political centers that has never been decisively solved.

Washington DC’s qualities as a multifaceted global forum and home to a continuous diplomatic machinery ready for use are even more strongly represented in Geneva and New York. These are also new political centers based on a different dynamic than the one that launched Berlin, Rome and Washington. Geneva has even become the most frequently used center for multilateral encounters with truly global representation. It has replaced Paris (and Brussels), which played a similar role in Europe in the second half of the 19th century. Between 1850 and 1914 Murphy (1994) counts 105 important multilateral conferences, about 1.6 per year. Just over 20% (23) were hosted in Paris and just over 15% (16) in Brussels. In the 12-year period from 1974 to 1985 there were 98 multilateral conferences, an average of over 8 per year (Murphy 1994, 255-257). Geneva hosted over 40% (42), and New York hosted over 20% (23). In pure numbers (if not in political importance) Geneva has become the most frequently selected multilateral venue at the global level.

How did a place that was not a national political center achieve this extraordinary position? Geneva was a longstanding resort with a proud urban tradition and an important religious heritage. Before 1920 Geneva had attracted only very few multilateral negotiations. Since 1863/64 the city had become a center of international charitable work with the founding of the Red Cross. This may have been the trigger for accommodating the arbitration tribunal to settle the ‘Alabama Claims’ between the US and the UK in 1871. It was the first tribunal of its kind meant to settle a dispute about damages incurred during the American Civil War. The arbitration was successful.

From very early on, Wilson had preferred the League‘s seat to be in a neutral country to alleviate the burdens of the divisions among the former warring parties on the postwar work of the League as much as possible. Probably he also quickly developed a preference for Switzerland and for Geneva but this is less certain. Reasons may have been his personal Presbyterian background and the tradition of international activity embodied in the Red Cross and the Alabama Claims Tribunal with which he professionally and ideologically sympathised (Rappard 1956, Ruffieux 1983).

When the League negotiations in the Paris Peace Conference of 1919 were underway, its permanent seat became an issue. The Americans, Wilson in particular, pressed for Geneva in preference of Brussels. They finally won the day. A complication in the negotiations was the official absence of neutral countries from the Paris Peace Conference. Another was the reluctance of neutral countries to become involved in great power politics as a result of eventual views and actions of the League. This was particularly important for Switzerland with its outspoken ‘passive neutrality’ policy.

From the other side there certainly was a lobby by important members of the Swiss Government, some of them clearly linked to Geneva. An important go between was William Rappard, a Swiss with a Genevan background raised in the US and later a League functionary and a professor of International Relations in Geneva. He was a constant participant in the Paris Peace Conference and had very close contacts with the major members of the American delegation and with Wilson personally (Seymour 1928, Rappard 1956, Peter 1983). It remains surprising that the Swiss did not opt for Berne which had grown into one of the main international multilateral diplomatic centers of the preceding period. It may well be that the position of Geneva as a French-speaking city was considered an asset under the circumstances. It is intriguing that Geneva’s candidacy was ultimately realized after a final diplomatic battle when the hope of US membership of the League had already vanished thanks to Wilson’s personal support despite many misgivings about the mediocre quality of its ‘hosting environment’ compared to Brussels (Rappard 1956).

Once Geneva had been selected as a ‘neutral place’, and perhaps also as ‘everyone’s place’ and a ‘safe place’, it became the epitome of an international diplomatic city. Hosting large numbers of multilateral conferences, it became home to a growing number of international organizations, a new kind of ‘our place’. Its rise was highly contingent and played out in the midst of a complicated international negotiation, but it proved to be consequential. If multilateral treaty-making is a secular trend, it should be no surprise that initial choices have long-lasting consequences.


Multilateral treaties have been negotiated for millennia, and certainly during the entire life course of the contemporary state system. They involve something more than additional partners, and are suggested to help create and prescribe global rules of behaviour. Over time the multilateral track has widened and ever further diversified the geographical range of interstate links, enlarged their scope and increased in numbers, particularly from the early 19th century to the 1960s. In this paper we considered the successive distribution of places where multilateral treaty-making has settled by looking at the venues where successful negotiations have been held over the past 400 years.

The growth of multilateral treaties has the attributes of a secular trend, and not a cyclical pulse. The phenomenon seems generally immune from economic (k-wave) or political (hegemonic) cycles. Such agreements have been organized in a large number of places, but only a few have been used for the purpose with significant frequency. These frequently used venues demonstrated some extraordinary continuities across the centuries, but also some significant shifts: some gradually declined and disappeared, others became prominent and remained so, a few made a flashlike appearance but did not survive.

Meeting places were largely concentrated in Europe. Initially, multilateral venues were concentrated in the semiperiphery, at that time still largely situated in Europe. Later, as the world-system expanded across ever larger parts of the world, meeting places concentrated in core areas. Variations in volatility between zones, plus improving transport and communication systems across the globe, account for some geographic shifts.

Hegemonic position in the world-system seemed to be only weakly related to participation and leadership in the hosting of multilateral conferences. A large proportion of the frequently selected meeting places consisted of national political centers. This can be explained by functional considerations. Such centers possessed the vital hosting environment to accommodate the conferences. Other functional considerations were the presence of resorts and existing secretariats of IGOs from which specialized expertise could be mobilized. But the preponderant choice for national political centers could particularly be explained by political considerations. They were more easily present on the mental maps of politicians and diplomats making the selection, and they are part of the rivalry for rank and status played out among states. The collection of frequently selected places suggested some role for three political considerations in particular: 1) being a member of the select number of capitals from where earlier (bilateral) diplomatic interaction was situated such that a city was traditionally seen as common ground for all, ‘our place’; 2) being somewhere on Europe’s urban spine as the main road to and from everywhere, well known and open to all, ‘everybody’s place’; and 3) being the political center of an upcoming or aspiring major power in the state system, ‘new show-place’ or a new global center: a renewed version of ‘our place’. Selection of places in individual cases, we derived from the example of Geneva, could be highly contingent, but has lasting implications.

There are three areas of future research that are strongly suggested by this work. The first concerns the apparent reduction in multilateral agreements of the last few decades reported in Figure 1. At the moment we cannot completely exclude that this is an artifact of the dataset, which ends in 1995. It can take well over a decade for a treaty to make its way into the compendia that were used to construct this dataset. But what if the decline is real?

As suggested earlier we might be witnessing an alteration in the nature of multilateral practice (see also Denemark and Hoffmann 2008, fig 7). Barnett and Finnemore’s (2004) work on the autonomy of international organizations suggests that many of the issues that had once been resolved through treaties that were sponsored by international organizations might now be undertaken by the international organizations themselves. If the international system is developing two mechanisms of multilateralism (through treaties and through autonomous international organizations) will this be accompanied by additional forms of specialization of the centers where these activities occur: frequent venues for multilateral treaty negotiations on the one hand, and concentrations of headquarters of international organizations on the other? Following other tracks to solve the initial puzzle, we should also test the strength of more familiar arguments about hegemonic unilateralism, or suggestions generated in the globalization literature regarding the decline of states (relative to private regulatory regimes) to account for a downturn in the overall number of multilateral treaties that are signed.

Secondly, the current shift of economic and social dynamism among regions raises questions about the nature of multilateral treaty-making and the places where it occurs. We earlier addressed such questions for past periods by way of world-systems analysis. We may now have to refocus our theoretical lenses and collect the necessary additional data for the period after 1995 in order to answer questions such as: Have the historic shifts that are now apparently underway altered the pattern of multilateral interaction e.g in terms of the venues used for their negotiation? In particular, do we finally see a decline of ‘Atlantic’ pre-eminence in this field? Are different patterns of multilateral treaty-signing evident across regions, or in the form of differential inter-regional adherence to global treaties? Such questions might be addressed with the addition of regional markers to the multilateral treaty data, and with the use of the techniques of network analysis.

Still another line of future questioning concerns the literature on the hypothesized difference between the behavior of large and small states in the global system. Small states are generally suggested to be more open to multilateral agreements. Their small size suggests a greater openness (and vulnerability) to global interaction of all sorts, while their lack of ability to impose their will on the shape of global systems suggests less of a reason to withhold participation in global agreements. Small states would be rational to attempt to minimize the costs of conducting foreign policy by engaging in more joint actions. Such arguments have been made with regard to general foreign policy behavior (e.g. East 1973; Cooper and Shaw 2009), as well as in the specific areas of economic interaction (e.g. Katzenstein 2003) and security concerns (Inbar and Sheffer 1997). Two questions arise. First, is the tendency of small states to engage in more multilateral agreements real, and if it differs over time or across issue areas, what circumstances are implicated? Second, does the general inclination of major powers to play a commanding role in the global system, be it through using multilateral channels or otherwise, result in the selection of multilateral conference venues in either small or large states?


ANDERSON, M.S. (1993) The Rise of Modern Diplomacy 1450-1919. London: Longman.

ARRIGHI, G and J. DRANGEL (1986) The stratification of the world-economy: an exploration of the semiperipheral zone. Review (Fernand Braudel Center) 10: 9-74.

BARNETT, M. and M. FINNEMORE (2004) Rules for the World: International Organizations and Global Politics. Ithaca NY: Cornell University Press.

BERRIDGE, G.R. (2002) Diplomacy. Theory and Practice. Basingstoke: Palgrave Macmillan

BOWMAN, M. and D. HARRIS (1984), Multilateral Treaties: Index and Current Status. London: Butterworths.

CHASE-DUNN , Ch. (1981), Interstate system and capitalist world-economy: one logic or two? International Studies Quarterly 25 (1): 19-42.

COOPER, A. and T. SHAW (eds.) (2009) The Diplomacies of Small States: Between Vulnerability and Resilience. Basingstoke: Palgrave Macmillan

DENEMARK, R. and HOFFMANN, M. (2008) “Just Scraps of Paper”? The Dynamics of Multilateral Treaty-Making. Conflict & Cooperation 43(2): 185-219.

DUCHHARDT, H (ed.) (1999) Städte und Friedenskongresse. Cologne: Böhlau Verlag.

EAST, M. (1973) Size and Foreign Policy Behavior: A Test of Two Models, World Politics 25:4, 556-76.

GOTTMANN, J. (1990) Capital cities, in: J. Gottmann & R.A. Harper (eds.) Since Megalopolis. The urban writings of Jean Gottmann. Baltimore:Johns Hopkins University Press: 63-82

HENRIKSON, A. K. (2005), The geography of diplomacy, in: C. Flint (ed.) The geography of war and peace. From death camps to diplomats. Oxford: Oxford University Press: 369-394.

HOPKINS, T., I. WALLERSTEIN and Associates (1979), Cyclical Rhythms and Secular Trends of the Capitalist World-Economy: Some Premises, Hypotheses, and Questions. Review 2 (4): 483-500.

IKENBERRY, G.J. (2003) Is American multilateralism in decline? Perspectives on politics 1: 533-550

INBAR, E. and G. SHEFFER (eds.) (1997) The National Security of Small States in a Changing World. London: Routledge.

KATZENSTEIN, P. (2003) Small States and Small States Revisited, New Political Economy 8:1, 9-30.

KEOHANE, R.O. (1984) After hegemony: cooperation and discord in the world political economy. Princeton University Press, Princeton NJ.

MIDDELAAR, L. van (2009), De passage naar Europa. Geschiedenis van een begin. Groningen: Historische Uitgeverij, in particular chapter 2: 31-60

MOSTECKY, V. (ed.) (1965), Index to Multilateral Treaties. Cambridge MA: Harvard Law School Library Publications.

MULDOON, J.P., J.A.F. AVIEL, R. REITANO, E. SULLIVAN (eds.) (2010 to be published) The New Dynamics of Multilateralism. Diplomacy, International Organizations and Global Governance. Boulder:Westview.

MURPHY, C. (1994), International Organization and Industrial Change: Global Governance since 1850. New York: Oxford University Press.

PETER, A. (1983) William E. Rappard and the League of Nations: A Swiss contribution to international organization, in: UN Library & Graduate Institute of International Studies , The League of Nations in retrospect. Proceedings of the Symposium 6-9 November 1980. Berlin/New York: Walter de Gruyter: 221-241.

RAPPARD, W.E. (1956) Woodrow Wilson, la Suisse et Genève. In: J. Freymond et al., Centenaire Woodrow Wilson 1856-1956. Geneva:Centre Européen de la Dotation Carnegie: 29-74

RUFFIEUX, R. (1983) La Suisse et la Société des Nations. In: UN Library & Graduate Institute of International Studies The League of Nations in retrospect. Proceedings of the Symposium 6-9 November 1980. Berlin/New York: Walter de Gruyter: 182-195

RUGGIE, J.G. (1992), Multilateralism: the anatomy of an institution. International Organization 46 (3): 561-598

SEYMOUR, Ch. (1928) The intimate papers of Colonel House. London: Ernest Benn .Vol. IV: 429-430, 534

TAYLOR, P.J. (1996), The way the modern world works. World hegemony to world impasse. Chichester: Wiley.

TAYLOR, P.J. and C. FLINT (2000, 4th ed.), Political geography. World-economy, nation-state and locality. Harlow: Prentice Hall.

TERLOUW, C.P. (1985) Het wereldsysteem. Een interpretatie van het werk van J.M. Wallerstein. Rotterdam: Comparative Asian Studies Programme Serie Werkdocumenten 8

TERLOUW, C.P. (1992), The regional geography of the world-system: external arena, periphery, semi-periphery, core. Utrecht: KNAG (Netherlands Geographical Studies 144).

VAN OUDENAREN, J. (2003) What is ‘multilateral’? Policy Review 117 (Feb/March): 33-47

WALLERSTEIN, I. (1974), The modern world system. Capitalist agriculture and the origins of the European world-economy in the sixteenth century. New York: Academic Press.

WALLERSTEIN, I. (1983), The Three Instances of Hegemony in the History of the Capitalist World-Economy. International Journal of Comparative Sociology 24: 1-2.

WIKTOR, Chr. (1998), Multilateral Treaty Calendar. The Hague: Martinus Nijhoff.

WUSTEN, H. VAN DER (2007), Political world cities. Where flows through entwined multi-state and transnational networks meet places, in: P.J. Taylor, B. Derudder, P. Saey & F. Witlox (eds.) Cities in globalization: practices, policies and theories. Routledge, New York & London: 204-220.



* Herman van der Wusten, University of Amsterdam, Department of Geography, Planning & International Development Studies,

** Robert A. Denemark, University of Delaware, Department of Political Science

*** Matthew Hoffmann, University of Toronto, Department of Political Science

**** Hasan Yonten, Neumann University, Division of Humanities and Social Sciences


Edited and posted on the web on 4th February 2011

Note: This Research Bulletin has been published in Tijdschrift voor Sociale en Economische Geografie, 102 (5), (2011), 499-514