Abstract

The governance of the political and economic world order builds on a complex architecture of international treaties at various geographical scales. In a historical phase of high institutional turbulence, assessing the stability of such architecture with respect to the unilateral defection of single countries and the breakdown of single treaties is important. We carry out this analysis on the whole global architecture and find that the countries with the highest disruption potential are mostly medium-small and micro countries. Political stability is highly dependent on many former colonial overseas territories that are today part of the global network of fiscal havens, as well as on emerging economies, mostly from South-East Asia. Economic stability depends on medium-sized European and African countries. Single global treaties have surprisingly less disruptive potential, with the major exception of the WTO. Our results suggest that the potential fragility of the world order seems to be more directly related to global inequality and fiscal injustice than commonly believed and that the legacy of the colonial world order is still strong in the current international relations scenario. In particular, vested interests related to tax avoidance seem to have a structural role in the political architecture of global governance.

1. Introduction

The global architecture of flows of people, goods, services, and information is regulated by a governance system that works through a complex web of institutions and international treaties, most of which have been developed in the postwar period to promote an enabling environment [1] to exploit opportunities of various nature, such as, for instance, gains from trade, optimal division of labor, free circulation of people and resources, reduction of uncertainty from local economic fluctuations, and multilateral conflict prevention and resolution arrangements. Taken together, such agreements form a complex, multilayered system, often associated with the notion of a “world order” [2].

The system has been mostly tailored to the challenges and needs of a bipolar, cold war world which called for an overarching institutional setting to prevent the escalation of conflict and mitigate frictional obstacles to trade, movement, cultural exchange, and so on [3]. The system further developed in the apparently unipolar moment that followed the collapse of the former Soviet Union, but has been increasingly under pressure ever since [4], for the concurrence of several critical factors. The first has been the rapid emergence of a new multipolar world order [5] in which several powers are competing for global influence [6], including at least the USA, the EU, China, India, and Russia [7], and paving the way to alternative conceptions of the world order itself [8], and to new variable geometries of influence [9, 10]. The second is the change of direction in US foreign policy which has gradually reduced the country’s pivotal role in the political and financial maintenance of the global system to increasingly prioritize the internal affairs agenda [11]. The third is major threats such as the climate change emergency and the global pandemic crisis [12, 13], and previously the great credit crash of 2007–9 which severely tested the resilience of the multilateral architecture of the global order [14]. In a multipolar order, there is not necessarily an interest in guaranteeing the multilateral functioning of an institutionalized global governance system [15], and some powers could rather benefit from disrupting it at least in part to gain more influence in strategic regions, including the US themselves in the light of the new political agenda.Moreover, the global governance principles that worked in a 20th-century setting need not be as effective in a 21st-century one [16], as proven by its basic inability to provide timely and effective responses to the new challenges [17], whose scale and complexity are unprecedented and call for radically new solutions and extreme levels of global institutional coordination and cooperation [18].

An eloquent proof of the new logic of global governance that is emerging in the multipolar order is being provided by the Russian invasion of Ukraine, which has led to an increasing level of sanctioning and suspension of Russia from major treaties and organizations, mostly from Western countries. In this situation, the role and efficacy of global institutions in preventing the invasion or in promoting diplomatic solutions have proven to be limited and have led other countries which previously maintained a long-standing neutral position with respect to defense agreements such as NATO to seek admission, as is the case of Finland and Sweden. Therefore, major and unexpected changes to the structure of international agreements and treaties have become a very concrete scenario.

In this new context, the possibility of a sudden collapse of some of the institutions and treaties that shape the governance system is no longer mere theoretical speculation [19], and likewise, it is possible that even major global powers contemplate their partial or total withdrawal from certain institutions or treaties [20], as it has happened for the US under the Trump administration, with the withdrawal from UNESCO and UNHRC as a form of political retaliation [21] and the threat to leave the WHO, which was not ratified by the incoming Biden administration. Even the historical alliance between the US and major European countries, which is at the root of the very definition of the West as the geopolitical cornerstone of the whole system [22], cannot be taken for granted anymore in the new scenario [23, 24]. Populist governments all over the world build their political agenda on extreme parochialism and opportunistic, case-by-case adherence to international agreements, often as revenge for perceived politically or economically marginal positions [25]. On the contrary, the pandemic crisis de facto temporarily suspended certain treaties, such as the free intra-EU mobility ensured by the Schengen Treaty during anti-Covid-19 lockdowns, as well as many other free circulation agreements, and such temporary suspensions could pave the way to larger and more stable institutional disruptions if normal functioning cannot be quickly restored.

It is important, therefore, to analyze the global governance system’s robustness against this kind of shock. Not all institutions and treaties have the same structural role in global architecture, and the same holds for the participation of a given nation in one or more institutions and treaties. Which are the truly critical ones? According to the current state of knowledge, there is no clear answer to this question, despite its undeniable importance. In this paper, we provide a first, systematic analysis of the global governance system against one-sided disruptions, both in terms of treaty suspensions and of countries’ unilateral withdrawals. This allows us to derive new risk indicators that may become very relevant in strategic and scenario analysis in the coming years.

In this increasingly uncertain environment, interpreting global changes in terms of linear processes of structural change is likely misleading. The tendency to interpret the evolution of the world order as a mere interplay between the goals and interests of the most influential countries and country blocks remains strong. However, a nonlinear science of networked international relations [26] is much more appropriate to deal with the complexity of international relations [27], especially so in response to major, unexpected shocks. We can conceptualize the world order as a multilevel network of alliances between countries, whose structural characteristics critically impinge upon its resilience. Inspired by biology, we propose a holistic approach to assess the resilience of the world order to the breakdown of geopolitical treaties due to unilateral moves by specific players. We call this approach system global policy, in analogy to systems biology [28], defined as the computational and mathematical modeling of complex geopolitical systems. Given the existing complex architecture of global relations, even minor changes in existing economic or political trade deals could, in principle, spark complex dynamic responses. One cannot rule out in principle that the collapse of relatively minor agreements or the withdrawal of second-tier countries from major agreements kicks off adjustment cascades whose consequences could be disruptive. We propose an innovative methodology based upon the structure of multilayer networks [29, 30] as a substrate for a nonlinear approach to the analysis of international relations, and we illustrate its potential by simulating the structural impact of simple shocks under the form of unilateral defections on the current world order architecture.

2. Results

2.1. Overview of the Data

We consider the full list of active (as of 2015) economic alliances, agreements and bilateral trades, and political alliances and agreements (see Table 1). The list of treaties has been extracted in March 2015. The source for the list of economic treaties has been the following two Wikipedia pages, as of March 15, 2015: list of multilateral free-trade agreements and list of bilateral free-trade agreements. WTO bilateral trade agreements have been extracted by the WTO database directly. For political treaties, the source has been the Wikipedia page (as of March 15, 2015): intergovernmental organizations established by treaty. The database with all the memberships for each treaty has been manually built from direct consultations of each treaty. The data set consists of 200 countries, with 4,733 deals in the political layer and 14,890 in the economic layer. Only 313 political interactions are not reflected in the economic layer. The resulting multilayer network is abstracted as two layers, political and economic, whose nodes represent countries and links represent deals in their respective layers (Figure 1).

2.2. The Damage Index

We test the world order’s resilience by evaluating the global structural consequences of a unilateral defection of one or more countries from their economic or political deals. For this purpose, we first compute the community structure of the multilayer network [31]. Communities are detected as groups of countries whose deals are denser between them than with the rest of the countries outside their group. Several publicly available algorithms allow us to carry out such computations. We make use of [32] while being aware of the available alternatives. Although it is possible in principle that the choice of the algorithm may influence community detection results to some extent, this is not the main concern here. The purpose of our paper is not that of offering a fine-grained classification but rather to explore the general picture of the structural stability of the world order, so that marginal differences in the results would not alter our analysis and conclusions substantially. Next, we simulate the disruption by eliminating a country and all its deals from one of the layers, and recompute the resulting mesoscale organization, including community structure and seeking for possible components disconnected from the system’s core. From this information, we can define a damage index measuring the level of disintegration of the original communities, the emergence of new connected components, and the loss of nodes in the largest connected component.

The mathematical definition of the damage index is easy and intuitive. Let be the original number of communities before any disruption, and indicate the number of communities found after removing a country (or block) from the layer . We define the ratio Similarly, we define the indicators and as the number of connected components in the system, and the size of the largest connected component, respectively. The three indices, separately, provide complementary information about how a country (or block) removal alters the whole structure, the size of its core, and the number of disconnected clusters (note that there can be just one connected component but several communities). The damage index is defined byand its normalized version as . It is worth noting that the damage index is high when, after disruption, the number of communities increases (more segregated network) or/and the number of connected components increases (segregation and isolation), or when the largest component of the system decreases (disaggregation). Note that these variables are not independent but truly correlated.

Thus, the damage index is a graph structural descriptor of network disruption and can be used to assess the resilience of the structural connectivity of the whole system. We have conducted several experiments on the world-order network structure. We have computed the damage index in the economic layer (EDI) and the damage index in the political layer (PDI) for all possible individual disruptions, i.e., by removing a single country and its deals from one specific layer. We can therefore assess their impact on the whole world order structure. The damage index lies between 0 and 1, reaching the unit value for the most disruptive countries. The results of our analysis are depicted in Figure 2. We have made use of distorted maps [33] to show the effect of every country’s defection from economic or political deals upon the rest of the world. The distortion is computed for the damage index, and the color indicates its value for every country.

It is important to stress that the purpose of the damage index is not that of making predictions on the future dynamics of the world order or to understand what the coalition (e.g., Shapley) value of a given country is in the current architecture of the world order and its implications in terms of bargaining power, etc. These are very important issues that require specific approaches and deserve to be pursued further. Here, we are essentially interested in highlighting how the structural organization of the world order is vulnerable with respect to a specific kind of threat -–unilateral defections of countries or the elimination of a single treaty from the global architecture. A better understanding of the effects of these perturbations may give us insights into the implicit forces that have shaped such architecture the way it is, not as a result of a top-down design but under the concurrent action of many different factors, opportunities, and events. On the contrary, considering this specific kind of threat has not only a purely theoretical value in a scenario in which the existing institutional setting is increasingly under pressure so that adherence of countries cannot be taken for granted any longer. If the structural stability of the world order architecture becomes problematic, understanding what could be the effect of unilateral deviations becomes important, not to make predictions (clearly, a unilateral defection by an important country would spark a train of chain reactions that could be very difficult to anticipate), but to find out where future policy redesign efforts should be directed to improve its resilience.

2.3. Independence of the Damage Index

To disentangle the complementarity of our structural descriptor, the damage index, compared to other indices, we have selected the fragile states index (FSI), which is a measure of a state’s internal sources of institutional, political, and socio-economic criticalities [34]. A statistical analysis based on Spearman correlation reveals that there is no direct correlation between FSI and the two defined economic (DIE; , value) and political damage (DIP, , value) indices, respectively. This result has important implications: it shows that intrinsically dysfunctional states need not be the most disruptive from a global perspective, as their turbulence may be inherently local, whereas we consider the global impacts of a perturbation of the world order. It also implies that the world order is not critically susceptible to the factors that typically cause states to fail, e.g., localized corruption and low levels of human and socio-economic development.

2.4. Geopolitical Assessment

The second important result is that the USA, as well as other major Western countries, are not disruptive, and the reason is that since Western countries tend to be tightly knit in their participation in international political or economic deals, the unilateral defection of a single country does not cause major structural damage as there is a significant level of redundancy. But the most surprising result comes from comparing PDI and EDI. In the political layer, the most critical countries are Myanmar and Thailand, even if there is a second group of countries that includes former colonial powers such as France and the Netherlands, and somewhat surprisingly, a number of tiny countries that are in most cases overseas territories, plus a few Northern African countries with important ties to Europe such as Algeria, Tunisia, and Egypt, and a third group which includes, among others, China and more South-East Asian countries. Here, we are not referring to “groups” as derived from a specific cluster analysis, but simply as the set of countries that share a common value of the damage index. It clearly looks paradoxical that a small island state such as, say, Saint Kitts and Nevis may have a (slightly) larger disruptive political potential than China. Moreover, the US damage index is extremely small. Overall, however, the size of the biggest disruption effects in the political layer is much smaller than in the economic one, although there is a large number of countries that are potentially critical. In the economic layer, instead, there are much fewer countries that are potentially critical, but the size of the associated disruptive effects is bigger than the political ones. Moreover, the critical countries include a G7 member such as Germany, and also a number of smaller EU countries, African countries, and Belize, which was also in the second group of politically disruptive countries; the same is true for Egypt, which is the only other country which is critical for both layers, and whose importance might be linked to its crucial mediation role in one of the most long-seeded, globally critical conflicts, the Israeli-Palestinian one [35]. Large disruption is, therefore, more likely to come from the global economy than from global politics. The politically most critical regions are South-East Asia, North Africa, and Caribbean overseas territories, whereas the economically critical regions are Europe and Asia. Belize is a notable, common exception.

Such striking differences between the EDI and the PDI may possibly reflect “grammatical” differences between the two networks, for instance in terms of degree distributions. But it would be misleading to think that such structural properties in themselves explain the differences. The political and the economic treaties considered do not present, in principle, any obvious structural feature that should differentiate them in terms of how many countries are involved in each and how membership in the various treaties is intertwined. Therefore, even if such structural differences emerged to some extent, it would be difficult to interpret them as an intrinsic structural property. It is more meaningful to interpret such differences in terms of the “semantics” of the networks, i.e., of the different roles that political and economic agreements have in tying up the global architecture of the world order. In the discussion, we further elaborate on this point.

3. Discussion

3.1. Country Disruption Potential

Many African countries are found to have major economically disruptive potential: Egypt, Benin, Ghana, Morocco, Burundi, Mozambique, Gabon, and Burkina Faso. The only G7 country with high economic disruption potential is Germany, accompanied by medium (Czech Republic, Austria, Portugal, and Sweden) and small (Malta) European countries. Finally, there is Belize. This is a surprising result, as most such countries are small or very small economies. The massive presence of African countries in this list seems to suggest that the economic world order still largely reflects the colonialist world order, a result that agrees with previous analyses carried out with different methodologies on different kinds of data [3638]. On the contrary, there is also strong evidence that the German EU area of economic influence also plays a key role in the global architecture, as both Austria and the Czech Republic are economically disruptive. The presence of Portugal also hints at the ties with the colonial world order, given the important past of the country as a colonial power. Finally, we have small former colonies such as Malta and Belize.

Even more interesting, though, is the country's disruption picture on the political side. Here, together with the already mentioned Myanmar and Thailand, we again find two past colonial powers such as France and the Netherlands. However, the most interesting feature of the list is the massive presence of micro-States, that is, colonial overseas territories that are currently well-recognized fiscal havens: Antigua, Dominica, Granada, Saint Lucia, Saint Kitts and Nevis, Saint Vincent. As recognized in the literature, micro-States with less than one million inhabitants have a comparatively large probability of being tax havens [39], and it is hard to think of other common features that might make them so salient for the stability of the political layer of the world order. There are in addition three North-African countries (Egypt again, Algeria and Tunisia), Guyana and again Belize (two more overseas tax havens), plus a heterogeneous group of countries with lower but still substantial levels of disruptive capacity. Overall, the political picture draws a significant overlap of remnants of the colonial world order and global networks of tax avoidance. It is remarkable that the disruptive potential of fiscal havens is political, rather than economical. These results present interesting relationships with recent research that reconstructs the global architecture of tax evasion [40, 41], suggesting that the political architecture of the world order could be still largely shaped by the primary capital accumulation of colonial empires [42], and the opaque system of tax havens which largely coincides with small territorial leftovers of previous, global territorial possessions [43], whose function was that of offshore sheltering of colonial riches during the uncertain decolonization process [44], and that have further consolidated their specialization ever since, also covering money laundering from illegal activities [45].

3.2. Treaties Disruption Potential

On the contrary, the disruption picture from the point of view of international treaties is considerably simpler, to a surprising extent. The only treaty which has a major (and very substantial) disruptive capacity from the economic point of view is the free-trade agreement administered by the WTO, whose multilateral governance role is tellingly undergoing significant changes [46]. From a political point of view, the critical treaties are instead the Commonwealth of Nations, the Association of Caribbean States, NATO, and the Black Sea Economic Cooperation. The major role of the Commonwealth and of the Association of Caribbean States points attention again toward the colonial world order-tax evasion nexus. NATO has, for once, a rather intuitive role in the global architecture of political relations. The fact that the Black Sea Economic Cooperation has a major disruptive impact in political terms seems to stress the importance of the socio-economic stability of the Black Sea area which not incidentally has been at the center of major tensions between the E.U. and Russia in recent years.

3.3. Continuing Legacy of the Colonial World Order

The nonlinear structure of the global architecture of economic and political relations is surprisingly complex and yet at the same time informed by a familiar logic, and its critical conditions of robustness do not reflect conventional wisdom. Despite its preliminary character, our analysis highlights how the legacy of the colonial world order, as reflected in the current vast network of overseas tax havens, seems to play a much bigger role in the shaping of the current world order than it could be imagined, to the point of turning the principle of country sovereignty of such micro-states into a pillar of a global system of tax evasion [47]. The countries with the most disruptive potential are not, with few exceptions, major market democracies and institutional agents of democratic peace, but less developed countries (LDCs) whose role in the current world order is still largely determined by their colonial past and overseas territories whose main economic specialization is the custody of large shares of the world’s financial assets whose property is undisclosed [43]. On the contrary, the disruptive potential of such countries is not merely related to corruption or a lack of socio-economic development; as we have seen, neither political nor economic disruption is correlated to the FSI. It depends on the structural role they play in the architecture of the global governance system, which, as we have seen, is still largely shaped by postcolonial logic. Interestingly, such logic also surfaces from the analysis of the disruptive potential of international treaties, with the additional driver of East-West relations as a legacy of the Cold War and Russia-NATO antagonism. These unexpected structural features of the world order might explain, among other things, the persistence of difficulties in dismantling the global network of tax evasion despite the ambitious commitments of major world powers [48, 49] and the tax havens’ capacity to flexibly adapt to, and essentially neutralize, more restrictive measures to control tax evasion, such as automatic exchange of information between tax havens and high-tax countries [50]. Our results suggest that the structural weaknesses of the current world order could be due to causes that are different from the ones generally considered. Specifically, they seem to be much more fundamentally related than commonly thought to the opaque mechanisms that ensure the perpetuation of global inequalities and postcolonial socio-economic divides. This result will be of vital importance in rethinking the world order after the COVID-19 crisis, as the pandemic has further and dramatically widened preexisting income gaps both at national and global scales so that a massive redistribution could be necessary to guarantee some minimal form of social justice [51]. So far, the incumbent global order has mostly adapted to the changing circumstances through tactical adjustment rather than major restructuring [9]. But the policy agenda of the postpandemic world might have to address fiscal injustice and its overlooked disruptive potential [52] much more explicitly while being called to move beyond the colonial world order, and its still looming legacy, for good.

4. Materials and Methods

A preprint has previously been published [53].

Data Availability

The data that support the findings of this study are available from public repositories, without restrictions.

Additional Points

The code used that support the findings of this study is available from the corresponding author upon reasonable request.

Disclosure

A preprint of a previous version of this paper has been published on arXiv: https://arxiv.org/abs/2203.00618.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

P.S. and M.D. performed the data analysis. All the authors designed the research and wrote the manuscript.

Acknowledgments

A preprint of a previous version of this paper has been published on arXiv: https://arxiv.org/abs/2203.00618. The authors acknowledge financial support from the Spanish Government through grant FIS2015-38266, the Spanish program Juan de la Cierva (IJCI-2014-20225), Ministerio de Economía y Competitividad (Grants no. PGC2018-094754-B-C21 and No. FIS2015-71582-C2-1), Generalitat de Catalunya (Grant no. 2017SGR-896), Universitat Rovira i Virgili (Grant No. 2019PFR-URV-B2-41), ICREA Academia, and the James S. McDonnell Foundation (Grant no. 220020325).