The institutional design of China-led international institutions: a qualitative comparative analysis

Abstract

Since its opening up in the late 1970s, China has gradually become an important stakeholder in international relations, and it has become involved in the establishment and leadership of more than 20 international institutions. Some of these international institutions follow Western templates of international cooperation, while others seem to have a distinct Chinese signature. To better understand China’s increasing role with regard to international institutions, this article focuses on their institutional design through qualitative comparative analysis (QCA). We find that the institutional design of China-led institutions is influenced by both the policy type of the institution and the composition of the rest of the membership. When other major powers and democratic states are included as members, such institutions tend to have higher levels of institutionalization but also Chinese control. We furthermore find that the design of China-led institutions is more consistent for those established during Xi’s presidency since 2013. These findings are important because they provide new interpretations of China’s uniqueness in institutional design while resonating with previous research findings on Western institutions.

Introduction

Ever since its opening up in the late 1970s, China has gradually become an important stakeholder in international relations and its engagement in international institutions has increased considerably.¹ Not only has China joined established international institutions, it has also initiated international institutions of its own. The Shanghai Cooperation Organization (SCO), rooted in the 1996 “Shanghai Five” mechanism, as a regional security cooperation platform,² and the Asian Infrastructure Investment Bank (AIIB) alongside the Belt and Road Initiative (BRI)³ are well-known examples. Yet overall, China has been involved in the establishment and/or leadership of more than 20 international institutions in the last three decades.⁴

These China-initiated international institutions can differ considerably. For example, the AIIB has a structure that resembles other Western-led multilateral development banks and shows a high standard of sustainability rules, meaning that China might be conforming to liberal international norms and principles.⁵ Instead, the SCO has a consensus voting system and upholds the rule of non-interference. The SCO has been accused, in this respect, of promoting authoritarian norms by invoking a domestically and internationally acknowledged standard that helps legitimize authoritarian governments.⁶

To better understand China’s increasing role with regard to international institutions, this article focuses on the institutional design of China-led international institutions. We assume that these institutional designs differ from established international institutions,⁷ and we are interested in studying the conditions under which they vary. We focus on the degree of institutionalization, Chinese control, and membership restrictions in China-led institutions. This helps to answer whether China is projecting its interests and principles into the design of international institutions. The implications can enrich our understanding of the evolving landscape of global governance.⁸ To explain complex configurations of institutional design across 20+ China-led international institutions, this article applies the qualitative comparative analysis (QCA) method which is suitable for our medium-size number of cases and the causal complexity of the institutional design.⁹

In this article, we show that the position of China compared to other member states in such international institutions informs design choices. Power asymmetries between China and the other member states and the democratic density of the membership are important conditions. If China does not have asymmetric power in a task-specific institution (an institution that deals with few issues, for instance, either trade or security or environment), the institution tends to have a high degree of institutionalization, an open membership but also a high-level of Chinese control. Conversely, when China possesses asymmetric power in a general-purpose institution (an institution that deals with many issues, for instance, trade and security and environment), institutions are designed to be more restricted to outsiders and less institutionalized. To conclude: China is careful about the design of its international institutions. This is particularly the case for institutions established during Xi’s presidency since 2013, which have a greater level of consistency in terms of design.

These findings are important, as the differences between China-led institutions have not been extensively examined with the literature being relatively scarce.¹⁰ A significant number of previous studies about the institutional design of China-led international institutions are single case studies and lack comparison between cases.¹¹ They overlook the question of what accounts for the differences between China’s multilateral behaviors and institutional outcomes. This article pays attention to the paths behind the different institutional designs and delves into the conditions that influence what China pursues in establishing and leading international institutions. China, of course, does not design these international institutions alone and does so through a negotiation process with other member states during which China also needs to make compromises—the type of other member states is indeed an important variable in terms of the eventual outcomes of the institutional design—but the medium-n QCA comparison allows us to explain design differences in institutions which were initiated or co-founded by China.

The article also contributes to the literature on institutional design of international institutions and the study of rising powers in global governance in several important ways. First, it moves beyond the prevalent dichotomy between norm conformity and norm contestation in studying China’s role in international institutions. Rather than treating China-led international institutions as either replicas of Western models or vehicles for the diffusion of authoritarian norms, this article demonstrates that China’s institutional design choices are contingent and context dependent. Second, by integrating power asymmetry and democratic density into a configurational framework, the article shows how different combinations of membership composition can shape institutional outcomes. This perspective highlights that China’s institutional choices are not solely driven by unilateral preferences but are conditioned by bargaining power and legitimacy considerations with other member states. In doing so, the article offers a more nuanced understanding of how China’s preferences, constraints, and strategic environment jointly shape the institutional design outcomes.

The remainder of this article is structured as follows. The article first gives an overview of the cases included in the research and the principles of data collection. It then explains the theoretical foundation for the chosen outcomes and conditions, and elucidates how to measure them. It subsequently presents the findings. It then discusses the result, and the final section gives the conclusion.

China-led international institutions

International institutions have been defined as “explicit arrangements, negotiated among international actors, that prescribe, proscribe, and/or authorize behavior.”¹² According to rationalist scholars, international institutions are “a set of rules” that are persistent and connected, which prescribe behavioral roles, constrain activities, shape expectations, and thus stipulate the ways in which states should cooperate and compete with each other.¹³ Because this article focuses on the intentional institutional design, the focus will be on the international institutions with a written contract or treaty that regulates the rules and prescribe state behaviors and activities such as regular meetings.

Stephen and Skidmore usefully provide a list of all multilateral institutions created from 1990 to 2016 in which China had a leading role in establishing the institution or was a founding member.¹⁴ We have verified this list and have updated it until the end of 2024. The major source is the list of international and regional organizations presented on the website of the Ministry of Foreign Affairs of the People’s Republic of China and is supplemented using Lexis Nexis Academic and searching in Chinese official media (e.g. The People’s Daily). The final list of 23 institutions presented here meets the following inclusion standards: First, China initiated the institution/China is the founding member of the institution; second, there is an official document/website proclaiming the establishment of the institution and there is at least one meeting by 2023; third, the institution is intergovernmental. Therefore, by these criteria, we rule out institutions in private sector, such as the Bo’ao Forum, Cross-border Interbank Payment System, Asia-Pacific Quality Network, and institutions that lack substantial operation after establishment (e.g. the Universal Credit Rating Group). Table 1 presents all the international institutions included in this study.

As with any dataset and inclusion criteria, there are always borderline cases. We have tried to be inclusive in this regard. For instance, we have included the BRIC Summit given that it has become an important institution over the years, despite not having a constitutive treaty, budget, or a permanent secretariat. Similarly, we have included the BRI. While the BRI consists of a series of bilateral agreements between China and BRI countries, it also includes several multilateral high-level and thematic forums.

A first glance at these institutions shows that China has been particularly active in the period around the turn of the Millenium (1999–2005) as well as the beginning of the Xi Presidency (2013–16). The former coincided with Chinese internationalism including, for example, China becoming a member of the World Trade Organization, while the latter period relates to Chinese assertiveness in the international arena. It is worth noting that there is a gap between 2016 and 2023, when the China-Central Asia Summit was established. During this period, some international institutions founded do not fulfill the inclusion requirements, such as the World Internet Conference Organization established in 2022. What is also striking is that China engages both through task-specific institutions as well as more general-purpose institutions with a basis largely in the Asia-Pacific region.¹⁵ In the remainder of the article, we will focus on the institutional design of these institutions and try to explain their differences.

Theoretical framework

This article studies the design of China-led international institutions. Their design is important as it determines, among others, which actors participate and have access, what topics they discuss, how they make decisions, whether they rely on centralized institutions to facilitate policy-making, and what happens in case of unexpected events.¹⁶ In other words, if international institutions are designed differently, we can expect that they will also differently prescribe, proscribe, and/or authorize actor behavior.

The academic literature on the design of international institutions is too extensive to fully summarize here. It is nevertheless important to point at the foundational special issue edited by Koremenos et al. who identified five design characteristics of international institutions: the membership (universal or restricted), the scope (few or many issues are covered), centralization (whether tasks are performed by the institution or the members themselves), control (how decisions are made), and flexibility (in case of new circumstances).¹⁷ This framework remains helpful, as it allows us to compare many different international institutions. More recently, Hooghe et al. have provided a clearer theory on the institutional design of international organizations.¹⁸ They importantly distinguish different purposes for international institutions. Some institutions are established to address specific tasks. These are task-specific institutions that have a small scope yet a large membership, for instance, the International Monetary Fund. Normally, those institutions take decisions by majority-voting. The second type concerns institutions in which a smaller group of like-minded member states seeks to address many issues at once. Such general-purpose institutions often take decisions by consensus. Regional institutions, such as the European Union, are examples of general-purpose institutions.

While this research agenda on the design of international institutions has been very prominent in the international relations discipline, it is also criticized for being too rational or insufficiently considering power and distributive outcomes.¹⁹ In this article, we examine and challenge the universal ambitions underpinning the concept of the design of international institutions. While Koremenos et al., for instance, consider distribution and enforcement problems, and uncertainty over behavior and the state of the world, they essentially propose 15 conjectures applicable to all international institutions worldwide.²⁰ We argue, however, that China may see the design of its international institutions, under certain conditions, differently, based on its own domestic and international experience and ideology.

Institutional design with “Chinese characteristics,” where international institutions do not fully follow a universal design template but are also informed by specific Chinese preferences, can indeed be traced back to the historical context of Chinese international relations and foreign policy strategies. Initially, during the Mao era, China maintained a profound skepticism toward international institutions.²¹ However, a pivotal shift occurred after the China–United States rapprochement, culminating in China’s recognition of sovereignty at the United Nations (UN) in 1971.²² Coupled with Deng’s Open Door Policy in 1978, this enabled China to increase its engagement. As a newcomer, China embarked on strategic learning to position itself as a legitimate member of international society.²³ Aligned with Deng’s proposition of “keeping a low profile” (), China initially exhibited a modest participation. China’s stance in multilateral settings became more pronounced with President Xi Jinping’s ascension to power in 2013. Xi redirected China’s focus toward “striving for achievement” (). Under this banner, China emerged as a prominent architect of institutions, exemplified by the establishment of institutions like the AIIB and various other international institutions during the Xi administration.

Considering both the institutional design literature²⁴ and the specific emphases of China when it comes to international institutions, in this article we study specifically (a) the degree of institutionalization, (b) China’s control, and (c) membership restrictions as three outcomes in institutional design. These are institutional design features where we can expect “Chinese characteristics” as we would expect China to favor lower degrees to institutionalization, higher degrees of control, and stronger restrictions on membership. We will assess whether these features vary depending on (a) policy scope, (b) the regime type of the membership, (c) power asymmetries in the membership, and (d) the timing of the creation of the institution (conditions). In other words, we wonder whether we will find Chinese characteristics to be present in certain types of China-led institutions. We will first discuss these three outcomes in terms of institutional design before specifying the conditions.

Outcomes: Institutionalization, China’s control, and membership restrictions

Firstly, we examine the levels of institutionalization in China-led international institutions considering China’s emphasis on sovereignty and non-interference alongside informality (see Table 2). The emphasis on sovereignty and non-interference is underscored by China’s Five Principles of Peaceful Coexistence, particularly highlighting “mutual respect for territorial integrity and sovereignty” and “non-interference in each other’s internal affairs.” The sanctions imposed by the UN on China during the Korean War created a profound aversion to interventionism among the Chinese population. Consequently, any intervention in the name of international institutions without the consent of the affected region immediately triggers suspicion in Beijing.²⁵ Driven by past experiences of imperialism, China is vigilant in ensuring that international institutions do not infringe upon its national sovereignty. This defensive stance could elucidate the design of China-led international institutions—an approach that prioritizes sovereignty principles to enable China’s participation in international affairs on an equal footing while safeguarding its domestic autonomy from external interference. We would thus expect lower levels of institutionalization and no significant degrees of delegation to secretariats or adjudication bodies, which could directly interfere with domestic affairs in China. The second critical factor of institutionalization level, informality, finds its roots in Confucian ideas of harmonious societal relationship.²⁶ Chinese values place significant emphasis on relationship over rule-based governance. In terms of Chinese institutional design, rather than enforcing rigid rules to dictate actors’ conduct, China manages member relationships through ongoing negotiations, stressing the importance of consultation, coordination, and adaptive cooperation within international institutions. This, once again, leads to an expectation of lower levels of institutionalization with a focus on plenary negotiation bodies rather than formally delegated powers.

Secondly, we are interested in examining China’s position in the institutions it leads or has created. China is initiating and creating its own international institutions that run parallel to the existing institutions, which is often seen as an act of “contested multilateralism” in order to pursue policies and practices different from those of existing institutions.²⁷ Instinctively, the creation of new international institutions connotes a certain intention of taking control, leading to the discussion of whether China intends to challenge the Western-led liberal international order. Following that framework, many scholars argue that China, as a revisionist power, is doing “competitive regime creation” by building new multilateral institutions to undermine the existing international rules.²⁸ On the contrary, other scholars hold the view that China does not seek fundamental changes to the existing rules and norms and does not pose a challenge to the intrinsic rules of the international order.²⁹ Therefore, it is necessary to measure how much control China has in the international institutions it creates. We expect that China will avoid setting up new institutions where other member states subsequently call the shots. Control over decision-making will therefore likely be a key institutional design feature.

Thirdly, we are interested in the membership restrictions in China-led international institutions; in other words, the extent to which other countries can join or leave China-led institutions. While rationalist scholars have previously linked restriction on membership with the severity of the enforcement problems (international institutions typically have a small membership in case it is difficult to enforce the agreed rules and norms), there is also increasing research on the democratic nature of the membership (democracies meet democracies), as well as clubs of autocracies getting together.³⁰ Furthermore, scholars have studied the closure of IOs and the geopolitical considerations of membership.³¹ It is therefore important to understand under which conditions China limits membership to its institutions: to what extent China is comfortable to open up its institutions to countries around the globe, or whether it is selective in terms of the member states it wants to engage with. Also, in line with the arguments previously made, we expect that China generally will be reluctant to set up new institutions and subsequently let all sorts of different member states in, as this risks diluting the purpose of the institution.

Conditions: Policy scope, regime types, power asymmetries, and timing

While the scholarship on the design of international institutions is rich and has provided a detailed list of relevant conditions, the purpose of this article is to provide a first overview of the conditions that help to explain specifically the design of China-led international institutions. It therefore proposes to look at four conditions derived from the major international relations theories which are of relevance for China: policy scope (institutionalism), regime type of members (domestic politics), power asymmetries (realism) as well as the year of establishment.

Firstly, this article considers the policy scope of the China-led international institutions (see Table 3). The logic here is that China may prefer to design different types of institutions differently. Form follows function in line with institutionalist theory. As mentioned before in discussing the general literature on institutional design, it is worth distinguishing between general-purpose and task-specific as two distinct types of international institutions.³² In general-purpose organizations, clubs of relatively like-minded countries come together to address a range of collective problems. These organizations are thus community driven. Task-specific organizations, on the other hand, deal with specific problems and include a range of member states. We therefore expect that China approaches its general-purpose organizations, in which it might seek to address a range of problems with a community of Asian states or a group of authoritarian states, differently than task-specific ones, where it might try to address a single specific problem. At the same time, it is also likely that within the task-specific organizations, China may pay more careful attention to some (e.g. security-related institutions) than others (e.g. trade-related institutions) depending on the extent to which they align with China’s core interests.

Secondly, we expect the regime types of the other member states to matter in how China designs its international institutions. As domestic politics argues, the nature of regime types affects how states act in world politics. Existing research has demonstrated that the average level of democracy in membership influences the level of openness.³³ China is also likely to engage differently in institutions with largely authoritarian regimes than with democratic regimes. Therefore, we propose that when establishing an international institution with similar states, China brings more Chinese characteristics into it. But when there are more democratic states, the institution is likely to be more institutionalized, have less restriction on membership entry, and China has less control in it. Establishing new international institutions ultimately requires negotiations with the other member states and different type of member states will likely have different types of preferences on design. China may thus have to compromise.

Thirdly, we consider power asymmetries within the membership of China-led institutions. The argument is that if China dominates with the other member states simply by virtue of its power, we can expect it to bring more Chinese characteristics in design. At the same time, if China faces rivals among the other member states, in other words a lower power asymmetry, it may want to include safeguards through the institutional design. The realist view, which posits that institutions are a reflection of the distribution of power in the world, emphasizes the importance of power in multilateral institutions.³⁴ International regime is not only shaped by the goal of resolving cooperation problems but also constrained by the distribution of power.³⁵ The model launched by Stone shows that when United States structural power declines, formal governance becomes more important relative to informal governance, and the United States retains more formal control.³⁶ We expect that China experiences similar situations and power asymmetries have influences on the outcomes.

Finally, we take into consideration the year of establishment of the relevant international institutions. Studies have confirmed that China has been more assertive, with a significant shift occurring around 2013 when Xi assumed office in March, marking a great strategic change in China’s foreign policy.³⁷ Xi has emphasized a lot on a shift toward a more “pro-active” foreign policy to “strive for achievement,” which is more ambitious and assertive than the expressions of the former leaders Jiang Zemin and Hu Jintao.³⁸ Departed from the “risk averse caution,” Beijing has forged a series of institutions built around China’s interests and guided by Chinese values in order to become a norm-shaper. Therefore, it is estimated that the timing of an institution’s founding influences its institutional design on the level of institutionalization and control.

In this theoretical section, we have suggested that China may design its international institutions differently and add “Chinese characteristics,” such as lower levels of institutionalization, larger Chinese control, and restrictions on membership. We noted that this may particularly be the case under conditions of certain policy functions (particularly for institutions close to China’s core interests), a more democratic membership and lower power asymmetry (which makes it more difficult for China to control such institution) as well as institutions created after 2013 (when President Xi came to power). Since this is an initial article on the institutional design of China-led international institutions, it is not possible to provide ex ante clear hypotheses on how these conditions precisely affect the three outcomes. The medium-n QCA approach, which we will present in the next section, is uniquely capable of linking different combinations of conditions to the outcomes. The 20+ cases where China (co-)created international institutions provide an important empirical basis for theory evaluation. Our purpose is, in this respect, to highlight some overall patterns how China approaches the design of international institutions.

Research design

After systematic research on different datasets, we identified a total number of 23 China-initiated intergovernmental international institutions (Table 1 above). As some of these institutions have significantly been institutionally revised over time (e.g. with the adoption of a different mandate or the addition of many more members), we include them as new cases, thus resulting in 27 cases in total. China was not solely responsible for initiating and creating these international institutions but was involved in the process, which allows us to theorize that China affected their institutional designs. Clearly the establishment of these institutions was a negotiation process involving various other member states as well, but if China held strong specific preferences on specific types of institutions, we should see variation in the ultimate design outcomes. In this section, we operationalize the outcomes and the conditions. For the three outcomes, we use qualitative data transitioned to fuzzy set while two conditions are dichotomous (crisp) and two are fuzzy sets transferred from quantitative data.

Operationalization and calibration of the outcomes

Firstly, the level of institutionalization measures the density of international institutions in terms of their institutional organs. Three indicators are used: a regular assembly, a permanent executive body with delegated decision-making powers, and a secretariat with delegated implementation powers.³⁹ The delegation-element here is important as it contrasts with the more traditional Chinese focus on sovereignty. If an institution has all three organs, it is considered to have an independent operating function. If none of these are present, this international institution is only an informal ad hoc forum or meeting. We calibrate these four levels of institutionalization into 0, 0.33, 0.67, and 1 (from no organs to three organs).

The second outcome is China’s control in these international institutions. We choose two institutional features that relate to institutional control and prestige, which are the location of the organization and the position of the organizational head.⁴⁰ The first indicator is whether the institution has a headquarters located in China. It is coded as 1 if the headquarters is in China, 0 if it is not in China, and 0.5 if it does not have a headquarters. The second concerns the leadership of the international institution. We measure whether the head of the executive body is Chinese national. This indicator reviews all the heads of the executive body in the institution’s history, and it scores 1 if all the leaders of the executive body in the history are Chinese, 0.5 if only partially are Chinese, and 0 if none are Chinese. However, some institutions do not have an executive body. We then code the secretariat general. If an institution has no executive body nor secretariat, it is coded as 0 because no executive body nor secretariat means that supernational control is impossible. The results are calibrated into four levels assigned as 0, 0.33, 0.67, and 1.

The third outcome is membership restrictions. The indicator is the entry of new members. If an institution sets high standards for accepting members, it is considered to be a restricted institution. An institution that allows members unconditionally is coded as 0, and one that needs to be approved by the consensus of all members is coded as 1. A simple majority of consent in order to accept newcomers is coded as 0.33, and a special majority is coded as 0.67.

Operationalization and calibration of the conditions

The four identified conditions draw on different IR theories and all have potential relations with the three outcomes, thus making them suitable for a QCA. The first condition is what functions the institution fulfills in terms of policy scope. On policy portfolio, it can be an institution focused on a comprehensive set of policies or one that concentrates on a limited set of policies.⁴¹ We divide all the international institutions according to their functions into general-purpose and task-specific and code them as 1 and 0.

The second condition is the democratic density of an institution which measures the regime type of the member states. For all of the member states, we take their polity5 scores and then calculate the average score for the institution. As a much used indicator of democracy, the Polity scores are integers that can be converted into regime categories in a suggested three-part categorization of “autocracies” (−10 to −6), “anocracies” (−5 to +5 and three special values: −66, −77 and −88), and “democracies” (+6 to +10).⁴² Therefore, we set the calibration threshold 6 as totally in the set and −6 as totally out of the set.

The third condition is power asymmetries between member states in each institution. This condition is assessed by measuring China’s power as a proportion of the total accumulated power of all member states in the institution. The measurement of member states’ power is based on the National Material Capabilities data and specifically the Composite Index of National Capability (CINC), which contains annual values for total population, urban population, iron and steel production, energy consumption, military personnel, and military expenditure of all state members.⁴³ This is the standard dataset to measure power in quantitative international relations studies. We use CINC in the specific year of the institution’s establishment or amendment. Since the data are updated to 2016, we use the data of 2016 for the cases after 2016 as well. We consider 75% as totally in the set (China has asymmetric power in the institution) and 25% as totally out of the set.

The final condition is the year of establishment of institutions. It is coded as 0 if the year of the founding time or the amendment of institution is before 2013 and 1 if it is 2013 or onwards. The year 2013 was significant because it marked a foreign policy transition when Xi came into power.

Findings

We study these outcomes and conditions through a QCA of 27 cases of international institutions. While the findings conform to a large degree to the expectations, there are also several surprises. First, as is generally understood in the literature on institutional design, we find a difference between the design of task-specific and general-purpose institutions with general-purpose institutions having lower levels of institutionalization, Chinese control, and more restrictions on membership.⁴⁴ Second, we find that the composition of the rest of the membership is critically important. When other major actors or democracies are included in China-led institutions, we see a higher level of institutionalization but equally also a higher level of Chinese control. Third, we find more consistency in the institutional design for institutions created during Xi’s presidency since 2013. In other words, there are clearly Chinese-specific considerations informing institutional design.

As noted, we rely on QCA as a method to study these 27 cases. QCA is a long-standing method that has specifically been designed to compare medium-n cases. When moving beyond a handful of cases, qualitative research methods become increasingly complex to apply, and it may be difficult for researchers to convincingly compare the significance of combinations of different variables. QCA, in this respect, provides a structured way of comparison. At the same time, QCA fills a gap below the threshold of quantitative statistical models (such as regression analysis), which normally require in the range of (at least) 100 observations to provide significance. QCA is, however, not without its limitations. It cannot provide as much case-specific context as in-depth qualitative studies and it cannot include as many variables as quantitative regression analysis due to the limited number of cases. For the purpose of this article, however, QCA works well: it allows us to compare 27 cases of institutional design beyond individual case studies and provide more generalized findings about all the new institutions that China initiated and co-founded.

To apply QCA to our cases, we investigate which combinations of conditions are necessary or sufficient for the outcomes following an Enhanced Standard Analysis process. Firstly, necessity tests were conducted on both the presence and the absence of the outcome. Then, we created two truth tables for the (non)presence of the outcome. Consistency scores indicate the fit between the QCA solution and the empirical evidence, and we set the consistency cutoff point to 0.75 for the inclusion of truth table rows.⁴⁵ PRI stands for Proportional Reduction in Inconsistency, which is used to prevent situations where a causal condition appears to explain both the presence and absence of the outcome. We set the threshold value for PRI at 0.51, which is a common good practice.⁴⁶ Then, we generated the solutions using all the possible logical remainders while excluding any contradictory simplifying assumptions in the solutions. The final solutions are intermediate solutions that partially use logical remainders by excluding some logical remainders and setting directional expectations, which situate in the middle between the conservative and the most parsimonious solutions. The intermediate solutions are chosen in order to achieve the most concise explanations while echoing with our theoretical evaluation. The software used for the analysis is the QCA package and SetMethods package in R.⁴⁷

Institutionalization

As a first step, in our QCA, we conduct the necessity analysis in order to determine whether any of the conditions or combinations of conditions is necessary to explain the degree of institutionalization. The criteria for considering a combination of conditions necessary are its consistency, relevance of necessity, and coverage. The good practices set the consistency value (inclN) at 0.9 and coverage (covN) and relevance of necessity (RoN) at 60% as the minimum.⁴⁸ After doing a necessity analysis for both the presence and the absence of high institutionalization, we found no condition that reached the standards and could be deemed necessary for the outcomes. In other words, there are different paths and combinations of conditions to explain the degree of institutionalization.

As a next step, we carried out the sufficiency analysis to examine whether a combination of conditions is sufficient to explain institutionalization. In other words, what pathways could lead to the outcome of different levels of institutionalization. We find indeed three paths with a combination of conditions that result in a high level of institutionalization (Table 4). Importantly, we find that institutions created from 2013 with low power asymmetry or high democratic density (paths 1 and 2) result in high levels of institutionalization (conditions marked with a black dot in the table need to be present, while conditions marked with a white do not need to be absent).⁴⁹ In other words, these are China-led institutions during the Xi Presidency that also include other major powers or democracies. We also find that task-specific institutions with a low power asymmetry (where China is not dominant) have high levels of institutionalization (path 3). The solution formula covers 11 of all the 15 cases and does not include any deviant in kind case, thus can be deemed reliant.

In terms of low institutionalization, our solution suggests three paths (Table 4). The solution shows that a general-purpose institution combined with other conditions can lead to a low level of institutionalization, which reckons with our expectation that China tends to approach many general-purpose institutions with informal mechanisms. The coverage for the solution is less than satisfactory, indicating that the lack of institutionalization is driven by multiple factors beyond the four conditions.

China’s control

As with the previous outcome variable, we conducted a necessity analysis for both the outcome and the non-outcome for China’s control, but none of the conditions passed the necessity thresholds (inclN >0.9 and RoN and covN >0.6). However, high democratic density comes closest to becoming a necessary condition for high control with its coverage and relevance of necessity both exceeding 0.6 (0.685 and 0.654) and its consistency score of 0.810. While this does not fully meet the strict threshold for necessity, it suggests that high democratic density is an important enabling factor for China’s control.

In our 11 cases that have an inclusion score above 0.5 in high China’s control, 10 of them exhibit high democratic density, with the only exception being the Shanghai Cooperation Organization (see Table 5). Contrary to our initial expectation that a high control is more possible in a group of autocratic states, this finding indicates that China is more likely to achieve dominance in an institution primarily composed of democratic states. It indicates that China does not want to give the initiative away and keeps the institution at a tight leash.

Path 1 and path 2 indicate that China is more likely to assume a central role in task-specific institutions characterized by a high level of democratic density. The coverage of the solution is a little less than satisfactory, which is not uncommon, especially when a wide range of factors influence how much China actually takes control.

For the absence of high China’s control, we identify two paths that are both the conservative and intermediate solution. They suggest that in general-purpose international institutions with either low power asymmetry or low democratic density, high control is less possible. The solution covers seven cases and does not include any deviant-in-kind case.

Overall, the findings indicate that the institutional design of China-led international institutions is shaped by both deliberate choices and environmental conditions. General-purpose institutions tend to rely on non-formal agreements and exhibit lower level of China’s control. While China can assert its control within a group composed mostly of democratic members, its ability to do so is constrained in general-purpose institutions, as they inherently provide fewer opportunities for institutionalization and formalized leadership structures.

Restricted membership

For the outcome restricted membership, 21 out of 27 cases have an inclusion score above 0.5, and only 6 score below 0.5. In the necessity analysis, we find that the conjunction T + PA (a general-purpose institution or a high power asymmetry) has a consistency value at 0.927, coverage rate of 0.854, and a relevance-of-necessity score at 0.605. These values indicate that this conjunction is a necessary condition for restricted membership. Accordingly, two SUIN conditions are considered necessary for a restricted membership: a general-purpose institutional focus or a high power asymmetry.⁵⁰

Among the 21 cases with high restricted membership, 20 meet the condition, with the sole exception being the AIIB, which is neither a general-purpose institution nor does China have a high level of power asymmetry within it (see Table 6). This exception can be explained as, despite requiring 75% approval for new members, the AIIB has generally been open to admitting new members, making it less restrictive compared to other cases. In the six cases that score low in restricted membership, all are either task-specific institutions or have low power asymmetries. This pattern suggests that restricted membership is more likely to be observed in cases where China holds significant influence, as this enables China to shape entry rules and control institutional dynamics. Without such influence, maintaining a restricted membership structure would be less feasible. In other words, to preserve China’s relative power within an institution, maintaining a restricted membership is crucial, or else China’s influence within the institution will consequentially decline.

The solution for a highly restricted membership identifies two paths. Path 1 is that institutions with a high degree of power asymmetry tend to have restricted in membership. Path 2 is that general-purpose institutions established after Xi’s presidency enforce restricted membership entry requirement. Regarding the second path, Xi has actively initiated new regional cooperation platforms with Asian and Latin American countries. Since the pool of potential member states in the regions is already fixed, these institutions inherently maintain a highly restricted membership structure. The first path focuses on the composition of members. Institutions following this pattern exhibit a clear power hierarchy, where China has the dominant power.

The solution for the absence of restricted membership is, unfortunately, more limited. Only one row in the truth table is included in the solution due to the small number of cases. However, it is noteworthy that five out of the six cases exhibit high democratic density, reckoning with our expectation that democratic institutions tend to be more open in terms of membership.

Robustness test

The robustness test is done in two ways (see also Appendix, Tables A1–A7). Firstly, we modify the calibration rules. For the calibration of the first outcome, we adjust the inclusion thresholds from 0.33 and 0.67 to 0.25 and 0.75. The same adjustments are applied to the second and third outcomes, as well as to the condition DD and PA, where we also change the calibration thresholds. Alternation in inclusion and exclusion anchors does not affect the solutions terms. The parameters of fit subsequently change but only have trivial variation.

Secondly, we change the consistency threshold to 0.8, following the suggestion that researchers should always run their analyses with at least two different raw consistency thresholds.⁵¹ The results show a decrease in coverage: when the consistency threshold is set at 0.8, the solution terms for high institutionalization become the subsets of the initial solutions, as they are limited to path 1 and path 2 in Table 4. The solution for restricted membership becomes more conservative, which is Y*T + Y*PA + T*PA, suggesting that high power asymmetry alone is not sufficient for a restricted membership. The solutions for O2 and the absence of O1 and O2 remain unchanged but for the ~O3, no solution exists. Overall, altering the consistency threshold does not lead to any substantial changes. The results confirm that the parameters of fit and solution terms remain stable, thereby demonstrating the robustness of the sufficiency analysis of all the outcomes.

Discussion of results

Based on theoretical expectations derived from the literature, our findings largely align with the initial assumptions. For example, we proposed that China approaches general-purpose organizations involving a community of regional states differently from task-specific ones in line with broader findings in the academic literature about all international institutions.⁵² The results indicate that in task-specific institutions, a high level of institutionalization and open membership are more likely to emerge when combined with certain conditions. In contrast, general-purpose institutions under the same conditions tend to exhibit the opposite outcomes. For example, INBAR and EAS have similar features in terms of democratic density, power asymmetry, and their roles as regional platforms for cooperation, with overlapping member states. However, they diverge significantly in all three outcomes. As EAS focuses on 16 areas of cooperation that cover peace, security, finance, health, and so on, INBAR concentrates solely on the use of bamboo and rattan, which makes inclusion and institutionalization easier for INBAR. In this respect, China-led institutions conform to the universal expectations of the rational design of international institutions.

Our second expectation was that China is likely to engage differently with institutions composed of predominantly authoritarian regimes compared to those with a high density of democratic regimes. The results support this expectation, demonstrating that a high level of democratic density is able to lead to a high degree of institutionalization (after 2013) as well as a high level of China’s control. Conversely, the conjunction with a low level of democratic density is likely to bring the opposite. An interesting comparison can be made between FOCAC, China-CELAC Forum, CASCF, and China-Central Asia Summit. They are all regional cooperation mechanisms initiated by China respectively with Africa, Latin America, Arab states, and Central Asia, where China dominates in terms of power relations. The first two mechanisms are mostly composed of democratic members while the latter two are not. Notably, the latter two exhibit lower levels of institutionalization and China’s control, reflecting preferences for flexible, bilateral engagement over formal multilateral mechanisms.

We also expected that if China faces rivals among the other member states, in other words a lower power asymmetry, it may seek to incorporate safeguards through the institutional design. The solutions of outcomes 1 and 2 indicate that low power asymmetry is associated with a high level of institutionalization (after 2013 or in a task-specific institution) and China’s control (in task-specific institutions with high democratic density), and the solution of outcome 3 demonstrates that high power asymmetry is linked to restricted membership. One good case example is the AIIB. Because many European countries were founding members, China’s relative power in the AIIB is comparatively low. Nevertheless, China still managed to secure a veto in its institutional design, as well as establishing a non-resident Board of Directors that enhances the importance of the President.⁵³

However, the solution associated with high control diverges from our initial expectations. We expected Xi’s rise to power signals a shift toward a broader strategic vision, and a more proactive and assertive foreign policy that can result in a greater willingness to manifest control.⁵⁴ However, our results do not support this assumption, as no single condition operates on its own due to the complex causal ambiguity inherent in the QCA method.

Also contrary to our expectation, high democratic density appears to be so vital for China’s control that it nearly qualifies as a necessary condition. Several explanations are possible. For instance, this may occur because democratic states exhibit greater flexibility in joining international institutions and are more inclined to delegate authority within multilateral settings. Another possibility is that China places higher importance on asserting its status within groups of democratic countries, where it strongly wants to maintain a prominent position. The solution for the absence of high control further suggests that China’s control is often constrained in general-purpose institutions, particularly when it does not have asymmetrical power or when the institution consists predominantly of non-democratic members. The finding is noteworthy because this observation raises the prospect of additional factors, such as the desire to uphold national dignity and the importance of having a stronger voice in the presence of democracies or developed states, which may warrant exploration in future research.

Furthermore, our analysis reveals that after 2013, China’s institutional design adheres to a more consistent pattern compared to prior to this period. Across all the solutions, we identify six paths after 2013 with an average consistency score of 0.889, whereas prior to 2013, only three paths emerge with a lower consistency score of 0.798. This suggests a more structured and coherent route in institutional design after 2013. This shift can be attributed to the absence of significant leadership changes after 2013, in contrast to the multiple transitions that occurred previously. It reflects China’s entry into a new era characterized by a firmer and more resolute stance on global governance. The future behavior in China’s foreign policy could be more predictable in the absence of significant change in leadership.

Conclusion

This article analyzes three key aspects of the institutional design of all the international institutions China initiated or joined as a founding member, using the QCA method. The total 27 cases exhibit considerable variation across multiple dimensions. This study explores the conditions that lead to three distinct institutional design features: the level of institutionalization, China’s control, and restriction on membership entry. Drawing on existing literature and empirical evidence, we identify four conditions that influence China’s institutional design choices: the type of the institution (general-purpose or task-specific), the member states’ power distribution, their regime types, and the founding period (before or after Xi’s presidency).

Overall, our analysis indicates that China does have preferences in the institutional design of international institutions, and that this design is significantly influenced by the conditions of the overall membership. For example, China’s emphasis on sovereignty makes it favor informal and flexible arrangements, and our findings on institutionalization reveal that a high level of institutionalization is likely to occur when China holds relatively less power, when democratic density is high, and in institutions established after Xi’s presidency. While China clearly needs to negotiate with other states in designing institutions, and compromise in the process, this article shows that there is clear variation in the eventual design outcomes across the different institutions that can be explained by the four conditions.

Indeed, our article reveals that the other member states are particularly significant. For instance, power asymmetry plays a crucial role in shaping China’s design choices. When China does not hold asymmetric power over other states, it tends to design an institution that has a formal structure (after 2013 or in a task-specific institution) and favors control over China (particularly in task-specific organizations with a high proportion of democracies). Conversely, when China possesses asymmetric power and the institution lacks democratic member states, it tends to design less institutionalized structures with more restrictive membership criteria. The regime composition of member states also influences the institutional design. In institutions with a high proportion of democratic states, China is inclined to develop more formal rules and emphasize its own control. An institution with a high democratic density fosters the potential for formal and binding rules; however, it may also signal China’s concern about maintaining its central position within the institution.

Our study underscores the prevalence of equifinality, that a particular outcome can be reached through multiple and distinct paths, in institutional design. The solutions associated with the presence of the outcome have an average coverage score of 0.61, while those of the absence of the outcome have a lower coverage score of 0.41, implying the existence of other pathways to the same outcomes beyond the four conditions considered. The presence of the outcomes exhibits more representative patterns, reflecting a rational choice to design institutions that are formal, emphasize China’s position, and impose selective membership criteria. In contrast, the absence of these features does not appear to be the result of intentional design but rather the outcome of other contextual factors. Given the complexity of institutional design, these findings suggest that further detailed examination is essential to untangle the broader range of influencing factors.

Through its findings, this article contributes to the broader theoretical literature on institutional design and Chinese approach to global governance. While confirming some rational-functionalist patterns of institutional design, this article also extends the literature by incorporating the interactions between regime types and relative capabilities. In doing so, the article demonstrates that the determinants of institutional design are not only functional but also relational. By focusing on China’s institutional building behavior, this article also contributes to the theoretical debates on China’s stance on global governance. We show that dualism is a viable strategy in China’s global governance approach: tendencies toward restricted blocs and revisionist designs are most likely to emerge in general-purpose institutions where China holds strong power, while at the same time China continues to present itself as a responsible stakeholder and to seek influence within the existing liberal international framework.

This study is the first comprehensive comparison of the institutional design of China-initiated international institutions using the QCA method. By bridging quantitative and qualitative data via QCA, it provides a structured framework for a better understanding of China’s approach to institution building. The findings offer a foundation for future research to investigate individual cases and explore the causal mechanisms that drive China’s strategic choices in creating international institutions. Our study has its limitations. In our effort to establish a consistent measurement standard across different types of international institutions, certain simplifications were necessary in the operationalization process. These simplifications may have affected the precision of the concepts we intend to reflect. Therefore, future research should aim to give a more detailed analysis of individual cases to provide deeper insights into the complexities of institutional design and China’s evolving role in global governance.

The findings of this study have important implications for understanding China’s future strategies of international institutional creation. The results suggest that China is more likely to initiate new international institutions in issue areas where existing institutions insufficiently reflect its preferences and where it could assume a leading role among a diverse group of participant countries. One example is the International Organization for Mediation (IOMed), established in May 2025 and headquartered in Hong Kong.⁵⁵ China has presented IOMed as a valuable supplement to existing global dispute resolution mechanisms, particularly in an area in which it claims to have been treated unfairly. Such issue areas are likely to include development cooperation, security, digital governance, as well as emerging domains such as artificial intelligence, space, and deep sea, where governance gaps persist and institutional rules remain relatively underdeveloped.

At the same time, China’s recent global governance initiatives—the Global Development Initiative (GDI), the Global Security Initiative (GSI), the Global Civilization Initiative (GCI), and the Global Governance Initiative (GGI), proposed at the SCO Summit in September 2025⁵⁶—reflect an ambitious effort to create a comprehensive Chinese version of global governance framework. These initiatives underscore China’s aspiration to assume a more prominent leadership role, particularly as a representative and leader of the Global South. The principles embedded in the new initiatives are likely to integrate and connect to other China-led international institutions such as the BRICS and the SCO. In this context, future Chinese institutional creation is likely to take hybrid forms, combining newly established platforms with existing multilateral institutions or embedding flexible arrangements within established institutions. This pattern is consistent with our findings that China’s institutional design choices are highly sensitive to member composition, power asymmetries, and regime types.

In sum, this article demonstrates that China’s future approach to institutional design under Xi Jinping is likely to remain broadly consistent, reflecting stable preferences regarding sovereignty, control, and leadership. At the same time, China is expected to pursue differentiated institutional strategies across different issue areas, balancing these preferences with the practical need to attract participation from diverse member states.

Acknowledgements

Z.W. gratefully acknowledges funding from the China Scholarship Council.

Footnotes