‘Formal Adoption – Informal Subversion’: Limits of International Constitutional Assistance

ABSTRACT

Peace processes involve reforming or rewriting constitutions as a pathway to establishing an inclusive state, often with international constitutional assistance (ICA). Examining Nepal’s post-conflict constitution-writing process, this article explores why ICA fails to deliver on inclusion. It argues that ICA’s prioritisation of formal aspects – inclusive institutional design and participatory process – makes it unable to influence informal politics through which elites informally subvert formal commitments on inclusive institutions in a bid to retain their power. This ‘formal adoption-informal subversion’ of inclusion-related commitments by elites is enabled by elites adopting four strategies, namely neglect, overwrite, trade-off, and exploit of formal commitments.

KEYWORDS:

• International constitutional assistance
• conflict-affected states
• inclusion
• Nepal
• peacebuilding

Introduction

Between 1990 and 2006, political transitions in 23 fragile and conflict-affected states (CAS) concluded in a new or revised constitution (Bell and Zulueta-Fülscher 2016).¹ Constitutional issues, including reforming state institutions to rethink how power is exercised, resources are shared, and the relationship between individuals and the state, are core to addressing conflicts, and thus central to peace negotiations (Berghof Foundation and United Nations DPPA – Mediation Support Unit 2020). New or reformed constitutions accordingly are seen to be pathways for addressing exclusion-related drivers of conflict and promoting inclusive institutions (Fiedler 2019) both by the international community but also by marginalised groups in CAS. This emphasis on inclusion is shared broadly across international support to peacebuilding, beyond constitution-writing processes in CAS. Based on the understanding that social, economic, and political exclusion is the key impediment to peaceful transitions, contemporary peacebuilding support is designed to foster inclusion of women, opposition political parties, and marginalised ethnic groups. The international community has promoted both inclusive processes, through encouraging participation of various marginalised groups in peace negotiations and decision-making structures and inclusive outcomes by instituting gender and quota systems, federalism, unity governments, and electoral reform, etc. to make state institutions more representative (Dudouet and Lundström 2016). Given such corelations, many CAS not only embed such constitutional issues into their peace agreements, but often undertake wholesale rewriting of a new constitution or a substantive reform of an existing one as part of their peace processes. A substantial degree of international support has been deployed to such constitution-writing or reforming processes, where the international community, including various multilateral bodies like the UN, but also bilateral and intergovernmental agencies, support CAS through financial support, technical assistance, and capacity building of domestic stakeholders (Bell 2017). While largely overlooked in peace studies, such international constitutional assistance (ICA), in practice, has become mainstreamed within the wider repertoire of peacebuilding tools used by donors.

Despite increased uptake, international constitutional assistance (ICA) as a policy arena represents a fundamental paradox. In this paradox, on the one hand, there is an increased focus on strengthening formal institutional structures through projects like ICA (Bell 2017). The increased number of CAS receiving constitutional advice from international experts, the number of international organisations like the UN Development Programme embedding constitutional assistance within their repertoire of work, and the attempt to coordinate international constitutional assistance signals this trend (United Nations Development Programme 2014). On the other hand, there is a wide acknowledgement that the preoccupation of international support with strengthening formal institutions through constitutions, democratic institution-building, elections, anti-corruption policies, and the rule of law as pathways to managing violence and organising politics in CAS is not working (Khan 2010; Menocal 2015). A broad body of scholarship in peace studies affirms that in CAS marked by weak formal state institutions, it is the political will of elites who rely on informal institutions, such as patronage rather than the design of formal state institutions provided for by constitutional provisions, that determine the outcomes of the peace processes (Barma 2017; De Waal 2009; Pospisil and Rocha Menocal 2017). Of course, scholars and donors engaged in international constitutional assistance have outlined the pitfalls of seeing constitution writing merely as a technical exercise and called for a politically informed approach to constitutional assistance (Bell and Zulueta-Fülscher 2016). This, however, is often difficult in practice. Donors do, for instance, acknowledge the relevance of understanding the role of elites and their reliance on informal institutions, but they continue to prioritise strengthening of formal institutions (van Veen and Dudouet 2017).

Distinct from other peacebuilding tools, a peculiarity of ICA is that its entire remit and objective is vested in crafting the ‘perfect’ formal institution: the constitution. Thus, the very raison d'être of ICA is largely limited to the formal sphere rather than the informal domain, which tends to be dominated by informal bargains and agreements between political elites, and which has the potential to dilute the formal constitution-writing process. Not surprisingly, despite the spate of international constitutional assistance and advice and formal commitments to international norms, CAS have rarely implemented them or have simply backtracked on them, as evident in contexts from Cambodia to Iraq (Lawrence 2021 on Cambodia). For instance, in Iraq, despite the international constitutional advice being centred on measures to decentralise the country, and such measures being encoded in the 2005 Constitution, the reluctance of successive federal governments to dilute their power at the expense of the Kurdistan region and the provinces has impeded attempts to devolve power (Alkadiri 2020).

This article engages with this paradox of increased focus on ICA at a time when the effectiveness of such formal institutions is being acknowledged and investigates the pitfalls in ICA with an in-depth study of international support to Nepal’s constitution-writing process. Nepal serves as an important case to understand the norms and practices underpinning ICA and its limitations in promoting inclusion. Between 2005 and 2015, Nepal undertook a well-institutionalised peace process, facilitated in some areas by the United Nations Mission in Nepal (UNMIN) as well as being supported by a host of peacebuilding organisations. The peace process not only sought to end the decade-long conflict waged by the Maoist rebels but also undertook comprehensive reforms towards making the state more inclusive. Writing a new constitution to institutionalise ‘inclusion’ was a demand made by the Maoist rebels, and accordingly embedded into the Comprehensive Peace Agreement. Constitution writing was also seen to be a primary pathway to addressing historical state-sanctioned exclusion. Despite Nepal being a multi-ethnic, multi-lingual state, a narrow elite base comprised of Nepali-speaking Hindu and High Caste communities from the hills, often abbreviated as CHHE (Caste Hill Hindu Elite), representing 31.25% of the population, have dominated the political settlements (Lawoti 2008; Pfaff-Czarnecka 1997). Accordingly, constitution writing in Nepal during the peace process was institutionally led by a Constituent Assembly whose members were elected by the people. This domestic momentum for an inclusive constitutional settlement was supported by substantial ICA in the form of financial, technical, and capacity-building assistance. Due to the challenges of ICA to foster broad-based inclusion despite a comprehensive peace agreement that prioritised a new constitution, the plethora of international support, as well as domestic demand for a new constitution, Nepal is an ideal case to explore the normative underpinnings, praxis, and limitations of ICA. At a time when the number of comprehensive peace agreements (since 2016) are declining and being replaced by local peace agreements, the generalisability of the Nepali case could be questioned. The short and largely issue-centred nature of local peace agreements often exclude broader questions of constitutional reform or writing process (Bell et al. 2021). However, it is important to note that even in contexts like Myanmar, where a national, comprehensive peace agreement failed to materialise, constitutional issues were an important focus for international actors (Myint 2020).

Exploring the case of Nepal, the article argues that, as attention by ICA providers is placed on the formal design and process of drafting a constitution, they tend to be unable to adequately influence informal political processes through which elites subvert the formal process. In this process of ‘formal adoption – informal subversion’, the formal aspects of constitution writing, centred on participatory processes and inclusive institutions, seek to curtail the elite hold on power, and thus are likely to be undercut by formal and informal negotiations and agreements between elites. The prioritising of the formal aspects of constitution writing leads ICA providers or donors to be dissociated from informal political power dynamics, often concentrated on political pacts and negotiations among elites. Such dissociation imperils the ability of international actors to understand informal politics and strategies, such as neglecting, overwriting, trade-off, and exploiting formal institutions. Such strategies lead ICA to fail and question the very faith that ICA providers put on the formal institutions like a constitution and its ability to beget inclusion. The article does note the counterargument that without ICA, the constitution in Nepal might have been more exclusive. However, the article seeks to understand processes and mechanisms as to why ICA does not deliver as it aims to, rather than ascertaining constitutional outcomes in CAS in the absence of ICA.

In making the argument, the article makes three contributions to the debates in peace studies and international relations more broadly. Firstly, it focuses on an underexplored aspect of the peacebuilding toolkit, constitutional assistance, which unlike others such as rule of law, transitional justice, or Security Sector Reform (SSR) has yet to be adequately analysed. Second, through a political-economic framework centred on the power and interests of elites, the article analyses why and how ICA fails. It categorises the four strategies through which elites undercut formal constitutional processes and issues, and shows how these strategies work against the rationale of ICA providers. Given the increased reliance on ICA by Western donors, these findings have a policy significance along with a scholarly one. Lastly, it contributes to the discussion on local agency in peace studies, by demonstrating the importance of informal bargains and processes.

The first section of this article discusses how international constitutional assistance providers, in their conceptualisation and execution in conflict-affected states, have emphasised formal processes and institutions. It explores how this advocacy for participatory processes and inclusive state institutions aims to dilute elite power and interests. The second section of the article highlights the growing consensus in peace studies, and the study of CAS in general, on the limitations of formal institutions, and the relevance of understanding and engaging with informal politics. The third section draws inferences from the previous two sections, and empirically applies them to Nepal’s constitution-writing process.

ICA: Prioritising institutional design and the process of constitution writing

The emergence of international constitutional assistance has seen the international community extending its peacebuilding role into constitution building (International Institute for Democracy and Electoral Assistance 2011). Promoting inclusion has been central to the ICA project for at least two reasons. First, constitutions help address conflict as they seek to mitigate exclusion-related drivers of conflict by designing inclusive state institutions through power-sharing mechanisms, changes in electoral law, and greater recognition of minority cultures (Berghof Foundation and United Nations DPPA – Mediation Support Unit 2020). Relatedly, a second rationale affirms that supporting constitutional processes in CAS to make it inclusive and participatory, with civil society organisations, marginalised groups, and elected representatives included in the drafting process, will deliver inclusive state institutions, which can herald sustainable peace (Castillejo 2014). With this mainstreaming of ICA into peacebuilding, a raft of international guidelines on constitution writing, as well as funding for constitution-related projects have sprung up. In 2006, a high-level committee in the UN endorsed a programme to review its constitution-making assistance, followed by a Guidance Note from the Secretary General on UN Assistance to constitution-making processes in 2009 (revised in 2020). Similarly, international (non-governmental) organisations, like the International Institute for Democracy and Electoral Assistance, the Berghof Foundation, and the Forum of Federations, have supported and issued recommendations on delivering ICA (International Institute for Democracy and Electoral Assistance 2011; United Nations Development Programme 2014).

Such international constitutional assistance has focused on two core aspects: content or design of constitutional provisions especially with regard to institutions, and the process of constitution writing in its endeavour to promote ‘good constitutionalism’ (Samuels 2006). In terms of the content or design, two themes have dominated. First, given the influence of a liberal peace approach in international assistance, constitutions in CAS have become vehicles for the transfer of liberal norms, ideas, and values, based on the understanding that such formal adoption of norms will ensure a shared commitment to such norms (Al-Ali and Dann 2006; Wallis 2014; Widner 2008). A push by ICA providers for ‘normative embedding’ of the constitution in the discourse, framework, and rules of international law is evident (Hay 2014). Increased references to minority rights, guarantees on human rights, and principles of equality reflects this ‘normativisation’ of post-conflict constitutions (Schoiswohl 2006). This is enabled by international law increasingly regulating post-conflict constitutions by setting benchmarks and standards (Bell 2017) but also by post-conflict states seeking to legitimate themselves as being compliant with international law, by encoding such liberal commitments in their constitutions (Grenfell 2016). This normative thrust is testified by the UNDP’s guidance on ICA, which outlines ‘encouraging compliance with international norms and standards’ as one of its six core principles (United Nations Development Programme 2014).

Relatedly, ICA providers have also encouraged CAS to incorporate institutional designs that promote power-sharing mechanisms designed to protect marginalised groups in the constitution in a bid to promote inclusion (Cheeseman 2011). Such institutional designs, usually through political, economic, military, or territorial power sharing – more tangibly in the form of national-unity governments, transition to federalism, quotas for marginalised groups, or proportional representation system in the electoral system (Sisk 2013) – are seen as a practical means of conflict prevention (Williams 2013). Different power-sharing mechanisms have been associated with the durability of negotiated settlements that end wars (Hartzell and Hoddie 2015). This emphasis on inclusive institutional design is underpinned by the rationale that institutional design in divided states can systemically favour or disadvantage ethnic, national, or religious groups. Despite the various critiques of power sharing perpetuating ethnic divisions (Rothchild and Roeder 2005) or being gender-blind (McCulloch 2020), in practice peacebuilding and ICA has tended to promote power sharing as a pathway for inclusive institutional design and conflict resolution. Accordingly, power-sharing institutions related to proportionality, federalism, and the mutual or minority veto have been marketed as levers of conflict management (Lijphart 2004). This preference of international actors for certain institutional models is transferred on the ground in CAS through projects which provide comparative knowledge-based support and expertise to the domestic constitution drafters, technical assistance to domestic groups to further discussions on power sharing, and support for the drafting of electoral laws, among others (Sripati 2008).

ICA has also promoted inclusive and participatory processes during constitution writing. The implicit rationale of ICA being that an inclusive constitution-writing process will breed inclusive institutional outcomes (Castillejo 2017). Multiple other justifications foreground this focus on the participatory process of constitution writing: the ability of participatory processes to mediate differences between conflicting parties, fostering constitutional ownership, enhancing accountability of constitution-makers, securing buy-in from all key social groups (Wallis 2016), and finally their ability to bestow legitimacy on both the process and the content of the constitution (United Nations Development Programme 2014). The faith in inclusive processes is inked into the UN constitutional assistance guidance, which has ‘supporting inclusivity, participation, and transparency’ as one of its six principles (United Nations Development Programme 2014). Accordingly, constitutional support programmes by the UN, as well as varied international non-governmental organisations, such as International IDEA and Interpeace, have included financial and logistic support for civic education, public debates, consultations, and public information and participation campaigns (Sripati 2008).

Through inclusion-related content or provisions and participatory processes, ICA providers seek to change the constitutional settlement in CAS through formal institutions and procedures. The push for inclusive state institutions and participatory process has a long-term impact on the nature of the political settlement or on how elite–society relations are structured in a country. For instance, the promotion of normative benchmarks, such as through the assertion of minority rights, inclusion, human rights, and civil society in the constitution, call for reconfiguring the state–society contract. Norms such as minority rights or civil society strengthening can be seen by elites in CAS as undermining the role of the state and the position of elites who have dominated the state (Lewis, Heathershaw, and Megoran 2018). Similarly, institutional designs promoted and preferred by ICA and peacebuilders more broadly, establish how power will be shared among the competing societal groups, often creating winners and losers. Given their distributional effects, the normative guarantees and institutions preferred by international constitutional assistance providers (Bormann et al. 2019) are not necessarily favoured by domestic elites (Feldman 2005). Likewise, the mantra of ‘inclusive’ constitution-writing processes challenges established power structures. Such inclusive peace processes often enable the voices of marginalised groups to exercise influence on the design of the constitution (Paffenholz 2015). The enabling effect on marginalised groups invariably constrains the role of elites, and thus resistance by powerful elites is likely. Not surprisingly, constitutions in CAS have either never been implemented, or they have been quickly modified or simply been ignored (Eisenstadt and Maboudi 2019), all of which calls for a reappraisal of prioritising constitutional assistance.

Formal institutions and the largely ‘informal’ elite-level politics

The rise of ICA (Tushnet 2008) as a global trend has paradoxically been accompanied by a rare consensus among scholars and practitioners that international assistance focused on supporting formal institutions in CAS can have limited utility in impacting inclusive constitutional orders or peace. Two distinct bodies of scholarship focusing on CAS, namely the local turn in peace studies and the political settlements literature, have come to a similar conclusion (Abboud 2021; Dodge 2021; Khan 2010; Mac Ginty and Richmond 2013; Westendorf 2018)

Multiple writings on the ‘local turn’ in peacebuilding highlight the agency of elites to resist, dilute, or co-opt liberal prescripts and formal institutions supported by the international community (Barma 2017; Barnett, Fang, and Zürcher 2014; Mac Ginty and Richmond 2013). Empirical analysis from multiple contexts reveals many forms of co-optation: formally committing to liberal norms but rarely enforcing them, paying lip service to these liberal norms in internationally sponsored peace agreements while subverting them in practice, formally establishing institutions but rendering them incapable of functioning; and adherence to commitments initially but reneging on or diluting them later. The resurgence of hybrid peace structures which embody liberal values promoted by international actors as well as illiberal traits are also seen to be the result of this capture of formal institutions by elites to suit their interests (Barma 2017; Pouligny 2006). For instance, in Cambodia, the elites superficially cooperated with international actors to institute democratic institutions, but left them as ‘political husks’, and instead subverted these institutions to entrench their clientelist and patronage networks (Roberts 2008). Likewise, in Kosovo, co-option was facilitated through the employment of Kosovo Albanian elites in the United Nations Interim Administration Mission in Kosovo (UNMIK), and their partnership with international actors and agencies, which enabled them to hijack the liberal peace agenda for pluralist and multi-ethnic Kosovo communities, and instead fomented an ‘Albanianisation’ of the state, thus serving the ethnic majority (Franks and Richmond 2008). Similarly, some scholars also assert that the legitimacy and resilience of the many very local and informal institutions, often driven by customary practices, are in competition with formal state institutions, rendering them alien to the societies in these states (Boege 2016).

Another body of work to have come to a similar conclusion is the scholarship on ‘political settlements’, originating from a political-economic perspective. This body of work stresses that formal state institutions need to be compatible with this distribution of power among different social or elite groups to function effectively (Khan 2010). If formal state institutions are out of step with the distribution of power or do not factor in elite interests, they are likely to be undercut by informal institutions like patronage, and religious, regional, and ethnic alliances (Putzel 2012). While definitions of political settlements vary (Behuria, Buur, and Gray 2017), there is an agreement firstly, on the importance of the role of informal institutions, and secondly, on the role of compromises between powerful groups or elites, that can either assist or obstruct the process of formal institutions or institutional reforms (Laws and Leftwich 2014).

Deployed in CAS, both bodies of scholarship caution international actors to assess if the institutions and constitutional reforms they promote are compatible with the underlying distribution of power and interests of elites (Pospisil and Rocha Menocal 2017). Here, prospects of peace are strengthened where the interests of powerful elites, in the form of allocation of benefits, opportunities, and resources (such as political positions and business prospects) are guaranteed (Cheng, Goodhand, and Meehan 2018). When international interventions do not suit their interests, not only can elites resist or co-opt even the most extensive international engagement, but they can also shape and control formal governance institutions, policies, and the distribution of development assistance to advance their interests (Parks and Cole 2010).

Despite the evidence on the centrality of elites and elite agreements, and their reliance on informal institutions in CAS, international peacemakers and peacebuilders do not pay sufficient attention to the functioning of such contexts, and do not understand the local ‘vernacular politics’ (De Waal 2009). This inability to understand is due to a disconnect from, and disengagement with everyday realities in conflict zones (Mac Ginty and Richmond 2013). It is also attributed to international actors following ‘an implicit hierarchy in which the formal, legal–institutional approach is superior to locally specific ways of conducting politics’ (De Waal 2009, 101). But as they focus on crafting such ‘perfect’ formal institutions, such as peace agreements and constitutions, based on the assumption that these will be binding, the less likely they are to be invested in understanding the informal mechanism through which national and local politics operate. This is further imperilled by the fact that international experts and donors often lack sufficient understanding of the local context, and its ways of ‘doing’ politics (Autesserre 2009). This absence of contextual understanding of local politics – centred on elite agreements through formal, but notably also informal mechanisms – facilitates the co-option and capture of international peace support by elites (Barma 2017). This article takes such insights on agency of elites in CAS to co-opt international engagement further to draw out four informal strategies that elites use to subvert formal institutions, and how these strategies relate to the underlying rationale of ICA.

Understanding ‘constitution’ as a formal institution – the case of Nepal

If the findings from critical peace studies and political settlements scholarship, discussed above, are applied to constitutional assistance, a core inference is that ICA’s prioritisation of formal aspects – inclusive institutional design and participatory process – leads it to overlook the informal politics through which elites formally commit to inclusion but informally subvert the process. This ‘formal adoption–informal subversion’ of inclusion by elites is evidenced in Nepal.

Drafting a new constitution

Since the start of the peace process in 2005, which sought to resolve the decade-long insurgency waged by Maoist rebels, the writing of a new constitution was a key priority in Nepal. A new constitution written by the Constituent Assembly (CA) was a central demand of the Maoists, who stated that their call to arms was caused by the state-sanctioned exclusion of many marginalised groups. A constitution written by ‘people’s representatives’ as opposed to previous experience where it was written by experts and political elites recommended by the King (Tamang 2011), was seen to be the hallmark of democratic legitimacy, promoting public participation, ensuring representation of excluded groups, and providing a pathway for engendering inclusive state institutions (Bhandari 2012). Accordingly, the 12-point agreement between the Maoists and non-Maoist democratic parties, which catalysed the Comprehensive Peace Agreement (CPA), agreed to elect a Constituent Assembly (Martin 2012). By then, Nepal had five prior constitutions.

The new constitution born of the peace process was to provide an institutional frame to the CPA’s core commitment of ‘restructuring the state’ to make it inclusive (Bell et al. 2017). Reflecting the spirit of the CPA, the Interim Constitution of 2007 promulgated to serve the transitional period committed to: a transition from a monarchy to a republican order, from an erstwhile Hindu state to a secular one, in addition to a commitment to human rights, as well as notions of respect for civilian supremacy (Jha 2014). Notably, to foster inclusion, the Interim Constitution outlined several paths, including: affirmative action (through proportional representation of marginalised groups in all organs of the state), changes in the electoral system from first past the post (FPTP) to a mixed system (with 40% of representatives elected from FPTP, 56% elected through proportional representation (PR), and a remaining 45 nominated by the cabinet), the transition from a Hindu state to a secular state (a long-standing demand of non-Hindu Janajati groups), constitutional recognition of all languages spoken as mother tongues, gender-friendly citizenship policies, and federalism (Lawoti 2016). Federalism was added to the Interim Constitution after protests from marginalised groups, including Madhesis from the southern plains of Nepal adjoining India, and Janajatis, or indigenous nationalities, who comprise 36.81% of the population and face religious and linguistic inequalities (Bell et al. 2017).

The Interim Constitution of 2007 also sketched provisions for electing CA members to prepare for the new constitution. ICA providers seized this domestic momentum created by the Maoist movement to craft an inclusive constitution (Jha 2014). In 2008, with the consent of the Government of Nepal, a donor consortium was formed under the auspices of the United Nations Development Program (UNDP) to support, coordinate, and facilitate participatory constitution making. Donors, including the UK’s Department for International Development (DFID), the Danish International Development Agency (DANIDA), the Norwegian Embassy, and the United States Agency for International Development (USAID) supported multiple projects for providing technical assistance to the CA members, strengthening parliamentary committees, expertise sharing for constitution drafting, and conducting civic awareness campaigns to enable citizen participation in the process of constitution drafting (International IDEA 2015). Programmes such as the Support to Participatory Constitution Building in Nepal (SPCBN) project were launched, which singularly saw an investment of $20.5 million between 2008 and 2015.

With the CA elections in 2008, a CA was formed, tasked with drafting the permanent constitution within two years. The CA also functioned as a parliament. However, between 2008 and 2015, there were four postponements to the CA’s tenure, along with elections to institute two different CAs (with CAI functioning between 2008 and 2012, and CAII between 2013 and 2015). In this political process, the primary agendas of the CPA and the peace process, such as inclusion (notably federalism), were all hard-fought.

Nature of international constitutional assistance

In line with the global advocacy, ICA providers in Nepal pitched a constitution that was written through a participatory process as one that would enjoy legitimacy, breed inclusive state institutions, and institutionalise liberal values of inclusion, human rights, and the rule of law (Clark 2011). This also suited the domestic context, where the entire peace process was centred on addressing historic exclusion. As noted earlier, despite being an multi-ethnic and multi-linguistic state, a narrow elite base comprised of Nepali-speaking Hindu and High Caste communities from the hills, had dominated all elements of the state, and all political parties, as well as civil society. This dominance of CHHE has historically fostered exclusion, based on multiple schisms of language, regional difference, ethnicity, religion, caste, and gender (Lawoti 2008). Of the many marginalised communities, prominent groups, which have mobilised socially and politically during the peace process, include Dalits, Janjatis, and Madheshis (Lawoti 2016).²

This process of exclusion has historically had a profound economic impact (Sharma 2021). The ritualistic Hindu order, fostered by the Shah monarchs who ruled Nepal until 2006, privileged the CHHE with access to state resources and power, thus marginalising others (Whelpton 1997). With the state promoting Nepali as the official language and as a requirement for state-sector jobs, non-Nepali-speaking groups like the Janajatis and Madhesis found themselves excluded from employment opportunities (Dong 2016). Additionally, state expenditure in this period was much lower in the Terai region than in the rest of the country, despite major industrial centres being sited there and the region accounting for two-thirds of national industrial output (Deysarkar 2015). A highly centralised state system and deep-seated patrimonialism furthered economic marginalisation even after the democratisation process in the 1990s by posing continued challenges for marginalised groups seeking employment in the civil service or access to public office (Bhattachan 2013). As a result, Janajatis, Dalits, and Madhesis continue to lag behind in lag on indicators of human development index (Gurung 2010).

In the process of supporting inclusive content or design, ICA providers offered technical advice on drafting the constitution, including: undertaking review and analysis of draft bills, facilitating dialogue among key political actors to help them reach a consensus on contentious issues, sharing comparative perspectives from other countries on constitutional designs, notably on federalism, and preparing for the transition to federalism (International IDEA 2015) – for instance, technical assistance to Nepal’s Election Commission on the electoral law which would ensure fair and equitable representation of marginalised groups. The very formula that Nepal adopted with respect to a mixed electoral system, encompassing FPTP and PR, and quota allocations for marginalised ethnic groups and castes, women, and people from backward regions, was based on a formula devised by international experts (Pokharel 2012; Slavu 2012).

To ensure the constitution-writing process was inclusive, international constitutional assistance providers supported processes within and outside the CA. Within the CA, elected in 2008, while there were no formal provisions for cross-party caucuses, international donors supported and generously funded the three caucuses in the CA: the women’s caucus (33% of CA members), the Dalit caucus, and the Janajati caucus (Malagodi 2019). These different caucuses were comprised of CA members from all parties, and aimed at promoting the interests of the respective groups, galvanising inclusion-related demands (Grävingholt et al. 2013). For instance, the Women’s Caucus, formed of 197 members from 19 political parties, was able to raise significant issues, including women’s right to inheritance, the right to proportional inclusive representation based on population, and the right to equality in citizenship regardless of gender (International IDEA 2015). Similarly, the Janajati caucus demanded ethnicity-based federalism, which was contrary to the stance of the several main parties, such as Nepali Congress (NC), (Bhattarai 2014) and the Communist Party of Nepal–Unified Marxist Leninist (CPN-UML), and also sought collective rights for indigenous people and their languages, in addition to affirmative rights (Khanal 2014). To strengthen these caucuses, international actors provided technical assistance and expert advice, study tours and workshops, and, further, training on such complex issues as federalism or the form of government. For example, International IDEA, UNDP, and DFID supported marginalised groups in the CA by providing technical expertise and skills training (International IDEA 2015), and sensitisation and awareness-raising sessions for political leaders and CA members on the issues of women, Dalits, indigenous peoples, Madhesis, and other excluded groups (UNDP n.d.). They also facilitated cross-party consensus on complex constitutional issues(Clark 2011).

Outside the CA, to ensure wider societal participation, many donor-funded NGOs supported civic outreach and participation through public education activities on the content of the constitution, as well as public opinion, among others (Breen 2018). The CA rules made participation of the people mandatory by allocating time for public consultations on the draft, and requiring that relevant suggestions from the consultations should be accommodated in the constitution (Khanal 2014) To this end, ICA supported such projects as the Support to the Participatory Constitution Building in Nepal Programme, which singularly reached out to 240 constituencies and 1,576 Village Development Committees and promoted debate on key constitutional issues through civil society groups (UNDP n.d.). The Centre for Constitutional Dialogue, supported by UNDP, saw 500,000 citizens participating in ‘democracy dialogues’, which highlights the efforts to promote wider societal participation in the constitution-writing process (Clark 2011)

Elite interests and ‘informal’ agreements undercutting the ‘formal’ constitution-writing process

The focus on ‘inclusion’ in the constitution-writing process and content was pursued in a bid to change the political settlements in Nepal. Inclusion in its content and process, by seeking to prioritise Dalits, Madhesis, and Janajatis, sought to devolve power away from the CHHE, and thus undermine their dominance. In the beginning of the peace process, all political parties, including traditional political parties like the Nepali Congress (Bhattarai 2014) and the Communist Party of Nepal–Unified Marxist Leninist, dominated by the CHHE, accepted demands for inclusion for pragmatic reasons rather than normative ones. Such pragmatism was due to several factors: responding to demands by civil society groups that were calling for greater inclusion, seeking international legitimacy, and the need for these traditional democratic parties to give in to the Maoists’ demand for inclusion in a bid to oust the monarchy (Jeffery 2017). But there were significant divides on issues of inclusion. On the one hand, there were traditional political parties, which had committed to federalism to stop the conflict, but were not keen for a systemic overhaul to accommodate marginalised groups, often described as ‘reluctant federalists’ (Bogati et al. 2017). On the other hand, there were the Maoists, Janajatis, and Madheshis, who despite their differences, saw identity-based federalism as the primary pathway to address exclusion and inequalities (Adhikari and Gellner 2016). A division thus emerged between the reluctant federalists, who believed that provinces must be defined based on economic viability, and the others, who sought to prioritise identity-based division of the country. Further, after losing the 2012 CA elections, the Maoists reassessed their position on federalism. They believed their failure was due to an overwhelming focus on ethnicity and exclusion without due consideration for poor people within the CHHE category (Adhikari and Gellner 2016), leading them to veer away from the idea of identity-based federalism, thus further weakening the momentum.

So, while political parties welcomed and participated in ICA-related programmes as well as formally embedded the inclusion-related content in the Interim Constitution, as the process drew to a close in 2015, political leaders, comprising largely of the CHHE elites, backtracked on inclusion-related pledges. To preserve their dominance and interests, the major political parties, dominated by CHHE leaders, crafted formal but notably also informal agreements between themselves, sometimes co-opting some Madhesi parties in the process, all of which undercut the formally agreed provisions and discussions of the CA (Jha 2017). In this process of ‘formal adoption–informal subversion’, Nepali elites employed four key strategies that weakened both the process and discussion on the design of the constitution, probing the very rationale on which ICA is grounded.

Neglect

While ICA providers prioritised the constitution-writing process, Nepali political elites largely neglected the deliberation process at the CA, or the discussion on constitutional provisions outside the CA. This created a divide whereby international constitutional assistance providers worked within the formal domain supporting the drafting of an inclusive constitution written through a participatory process, while the political elites were rarely invested in the CA proceedings. For instance, senior leaders from the main political parties were largely absent from deliberations with the CA, or the public deliberations on the constitution outside the CA. Even a cursory glance at the attendance of the CA highlights that, on average, over a third of CA members were regularly absent, and prominent senior leaders of the main political parties rarely attended. For instance, former Prime Minister Girija Prasad Koirala, the chief architect of the peace process, attended the CA proceedings only three times (Martin Chautari 2010). Despite an elaborate institutional framework, with varied thematic committees, standards, and procedures, senior party members were barely even aware of the many choices made at committee level by CA backbenchers (Adhikari 2014). Similarly, many senior leaders from major political parties did not take part in the public opinion drive which saw millions of dollars invested by ICA with an aim to make the constitution-writing process participatory, and hence inclusive (Martin Chautari 2009)

In turn, the political parties were occupied with informal political negotiations and bargains on government formation. The centrality of elite pacts centred on government formation rather than the CA also ensured that the ideological ambition, of whichever coalition was in power, dictated the fate of such commitments as inclusion. The changes in the debate on federalism confirm this. A pro-federalist stance was adopted after the Maoist-Madheshi win in the 2008 elections. This pro-inclusion alignment suffered when some Madheshi parties left the coalition, resulting in its collapse in 2009. Further, the political momentum shifted to a largely ‘anti-federalist’ mode when Nepali Congress and the UML won the elections in 2013, reversing the electoral outcomes of 2008 (Jha 2014).

Similarly, public views and suggestions collected as a part of the constitutional outreach campaigns funded by ICA were barely incorporated in the final constitution. For instance, it was reported that ‘as many as 184,674 people had participated in public feedback collection, and mostly favoured a directly elected executive, religious freedom, the demarcation of federal provinces before the constitution, and citizenship either in the name of father or mother’ (Nepali Times 2015). Yet the content of the constitution of 2015 does not entirely reflect, or even contradicts, these aspirations on issues like gender rights (Malagodi 2018). Even a scholar like Breen, who has highlighted that public feedback on aspects like federalism was reflected in the final draft of the constitution, does note that this process was limited, as ‘key decisions were made by the elite in another place and often through bargaining, rather than deliberation’ (2018, 420). He further notes how the public feedback process became side-lined as contentious issues, especially with regard to inclusion, came to the fore. After the formation of the new Constituent Assembly post-2012 elections such formalised mechanisms to elicit feedback were discontinued (Breen 2018).

Overwrite

Many informal agreements between political elites of the major parties contravened and overruled the discussions in the CA. As a formal avenue of constitutional deliberation, the CA was the focus of ICA’s support. This subverted the status of the CA as the deliberative forum on constitution writing. For instance, despite formal bodies like the Restructuring of the State and Distribution of State Power Committee being formed to recommend principles to determine the delineation of federal boundaries, its structure, and the distribution of power between levels of government (Constituent Assembly-Nepal 2010), the final decision on federalism did not come from the CA, or the many campaigns by different identity groups. Rather, a 16-point elite pact between the leaders of the four largest political parties, namely the Nepali Congress, the Communist Party of Nepal–Unified Marxist Leninist), the Maoists, and one Madheshi party (Madhesi Janadhikar Forum-Loktantric) agreed on all contested issues of the constitution including federalism, the form of judiciary, and the form of government, which was then adopted by the constitution of 2015. Further, this public consultation was ‘fast-tracked’, giving only two weeks for citizens to discuss provisions of the draft constitution.

The 16-point pact and eventually also the constitution of 2015 diluted all the inclusion-related provisions enshrined in the Interim Constitution of 2007 (Jha 2017). For instance, based on the provisions of the Interim Constitution, the Constituent Assembly in 2008 was elected through a mixed system, with 56% by proportional representation, 40% by first past the post , and a final 26 members being nominated by political parties (International Crisis Group 2008). Political parties were required to ‘ensure proportional representation of the women, Dalit, oppressed communities/indigenous peoples, backward regions, Madhesi and other classes’ as they nominated candidates (Bell et al. 2017). This focus on inclusion in the electoral system led the CA elected in 2008 to emerge as the most diverse state institution in the history of Nepal. Women made up 33% of the seats, Janajatis 36%, the Madheshis 23%, and Dalits 8% (Dudouet and Lundström 2016). Despite its immediate impact on inclusion of marginalised groups in state organs, the constitution of 2015 reversed these provisions with 40% representation from PR and 60% by FPTP (Rai 2018). Likewise, affirmative action in the Interim Constitution had embraced policies such as 33% female representation in all organs of state, as well as the allocation of 45% of civil service seats for marginalised groups. The 2015 Constitution, which arose from the 16-point agreement, weakened this provision by adding the already over-represented category of ‘Khas Arya’, or CHHE, as one of the ‘communities’ entitled to reservation. Similarly, while the new constitution declared Nepal a secular state, it defined secularism through an explanatory clause as ‘protection of religion and culture being practised since ancient times and religious and cultural freedom’, which, given Nepal’s history of being a Hindu state, facilitates its protection but is also interpreted as excluding other religions like Christianity and Islam (Gellner and Letizia 2019). Further, in line with the 16-point agreement, the 2015 Constitution did affirm Nepal’s transition to a federal state with seven provinces (the agreement had eight) but only one of the provinces was created in such a way that it could have a non-CHHE majority: Province 2, where Madheshis are dominant (Jha 2017). This gerrymandering of territory – where the ethnic composition of the provinces did not match the electoral constituencies – left aspirations for identity-based federalism unfulfilled. Not surprisingly, despite the fact that the constitution was ratified with 90% approval in the CA, the contention over demarcation of federal provinces saw its promulgation give rise to violent protests, during which over 40 people in the southern plains were killed (Phuyal 2015).

Likewise, the formula that a participatory constitution-writing process would elicit inclusive content in constitutional provisions was unsuccessful in Nepal. For one, despite the electoral reform and the rather diverse CA being elected in 2008, the diversity of physical representation did not translate into substantive representation (Tamang 2011). Instead, larger parties, dominated by CHHE leaders, continued to dominate the CA in terms of an allocation of speaking time, seat allocations, and representation and leadership in prestigious committees within the CA (Martin Chautari 2009). Likewise, in order to weaken the influence of the cross-party caucuses of women, Janajatis, and Dalits, major political parties used party whips to get their respective CA members to vote along party lines, often on policies detrimental to broad-based inclusion (Tamang 2017). So much so that when the first CA failed to agree on a constitution, the second CA elected in 2013 banned caucuses (Malagodi 2019).

Trade-off

In yet another strategy, the formal agreements already made on constitutional issues were traded off for informal agreements between political elites, centred on which political party would form the next government (Thapa 2015). This rendered many of the areas supported by ICA, including technical assistance and capacity building on governance structures, and human rights compliance to succumb to informal deals between elites. Political leaders focused on formal and informal agreements between a small coterie of political leaders from the main political parties, where agreements on key constitutional issues were bartered for their support for which party leads the government, ministerial positions, and ambassadorships. While political leaders sold these largely informal agreements to the public as generating cross-party ‘consensus’ on constitutional issues, they actually focused on political coalition making, and often did so at the cost of constitution writing.

In the four-year period of the CAI, there were four coalition governments, two of which were headed by the Maoist and two by Communist Party of Nepal–Unified Marxist Leninist. This diluted the focus of party leaders, as they put their political and private interests over constitution writing (Miklian 2009). For instance, in May 2010, when the CA’s initial two-year term would end, and the only way to extend it was for two-thirds of the CA to support it, the Maoists stated that they would only support the extension if the incumbent Madhav Nepal from the CPN-UML resigned, and made way for a Maoist-led government (Jha 2014). Similarly, in 2011, as another deadline for the CA loomed, the NC asked for the CPN-UML-led Jhananath Khanal’s resignation before agreeing to extend the CA deadline (The Nepali Times 2011). In August 2011, as the Khanal government resigned, a four-point agreement was hammered out between an alliance of the Madhesi political parties, the United Democratic Madhesi Front (UDMF), and the Maoists, which elected the Maoist leader, Dr Baburam Bhattarai, as Prime Minister. While the four-point agreement pledged to complete the peace process, and integrate and rehabilitate former combatants, it also agreed to withdraw criminal cases against individuals affiliated with the Maoist party, the Madhesi, Janajati, Tharuhat, and Dalit movements, and to declare a general amnesty, which could include serious crimes and human rights abuses (Human Rights Watch 2011).

Exploit

Formal rules of the CA that the ICA providers had so firmly promoted to foster inclusion were manoeuvred informally to suit political elites. This exploitation was evident most prominently in proportional representation in election laws. In the 2008 election, a number of senior party leaders from traditional parties, Nepali Congress and the Communist Party of Nepal–Unified Marxist Leninist, lost their seats, in FPTP constituencies. However, exploiting the electoral law, many, including former Prime Ministers Sushil Koirala and Madhav Kumar Nepal, managed to secure seats in the CAI as part of the 26 cabinet-appointed posts (Malagodi 2019). While a PR system was supposed to give marginalised Nepalis a say in the state, political parties have tended to tap into businessmen close to, and funding the political apparatus, as representatives in the Parliament through the PR system (Kharel 2021)

Further, elites also exploited the support of ICA on inclusion, by projecting ‘inclusion’ as a foreign-funded project rather than a movement driven by Janajatis, Madhesis, and Dalits for decades. ICA became a scapegoat, spurring a backlash against donors and the very idea of ‘inclusion’ as the peace process was drawing to a close (Drucza 2017). ICA and the peacebuilding community in general were asked to refrain from commenting on contentious domestic issues and to reduce their support for ‘political’ issues like inclusion (Drucza 2017). In fact, in 2018, Nepal decided not accept any funds from donors for legislative processes, which were key sources of international support during the peace process (Ghimire 2018) Amid the elite backlash, peacebuilders diluted their support for marginalised groups and made an abrupt reversal of policy on their promotion of the liberal peace. Due to the backlash, DFID withdrew its funding for the Nepal Federation of Indigenous Nationalities (NEFIN) to work on Janajati rights. Further, even the constitutional provisions for inclusion prompted unintended consequences. Policies on inclusion, pushed for by ICA and donors in general, generated the feeling amongst elites that marginalised groups would benefit immediately and at their own expense, while the actual transformation would take generations, if it was implemented at all (Rai and Shneiderman 2019).

The four strategies used by elites, namely neglect, overwrite, trade-off, and exploiting of formal commitments, helped reinforce their hold on power. Such strategies also enfeebled the formal institutional design and processes of constitution writing that was core to ICA.

Conclusion

Articulating a framework of ‘formal adoption–informal subversion’, this article unpacks how elites in CAS formally adapt institutions that seek to foster inclusion to obtain international legitimacy and pacify domestic demands but informally subvert these formal institutions to prevent them from undermining their hold on power. ICA, with its focus on the formal domain – the content and process of writing an inclusive constitution – falls prey to such informal elite politics. ICA fails because its focus on the formal domain makes it unable to influence the informal strategies that elites use to subvert the formal constitutional commitments to inclusion in which it so heavily invests. Charting out four informal strategies of neglect, overwrite, trade-off, and exploit, this article also questions the core mantra that guides ICA: championing inclusive constitution-making processes and incorporating protections for marginalised groups in the institutional design will foster an inclusive post-conflict state and address exclusion-related drivers of conflict. With an in-depth discussion of Nepal’s constitution-making process, the article established how political elites in Nepal have employed strategies to renounce commitments for inclusion and even breached international legal standards on citizenship and women’s rights (Malagodi 2018), despite substantive ICA to promote inclusive institutional design and participatory processes.

The inferences from this article build on and supplement three distinct bodies of work in peace studies. First, while the article builds on the extant work on international constitutional assistance, it also calls into question two of its central assumptions. One, while scholars like Widner and Wallis’s have advocated for a participatory approach to the constitution-writing process (Wallis 2014; Widner 2008), Nepal demonstrates that well-instituted participation mechanisms can be co-opted by elites, and participatory processes will not always lead to inclusive constitutional provisions. The inability of inclusive processes to effect inclusive constitutional outcomes can, instead, raise expectations and cause marginalised groups to resent ICA for not being able to deliver on inclusion. The ineffectiveness of such participatory processes can be a cautionary tale for other peacebuilding projects where international actors have advocated for amplifying local participation. Two, the Nepali case also questions the core rationale of ICA that participatory processes and inclusive institutional design will eventually deliver on inclusion of marginalised groups. Rather, as this article highlights, even when inclusive institutions are guaranteed in interim constitutions or peace agreements, they can be rolled back when elites resist such shifts – evidencing the fragility and limitations of formal legal instruments.

Second, the article calls upon scholars and practitioners to embed a political-economic approach to international constitutional assistance. It does so by bringing the literature on political settlements and elite agency into conversation with the scholarship on ICA. It highlights that when formal constitutional arrangements – focused on the inclusion of marginalised groups – threaten to decentre the power away from elites, then elites can neglect, trade-off, exploit, or overwrite these formal provisions. It thus cautions that unless ICA are attuned to understanding informal political institutions and agreements, any success in the formal adoption of inclusion-related norms in post-conflict constitutions is likely to be temporary at best. The case of Nepal also outlines the inherent limitations of ICA to understand informal elite dynamics as they focus on formal channels, yet the elites in CAS are not invested in these formal processes and rather focus on informal political deals to secure their power.

Third, the article also supplements the extant scholarship on the agency of elites in CAS by discussing how elites co-opt and instrumentalise international assistance in their favour. It extends the analysis to argue that such co-option by elites can undermine faith in constitutionalism. The Nepal case empirically reveals that the political elites might be only superficially invested in formal constitutional deliberations, and that formal adoption of inclusion does not equal an internalised commitment to it. Indeed, the article argues that the very modality of international constitutional assistance can enable elites to ‘game’ or co-opt ICA, given its preoccupation with the formal institutional design and processes of participation.

While inferences from this article question the effectiveness of ICA, it does highlight the role of such assistance providers in advocating precepts of inclusion, human rights, and international law. Relatedly, the article also does not seek to dispute or trivialise the gains that Nepal saw during the peace process on human rights, inclusion, and democratic consolidation. It outlines how the initial ambition of a ‘New Nepal’ and ‘radical restructuring of the state’ embedded within the Comprehensive Peace Agreement were reduced in scope over time. In doing so, it calls for balancing advocacy for inclusive constitutions with understanding elite interests and approaching the terrain with a realistic assessment. As international constitutional support widens and is mainstreamed into different organisational and peacebuilding mandates like that of the UN, all these inferences are likely to not only fill academic gaps but also ensure policy traction.

Acknowledgements

I would like to thank Kate Weir, Dr. Sanja Badanjak, Dr. Jeevan Sharma, and the reviewers for their close reading and comments to this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This research is supported by the Peace and Conflict Resolution Evidence Platform (PeaceRep), funded by the UK Foreign, Commonwealth & Development Office (FCDO) for the benefit of developing countries. The information and views set out in this publication are those of the authors. Nothing herein constitutes the views of FCDO. Any use of this work should acknowledge the authors and the Peace and Conflict Resolution Evidence Platform.

Notes on contributors

Monalisa Adhikari

Monalisa Adhikari is Lecturer in International Politics at the University of Stirling. She is a Co-Investigator of the Peace and Conflict Resolution Evidence Platform (PeaceREP) project at the University of Edinburgh. Prior to coming to Stirling, she worked at the University of Keele. She currently serves as Associate Editor for the Journal of Intervention and Statebuilding. Her research interests include peace processes, international interventions, and conflict management efforts by Asian states including India, China, and Japan.

Notes

1 The list includes conflict-affected countries that experienced civil wars, like Nepal, Bougainville (Papua New Guinea), Rwanda, South Africa, and Zimbabwe, as well as others like Kenya and Zimbabwe, where constitutional settlements emerged to address post-election violence and crisis. This article focuses on the former.

2 Dalits, designated as untouchables according to Hindu norms, live across the hills and the southern plains and have been most marginalised culturally, socially, and economically. The Janajatis, or indigenous nationalities, speak non-Nepali languages, practise a variety of religions, and are marginalised on grounds on language and religion. Lastly, the Madheshis, who reside across the southern plains (Terai region) adjoining India, share deep cultural ties with people in Northern India, and, as noted above, have been excluded on grounds on race, region, language, and citizenship.