Land Governance in Post-Conflict Settings : Interrogating Decision-Making by International Actors

Humanitarian and development organizations working in conflict-affected settings have a particular responsibility to do no harm and contribute to the wellbeing of the population without bias. The highly complex, politicized realities of work in conflictand post-conflict settings often require quick, pragmatic and results-oriented decisions, the foundations of which remain frequently implicit. Such decisions might follow an intrinsic logic or situational pragmatism rather than intensive deliberation. This paper reflects on the realities of working on land governance in post-conflict settings shaped by migration, ethnic division, power struggles and limited statehood. Using case examples from the Democratic Republic of Congo (DRC) and Burundi, this paper reflects on the drivers of decisions around land governance in such contexts in a structured, theoretically informed way. Drawing on the author’s own experience with supporting land rights work and utilizing Giddens’ concept of the Duality of Structure, this article provides an analysis of actors and structures that sheds light on the factors that affect the decision-making of practitioners relating to land rights in post-conflict areas of limited statehood.


Introduction
In the African Great Lakes Region, land, conflict and displacement are connected in a multi-faceted relationship.Questions of identity and belonging add an additional layer of complexity [1,2].In the region, violent conflict causes displacement, while displacement also constitutes a driver of conflict.This is the case where displaced people migrate to areas where resources are scarce and tenure rights are insecure.In other instances, the return of people to their places of origin causes conflicts with the current occupants [3].Limited capacities of state actors or competing (power) interests frequently exacerbate such conflicts.NGOs operating in these complex environments often take operative decisions with far-reaching consequences.Such decisions relate to questions regarding if and how to engage with land governance, how to address land conflicts, in what way to deal with tensions between statutory and customary land management systems, as well as how to engage with opposing actors.
While there is increasing global attention for land rights issues 1 , underlining the importance of the factor land for development and peace, there is little research on NGO decision-making that explains why, when and how specific issues are engaged with [5] and possibly even less explicit reflection by NGOs themselves on this matter.This article provides a practitioner's perspective on the different 1 The adoption of the Voluntary Guidelines on the Responsible Governance of Tenure (VGGT) in 2012 was a milestone in this regard.Since then, various actors work towards the implementation of the VGGT [4].
factors that influence decision-making based on experience with land rights work in the African Great Lakes region.

Post-Conflict (Land) Governance
In Burundi and the Eastern Democratic Republic of Congo (DRC), tenure insecurity, violent conflict, and displacement present an explosive mixture.Numerous actors have started to address land and conflict issues, resulting in a variety of interventions and approaches that represent a land governance patchwork.It is widely accepted that governance is supposed to contribute to certain standards in human rights, democracy, the rule of law, the provision of common goods and others [6].Land governance is no exception to this. 2  In post-conflict settings, the capacities of the state are limited, as is the case in the DRC and Burundi.Furthermore, any existing documentation of legal rights to land is likely to be incomplete.Adequate policies and legislation form part of the basis for the creation of a functioning state.Unfortunately, land governance is generally not receiving sufficient attention in post-conflict settings [7].In most sub-Saharan countries, there has never been a comprehensive documentation (let alone registration) of land rights.In the DRC and Burundi (as well as many other countries), there is an ongoing tension between statutory and customary land management systems [8].Particularly in the DRC, the legal system in itself is incoherent regarding the existing rules and regulations on land tenure registration and management [9].This results in an uncertain relationship between customary and statutory tenure systems.Moreover, land represents not simply an economic asset or a place of residence but a source of identity, power, and social status.
It is widely assumed that resolving land rights issues is fundamental for sustainable peace, economic-and food security as well as state-building [7].There is also increasing agreement that securing land and other property rights needs to be at the center of global efforts towards sustainable development [4].A range of global standards has been developed that are intended to deal with questions around land governance and the rights of displaced or returning people to access land.Prominent among these are the Voluntary Guidelines on the Responsible Governance of Tenure of Land, Fisheries and Forests in the Context of National Food Security (VGGT), the Pinheiro Principles on Housing and Property Restitution for Refugees and Displaced Persons (Pinheiro Principles), and the African Union Framework and Guidelines on Land Policy in Africa.While these standards are oriented towards state responsibility, inclusive and participatory practices and intend to strengthen the rights of marginalized and vulnerable groups, their generic nature leaves implementers with serious challenges in adapting these standards to specific contexts.There is little concrete guidance for actors involved in land governance in areas where the state is largely absent.
This paper reflects on the day-to-day realities of working on land governance in post-conflict settings that are affected by different types of migration, identity-based divisions, power struggles, and symptoms of limited statehood.In fluctuant, changing contexts where violent conflict frequently occurs and state functions are limited; practitioners need to adhere to high standards without becoming paralyzed by uncertainty in the face of complexity.Using concrete case examples, this paper reflects on the drivers of practical decisions taken by NGOs.Drawing on the author's own experience with supporting land rights work in the region and utilizing Anthony Giddens' concept of the Duality of Structure [10], an analysis of actors and structures is provided that sheds light on central factors that affect NGOs' decision-making and the success of multi-actor intervention strategies.

2
The governance definition used here deviates from the common understanding of the term that assumes a modern state to fulfill governance functions, usually on the national level.Instead, I follow an understanding of governance that accounts for the context of limited statehood found in post-conflict settings where governance is much more decentralized and a diverse set of actors fulfills governance functions [6].

Materials and Methods
The data used for this paper derives from a range of sources.A primary resource is the author's own involvement in the projects and issues described.Through various field visits over the past years insights into the decision-making relating to land rights work were gained.A second source of information was a process to develop land rights guidelines that inform practical work in post-conflict contexts.This process stretched over a total of 12 months and involved extensive discussions with peers, partners, and colleagues including three feedback workshops.Much of the information gathered exists in the form of written feedback, notes and e-mails.It affords valuable insights into the reasoning and justification behind practical and programmatic decisions.The strongest influence on the conclusions nevertheless comes from personal experience, which is involved in the issues described here.This certainly makes the conclusions subjective.However, by freely admitting the influences that inspire them and by using an analytical framework that allows for probing of conclusions using different types of data, I enable a critical discussion of the findings and the analytical framework.Ideally, this will enrich both the academic and the practical debate around decision-making and land rights work in the non-governmental sector.

Structuration Theory
The analytical framework used for this paper is based on Anthony Gidden's Theory of Structuration [10].This theoretical approach provides a basis for analyzing the complex relationship between actors and structures.Giddens proposed a recursive, dualistic relationship between action and structure.The basic notion behind this is that existing structures, consisting of norms and rules limit and enable action, while action can change the existing structures over time.Giddens called this principle the Duality of Structure.He used the term institutions to refer to norms that have acquired the status of rules through their widespread application over time (the process of institutionalization).Long-term engagement of actors within a certain context can thus lead to the purposeful or respectively involuntary changing of the structures that shape their own and others' actions.To give a concrete example: An NGO working towards an improvement of the legal and social status of women's land rights in a particular context will be bound by the legal norms that define these rights, as well as to some degree by the cultural norms that shape social relationships.Simultaneously, the presence and actions of the organization are likely to change these factors in the long term, for example if effective lobbying leads to legal changes or if certain cultural restrictions on women's inheritance of land change through discursive engagement.The existing rules and norms will also be likely to influence the methods chosen by the NGO engaging in this context and possibly also the concrete objectives (which outcomes are seen as feasible).The analytical challenge lies in explaining the effects of actors and structures on each other.Useful analyses at a more general level become almost impossible if we concentrate on the level of individual decision-making.We would have to assume that in every situation and every context the decisions taken depend on the individual taking them, making not only predictions and generalizations impossible but also preventing useful comparisons across different actors and contexts.The argument made here is that we can find explanations for decision-making beyond the individual level, which are useful for organizational self-reflection as well as academic analyses.An analysis based on this assumption requires a concept of collective actors or rather composite actors within which individuals act on the basis of some form of collective identity.NGOs constitute such composite actors, composed of a number of individuals who act towards a collective purpose, based on shared norms and rules and bound to decision-making structures that eliminate individual preferences at least to a certain degree.While NGOs as composite actors cannot be assumed to be fully rational actors in the sense of a profit-maximizing rational actor paradigm [11], they can be assumed to act rationally within the limits of the information they have and the outcomes they expect to result from their actions.This is because they have clear constraints in terms of rules and clear normative orientations according to their mandate and identity.In the context of third-party financed projects, they are also bound by deliverables and pre-defined frameworks for action.NGOs act purposefully, bounded by norms and rules.Their own actions, and in particular their interactions with other actors, have an impact on the structures within which they act.

Actor-Centered Institutionalism
For analytical purposes, this understanding of action and structure is too vague; Giddens never fully specified the concepts used in Structuration Theory towards a practicable analytical framework.An attempt towards such a modification was made by Renate Mayntz and Fritz Scharpf in their analytical framework Actor-Centered Institutionalism [12].Mayntz and Scharpf propose that, at the level of the composite actor, it will be possible to make generalized assumptions about actors' preferences based on the analytical concepts of self-interest, identity, prevailing normative frameworks, and explicit rules.Through this, one can explain the actions of different actors as the result of interdependent choices of a plurality of actors that have specific assumptions regarding the results that can be obtained in a specific setting [13] (p. 69).This means that structures will affect an actor's actions in the form of institutional settings consisting of institutionalized norms and explicit rules.At the same time, an actor's preferences will further determine which actions are feasible to them.This is not only determined by self-interest and identity, but also by assumptions about the reactions that other actors will show towards specific actions.This means, e.g., that social norms constrain actors because a violation of such norms may result in social disapproval, withdrawal of cooperation or even social exclusion [13] (p. 38).However, this constraint only comes into effect if the actor is conscious of the risks and unwilling to accept the consequences.
Actor preferences are understood to consist of four different components: interests, norms, identities, and interaction orientations.Interest can be deducted from the purpose of a composite actor; in other words: what is the organization's self-interest based on its nature and purpose?Norms drive decisions and can be the purpose of action (achieving norm adherence or establishment of a norm).The role of norms in action relates strongly to the cause-and-effect assumptions that an actor adheres to: which action is likely to create which effect?Norms and interests can conflict and the choices made in such conflicting situation will largely be determined by an actor's identity.This identity is composed of the selective emphasis on aspects of self-interest, norms, and rules.This means, the choices made by an actor will provide a picture of this actors' identity, which in turn will allow us to explain action by taking into account the actor specific identity.Lastly, an actor's interaction orientations affect action.This essentially means the relationship with other actors.Are others seen as opponents, partners, potential beneficiaries or target groups of benevolent action?This will for example determine whether own losses will be accepted on behalf of the benefit of a counterpart or if gains to the counterpart are viewed as negative, independent of one's own gains or losses.To give an example: imagine a situation where a violent armed group controls large tracts of land, which small-scale farmers also use.The cadastral registration of that land, guaranteeing use-and ownership rights to both the famers and the armed actor might be seen as not acceptable to an NGO that intends to decrease the influence of armed groups and as a matter of policy does not cooperate with such actors.The conception of the armed actor as an opponent with negative influence on the larger objectives (e.g., ending armed conflict) could thus determine which actions are perceived as feasible.
Therefore, to analyze the (inter)actions of actors, information is needed on their preferences as well as the institutional setting of the specific context.This enables an analysis that breaks down the range of options an actor has into a limited number of choices based on preferences and institutional framework.The choices made will largely depend on the assumed (re-)actions of the other relevant actors.This can lead an actor to choose an option that is not the most-preferred one but that is assumed to produce the most sustainable outcome or is least likely to have negative consequences for that actor or third parties. 3The institutional framework (rules and norms) also structures the nature of interactions between different actors.The Figure 1 below gives a schematic overview of the elements of Actor Centered Institutionalism.

Women's Land Rights in Burundi
The Dutch NGO ZOA is working on land tenure registration and agricultural development in Burundi.The context is marked by high numbers of former refugees returning to the area after years in exile, often in neighboring Tanzania.Land is scarce due to high population growth and the return of people contributes to land-related conflicts.These land conflicts are highly complex and range from intra-family disputes on heritage rights to disputes between repatriates and residents.Customary and state authorities are involved in conflict resolution on the local level but in many instances, these actors are overburdened with the number as well as the complexity of the cases, particularly because the property rights are often not formally registered and conflicts between formal and customary rights exist.The formalized land registration in Burundi is based on the land law (code foncier) of 2011, which introduced land certificates as alternatives to titles for the registration of customary land rights and decentralized the land administration [16].In this context, ZOA decided to support the establishment of local land offices and the registration of land on the basis of resolving conflicts via on-the-spot as well as long-term mediation facilitated by local partner MiParec.The longer-term objective was to support farmers with agricultural activities after they secured their land rights.The project was initially funded by the Netherlands' embassy.
At the beginning of this work, it was realized that the chosen activities could have potential negative consequences for the land rights of women.The main reason for this was that women's land rights in Burundi are generally so-called secondary rights.This means that land rights of women derive via the rights of male relatives and constitute use rights rather than ownership rights.The customary context in Burundi is dominant in the day-to-day management of land and most women have no resources with which to acquire land of their own.In the early phase of the project, questions were raised in the Dutch Parliament with regard to the possible consequences of the work on the outcome of institutional reforms.For an application of Actor Centered Institutionalism in the analysis of land reforms, see: Betge 2017 [15].

Women's Land Rights in Burundi
The Dutch NGO ZOA is working on land tenure registration and agricultural development in Burundi.The context is marked by high numbers of former refugees returning to the area after years in exile, often in neighboring Tanzania.Land is scarce due to high population growth and the return of people contributes to land-related conflicts.These land conflicts are highly complex and range from intra-family disputes on heritage rights to disputes between repatriates and residents.Customary and state authorities are involved in conflict resolution on the local level but in many instances, these actors are overburdened with the number as well as the complexity of the cases, particularly because the property rights are often not formally registered and conflicts between formal and customary rights exist.The formalized land registration in Burundi is based on the land law (code foncier) of 2011, which introduced land certificates as alternatives to titles for the registration of customary land rights and decentralized the land administration [16].In this context, ZOA decided to support the establishment of local land offices and the registration of land on the basis of resolving conflicts via on-the-spot as well as long-term mediation facilitated by local partner MiParec.The longer-term objective was to support farmers with agricultural activities after they secured their land rights.The project was initially funded by the Netherlands' embassy.
At the beginning of this work, it was realized that the chosen activities could have potential negative consequences for the land rights of women.The main reason for this was that women's land rights in Burundi are generally so-called secondary rights.This means that land rights of women derive via the rights of male relatives and constitute use rights rather than ownership rights.The 3 Clement and Amezaga (2013) use a similar approach to analyze land and resource management in Vietnam [14].They develop an analytical framework that accounts for structural and discursive factors, which shape the outcome of institutional reforms.For an application of Actor Centered Institutionalism in the analysis of land reforms, see: Betge 2017 [15].customary context in Burundi is dominant in the day-to-day management of land and most women have no resources with which to acquire land of their own.In the early phase of the project, questions were raised in the Dutch Parliament with regard to the possible consequences of the work on women's land rights.The result of this interrogation was that the effects, at that point in time, were unknown but that there was a possibility the formalization of customary rights would lead to a weakening of women's rights because these would not be formally registered due to their secondary (dependent) nature.It was clear that this posed a danger to the long-term sustainability of the project, not least because it also increased the likelihood of intra-family disputes.The International Development Law Organization (IDLO), one of the project partners, was responsible for in-depth impact studies of the project with a specific focus on conflicts and gender relationships.IDLO proposed a pilot project with particular attention for women's land rights, with specific activities that would bring men and women together to discuss perceptions and fears around land registration in order to increase the number of registered land rights of women.There were also discussions around stronger support for the capacities of the formal justice sector to resolve land issues and options to advocate for stronger regulation from the political level to improve the land rights situation of women.A major challenge in the Burundian context is that inheritance law is not formalized, which means that inheritance issues are regulated by customary practices [17].The required changes in law have been pending for a long time and it appears that the issue is not high on the political agenda.Some decision makers view a formal inheritance law to be a potential fuel for further land related conflict.Table 1 below summarizes the actor constellation and options for ZOA in this situation.ZOA eventually decided to follow an approach that facilitates intra-community discussions on women's land rights and brings men and women together to allow for open discussion around the issue of land registration.The approach provides detailed information on the process and results of land registration and its implications for women's and men's land rights.This concept was applied in a pilot area, with significant results relating to the increase in registered female land rights.Through comprehensive engagement within communities, fears and prejudices around registering women's rights on the certificates could be alleviated in many cases.Men proved much less reluctant to register their spouses along with themselves once the concrete implications of this decision had been sufficiently addressed and women had been given the opportunity to voice their own concerns and needs regarding the protection of rights.While these are clearly positive results, there are also limitations to the chosen approach.The strategy is time-consuming and relies on the voluntary cooperation of various actors.The lack of legal requirements or political backing for this specific work is a threat to the sustainability of this approach and the achievements.
One central question here is why this specific approach was chosen, while a number of different, more-but also less far-reaching options are available.It could also be asked why ZOA engaged with a complex issue such as women's land rights in the first place.Regarding this latter question, the answer relates to the institutional framework as much as to the actors involved.The table above highlights that a vocal, advocacy-oriented approach could have endangered relationships with third parties.At the same time, ignoring the issue would almost certainly have threatened the relationship with the donor.Furthermore, ZOA's own gender policy commits the organization to gender equality and ZOA has signed the Dutch NAP 1325, joining a consortium of organizations working towards the objectives of UN Security Council Resolution 1325 on Women, Peace, and Security.In addition, the scrutiny via the Dutch parliament and the engagement of IDLO, offering concrete options to improve the gender related aspects of the work increased the need to pro-actively engage with the issue.Thus, internal and external norms as well as external pressure and incentives are all factors in this decision.
The concrete steps taken by ZOA present an interesting point of interrogation.There have been questions from peers as to why the project was not abandoned or completely re-designed given its possible negative consequences.There have also been challenges regarding the limited scope of the gender activities, asking why not a more vocal, policy-oriented approach was chosen.The first question can be answered by looking at the identity of ZOA as an organization.The core mandate of ZOA is to support people affected by crises, founded on a Christian identity that first and foremost obliges the organization to serve the most vulnerable.This is also strongly related to the first of the organization's core corporate values: Loyalty (Faithfulness).To management staff this means any initial commitment to a community is seen as highly binding.This self-perception of being an organization that provides support in the most dire and therefore often most complex contexts, with vulnerable and hard-to-reach groups as the main beneficiaries makes it extremely problematic to abandon a project.This issue has frequently been discussed internally and is seen by many of the staff as a core aspect of what defines the organization.At the same time, this also provides part of the answer to the second question: why not a more vocal, advocacy-oriented gender rights approach?In this regard, the anticipated reactions of various other actors come into play.Local customary leaders as well as political actors on all levels were expected to react negatively to a more outspoken approach to land rights for women with possible consequences for the non-land rights work of the organization in Burundi.Furthermore, a more vocal approach might also lead to stronger polarization of the positions on the ground.While in the end, this could lead to a transformation of the status-quo, it could lead to conflicts and maybe even violence in the short run.From a do-no-harm perspective as well as with a view on the objectives of the project (reduce land conflicts, agricultural support, livelihoods) this was not regarded to be an acceptable risk.Furthermore, limited experience with this specific type of work within the organization also limited the range of what was viewed to be possible.Thus, while internal and external norms required an engagement with the issue of gender relations in the context of land rights and while the organizational identity strongly supported a continuation of the project, the expected reactions of third parties limited the chosen approach to localized and rather 'soft' measures to improve the level of registered women's land rights.Recent impact evaluations have indeed shown an increase in the percentage of registered rights, but also indicated that there is a further need for improvements, warranting further adaptions of the approach.This might also lead to a re-evaluation of what is possible because further analysis could show that the previous activities have increased the scope of what can be done (and in that sense they might have changed the institutional context).

Land Governance in the DRC
In the eastern Democratic Republic of Congo (DRC), ZOA works on a project that addresses land rights issues as a root cause of conflict, poverty and instability 4 .The project is financed by the Netherland's Ministry of Foreign Affairs.One of the central challenges around land rights in the specific area is conflict between local communities and large concessionaires.Many smallholder famers have inherited their land rights without ever receiving any formal document to prove their ownership or use rights.In some cases, this land is part of a concession given to an agricultural or other enterprise by the national government.Other farmers have started farming on the land of existing concessions out of necessity.In many cases, they have entered sharecropping arrangements with the concession holders.In these contexts of semi-formal or informal land use, conflicts are widespread.Common types of conflicts revolve around forced displacement and changes in land-use conditions from the side of concession holders.Farmers might be forcibly removed from their land because the concessionaire fears long-term claims on the land or because the land is eroding due to the farming activities.In other cases, farmers might be unwilling or unable to agree to changes in the sharecropping agreements.Identity issues often exacerbate these conflicts, with concessionaires and land users belonging to different ethnic groups.The principle of large-scale concessions is a remainder from colonial times and concentrates large tracts of land in the hands of few.
Land degradation through intensive use makes some landowners reluctant to lease out their land, which in turn enhances tensions around land access.Forced removals of small farmers by concessionaires have happened even in cases in which the concessionaires themselves did not adhere to the standards required for them to keep possession of the land.The Congolese land law states that land has to be put to use within a period of 18 months by the concession holder.In many cases, only part of the concessions are actually being used, making the concession holder's own legal claim fragile at best.Nevertheless, the often well-connected concessionaires rarely face the risk of losing their land.This is also because the local land administration authorities are challenged to keep up with their tasks, facing severe funding constraints and staff shortages.Adding to the already difficult land situation are conflicts relating to tensions between pastoralists and farmers, inter-village disputes regarding borders of agricultural land and intra-family disputes, e.g., because of inheritance questions.Women face particularly strong challenges in many cases because their land rights often depend on their male relatives and they lack inheritance rights as well as decision-making powers related to land and other assets.
ZOA and its partners developed a multi-pronged approach to address these challenges.The approach focuses on different levels of the social and political sphere.It takes into account existing formal and informal governance systems and aims to facilitate dialogue between central stakeholders.The approach requires ZOA and its partners to engage with existing civil society actors while establishing links between the formal and the informal levels of governance.These locally oriented efforts link to higher governance levels.Thus, locally improved governance and inclusive conflict resolution are envisaged to inform stakeholder engagement on higher levels (e.g., territory, province, national).On the local level, four central pillars provide the foundation of the work: 1.
Community-Based Sociotherapy (CBS): Empowering individuals to engage in positive social change [18].Implemented by a local partner.

2.
Cadres de Dialogue et Mediation (CDM): Engaging in conflict mediation and in negotiation with large landholders so that farmers with no or limited access can rent land on the medium or long 4 The partner organization facilitating het conflict resolution work is APC-Action Pour la Paix et la Concorde.
term and are protected through clear lease agreements.Implemented by local partner APC with long-term experience in peacebuilding and conflict resolution.

3.
Civil Society Engagement: Mobilization and empowerment of existing civil society structures based on the CIVICUS approach [19], allowing concerted action on land rights.4.
Improved Governance: cooperation with/support to formal and informal authorities to improve local governance by increasing transparency and service delivery.
Each of these pillars presents an individual focal point of the work while at the same time opportunities to create synergies are systematically pursued.Particularly, linking the three community-based components to formal governance structures and customary authorities is required.The experiences from working on the local level need to be fed into frameworks on higher levels of government to create sustainability.Regular information exchange between relevant stakeholders is viewed as a minimum requirement for an inclusive and sustainable approach.The necessary policy and administrative changes identified through the different activities and by the different actors (through respectively CBS, CDMs, and broad civil society engagement) would be explained and advocated to higher levels of government and actors engaged in efforts for regional stabilization.In collaboration with third actors (NGOs, CSOs, administration) and through active participation in an existing regional land tenure-working group, necessary improvements in administrative and legal framework are lobbied for on the provincial and national level.

Rapid Appraisal of the Four Pillars
The CBS process relies on 15-week cycles during which small groups of volunteers come together on a weekly basis and are guided through the six phases of sociotherapy: 1. Safety, 2. Trust, 3. Care, 4. Respect, 5. New rules, 6. Memory of emotions.The assumption relating sociotherapy work to land governance is that the increase in psychological wellbeing and trust achieved through sociotherapy enables people to constructively deal with land issues.So far, there is no hard evidence from the DRC context that this bears fruit.However, anecdotal evidence from the implementation suggests that community cohesion is already increasing after about one year of applying the approach.
CDMs are a proven way of addressing local conflicts and have received much praise in the past [20].They operate based on a principle of cooperative and consensus-oriented problem resolution on the local level.They enable dialogue and trust between different groups within a community and provide an alternative to cumbersome and often biased judicial processes.Nevertheless, a stronger integration with formal justice e.g., by having outcomes of CDM processes validated by courts could improve the sustainability of the work.Furthermore, CDMs face strong challenges when engaging with big concessionaires and require effective training on how to act strategic in these situations.
Civil Society Engagement poses a central challenge to ZOA and its partners.There is a broad range of actors that can potentially be involved but among other issues logistical challenges impede this work.The idea behind the Civil Society Engagement is that existing civil society actors are guided through a process of formulating a common agenda to improve local governance structures and service delivery related to a range of issues including land rights.This is a new way of working to ZOA and requires strong diplomatic and organizational efforts.So far, tangible results are limited.
Improved governance on various levels by increasing the capacities of (local) actors is a necessary condition for long-term success of the strategy.ZOA works with actors on the provincial level as well as local actors to achieve this.The organization also engages in cross-level structures (including the national level) aimed at improved governance.This work stretches the (staff) resources of the organization and goes beyond the usual activities conducted.There are some visible successes so far such as an upcoming multi-stakeholder conference with national, provincial, and local actors aiming to produce a roadmap for improved land governance.This conference is supported by ZOA and responds to urgent needs identified in the current work.Coordination with provincial miniseries has also led to ongoing attention for land governance issues on the political level.

Analysis of the Choices Made in DRC
Without going into too much detail on the individual components of the land rights strategy, it is obvious that the chosen approach differs in significant aspects from the way of working chosen in the Burundian context.The approach explicitly targets different governance levels and is oriented towards high-level coordination and lobbying for policy measures that support the envisaged outcomes.This implies a much more direct involvement in politically contested and sensitive issues than the approach chosen in Burundi.So why did the organization choose a strategy that clearly involves a risk of stretching limited staff capacities and becoming strongly involved in a highly politicized context?This question is even more relevant knowing that ZOA defines itself as a rather locally oriented and anchored organization.While the organizational identity of ZOA can be assumed to be stable across country contexts, the central factor that influenced the unusual decision in this case is an institutional one.
Humanitarian and development efforts in eastern DRC are generally required to align with the International Security and Stabilization Support Strategy (I4S) of the international community and the Congolese government in the context of Stabilization and Reconstruction Plan (Starec) for the region [21].In particular, large projects with peacebuilding objectives need to be coordinated and aligned with these efforts and generally require multi-level approaches.However, it would be wrong to view the strategic approach simply as the result of adhering to external requirements.Internally, the need to coordinate and cooperate on different levels is also seen as an opportunity.The fact that multi-actor coordination mechanisms already exist is understood to be a challenge as well as a possibility to create the necessary conditions to reach all objectives of the work, not least relating to land rights, as well as to establish structures and support for long-term sustainability.At the same time, without the external requirement to align with the I4S such a complex, multi-level strategy would likely not have been chosen as it goes against the standard modus operandi of the organization.This poses a challenge to the outcomes, as it also means that within ZOA, the opportunities to purposefully change the institutional framework might not be fully realized because this is a new way of working for the organization.The fact that a strategy was chosen that diverges from the organizational standard has implications for the ability to realize certain targets.Table 2 below summarizes the positions of different actors in this context as well as the relevant rules and norms affecting ZOA's decision-making.While ZOA would always prefer a localized approach, there are no internal norms or rules prohibiting multi-level approaches.The table above highlights that the specific institutional framework within which ZOA operates in the DRC requires engagement beyond the local level.The coordination and function of international actors under the I4S strategy can be seen as an acknowledgement that in areas of limited statehood 5 , governance functions need to be fulfilled (also) by non-state actors.While many non-governmental organizations implicitly recognize this, it conflicts with the self-perception of an actor like ZOA.There is broad internal agreement that it is not the role of the NGO to assume quasi government functions.While it can be argued that this position is based on a misconception of the factual role of NGOs, this self-perception makes the organization even more reluctant to engage in higher-level governance work.
Within ZOA, there is agreement that the most important steps towards long-term stability and positive peace need to be taken on the local level, creating trust and a basis for cooperation as well as functioning local governance mechanisms.While it is realized that higher-level actors need to create certain institutional (legal, political) frameworks for enabling this locally oriented work, the perspective of a locally-embedded actor results in a strong orientation towards communities such as villages, households or congregations and other forms of communal organization.This also means that cooperation partners are usually sought on the local level.For an actor that has its roots in the humanitarian sector such a focus and prioritization seem quite natural.However, this orientation makes strategic cooperation and purposeful interaction around land governance issues a potential challenge.The reason for this is that such work might be perceived as mainly an externally determined task.This perception is likely to limit the overall commitment towards land governance efforts.Similarly, actors working mainly on governance issues beyond the local level are also more likely to focus on what they perceive to be peers and actors with the power to make substantial, structural change happen.They might view cooperation with locally oriented NGOs to be a requirement rather than an opportunity.This does not mean that cooperation between these different kinds of actors is uncommon.However, there is usually a clear division of tasks.In the case of land governance in the DRC, strong cooperation is required regarding policy issues and changes in the institutional framework, which means cooperation across different levels and types of actors in the form of a specific private-public partnership model 6 .Such direct cooperation between state and supra-national actors and local NGOs with the objective of shaping policies and politics is much less common.Furthermore, it happens in a space where different kinds of actors fulfill governance functions without necessarily having a clear mandate for this.At the same time, the capacities of mandated actors might be overestimated.In such a setting of limited statehood standard setting partnerships need a high degree of institutionalization.If obligations and the precision of norms are not binding, multi-actor partnerships are likely to be less effective [21].
The orientation of supranational and governmental actors towards perceived peers can result in a bias and misconception of their counterparts by assuming a degree of implementing power and political will that are unrealistic in a context like the Congolese where local actors often have the power to act as spoilers to higher-level political decisions.It can also lead to an under appreciation of the potential that lies in direct cooperation with more locally rooted or focused actors.The bias of international actors towards high-level initiatives has been strongly criticized in the past [22] and the I4S approach can be seen as a reaction to the realization that top-down efforts will not produce sustainable outcomes.Unfortunately, this does not mean that the holistic approach is successfully translated into practice.Land governance is a good example of the challenges of facilitating effective cooperation between grassroots and top-level oriented actors.While the I4S framework requires cooperation, the aspects that actors focus on in their daily practice are likely the ones that are close to their standard way of working and shaped by internal norms and rules.This means that external factors can influence decision-making, but the degree of commitment to these decisions and the likelihood of success depend on a number of internal and external factors (see Table 3  No engagement with the issue.

Discussion
The analysis in this paper sheds light on important factors that help to explain successes and failures of humanitarian and development work.More specifically, it points towards factors that influence decision-making in the context of areas of limited statehood where (land) governance is not solely a task of the state.
The first finding is that, unsurprisingly, institutional factors have an impact on decision-making and can push actors to work out of their comfort zone, addressing issues they might otherwise not have engaged with or choosing strategies that deviate from their standard way of working.The second and less obvious factor is that an actor's preferences (identity, interests, actor orientation, and internal norms) are likely to have a strong impact on how the actor fulfils its institutionally determined role.This can lead to a situation in which miscommunication among different actors occurs with significant consequences and positive effects of structural frameworks are overestimated.It can also lead to an under-utilization of available opportunities.In the case of ZOA's engagement in Burundi, there might be options to influence policies and politics around land issues through subtle lobby and advocacy work.The mandate and self-perception of the organization nevertheless do not make this a default way of working.Lobbying for specific policy objectives is not a standard tool and advocacy not part of the organization's mandate, according to internal rules.Lobby and advocacy work is also only to a limited degree driven by internal norms.This leads to a bias towards locally oriented action with limited effect regarding policy changes and norm diffusion.
In the Congolese case, there appears to be a gap between the complex, holistic strategy chosen and the manner of implementation.This gap is likely to be a result of diverging preferences and particularly the interaction orientations that different actors have.Furthermore, limited time and resources for a critical reflection of multi-level processes in the realities of day-to-day work stand in the way of improving the modes of cooperation in the context of land governance.It appears logical that cooperation tends to be stronger among organizations that perceive each other as peers (same type of actor), and if internal and external norms and rules require engagement with a particular issue.It can also be assumed that commitment towards an objective is highest if actors share a normative basis.
Nevertheless, effectiveness often depends on the existence and quality of rules to guide decisions and follow-up processes.The complex challenges of land governance cannot be tackled by simply producing a number of outputs.There needs to be strategic and constructive cooperation of actors on different levels.If normative commitment is lacking on one side, this is likely decrease the effectiveness of cooperation.This is exacerbated if actors do not invest sufficient time and resources into their cooperation, for example in cases where they under-value the benefit of working together.In the context of land governance, this leads to a gap where the legal issues and policy requirements are widely known, local challenges are identified, but concerted efforts to reconcile these spheres are limited in practice.Openly addressing the requirement for different actors to assume governance functions in a cooperative setting can be a first step to create the necessary rules and guidance to improve on these processes.
The Actor-Centered Institutionalist approach underlines that rule-based institutional frameworks change over time if repeated norm-driven action changes the shared understanding of acceptable or desirable behavior.Similarly, rule-based actions can change normative orientations over time.This means that, based on an analysis of existing rules and norms, the right strategy for action can be adopted.Windows of opportunity for norm or rule changes might open up over time, but a conscious strategy to achieve this requires long-term efforts based on cooperation.ACI highlights the importance of actor orientations.The examples above have shown that the way third parties and their preferences are perceived is an important factor in decision-making.This not only calls for thorough analyses of third-actors on the part of NGOs.It also underlines that continuous interaction is likely to change the perception of third actors by providing better information on their preferences.Actors can enhance the effectiveness of their interactions through regular analyses.Conflict sensitivity tools for self-analysis and context analysis can support such processes of critical review [23].Table 3 below provides an overview of the possible combinations of drivers of decision-making and their respective effects.
The table above shows that intrinsic motivation of actors reinforced by external norms is likely to produce strong commitment towards an issue.Nevertheless, external guidance through rules might be required to make such commitment effective.The Burundi case shows that intrinsic motivation and a normatively based demand alone are not sufficient to create effective engagement if the rules of engagement are contradictory or there is normative incoherence across different actors.At the same time, the DRC case underlines that even a strong external requirement to commit towards an issue is not sufficient if actors with diverging preferences engage in a highly complex field.

Conclusions
The theoretical concept of the Duality of Structure draws attention to the mutual influence between actors and structures while the analytical concept of Areas of Limited Statehood underlines the specific governance challenges actors face in post-conflict contexts.In the absence of strong state actors and good institutions the decisions of non-state actors potentially weigh heavier than in contexts were states ensure the rule of law and the functioning of institutions.Moreover, knowing that land governance is a crucial factor for peaceful and sustainable development but one that is not receiving sufficient attention in post-conflict settings places a heavy burden on the actors engaged in this field of work.Recent studies re-confirmed the crucial role that multi-level partnerships play when addressing land and conflict [24].
The two case examples discussed above highlight the complex web of factors that influence the decisions of NGOs and other development actors in the context of land governance.They underline the need for actors engaged in land governance in post-conflict contexts to adapt their ways of working and cooperating as well as the challenges that this brings.Default modes of working will fail to provide durable solutions for land rights work.However, in the absence of clear guidance regarding the issues on which to focus and the ways to structure effective cooperation, many actors will fall back on standard operating procedures.In the Burundian case, ZOA chose an approach to address women's land rights that accommodates the diverging preferences of actors, which the organization assumes to be key for its future operations in the country.While under the given circumstances this was a logical choice, there is a need to keep interrogating the contextual and normative factors underlying this choice in order to identify opportunities or needs to adapt the chosen approach.This requires critical, conflict sensitive and multi-level context analyses as well as critical self-interrogation.Such analyses will serve to support the faithful fulfillment of the commitment made by the organization through engaging with land governance in Burundi.They could also help to address current impediments to sustainable land governance such as ill-defined policy frameworks and unclear rules [25].The Congolese case shows that institutional requirements can push an actor to work out of their comfort zone.This will often be necessary to address land rights issues in a coherent and sustainable way.ZOA DRC adopted a complex strategy to engage with land rights issues.However, without clear guidance on how to structure multi-actor cooperation and without binding norms and rules, actors are likely to struggle to cooperate effectively.
Effective cooperation requires well-founded analyses of the issues to engage with, strong agreements and concise normative frameworks.Furthermore, cooperation requires an understanding of the decisions other actors take based on their specific preferences.The experience of ZOA shows that an actor's specific identity plays a key role in decision-making.This analytical factor is potentially under-appreciated-not only by third parties but also by actors themselves, in processes of self-evaluation.An analytical focus on the interplay of actors and structures including actor identities and their effects on choices regarding how to work and who to cooperate with can help to explain action and understand difficulties in practical day-to-day cooperation.Actors (not only) in international development need to reconcile the structural factors defining their options for action with their own preferences, including their identity and normative role orientations.An increasing focus on fundamental and complex challenges such as land rights requires better cooperation based on comprehensive self-analyses.Development and humanitarian actors need to interrogate how they take decisions, focusing on contextual factors (institutional framework) and their own preferences.Through this, they will better understand the effects that their choices and actions have on the context within which they operate and vice versa.

Land 2019, 7 ,
x FOR PEER REVIEW 5 of 16 nature of interactions between different actors.The figure below gives a schematic overview of the elements of Actor Centered Institutionalism.
Ignore donor concerns.

Table 1 .
Schematic representation of Burundi case: women's land rights.

Table 2 .
Schematic representation of DRC case: land rights strategy.

Table 3 .
Schematic representation of drivers of decision-making and effects.