Seeing like a donor: the unintended harms of rendering civil society legible

ABSTRACT

Following the Grand Bargain, there has been increasing focus on aid localisation and partnerships between international and local aid agencies. Yet there has been less scholarly attention on how and why international agency policies and partnerships can cause unintended harm to civil society organisations and their staff. Drawing on James Scott’s seminal work Seeing like a State, and interviews with Myanmar civil society organisation leaders in 2023, this article argues that international agencies often attempt to render civil society “legible” through processes of systematisation and codification. However, these processes can in turn sideline accrued experiential and contextual knowledge, or metis, which is necessary for local organisations’ survival, especially in times of instability. The article highlights several instances in Myanmar where the marginalisation of this more contextual knowledge results in unintended harms. The article concludes that international agencies’ acknowledgement of metis is a crucial and yet still under-recognised pillar of aid localisation.

KEYWORDS:

• Humanitarian aid
• development
• localisation
• civil society organisations

Introduction

In 2023, a local village leader in Karen State, Myanmar, died after being shot by members of a revolutionary armed group. The story of his death was recounted in our interview with the leader of a Myanmar civil society organisation (CSO) that was providing humanitarian assistance (Int 12, October 2023). The incident took place in an area of active military conflict between the Myanmar military forces and ethnic revolutionary organisations, and there had been recent tensions in the village tract over processes for distribution of humanitarian aid. An international donor agency – that was supporting the local CSO – insisted on targeted beneficiary selection, prioritising only the poorest households. Guided by knowledge of local community tensions, the preference in the village development committee was to share the aid evenly across all households in the village tract. After some attempts to negotiate, the donor representative ultimately pushed the village committee to do beneficiary selection. This selection was done “carefully” by the committee, according to the CSO leader (Int 12, October 2023); however, the targeting angered some local members of a revolutionary armed group, inflaming existing tensions. In this tense environment, the head of the village development committee had been shot.

It is not our intention here to speculate forensically on the flow of events or to attribute responsibility for the death. What we are interested in exploring is the interpretation of events by the CSO leader – who clearly linked international donor agency rules about beneficiary selection to the tensions that led to the village leader’s death. In his interpretation of events, he emphasised the unintended impact of donor systems and policies when they override contextual knowledge and decision-making. Such narratives of unintended harms caused by donor systems were common in our interviews with CSO leaders in Myanmar in 2023 and deserve further scholarly attention. In this exploratory article we draw on this case of CSOs operating in the unstable context of Myanmar since the attempted military coup in 2021. We seek to examine the ways in which Myanmar’s humanitarian and development organisations draw on specific accrued contextual knowledge and survival strategies to operate in a repressive and unstable context. We then also seek to explore unintended negative consequences when donor organisations standardise and homogenise partnership policies and undermine local organisations’ abilities to draw on these embedded local knowledges and strategies.

Here, we focus on systems within both bilateral donor agencies and in international intermediary agencies – United Nations agencies and international non-government organisations (NGOs) – which play a bridging role between institutional back-donors and CSOs or community groups that are implementing programs on the ground. Importantly, from the perspective of the CSO leaders whom we interviewed, bilateral donors and international intermediary agencies were often considered part of a single system of “donors”, where international intermediary agencies replicated, or at times even amplified, the same stringent compliance measures as bilateral donors. In this sense, while we acknowledge that there are complex differences in the experiences and roles of bilateral donors and intermediary agencies, when we refer to “seeing like a donor” in this article, we include both of these sets of actors.

At a time when the agenda of aid localisation is prominent, bilateral donors, UN agencies, and international NGOs are seeking to engage more with local and national CSOs around the world. Yet “civil society” is a complex, diverse, and often confusing web of informal and formal connections. Therefore, in order for donor agencies to engage systematically with civil society, we argue that they seek to render local social groupings legible, or more easily comprehended.

Here, we draw on James Scott’s (1998) concept of legibility from his seminal work Seeing Like a State in order to explore why – as noted by analysts of humanitarianism – international aid can at times have unintended consequences, which can in turn cause harm for local actors. In the first section of the article, we explore the potential and limits of Scott’s arguments. We contend that, like states, donors within the aid sector can seek to establish the legibility of civil society through standardisation and classification of inherently complex social entities and relationships. Through a myriad of processes, international donor agencies attempt to categorise and homogenise their interaction with civil society groups and shape the groups themselves. Though such processes can be used overtly for purposes of manipulation or control, our argument here is that they are primarily employed by donors practically as means by which the bureaucratic aid agency can engage in a standardised way with a messy array of diverse local organisations. Scott emphasises, however, that even when well intentioned, such schemes to render society legible often fail, or at least have significant unintended consequences. This is because they suppress or marginalise accrued knowledge of contextually embedded solutions to local problems – or what Scott (1998) describes as metis.

In part two, we draw on interviews with CSO leaders in Myanmar in October 2023 and argue that in the unstable context since Myanmar’s attempted coup, civil society groups have used subtle practical strategies to manage risk, avoiding unwanted attention from authorities, and continuing their programs. However, in their efforts to make civil society organisations legible through applying various standardised processes in partnerships, donor agencies can suppress local CSOs’ ability to rely on tacit experiential knowledge or metis. These dynamics are found elsewhere in aid-recipient countries. Yet what makes the Myanmar case unique is the way in which the context of conflict and uncertainty has accentuated international agency efforts toward rendering civil society legible, while at the same time increasing the risk of harm to CSOs. The context of revolution in Myanmar therefore illuminates these dynamics of “seeing like a donor”, and their dangers, more clearly.

In part three, we again draw on interviews with CSO leaders to outline several key instances of program documentation, security protocols, and beneficiary selection – where donor systems have had unintended negative consequences. These key areas of analysis of unintended consequences were selected inductively from the study data, as participants identified key areas of concern about donor systems. We conclude that, while funding flows and decision-making power are key considerations in progress toward localisation, from the perspective of CSO leaders in Myanmar, localisation is also about international agency recognition of alternate knowledges.

Part 1. Seeing like a donor – theory and literature

Since the Grand Bargain in 2016, the localisation of aid has been a widely discussed goal among international donor agencies, UN agencies, and NGOs. It has also placed partnerships between international agencies and local civil society groups in the spotlight (Décobert 2023). Within humanitarian studies, an important body of literature has emerged examining the goal of localisation and progress, or lack of progress, toward it. Much reflection on aid localisation within this literature focuses either on funding flows and effectiveness – how much money is going to local organisations and how well is it used (see Barbelet et al. 2021; Manis 2018) – or on aid decision-making – how much power is given to local organisations in aid systems (see Milasiute 2023; Pincock, Betts, and Easton-Calabria 2021). These are, of course, important considerations within debates about localisation. Yet the danger is that these types of analyses overlook more fundamental contests between different forms of knowledge or “ways of being” (Baquios 2021) International donor agencies develop systems to standardise and codify their relationships with local groups – in other words, to make them more legible – yet what forms of knowledge does this approach suppress? And what impacts might there be in relation to the goal of aid localisation?

Within development literature, there has been long been analysis of the ways in which technocratic, standardised aid compliance systems structure international partnerships with local organisations (see Bryld 2000; Ferguson 1990) – including in Myanmar (Bächtold 2015). There has also been focus on the unintended consequences of foreign aid (Davidson et al. 2022; Koch 2024; Wood and Sullivan 2015), and particularly on the ways in which aid interacts with dynamics of conflict, governance, and security.¹ Yet these studies on unintended consequences of aid systems do not shed light on the kinds of tangible harms to civil society organisations described in this article. Nor do these studies explore these consequences through the lens of competing forms of knowledge. Here, we argue that James Scott’s (1998) seminal work Seeing Like a State is valuable in shedding light on the types of underlying dynamics that can lead international aid systems to cause unintended harms, by focusing on competition between different forms of knowledge and by highlighting the negative consequences of prioritising some types of knowledge (and associated practices) over others.

According to Scott (1998, 2), what states have done is take “exceptionally complex, illegible and local social practices … and created a standard grid whereby it could be centrally recorded and monitored”. This approach is driven by a “high modernist ideology” (Scott 1998, 2), where there is faith in progress through technocratic processes. Importantly Scott introduces the concept of metis, or contextual knowledge, which is in contrast with techne, or technical knowledge (Scott 1998). Scott explains: “Broadly understood, metis represents a wide array of practical skills and acquired intelligence in responding to constantly changing natural and human environment” (1998, 313). Metis is therefore knowledge developed through lived experience rather than being derived deductively through “scientific” or standardised principles. Part two of this article highlights the unique ways in which Myanmar organisations use metis to navigate the current context of instability and authoritarian rule.

Using examples of China, Russia, Brazil, Tanzania, and other contexts of highly centralised planning, Scott (1998) highlights how efforts to standardise and homogenise society based on a “high modernist ideology” often result in failure, and unintended consequences, as they sideline locally embedded decision-making. Similarly, we argue in the third section that international aid systems play a role in standardising donor partnerships with Myanmar organisations. We highlight how suppressing the use of metis, in favour of homogenised donor processes, can have negative outcomes on personal security, organisational effectiveness, and the autonomy and agency of CSOs.

These key concepts from Seeing like a State have been applied in a myriad of other fields, for example, in relation to urban governance (Valverde 2011), forestry services in Senegal (Blundo 2014), and COVID-19 contact tracing (Liu 2022). Scott’s work has also faced criticism, however. For example, Laitin describes his historical method as being “selective and eclectic” and argues that there are counter examples where high modernism has brought “stunning successes” (Laitin 1999, 177). International aid agencies’ standardised policies have, in many cases, served to increase participation of women, people living with disabilities, and religious minority groups. And, as Seabright (1999) notes, infant mortality rates in poorer countries have fallen substantially over the last century largely due to access to systematic modern medicine based on techne, alongside a rolling back of reliance on locally developed health remedies. In this sense, we caution against a simplistic narrative where metis is always valuable and techne always constraining and ineffectual. While avoiding binary or simplistic analytical frameworks, however, we argue that Scott’s concept of legibility does hold relevance for the aid sector – in which techne often holds a privileged position.

Some scholars have attempted to extend Scott’s work in Seeing like a State to humanitarian and development sectors. In their special issue, “Seeing like an international organisation”, Broome and Seabrooke (2012) apply Scott’s framework to approaches of large multi-lateral organisations in aid-recipient countries – including the World Bank, International Monetary Fund, and World Trade Organisation. For example, Vetterlein (2017) argues that in efforts to make developing countries “legible”, the World Bank has provided standardised poverty reduction advice and responses that ignore local social knowledge.

Heathershaw (2008) makes similar arguments applying Scott’s notion of Seeing Like a State to the field of peacebuilding. Using the example of Tajikistan, he suggests that, led by the international community, peacebuilding took the form of a rational, high-modernist design and vision. Yet crucially, the formal peacebuilding project ultimately depended on informal, locally embedded processes that did not always align with the overall technocratic design. The formal design was, in Scott’s (1998, 310) terms, “parasitic on informal processes”. While the formal peacebuilding design did not acknowledge the role of metis, it was nevertheless dependent on it. This brings a paradox where formal planning is dependent on informal processes. In the third section of this article, we explore a similar paradox where the implementation of international humanitarian programs in Myanmar is dependent on CSOs’ use of opaque and informal processes based on local knowledge. Yet many of these processes are not recognised or even sanctioned within codified aid partnerships.

While these analyses valuably extend the work of Scott in relation to the field of development, they do not examine the approaches of bilateral donors, NGOs, and United Nations agencies and their efforts to render “civil society” more legible in recipient countries. Cowling’s (2020) thesis “Seeing like a humanitarian” moves further in this direction by applying Scott’s concept of legibility to systems of humanitarian programming in Lebanon. Cowling (2020, 15) explains: “though humanitarian organisations are usually non-state actors, they must nevertheless make sense of complex social realities when designing and implementing emergency programmes, scaffolding them with systems designed to simplify and codify”. Cowling (2020) does not, however, address the relationship between humanitarian donors and civil society organisations. In this article, we focus on specific donor processes, including documentation and financial compliance, security protocols and targeting of recipients. Such systems of accountability play a key role in standardisation and homogenisation of CSO practices and relationships.

Methodology

Where Scott takes a historical approach, in this article, we draw primarily on extended semi-structured interviews with civil society organisation leaders in Myanmar. This article draws from 12 interviews² with Myanmar CSO leaders – conducted in Burmese language by one of the authors – as part of a fellowship program led by Myanmar aid practitioners and researchers. These interviews were part of broader data set of 64 semi-structured interviews with Myanmar CSO leaders and international aid agency representatives, which were conducted by four researchers (two Myanmar and two international) between October 2022 and November 2023. This broader data contributed to contextualisation and validation of the findings.

The 12 interviews were conducted with Myanmar leaders from eight local humanitarian and development organisations and focused on participant perspectives on aid localisation and partnership. The eight participating Myanmar CSOs all had multiple project partnerships with international NGOs and donors. There was, however, also diversity in the selection of participants. Six of the local organisations involved are unregistered with the Myanmar military junta, while two are registered.³ The organisations also varied in terms of where they are based – some are primarily operating and based within Myanmar’s military junta-controlled areas, some are in areas controlled by other armed groups, and others have a management base in Thailand. While seeking to include diverse Myanmar civil society participation, we do not claim that this interview data are representative of all CSOs in Myanmar. We do argue, however, that the data raise valuable questions about the dynamics of partnerships between local and international organisations, and the unintended consequences of these dynamics. This article also draws on the extended practitioner experience of each of the authors in the aid sector in Myanmar, including work experience in local and international NGOs and UN agencies. Having a Myanmar aid practitioner as lead researcher in the study allowed links to participants through established relational networks of trust, which was crucial in the current context of security risks for civil society leaders. The lead researcher was not affiliated with any of the participating local organisations yet had pre-existing relationships with participants through a role with an international aid agency. The other authors had no pre-existing links to the participating agencies or individuals.

At this point, it is also important to acknowledge a key dilemma in the writing of this article. On one hand, we seek to examine practical challenges faced within international and local aid partnerships in the context of Myanmar, and to apply these findings to broader development scholarship. On the other hand, all participating agencies in this study are highly vulnerable and operating in contexts of considerable instability. We have therefore sought to protect the identities of the participating agencies, and also to limit our description of the nuanced strategies that they use to evade scrutiny from the Myanmar military regime, or from other actors.

Part 2. Metis and the adaptations of Myanmar organisations since the 2021 coup

The early 2010s saw a slow building of freedoms for civil society organisations in Myanmar as the longstanding military regime gradually liberalised under the semi-civilian Thein Sein government. This period was one of significant growth in the number, scale, and scope of civil society organisations, while also revealing the illiberal potential of civil society networks through anti-Muslim movements such as MaBaTha (Décobert and Wells 2020).⁴ The election of Daw Aung San Suu Kyi’s National League for Democracy government in 2015 then saw further expansion of the role of CSOs in some areas of service delivery and policy making. The most dramatic change for Myanmar CSOs in recent years, however, has been the attempted military coup of February 2021 and the authoritarian rule that the junta’s State Administrative Council (SAC) has imposed in the areas which it controls. The three years since the coup have been tumultuous and many civil society leaders have been arrested or forced into exile.

With personal and organisational risks increasing dramatically after the 2021 coup, Myanmar humanitarian and development groups have had to adapt in creative ways to survive and continue to be active. These adaptations have been well documented in other studies (Protect 2021; Wells and Maung 2024) but here we seek to emphasise how these nuanced strategies are only possible through use of specific accrued contextual knowledge, or in Scott’s terms, metis. This section gives several examples of the complex and often opaque ways in which CSO leaders negotiate with communities and the varying levels of decision-makers to be able to continue essential activities.

It is important here to clarify the relationship between different forms of knowledge (metis and techne) and tangible practices of donors and civil society groups, such as the examples of program documentation or security protocols that we discuss in this article. In the next section, we emphasise contests over the appropriateness of these practices in contemporary Myanmar. Yet interpretations of the appropriateness of these practices are embedded in contested knowledges. In particular, our interviews revealed differences in the expectations of what can be trusted within a context of conflict and uncertainty.

Technocratic processes of compliance introduced by most international agencies placed an emphasis on standardised documentation and assessments against externally determined technical criteria as crucial in making sense of the context and of the actors involved in aid programs on the ground. This included target communities as well as local aid agencies, whose “accountability” (and, therefore, also trustworthiness) was also assessed in relation to such technical criteria. In contrast, participants from local organisations typically emphasised trust embedded in experiential knowledge as well as interpersonal relationships and networks as central to understanding. This distinction between trusting technical standardisation or relational connection resonates with the analysis of aid partnerships in the Start Network’s (2022) “Anti-racist and decolonial framework” – which emphasises “technical” versus “contextual” expertise, which is in turn linked with “standardised” versus “relational” practices.

We caution here against essentialising different forms of knowledge and a simplistic division between techne and metis, or between “international” and “local” knowledge and actors.⁵ Yet research in Myanmar highlights important connections between the types of knowledge underlying differing practices of local and international aid organisations, and this clarification of relationship between knowledge and practices is important in setting the scene for the third section where we explore examples of unintended consequences, when local knowledge is suppressed through donor agency efforts at project homogenisation and systematisation.

In areas controlled by the SAC, local organisations draw on their experiential knowledge and use specific local relationships and structures to gain permission and protection for their activities. Some have opted for registration with the SAC⁶ to be able to continue humanitarian activities, though this is also deeply challenging due to the corruption and violence of the regime. For others, working completely under the radar is possible. A local organisation leader in the southeast of the country (Int 4, October 2023) said that they have no registration and work at a local level without mentioning their organisation name. Other CSOs may not have SAC registration, yet they seek to establish informal relationships with local-level authorities and local village elders.

In Myanmar’s system of government, the backbone of local-level public administration is the General Administration Department (GAD). A CSO leader from Kachin State explained the importance of creatively gaining GAD permission for activities and working “smoothly on the ground” while avoiding the attention of higher levels of SAC governance (Int 10, October 2023). The leader of a local women’s organisation in another area of the country, meanwhile, explained how they have informal agreements with the GAD about possible security risks. She said “if the SAC forces are coming, [the GAD] will inform us at least one hour in advance. We’ve packed everything so that we can run when the time comes” (Int 8, October 2023). Rather than ceasing their activities, or seeking explicit regime permission, many Myanmar CSOs have used their accrued contextual knowledge of governance and local relationships to make pragmatic decisions – avoiding higher levels of the regime, while gaining local permissions where necessary.

In response to the SAC’s national Registration of Associations Act, many Myanmar CSO leaders have developed creative strategies to continue to operate. Some have renamed their organisations; others have applied for registration as other types of private entities or have set up non-traditional means of receiving funds. One local CSO leader explained how they manage a portfolio of identities so that they can continue to receive funding from international donors (Int 12, October 2023). These CSO approaches to maintaining operations are intentionally opaque and often do not align with international donor expectations of transparent and formal registration.

Meanwhile, in areas of mixed, or other armed group control, CSOs also need to find subtle ways to gain the permission to operate. Local CSOs must work within revolutionary organisations’ expectations about the implementation of projects, and particularly, the distribution of humanitarian aid. A CSO leader explained how “some ethnic armed organisations have … set up guidelines … if you are to give emergency support, you have to support the whole village tract or village” (Int 12, October 2023). CSOs on the ground face pressure to adapt to these guidelines made by local armed group administrations, which may not always align with international aid agency policy.

CSOs engaged in humanitarian aid distribution in ethnic revolutionary organisation-controlled areas must also use contextual knowledge of local power relations to navigate different codifications of administrative boundaries. The Myanmar Information Management Unit (MIMU) was established by aid agencies in 2007 to standardise mapping and place naming in the country to facilitate coordination among international and local aid agencies. The codification of maps under MIMU was aligned with previous quasi-civilian and civilian Myanmar governments. Since the coup, this codification has also been adopted by the military’s SAC. Yet the administrative boundaries drawn by ethnic revolutionary organisations do not always align with Myanmar’s official administrative boundaries. As described in the next section, this becomes an issue when CSOs are attempting to both gain permission from local armed groups (who want aid projects to cover particular areas) and also receive support from international agencies (who may not recognise those areas as being within their target region). CSOs must therefore nuance their communication with different donors and armed groups, using their knowledge of mapping classification differences and local authorities to continue their programs while avoiding disagreements and attention.

These examples from the Myanmar context highlight the ways that CSO leaders draw on accrued experiential knowledges of local dynamics and political relations. Though often complex and opaque, the knowledge and strategies used by CSOs allow them to continue their operations within extremely challenging circumstances, both in areas of regime control and areas of control by other armed groups. In the next section, however, we explore how this reliance on metis in local survival strategies can be suppressed by international donor agency efforts to homogenise and categorise partnerships.

Part 3. The legibility of civil society and unintended harms

We contend, in this article, that international donor efforts to render Myanmar civil society more legible – through standardisation of policy and processes – are having unintended consequences in a range of areas. In the previous section, we highlighted the contextually embedded knowledge that CSOs draw on to manoeuvre through challenges, both with the regime, and also in areas controlled by other armed groups. In this section, we explore the harms caused when CSO reliance on this experiential knowledge, or metis, is suppressed due to international donor policy and processes. We draw on examples related to program documentation, security protocols, and targeting of beneficiaries. As clarified earlier, however, it is not our intention here to argue that all efforts to homogenise and systematise aid projects – or, in Scott’s (1998) terms, all efforts toward legibility – are negative. Our intention here, rather, is to illuminate the ways in which the tendency toward codification and systematisation can cause unintended harm.

Documentation and financial compliance

A key strategy for rendering CSOs more legible in Myanmar is through documentation of program activities and spending. Documentation is, of course, a necessary part of aid partnerships. However, certain policies and rules have exacerbated risks for many CSOs. A key area of policy identified by CSO leaders in interviews was aid program record retention. Within SAC controlled areas, record retention is problematic as, if found by authorities, program documentation can be used as evidence of humanitarian activities that could be aligned with revolutionary forces. Staff in CSOs are vulnerable to crackdowns by the SAC and, as noted earlier, many CSO staff have been arrested and some killed. Tensions have arisen when donor agency policy requires systematic retention of program document hard copies – a requirement that is linked to the dynamics of knowledge and trust outlined above.

A women’s group leader explained the danger of carrying hard copies of program records and how they “have to save the documents underground, digging in the earth” (Int 11, October 2023). She went on to say that this issue of documentation and the associated risks was a repeated topic of discussion with their project funders who said that they had to retain hard copies of records for five years. “We have to keep fighting with [the donor] about this issue”, she reflected (Int 11, October 2023). Eventually, after their “fighting” with the donor, they were successful in making some changes to the policy of their international partner. “Now, they have become more flexible”, she concluded, “and they let us keep soft copies” (Int 11, October 2023).

Despite efforts to hide records, our study revealed examples where local groups have experienced repression by the SAC because of donor requirements to keep hard copies. One local organisation leader recounted how their organisation’s office had been destroyed by the authorities as a result of finding program documentation. The women’s organisation runs safehouse accommodation for survivors of violence and they had a working relationship with some members of the local-level authorities. However, in order to comply with donor rules about project records, they kept a guest list of people in the safehouse. One day, the house was searched by other military authorities, and they found the guest list. These authorities realised that it was a safehouse for survivors of violence and they then went to search the organisation’s office. The CSO leader described the situation: “Our staff had to run away when they got the information in advance … the SAC authorities destroyed the office door and came into the office. Our staff had to flee … ” (Int 8, October 2023).

It should be noted that while this study did not explicitly explore the gendered dimensions of unintended harms to civil society organisations, there were indications from our interviews, that women’s organisations were more vulnerable to attention and repression from the regime. Anecdotal evidence also suggests that, as observed in other contexts, women-led organisations and organisations working on gender issues in Myanmar commonly face greater exclusion from mainstream humanitarian systems than do men-led organisations, in turn, affecting their ability to influence decision-making.

A further example of international donor compliance policies overriding use of accrued contextual wisdom was in relation to records of car rentals. Staff from an organisation that works on conflict transformation would frequently travel between different ethnic armed organisation-controlled areas. Their international agency partner required, however, that they obtain three quotations for any car rental. Such requirements of three quotations are common among international agencies as a way of systematising the use of services and reducing fiduciary risks – and, at a deeper level, as a technical approach to developing knowledge about aid systems. Yet in conflict areas in Myanmar, this standardised system created risks. “We rent cars to travel to the program areas”, said the organisation leader, “but we cannot get three quotations on the ground” (Int 12, October 2023). Private car drivers in conflict areas were often unwilling to have their names on written records as it would draw unwanted attention. Unable to obtain official quotes, staff from the CSO had to explore other, more risky transport options.

Standardised security protocols

Another concern for CSO leaders was international agency security protocols and the ways in which they can adversely affect the local organisation’s humanitarian work. A crucial example of this was in relation to UN agency protocols on visits to field sites. The leader of a women’s organisation explained how their UN agency partner was required to visit field sites twice per year for monitoring purposes. However, UN staff were always required to travel to the site visits in UN cars. Such a standardised policy is understandable in international agencies to control the risks for UN staff around the world. However, as the CSO leader described, it created other risks for them. “We are working low profile”, she explained, “and those UN cars would expose [us] in the field. Then, we’ll be closely scrutinised by the SAC. This could do harm to our local CSOs” (Int 8, October 2023). While the intention of the standardised UN policy was to reduce risk for UN staff and their partners, by suppressing the embedded knowledge and practical perspectives of local CSOs and communities, the policy, in fact, increased the risks for all, especially local aid workers and communities.

Another CSO leader expressed frustration with such security protocols of international agencies and how they presented unmanageable restrictions for CSO operations, which, at times, made programming impossible. “If they keep on highlighting … security updates”, he complained, “we won’t be able to work anymore” (Int 1, October 2023). Local organisations constantly draw on their accrued practical knowledge of the context to weigh risks and possibilities for continuing their programming. However, standardised security protocols of international organisations, such as the example above, can mitigate against the application of these knowledges, thereby undermining CSO program access and effectiveness. This resonates with Heathershaw’s (2008) analysis of the paradox of peacebuilding in Tajikistan, outlined in the first section of this article. The implementation of humanitarian programs in most parts of Myanmar is highly dependent on CSO use of informal and often opaque approaches. Yet international agency systematisation and codification of security protocols can sideline the very processes, and knowledge and associated practices, which make the humanitarian action possible. As the CSO leader above stressed, if they actually followed international agency policy, they could “not work anymore” (Int 1, October 2023).

Targeting of beneficiaries

In his work on the unintended consequences of aid, Koch (2024) describes a range of “conflict effects” in which humanitarian aid sparks violence. In the example in the introduction of this article, a village committee leader was killed, potentially as an indirect consequence of aid targeting. In Myanmar, donors often require that humanitarian aid distribution is targeted towards internally displaced or especially vulnerable people only, but local ethnic armed organisation rules often stipulate that aid should be distributed evenly throughout the village or village tract. Given this tension, in some cases, ethnic revolutionary organisations manipulate donor policy by classifying all the village populations in their identified conflicted areas as “IDPs” (Int 12, October 2023).

The issue of targeting also extended to conflict over administrative boundaries and codification of land. As noted above, in some places, there are key differences in administrative boundaries and place names between the SAC, donors, and ethnic revolutionary organisations. The CSO leader explained the challenge of operating according to MIMU’s system of map codification. “If we’re going according to MIMU as donor requirements”, he reflected, “it is very difficult to work for us on the ground” (Int 12, October 2023). He explained:

The names of the villages on the ground are different from MIMU … but when we talk about locally defined areas that were defined by local people, that was not accepted by MIMU as they could not go against the demarcation of SAC. (Int 12, October 2023)

These differences between MIMU’s SAC-aligned codification and those of local armed groups presented practical challenges for his organisation. He explained that a particular town is, according to MIMU, not within Kayin State. But he said that

it was part of Kayin State according to KNU’s⁷ land demarcation. Therefore, the KNU would request us to include the town in Kayin State while the donor would not accept it as project area since it was not part of Kayin State according to SAC’s [and MIMU’s] land demarcation. (Int 12, October 2023)

International donor efforts to codify places and names for the purposes of coordination had the unintended impact of creating tensions for CSOs in their programming. Rather than being a purely technical dimension of systematisation, mapping has become highly political.

The examples in this section demonstrate the ways in which metis can be sidelined by donor efforts at systematisation and codification. A CSO leader concluded that “localisation is not about working only with local organisations … the main thing is the [international] organisation needs to understand the locally accepted norms and culture … They need to listen to local voices” (Int 12, October 2023). Local organisation leaders commonly expressed this desire for greater awareness of, and respect for, the accrued local wisdom that CSOs would draw on for their survival in times of instability and vulnerability. CSO leaders also warned against forms of international–local partnerships that undermined recognition of local knowledge and associated practices. Such dynamics are common in other aid-recipient countries, yet, in Myanmar’s context of instability and conflict, they are accentuated – with donor agencies ramping up efforts to systematise relationships with CSOs, and CSOs relying more heavily for their survival on informal practices informed by contextual knowledge. In this sense, the Myanmar case illuminates the dangers in rendering civil society legible.

Conclusion: localisation and “ways of being”

CSOs in any context draw on accrued locally developed knowledges in order to operate. In Myanmar, this is particularly the case, as local groups have sought to survive in a high-risk environment since the 2021 coup. As we have seen, however, whether through document keeping, beneficiary selection, security protocols, or seemingly technical details such as delineation of administrative boundaries, donor policies play a role in homogenising and standardising partnerships in a way that can marginalise the use of embedded local knowledge. Analysing international–local partnerships through Scott’s notion of legibility illuminates how donor efforts at homogenisation can have unintended consequences, therefore shedding light on some of the dynamics underlying the widely documented tendency for international aid systems to cause harm to local groups.

In turn, this also has important implications for the way in which localisation is conceived. Influenced by the framing of the Grand Bargain, much consideration of localisation focuses on aid effectiveness (Barbelet et al. 2021; Slim 2021), through changes to funding flows and decision-making roles given to local actors in aid forums. Yet less attention is given to the valuing of tacit embedded knowledges. By highlighting the importance of metis to local organisation leaders in Myanmar, this study suggests that localisation is as much about, as Baquios (2021) suggests, recognition of “ways of being” as about financial flows and decision-making.

Future research in aid-recipient countries could valuably explore aid localisation in terms of the acknowledgement and centring of alternate knowledges in international–local partnerships. A more circumspect approach to partnership, which acknowledges the deep value of accrued practical knowledges, can help donor agencies in Myanmar, and more broadly, to avoid unintended harms to civil society.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by Norweigan Institute of International Affairs.

Notes

1 Koch and Rooden’s work explicitly explores the unintended effects of the aid localisation agenda on partnerships, including through competition, administration, and values.

2 There were eight women and four men interviewed.

3 The SAC’s 2022 Registration of Associations Act places considerable restrictions on civil society groups.

4 MaBaTha, or the Association for Protection of Race and Religion, was a Buddhist ultranationalist organisation active between 2014 and 2017.

5 Although we use the terms “local” and “international”, we also recognise that critiques of aid relationships can essentialise – and reproduce an overly simplistic binary between – these categorisations, when the “local” and “international” are much more interconnected, complex, fragmented, and multifaceted than such binaries imply (Roepstorff 2020).

6 In 2022, the SAC introduced the Registration of Associations Act, which mandates registration for all associations regardless of size or location, under penalty of criminal sanctions. Many local organisations have refused to register, seeing the law as an attempt to further oppress civil society.

7 That is, the Karen National Union – largest ethnic revolutionary organisation in Kayin (Karen) State.