Keywords

The analysis of relationships between providers and recipients of aid inevitably leads to the discussion of power. One can identify two approaches in the development aid literature. The conventional perspective on power primarily emphasizes inequality among actors and aid providers’ predominance. In contrast, the alternative perspective highlights aid recipients’ agency and suggests that inequalities among stakeholders are not constant.

The conventional perspective builds on the discussion of three approaches to power in development aid by Eyben (2008, pp. 36–37), who differentiated between the differences in powers enjoyed by actors, power distribution as a historical legacy, and power as a “process that enables and constrains action.” All three approaches outline specific aspects that sum up to the assumption that inequality among stakeholders is inevitable and is led mainly by donors.

First, the differences in actors’ powers suggest that providers enjoy more power than recipients. As the source of power, aid provides the means for donors to hold the recipients accountable (Hinton & Groves, 2004). This accountability, however, works only one way (Renzio, 2006). There are cases of development aid used by donors as “sanctions” against the recipients (Feyissa, 2011, p. 801), but the recipients do not hold donors responsible for breaking their promises due to the fear of not receiving assistance (Eyben, 2008).

The second approach views power distribution as an outcome of historical legacy. Unequal settings between the global “north” and the “south” created the basis for development assistance. The meaning of “development” traces back to the colonization period, when the initial ideas of what “development” is and who defines it were established. This is reflected in, for instance, the underdevelopment of recipients and donor obligations to bring “progress” into these countries (e.g., Schafer et al., 2012). “Development,” as defined by donors, was imposed on the recipients.

The third approach examines power by viewing development aid as the process that enables and constrains stakeholders’ actions. It suggests that aid per se implies inequality (Robb, 2004) because it underlies “gift-giving” and “gift-obligation dynamics” (Hinton & Groves, 2004, p. 12). Following this approach, inequalities between actors are unlikely to be changed because development aid defines or even preassigns the roles, responsibilities, and opportunities of each actor. Overall, the three approaches above are cornerstones of what I call conventional power dynamics in aid, characterized by inequality in resources and hierarchical roles.

The alternative perspective on power is based on another strain of the literature highlighting the recipients’ roles, the interdependence of the actors involved in giving and receiving aid, and the changing nature of power throughout the development assistance process. Accordingly, recipients may depend on donor assistance, but after receiving it, they weigh the “pros” and “cons” of the objectives of the aid provider and decide accordingly. Following this perspective, they are not “passive” recipients but discuss the terms and conditions of receiving development aid to maximize “their welfare in the face of budgetary constraints” (Lamothe, 2010, p. 5). Recipients may change their behavior if the incentives and benefits offered by donors are higher than the costs of required changes (ibid., p. 19). If not, recipients retain the status quo. Thus, the reforms anticipated and promoted by development aid take place if the recipient is committed to them.

Furthermore, aid relationships between donors and recipients are characterized by interdependence. Development aid involves actors other than only the direct providers and recipients of the assistance, such as parliaments, governments, constituencies, and local municipalities. Both donors and recipients are accountable for the aid they spend. Although the level of accountability varies depending on the role of the public and the political system of the country, it nevertheless ensures the interdependence of donors and recipients. The actors are mutually dependent because the recipients need the donors’ financial resources, and the donors need the recipients’ support to show the “success” of their activities (Shutt, 2006a, p. 154; Swedlund, 2017, pp. 75–76). This interdependence outweighs the hierarchies, as both actors are interested in maximizing the output of the assistance and, therefore, are interested in interacting with each other.

Last, there is an evolving or changing nature of power at different stages of the assistance process. Although they exercise more power during the allocation process, donors nevertheless have limited influence over the outcomes of an aid project. As they provide the project finances (in some cases also ideas), donors are important during reform initiation, but their role decreases during the implementation stages (Andrews, 2013, pp. 209–210). In contrast, the role and power of the recipient (state, civil society organizations (CSOs)) increase. Although nonachievement of the outcomes could result in aid suspension, this is not always the case, and it could also be justified by domestic politics, the pressure of constituencies, or reform opponents (Swedlund, 2017, pp. 73–96).

Overall, the agency provided to recipients, stakeholders’ interdependence, and the evolving nature of power suggest that the inherent inequalities between donors and recipients underlined in the three interpretations of power are not constant. Following these insights, I suggest a framework composed of the following four steps that are intended to provide a comprehensive basis for grasping the aspects highlighted in the alternative approach:

  1. 1.

    Inspired by scholars in political theory, the first step commences with a reflection on the meaning of power and the common terms associated with it, such as resources, consensus/conflict, and interests. It is intended to provide a necessary conceptual basis for understanding the types of power in the context of inequality in development aid. I intentionally focus on classic political theorists, as they, in my opinion, reflect the aid hierarchy best.

  2. 2.

    This book emphasizes the relevance of both stakeholders and the context in which they interact, which is consonant with the agent-structure approach to aid relationships. In so doing, the second step aims to expand on the relevance of individual and collective agency (e.g., organizational level) of abstract categories, such as “donors,” “CSOs,” and the “recipient state.” Finally, in terms of structure, this book focuses on the frequent issues associated with inequality among actors, namely the recipients’ capacities, their dependency on aid, and the flexibility and volatility of aid.

  3. 3.

    The third step calls for a project-level analysis differentiating the following phases of the project cycle: initiation, design, implementation, and evaluation. Empirical analysis at this level offers a detailed yet standardized analysis of development projects, which is beneficial to cross-project comparison.

  4. 4.

    The fourth step culminates the analytical framework by linking the empirical insights from step 3 and the conceptual basis defining stakeholders, power, and the context in the first two steps to a theorization of power dynamics and aid relationships. This step is necessary to understand the empirical cases by placing them in a broader theoretical framework. I built on the seven ways of creating power by Haugaard (2003) because they provide a suitable basis for comprehending the roles and means stakeholders use and the types of power they exercise in relation to each other.

2.1 Conceptualization of Power

To a certain extent, the discussion of power in the context of development leaves the following two impressions: providers and recipients stand in opposition to each other, and their powers are inversely related because if the recipients have more power, donors are presumed to have less power. These impressions recall the perceptions of power as a “zero-sum game,” in which more power for one actor equals less for another. However, this conceptualization of power was criticized by Parsons as early as the 1960s as inapplicable to all cases (Parsons, 1963). Correspondingly, scholars such as Arendt (1970, p. 44) viewed power as “acting in concert” and, therefore, not antagonistic in the relation of one stakeholder to another. Relying on these insights from scholars in political theory, I define the conceptualization and theorization of power that are essential to defining the types of interaction among stakeholders in step 4 of the analytical framework.

The analysis of any complex phenomenon is associated with multiple issues, and power is not an exception (see Dahl, 1957). There are disagreements about its definition, measurement, and nature. While some scholars defined power as a “circulating medium” (Parsons, 1963, p. 236), others denied its existence as an independent entity, viewing power as “a mode of action upon the actions of others” instead (Foucault, 2002, pp. 341–342). Similarly, the essence of this phenomenon, including its directions (bilateral vs. unilateral) (Goldhamer & Shilds, 1973), interpretations (power “over,” “to,” and “with”) (Pansardi & Bindi, 2021), and forms (dispersed or concentrated), remains contested, along with its measurement and feasibility of empirical observations (Dowding, 2017, p. 4). Overall, there is a tendency toward a multidimensional interpretation of this phenomenon that involves synthesizing different approaches (Ledyaev, 2021).

Indeed, the analysis and operationalization of power are inevitably normative (Lukes, 2005, pp. 37–38). Following the focus of this research on the implications of relationships among stakeholders on the sustainability of health aid, I approach power as a socially constructed phenomenon (Dowding, 2017) and a product of a “set of interacting individuals” (Barnes, 1988, p. 61). I differentiate between the power “over,” “to,” and “with” due to their relevance to understanding the power dynamics between stakeholders. Power over is among the first forms, and is defined as A having power over B or as relations among controlling and dependent units (see Dahl, 1957). The “power over” form is often associated with hierarchical relations, whereas the “power to” form closely relates to altering these relations. This difference was introduced by Pitkin (1972) and further reemphasized by feminist scholars such as Allen (1998). The “power with” form was introduced by Barnes (1988) and is based on the presumption that power is not attributed to a single entity, which was further strengthened by Arendt’s (1970, p. 44) view of power as an “act in concert.” There is still an ongoing discussion about interpretations of power and the validity of these differentiations (see Pansardi & Bindi, 2021). However, in the context of development aid, these distinctions are relevant, as they are the key to understanding whether the hierarchy among stakeholders, as in conventional power dynamics, remained or was altered in the course of providing aid or was not present at all.

The “power with” form lies at the core of the analytical approach. This book approaches interaction and sustainability both as an individual and as a collective endeavor. This approach coincides with the perception that power is not something that, as Barnes (1988, pp. 61–62) aptly noted, “radiated from heroic figures; they have glowed with it and illuminated everyone else.” Individuals or entities may enjoy power, but it is nevertheless “embedded” in society (ibid.), and the supportive group enables the presence and exercise of power (Arendt, 1970). In this way, power is not attributed to a single entity or an individual but to a broader constellation of stakeholders and structures.

I approach “power with” as an overarching perspective toward the interaction of all stakeholders and their joint impact on the sustainability of development aid. However, for precision and practical reasons, I assign interaction in a dyadic manner by delineating two broader categories of stakeholders (donor–recipient state, recipient state–CSOs, donor–CSOs, donor–donor). Identifying the interaction types of all stakeholders at once would be practically challenging but also possibly analytically meaningless, as this would not allow the precision necessary for grasping the power dynamics. I acknowledge that both aid relationships and aid sustainability are the outcomes of the “power with” form and not individual dyadic interaction types defined in this book. However, the dyadic focus provides a meaningful basis for grasping how the “power with” accumulates and potentially changes, although this discussion falls beyond the focus of this book.

In addition to this conceptualization of power, the analytical approach introduced in this book ingrains the following phenomena associated with power and relevant to defining aid relationships: resources, (in)compliance, and interests.

First, resources are commonly associated with power, with the premise that more resources imply more power (Hinton & Groves, 2004). These are not limited to material resources and include knowledge and access to them. This assumption underlies the unequal relationship among actors in development aid. Nevertheless, although closely related, resources do not equate to power (Giddens, 1984, pp. 15–16). The way actors approach their resources makes a difference, as stakeholders with the same resources may use them dissimilarly (Dahl, 2005, pp. 273–276). Accordingly, the empirical analysis in this book shows the relevance of resources in understanding aid relationships, for instance, in relation to the incentives that one stakeholder may offer to another. However, it also demonstrates that the difference in resources does not necessarily equal hierarchical relationships among actors.

Second, the (in)compliance of stakeholders with the recommendations and regulations of the other stakeholders is another aspect that is essential to grasping the power dynamics in aid. Here, I focus on sanctions as a “reprisal for nonconformity with a prior act of power” (Goldhamer & Shilds, 1973, p. 300), following the act of incompliance (Parsons, 1963, p. 238). It is important to note that sanctions can be positive or negative (e.g., Baldwin, 1971), the main difference being if the change in the situation is for the benefit or disadvantage of the stakeholder to whom sanctions are applied (see Parsons, 1963). The empirical analysis in this book mainly showed the presence of negative sanctions following the act of incompliance. One of the reasons was the visibility of the conflict. However, this may not always be the case, as conflicts rooted in contradicting interests may be latent and never realized from the outside (Lukes, 2005, pp. 28–29). Similarly, consensus among stakeholders could be implied and is not always expressed explicitly (Dowding, 2011a). The empirical analysis was limited to visible conflicts due to the objections stakeholders expressed in relation to actions taken by the other stakeholders. Overall, both sanctions and consensus/conflict provided a useful basis for examining acts of (in)compliance.

Third, closely associated with power, interests are also essential to defining and validating aid relationships. Power is often defined in relation to forcing one to act contrary to one’s interests or the capacities of stakeholders to realize their interests (see Lukes, 2005). This reference to power and interests further presumes that power is intentional or in pursuit of specific interests. The definition of interests also closely relates to the costs and benefits that those using and are subject to power face and gain (see Dahl, 2005). This accords with the underlying idea of why stakeholders participate in development aid or choose not to do so. However, scholars disagree about the (un)intentional character of power (e.g., Allen, 1998), its relation to objectives (Giddens, 1984), and the ability of stakeholders to comprehend their interests (Lukes, 2005). The empirical analysis encompassed subsections on stakeholders’ interests in pursuing a specific aid relationship form. I argue for stakeholders’ abilities to define and voice their interests, noting that the emphasis on the opposite may unintentionally cause unnecessary victimization of stakeholders. Indeed, actors vary in their access to information and capacities, and yet, as the empirical analysis shows, they have pursued their interests by explaining their compliance with specific aid relationship forms.

I follow the simplistic definition of interests related to the realization of personal and organizational objectives due to the different levels of abstraction pursued in the theoretical approach to the operationalization of actors. Individuals and organizations representing donors and recipients operate in conditions of uncertainty since they are insecure about each other’s actions and the amount as well as the duration of development assistance (Swedlund, 2017). Furthermore, the complexity of development assistance, which is related to a multiplicity of actors, interests, and the areas involved, results in the actors receiving incomplete information. Therefore, I suggest that stakeholders have limited or “bounded” rationality in maximizing their personal as well as organizational interests. “Bounded rationality” means that actors are constrained in their “information-processing” abilities by risks, uncertainty, limited awareness of other options, and the “complexity” of the setting, resulting in an inability to choose “the best course of action” (Simon, 1972, pp. 162–164).

In addition to rationality, interests relate to actors’ perceptions of what is “important” and acceptable from their personal and organizational perspectives, as well as in relation to other stakeholders. The actors’ “mental image of the world” frames their perceptions of and reactions to the ongoing processes and preferences for certain decisions, or what Scharpf (1997, p. 62) defined as “subjective preferences.” Personal perception is also shaped by what is “acceptable and legitimate” from both individual and organizational perspectives (Campbell, 2004, p. 96). This interrelation between individual preferences and acceptability is vivid, particularly in the cases of politically and culturally salient issues. Equally, the actors’ choices are guided not only by personal perceptions but also by a “relational” aspect of the actors to each other (Scharpf, 1997, pp. 69–84), which emphasizes actors’ responsiveness to ongoing processes and others’ reactions to these processes, which also shape their perceptions. This once again reemphasizes the assumption that decisions do not take place in isolation but in the context of not only structural factors but also in relation to other stakeholders involved.

2.2 Conceptualizing Agents and Structures

Following the long-standing discussion on the roles of actors and the relevance of the context in development aid, I emphasize the significance of both actors and structures in understanding power dynamics. This approach corresponds with a meso-level theorization of power as a conceptual tool for specific purposes advocated by Haugaard (2002).

In this book, actors and stakeholders refer to organizations and occasionally individuals whom I approach as agents that act depending on incentives provided in the relevant structures and the roles assigned to these agents (Dowding, 2017, p. 22). An action is defined as changing “the pre-existing state of affairs or course of events” (Giddens, 1984, p. 14). I acknowledge the significance of both individual and organizational levels of analysis but largely keep to the organizational level, except for cases in which individual actions explicitly emanated from individuals and their specific backgrounds, contributing to actions beyond the organizational perspective. This attribution to roles relates to practical concerns, namely, “collective” agency is “easier to comprehend” than that of an individual (Dowding, 2011b, p. 9). Furthermore, I believe that individuals are shaped by the organizations they represent as well as the roles they are assigned to, particularly in the context of aid. This assumption accords with Scharpf (1997, p. 12), who suggested that individuals were “much less free in their actions” but represent certain entities and act on behalf of them. Indeed, individuals also pursue personal interests shaped by their comprehension of reality. The impact of self-interest is specifically relevant to leadership positions, where individuals have fewer organizational constraints (ibid., p. 62). However, even with these positions, individuals are censored by their positions and organizations. In addition to individual and organizational perspectives, agency in the context of development aid closely relates to the roles assigned to “providers” and “recipients.” Therefore, I link individuals and organizations to broader analytical entities, donors, recipient states, and civil society organizations.

Structures are equally significant to power dynamics. “Recursively organized sets of rules and resources” enable and constrain stakeholders (Dowding, 2011c, p. 10), shaping their action and inaction (Lukes, 2005, p. 26). The structures encompass a number of phenomena, but in the context of development aid, some are regularly of specific relevance. For example, although common to development assistance in general, the inequality between the providers and recipients of aid varies across cases. I suggest that aid dependency and the capacity of the recipient are vital to understanding these variations. Furthermore, actors dealing with development assistance face the problems of aid volatility (uncertainty) and aid (non) flexibility. Similar to inequality, these phenomena are common to development aid, although donor policies on these issues vary; therefore, I attribute these factors to the providers and not the recipients of assistance (Table 2.1).Footnote 1

Table 2.1 Structures and their relevance to agents
Full size table

First, aid dependency is critical to understanding the actors receiving aid. A country (in this framework, a recipient state and a CSO) is aid-dependent when it cannot “achieve objective X in the absence of aid for the foreseeable future” (Lensink & White, 1999, p. 13). For example, if stakeholders are interested in conducting specific reforms, aid dependency means that the recipient cannot implement the reforms without the donor. Obviously, financial and institutional constraints may prevent the recipient state or CSOs from implementing the desired reforms or policies independently. However, we need to distinguish between the necessity for “additional” support and the “sole” reliance on it. A recipient country or a CSO seeking donor support in addition to its own resources is not aid-dependent; however, the country fully relying on the assistance is aid-dependent. Although reflected in individual actors, aid dependency remains a structural issue because it is rooted in a broader country/region/global context beyond these actors.

There are different measurements of aid dependence, but this study suggests a sector-specific definition. Glennie and Prizzon (2012) propose a quantitative indicator of dependence, calculated by the ratio of aid to the gross national income of the recipient country. For civil society organizations, this could refer to the ratio of “external” funding to the resources of the organization. These types of indicators are useful for the general ranking of recipient countries/organizations, but they are not helpful for understanding the power dynamics within specific sectors, for example, health care. Generally, a country’s dependence on aid is not equal to its sectoral dependence. The state may receive a large amount of aid but no health care aid. The sectoral division of the assistance provides a more accurate picture, but even in this case, the numbers might be misleading. For instance, the share of “external” health expenditure as part of current health expenditure in 2016 in Kyrgyzstan was approximately 4% (World Bank Group, 2023). One may assume that the country is relatively “independent” from aid because public (state) and private (patients) contributions to health care are much higher than those from donors. However, the empirical chapters in this book demonstrate the opposite. Although independent at the general level, the country relies on technical assistance in health care reforms or financing in terms of access to treatment. Therefore, the analytical framework presented here suggests that a more specific sector or subsectoral focus provides a better understanding of the aid dependence of recipients.

The second factor relevant to understanding the differences in power dynamics is capacity. Broadly defined as “the ability of people, organizations, and society as a whole to manage their affairs successfully” (OECD, 2011, p. 2), capacity in a narrow sense refers to the individual, organizational, and systems’ abilities/“competencies” to implement their functions (European Centre for Development Policy Management, 2008, p. 2). Based on these definitions, this chapter operationalizes capacity as a recipient’s ability to perform its functions and administer its activities with a focus on the availability of human resources. Human resources are essential to negotiations, implementation, and the evaluation of development assistance. A limited capacity, reflected in an insufficient number of staff members and their qualification issues, causes communication problems with donors.

However, again, critical to understanding stakeholders, “capacity” is still an outcome of broader issues in a given sector or country and is not limited to an individual actor. Thus, in her interviews with donor representatives in Sub-Saharan Africa, Swedlund (2017, pp. 92–93) highlights the staff shortages and computer literacy problems of the recipient countries. Limited capacity is related to and caused by a “brain drain” from public institutions. Qualified staff members are often recruited by donors offering better remuneration and advancement policies (Swedlund, 2017; Toornstra & Martin, 2013). Similar issues with staff retention are noticed in the case of CSOs (Frontera, 2007), although there are differences within this group. The level of staff rotation in community-based organizations (CBOs), where members work on a voluntary basis, might be higher than in a nongovernmental organization (NGO), which pays its employees and provides additional nonfinancial incentives, such as training and travel.

Third, stakeholders operate in conditions of uncertainty related to aid appropriation procedures and relatively short development program durations. Aid volatility varies depending on aid modalities, with budget support being more predictable than project-based assistance. There is a general acknowledgment of the need for increased aid predictability (e.g., Menocal & Mulley, 2006). Correspondingly, the emphasis on long-term partnerships in institutional capacity-building in the 2000s, for instance, contrasts with the ad hoc assistance provided in the 1990s (Leitch, 2016, p. 195). Nevertheless, donors have different aid appropriation procedures and opportunities to make commitments before their partners. Bilateral aid from OECD countries often depends on the annual appropriations approved by their parliaments on the basis of their governments’ proposals (Isabekova, 2019). Multilateral aid, by contrast, depends on the contributions of funding countries and organizations. Overall, making long-term commitments beyond the scope of the specific project is problematic in cases of both bilateral and multilateral donors.

Fourth, aid flexibility is equally important to understanding stakeholders. The flexibility of donors has been emphasized in relation to the ability to adjust to local priorities and contexts (Hirschhorn et al., 2013). Strict regulations from the parliament or the government negatively impact the flexibility of the assistance by assigning it to certain purposes. Thus, driven by the goals defined by the “central” authorities of donor agencies and not necessarily by the recipients on the ground, tied aid is commonly viewed as the opposite of flexibility. However, in addition to preset objectives, flexibility is also closely related to responsiveness to changes taking place on the ground. This relates to what Leitch (2016, p. 215) calls the “institutional factors on the donors side.” For this reason, in addition to adherence to recipients’ priorities, I explore the decision-making authority held by the field offices of donor organizations. Authorities delegated to the “field” offices of donor organizations contribute to the ability to respond to changes on the ground. In contrast, a highly centralized organizational structure means the concentration of decision-making authorities at relevant institutions or headquarters. As a rule, this affects aid responsiveness due to bureaucratic delays.

Overall, the reflections on the roles of stakeholders as individual and collective agents and structures in the form of key issues attributed to aid relationships complement the conceptual definition of power in the first step. A combination of these two steps offers a valuable guide to the empirical analysis in the following step.

2.3 Project Life Cycle

To grasp the power dynamics in development assistance projects, I suggest a project-level analysis by differentiating the four phases of the assistance, namely, initiation, design, implementation, and evaluation. These stages are not consecutive since evaluation, for instance, can take place before, during, or at the end of the project. However, differentiation into phases allows for the roles of actors, as well as the division of labor between them, to be analyzed throughout the assistance process. Following Bachrach and Baratz (1962), power closely relates to agenda-setting and prioritizing the issues one wants resolved, irrespective of their relevance to the subject. Thus, a detailed analysis of project phases allows the analysis of power dynamics, including but not limited to, agenda-setting and observing stakeholder participation throughout the project realization process.

2.4 Uniting Theory and Empirical Findings

This step links the empirical insights from step 3 and the conceptual basis defining stakeholders, power, and the context in the first two steps to a theorization of power dynamics and aid relationships. It builds on the seven ways of creating power, defined by Haugaard (2003), as it offers a suitable basis for grasping the “power over” and “power to” forms in aid projects.

His theorization intends to organize and explain the analysis of and insight into power by scholars such as Parsons, Luhmann, Barnes, Clegg, Giddens, Bachrach and Baratz, Foucault, Lukes, Weber, Dahl, Mann, and Poggi (see Haugaard, 2003). In his theorization, Haugaard provides some structured way of understanding the power dynamics between stakeholders. His overview was utilized by Shutt (2006b) to analyze aid relationships, inspiring the inquiry and application of this approach in this book as well.

The first way of creating power is through social order, which derives its essence in the predictability assured through the intended reproduction of meaning accepted and emulated by others (Haugaard, 2003, pp. 90–91). Two practices are highlighted in this regard, namely, “structuring,” which occurs through attributing a similar meaning to actions irrespective of time and place, and an agent that intentionally exercises these actions (ibid.). In development assistance, for instance, individuals and organizations aim to demonstrate the ownership of aid recipients by negotiating the objectives and adjusting them accordingly. These actions, conducted in numerous couloirs, represent the intended reproduction of this principle. The similarity and impersonalization in the reproduction of meaning contribute to predictability, the foundation of the social order, which also requires its broader acceptability, which is ensured by the second element. The “confirm-structuring” embodies the “public” and “willing” reproduction of the meaning, pointing to the acceptance and consensus regarding the meaning (ibid.). Accordingly, the practice mentioned above is broadly acknowledged and adhered to (more or less) by various stakeholders in multiple countries and sectors. However, consensus is not always the case, or as Haugaard notes, acceptance does not preclude conflict, as the social orders taken for granted today may not have been prevalent but were “fought for” in the past (ibid., 96). Similarly, ownership and actions regarding aid were not common throughout the history of aid.

The second way of creating power, that is, structural bias, is closely connected to the former, as it ensures the predictability and stability of the social order. Nevertheless, system bias represents a different source for creating power through structural constraints imposed by one actor upon another (Haugaard, 2003, p. 94). It is characterized by the process in which stakeholders or initiatives inconsistent with and aiming to change the prevalent social order face the “noncollaboration” of those who reproduce it (ibid.). Thus, the constraints would ensure stakeholder compliance with the principle of ownership and the unacceptance of those unwilling to implement it upon receiving assistance. This supports the premise that structural biases are not necessarily negative but rather essential to the stability of the social system (Haugaard, 2003, pp. 93–96). However, constraints do not empower stakeholders to the same extent. Predictability is ensured at the expense of certain forms of interaction and power that might have benefited stakeholders disadvantaged by the social order (ibid.). However, actors comply with constraints for reasons that are not necessarily “consciously chosen” and could be an outcome of structural and cultural patterns and other reasons (Lukes, 2005, pp. 25–26). In addition to consensus regarding the order, social biases may also involve conflict in the form of destruction or evasion from supporting the system (Haugaard, 2003, pp. 95–96).

The third way of creating power relates to social consciousness, which is in the foreground of the reproduction of specific order and biases. Predictability here derives from stakeholders’ perceptions that specific meanings are not “arbitrary” and exist “out there in the world” but are consistent with their interpretations of the matter at stake and the world in general (Haugaard, 2003, pp. 97–99). Thus, the critical aspect of reproducing meaning and biases lies in their relation to the actors’ systems of thought. For instance, gender equality and women’s empowerment in development assistance relate to the consciousness of gender equality in a broader context. Prevalent among members of the Organization for Economic Co-operation and Development’s Development Assistance Committee, this explicit emphasis today may not have been common some time ago or in contexts with different perceptions of gender. Similarly, the reasoning behind some health care intervention programs relating the cognitive abilities and behaviors of individuals to their ethnic background would be unthinkable today (see Morgan, 1993). In this way, changes in the system of thought result in practices common to the previous system becoming obsolete (Haugaard, 2003, pp. 98–99).

The fourth way of creating power is the relationship between tacit and discursive knowledge. It is based on Giddens’ (1984, pp. 4–5) differentiation between the “practical” or unconscious knowledge actors use in their social lives and the “discursive” or conscious knowledge underpinned by their reflection, rationale, and reasoning. Haugaard (2003, pp. 100–102) notes that most knowledge in social life is tacit to ensure stakeholders’ comprehension of reality. However, a transformation of this tacit knowledge into discursive knowledge allows distance, evaluation, and recognition (ibid.). An instance of this transformation could be, for instance, reflections on the meaning of development, which changed from an initial economic focus to a broader operationalization of development to include social, environmental, and other factors. Haugaard notes that tacit knowledge may benefit those in power, but its transformation into discursive knowledge may change the order if actors realize that the reproduction of social structures disadvantaged them (ibid.). Following the example mentioned above, critical reflections on development, in combination with the evidence for the noticeable shortcomings of a solely economic reform focus (see Cornia et al., 1987, 1988), offered favorable circumstances for targeting social needs neglected before.

The fifth way of creating power, that is, reification, stabilizes existing power relations and structural reproduction on the premise that they embody something “more than social constructs” (Haugaard, 2003, pp. 102–103). This source of power resembles the system of thought but differs from it in one dimension. Stakeholders conform to structures and related practices, not because of their relations to actors’ perceptions of reality but for other reasons. This resemblance and distinction are not explicitly stated in the theoretical framework but are nevertheless critical to its valid application. Following Haugaard (2003, pp. 103–105), “nonarbitrariness” in reification could be based on multiple grounds, including religion, nature, truth, and science, with practices incompliant with existing structures as going against these concepts. For example, health care programs frequently appeal to science to support their objectives and activities, which could be a vivid example of reification in practice.

The sixth way of creating power, that is, discipline, is manifested through socialization aimed at establishing a routine based on practical consciousness (Haugaard, 2003, pp. 105–106). This, in a way, follows the logic opposite to the one described in the relationship between tacit and discursive knowledge. It does not result in predictability but originates from it (ibid.). The assumption consonant with what Barnes (1988, p. 58) referred to as social power is at the disposal of those able to judge and decide upon routines. Haugaard (2003, pp. 106–107) notes that as a relatively modern phenomenon, this power is based on the premise that discipline is not arbitrary, and practices inconsistent with it are “irregularities” and foes of the social order. In the context of development assistance, disciplinary power could relate to functions agreed upon and assigned to stakeholders in projects and the relevant training they received. However, closely related to stakeholders’ socialization and use of practical knowledge, disciplinary power is also limited to the extent to which stakeholders “internalize” the suggested routines (ibid.). These limitations are explicitly discussed in Chap. 8 of this book.

The seventh way of creating power is coercion, which represents the last resort measure. Haugaard (2003, pp. 108–109) builds an analogy with physical power and suggests that it comes into play when other sources of power fail, approaching coercion as neither an “ultimate” nor “effective” source of power. This perception is consonant with that of Arendt (1970), who defined coercion as the weakest form of power, as well as that of Parsons (1963), who stated that coercion was not equal to power. Coercion is often discussed in relation to freedom. However, presuming that stakeholders have freedom and make decisions depending on their interests and structural factors, I approach coercion as an act of compelling that is contrary to the will and interests of stakeholders. Although associated with physical power, coercion takes other forms in stakeholder relationships, as discussed in Chap. 8 of this book.

Overall, the seven ways of creating power help theorize the roles of stakeholders and their actions in relation to each other, as empirically discussed in the third step. This theorization is necessary to grasp how the conceptualization of power, agents, and structures unfolded in selected cases. More specifically, by defining ways of creating power, I expand on the types of power, how actors used resources, or whether a conflict/consensus accompanied this process. This theorization is critical to validating the empirical findings in relation to the type of power (“over/“to”) stakeholders created.

The type of power is essential to understanding whether an alternative approach to stakeholder relationships, characterized by interdependence, recipient agency, and changing power dynamics, occurred. In contrast, it could also be the case that the conventional approach to stakeholder relationships, mainly characterized by inequality, took place. However, I refrain from attributing specific ways of creating power to any of these two approaches or types of aid relationships discussed in the following subsection. The discussion of the seven ways clearly showed that the “power over” or the “power to” forms could equally be produced by some stakeholders, depending on how they use their resources or whether there is a conflict/consensus. Nevertheless, the seven ways of creating power provide a good indication of whether “power over” and/or “power to” emerged between stakeholders. Assuming that the former is a characteristic of the conventional approach to stakeholders and their relationships with each other, I suggest that the types of relationships associated with inequality have “power over” as the prevailing form. In contrast, aid relationships associated with equality have “power to” as the prevailing form of power emerging between actors.

However, given that power may be changing throughout the project cycle, I acknowledge that in some cases, it could be a combination of the “power over” and “power to” forms. Here, the roles of stakeholders throughout the project cycle and the analysis of structures (aid dependency, capacity, aid volatility, and flexibility) may help. If stakeholders were equally engaged throughout the project life cycle and structures changed in favor of the aid recipient, the “power to” prevailed; otherwise, the “power over” form prevailed. Why does this matter? Equal engagement of stakeholders addresses the problem of limited involvement of aid recipients and aid fragmentation (Chap. 1). It also demonstrates the recipients’ agency and abilities to raise issues or participate in decision-making. Structures, in turn, demonstrate whether the context in which aid relationships took place was favorable to changing the hierarchy underlined in the conventional approach. I argue that aid volatility and inflexibility of providers, as well as aid dependency and limited capacity on recipients’ sides, are favorable to retaining the hierarchy among stakeholders. In contrast, aid predictability and flexibility, accompanied by relatively limited aid dependency but necessary capacity, are beneficial for altering the hierarchy.

I argue that considerations regarding engagement in the project cycle and structures are case-dependent and not attached to a specific type of aid relationship. However, inequality and hierarchy are attributed to certain types, as discussed below.

Aid relationships in this book encompass a variety of relationships between stakeholders, including noncoordination, coordination, unequal cooperation encompassing recipient/donor-driven cooperation, and a “utilitarian approach toward CSOs, and equal forms of cooperation, such as partnerships, and an “empowerment” approach toward CSOs. I acknowledge that my findings in this book demonstrate the power dynamics limited to the time and space covered in the empirical chapters. However, further theorizing regarding the power dynamics in aid relationships helps us abstract from project phases to a broader conceptualization of stakeholder interaction. Therefore, this section operationalizes different types of relationships, building on the findings and filling the gaps in the relevant academic literature and international agreements. The suggested types of interaction are Weber’s “ideal” types, which combine the features found and relevant to them in reality but in a manner that is coherent with each type of interaction (Oxford Reference, 2023).

First, noncoordination may range from the noninteraction of actors with each other to the noncompliance of one actor with the priorities of another. Among donors, this means no exchange of information, resulting in unawareness of each other and a subsequent possible duplication of activities. For interactions between donors and recipient states, noncoordination may refer to a donor(s) pursuing activities without exchanging information with the state or without complying with its priorities. Regarding the interaction of donors and the recipient state with CSOs, noncoordination is expressed by the noninvolvement of the latter in development assistance. Since the coordination of activities is time- and resource-consuming, noncoordination could be beneficial to actors from the short-term perspective, as this does not require time and an additional workload in contrast to coordination or cooperation.

I suggest that noncoordination takes place in cases of inequality between donors and recipients. The recipient is too aid-dependent to raise the issue of noncoordination or has no capacity to require/implement the donors’ compliance with its requirements. Donors, in turn, are disincentivized by the time and resources needed for coordination to initiate this voluntarily. These disincentives may further be exacerbated by limited aid flexibility and certainty. As coordination and cooperation presume a certain level of adjustment and awareness of aid flows, aid volatility and inflexibility may be accompanying factors of noncoordination. Potential incentives for donors to coordinate with each other could relate to increasing their influence over the recipient. However, as the recipient is aid-dependent, each donor may already have leverage over it and may not understand the benefits of coordination.

Second, coordination among stakeholders is expressed by the parallel implementation of activities with information exchange. It is characterized by an agreement on priorities, but actors pursue their activities toward these priorities without involving each other. This agreement on priorities entails the recipient’s capacity and/or the donors’ willingness to engage in coordination. The recipient’s capacity is essential since the recipient is expected to request and, most importantly, ensure donor coordination and compliance with its priorities. However, coordination could also be the outcome of donor initiatives. The reasoning behind the coordination taking place as a result of the donor’s initiative could be (in addition to altruistic motives) the donor’s interest in increasing their influence over the recipient. Thus, the recipient may still be aid-dependent, but the influence of the individual donor may not be sufficient or as high as in the case of noncoordination.

Coordination is beneficial to both recipients and providers of aid in the long-term, as it decreases transaction costs. Therefore, similar to cooperation, coordination has been emphasized in a number of international documents (e.g., OECD, n.d.). However, the costs stakeholders face overshadow its long-term benefits in the short term. Its establishment requires staff involvement and time for negotiations that may take up to several years (Lawson, 2013). Facing a trade-off between long- and short-term benefits, the actors may favor the latter.

Third, in the context of development assistance, I define cooperation as a joint realization of aid, which may take unequal and equal forms. The former takes place when one of the actors, be it a provider or a recipient, dominates in the aid realization process, while the latter means that actors are equally engaged in the process. This equality is not only about the stakeholders’ responsibilities but is also about their roles and “say” in the process. Although “equal” and “unequal” labels of cooperation inevitably recall normative connotations, I approach them as mere variations of cooperation. Following Weber (1986, pp. 28–29), I acknowledge that domination, in its general meaning, is an essential component of social action and can take different forms in regard to how one imposes their “own will upon the behavior” of others. Dominance in unequal cooperation takes different forms, depending on the stakeholders involved.

Between donors, cooperation means that one donor relies on the operational procedures of another and complies with its regulations and approaches. The donor may have a “leading” role due to a larger share of finances or taking over the responsibility for the outcomes. This emergence of the “dominant” actor could be the outcome of other donors being “less motivated” to compete or “much poorer” (Bueno de Mesquita & Smith, 2016, p. 2). Cooperation presumes a certain level of flexibility and certainty necessary for negotiations among stakeholders and relevant adjustments that are possibly more demanding than that required in coordination, in which stakeholders agree on priorities and not on operational procedures and approaches.

In relationships between donors and recipient states, unequal cooperation takes two forms, namely, cooperation driven by donors and their conditions and aid driven by the recipient state and its priorities. There is an extensive discussion of conditionality in the development aid literature concerning the requirements the recipient was expected to fulfill in economic, political, and other terms (e.g., Crawford & Kacarska, 2019; Molenaers et al., 2015). This type of interaction is probably characterized by aid dependency and capacity issues on the recipient’s side that also define the flexibility and predictability possible and provided by the donor. Thus, if the recipient is relatively aid-independent and/or has the capacity, the donor may be forced to be more flexible and predictable to enforce its conditions.

Another form of unequal cooperation between the donor and the recipient is when aid is driven by the priorities of the latter, with donors adjusting their activities accordingly. The idea of the recipient state being the “driver” of interaction complies with the notion of “ownership,” as emphasized in the Paris Agenda, by increasing its role in the process and in achieving assistance results. This type of cooperation presumes high capacity and relative independence of recipients from external aid, accompanied by flexibility and predictability of assistance, which may not be immediately offered but are achieved during the negotiation process.

In civil society organizations’ interactions with other stakeholders, unequal cooperation refers to a “utilitarian” approach. This approach was initially used to discuss community participation in development aid (Morgan, 2001). The analytical framework presented here extends to explaining the interaction of CSOs with other actors. Following the “utilitarian” perspective, communities (in this book CSOs) are involved in development assistance through “passive means” used to reach the project objectives (Rasschaert et al., 2014, p. 7). According to the “utilitarian” perspective, CSOs are dependent on “external” assistance due to their low capacity and structural barriers (ibid.). These include, for instance, illiteracy (Jana et al., 2004), gender-related biases (WHO, 2008), the political situation in the country, and poverty (Fawcett et al., 1995). In these conditions, donors may not have sufficient incentives to offer the predictability and flexibility regarding their assistance. Moreover, even if provided, CSOs may not be able to negotiate for these characteristics because of the fear of losing access to funding.

Fourth, equal cooperation takes place when all the actors are involved in the aid realization process and have an equal say in it. None of the parties dominate aid realization. This notion of equality recalls partnerships that are founded on equality, trust (Hyden, 2008), nonconditionality (Abrahamsen, 2004), and shared responsibilities and authority. Trust is ensured in partnerships in which stakeholders fulfill their commitments (Del Biondo, 2020); however, this is often problematic, as actors may break their promises in the face of pressure from their constituencies, parliaments, and so forth. Ideally, a partnership has no conditions, meaning that stakeholders fulfill their responsibilities voluntarily. However, conditionality may be inevitable and still be present in a partnership as long as it applies to all parties. Furthermore, I also define shared responsibilities and authorities that are essential to partnerships, as these provide a necessary underpinning for equality in the aid realization process.

In general, equal cooperation or partnership rarely takes place in practice, although there are some exceptions. The most pressing issue for donors in equal cooperation contexts is the adoption of joint procedures, which is essential to the joint implementation and evaluation of the assistance. These aspects require lengthy discussions and a high degree of flexibility and predictability. Given the complex structures of donors and their adherence to their own rules, this might be problematic to implement in practice. Between donors, equal cooperation seldom occurs because of “harmonization” issues, although there are exceptions among medium-sized organizations sharing similar perspectives (see Isabekova, 2019). In the relationships of donors with recipient states, equal cooperation similarly assumes the presence of equality, trust, nonconditionality, and shared responsibilities between the parties. This is problematic due to power dynamics and the inherent inequality between the actors, as discussed in the beginning of this chapter. The recipients might be reluctant to participate or criticize the donor because of the fear of donors cutting funding (Hinton, 2004). As the agenda is still set by donors (Nissanke, 2008), partnerships (or equal cooperation) might be viewed as “little more than rhetoric” (Abrahamsen, 2004, pp. 1455–1456). Because of aid dependence and limited capacity, equal cooperation rarely takes place between recipient countries and their donors.

Finally, the relationship of donors and recipient states with the CSOs’ definition of equal cooperation in the analytical framework presented here is based on the “empowerment” approach. Similar to the “utilitarian” perspective described in unequal cooperation, this approach was initially suggested for community involvement (Morgan, 2001, p. 223). This chapter extends it to cover CSOs. Empowerment is a “process of gaining influence over conditions that matter to people” (Fawcett et al., 1995, p. 679). In development aid, CSOs are able to express their concerns, set priorities, and participate in negotiations and the decision-making process. They equally cooperate with other partners by participating throughout the assistance process. Following this approach, CSOs are viewed as the source of initiative rather than “passive” aid recipients (Morgan, 2001, p. 223; Rasschaert et al., 2014, p. 7). However, there is inherent inequality between donors, recipient states, and CSOs because of the differences in resources and the structure of development assistance (see the section on power dynamics). The power dynamics further vary among CSOs. CBOs are relatively aid-dependent and require more capacity-building activities. There is evidence that at the end of a development project, CBOs will continue its activities if it continues to receive funding from another donor; otherwise, they will cease or decrease their activities (Ahluwalia et al., 2010). The dependence of CBOs on donors is clearly illustrated by the statement of one CBO member in Central Asia: “getting a grant is similar to receiving money from God” (Earle et al., 2004, p. 34). In contrast to CBOs, NGOs might also be aid-dependent but have a relatively higher capacity, although there is variation between local, national, and international NGOs. An organization with several branches across the country or in several countries has more human and financial resources than one operating in a village or a town.

Referring to aid relationships, I intended to synergize the narrow and general approaches in the international documents and development aid literature. The classification of the “ideal” types of interactions among stakeholders presented above serves two purposes. First, it provides the level of abstraction necessary to observe the link between interaction among stakeholders and the sustainability of development aid. Second, this level of abstraction is critical to our comprehension of complex relationships and learning beyond specific cases. In other words, how cases are selected for empirical research may help us understand both aid relationships and their connotations of sustainability in other contexts.