Organizing in the Shadow of Donors: How Donations Market Regulates the Governance Practices of Sponsored Projects in Non-Governmental Organizations

The purpose of this study is to analyze the influence of donors in the organization of governance practices of sponsored projects in non-governmental organizations (NGOs) from an agency theory perspective. For this, the sponsored project was understood as a temporary organization. The study relied on data collected through document survey in contracts and call for tenders for project support, and on semi-structured interviews with executives from donor organizations in a Latin America country. The study found that the governance of NGOs is impacted by temporary governance derived from projects. In turn, the projects are circumvented by aspects of compliance and enforcement through pressure from donors, and transferring donor management processes to NGOs produces two shadows over NGOs: (a) there is a public organizations ’ influence of laws, norms and actions of state decentralization that demands NGOs to organize themselves like the State; and, (b) on the corporate donors side, an organization aligned with business is required.


Introduction
In the last 30 years, Latin America countries underwent several changes after the end of authoritarian regimes and the rise of democratic governments in the context of market economies (Dagnino, 2002;Weyland, 2004). Among the changes, we highlight the development of civil society and multiple reforms of the State. In the context of civil society, the 1990s were marked by the phenomenon of NGO-ization (Alvarez, 2009), which led social movements to structure themselves as organizations, depoliticizing their discourse and engaging in social projects financed by private agencies or by governments in the implementation of public policies. When structuring themselves as organizations, Latin American NGOs adopted practices of planning, strategymaking and managing that are typical of the private sector, in search of improving their performance in the access to resources, whereas this also put them in ambiguous positions before public or private interests (Landim & Thompson, 1997).
At the level of the State, reforms in public administration in countries such as Argentina, Brazil, Chile, Colombia, Dominican Republic, Guatemala, Honduras, Peru and Venezuela have increased the capacity of governments in the execution of public policies through administrative decentralization, as well as restructuring public service delivery (Saravia & Gomes, 2008). Despite the often frustrated efforts to implement reforms in Latin America (Pereira, 2002), countries such as Brazil and Chile have been relatively successful in their reforms by adopting state reform models more focused on a managerial approach inspired by the New Public Administration (NPM). In this context, the influence of the NPM agenda promoted a paradigm shift in philanthropy (Dangino, 2011). The traditional approach of almost granting social services to NGOs in a non-competitive environment has given way to an environment in which donors awarded contracts to NGOs "not because of what they are but what they can do" (W. P. Ryan, 1999, p. 129). This extended the external responsibility of NGOs, which were increasingly forced to demonstrate performance in terms of efficiency, effectiveness and transparency (Arvidson & Lyon, 2014;Silva, 2010) and to introduce management models for this purpose (Smith, 2010). Therefore, for NGOs to deal with donation market constraints, they have increasingly adopted private, for-profit market approaches, leading to what Salamon (1997) termed as the marketization of a non-profit sector.
Partnerships with business organizations, state-owned companies, corporate foundations and other NGOs, resulting from fundraising processes that trigger organizational arrangements to coordinate the activities rewarded from partnerships, in addition to the actual competition for public and private resources among NGOs, have put pressure on NGO managerial professionalization. Such situation has driven them to migrate from a philanthropic model to a corporate model (Lacruz & Cunha, 2018). Under the assumptions that NGOs need resources to survive and that there are requirements they must conform to in order to receive financial support, the logic is that NGOs are led to act in adherence to those who control the resources in the donation market, i.e., corporate foundations and public and business organizations. Thereby, they are driven by prevailing practices and procedures of organizations that make the donation market.

OPEN ACCESS
NGOs have thus adopted management practices similar to those of for-profit business organizations (Eikenberry & Kluver, 2004), among which practices of governance. There is a reasonable framework of study on NGO governance. The general argument is that certain NGOs can differentiate themselves with good governance. Such organizations would be considered more attractive to donors of resources, which would lead them to receive more resources in donation (Harris, Petrovits, & Yetman, 2015). Studies have addressed processes related to internal governance control mechanisms, in particular within the limits of the board of directors (Willems, Andersson, Jegers, & Renz, 2017). Governance in NGOs has been analyzed through the presence of governance mechanisms in its organizational structure (e.g., Bromley & Orchard, 2016), but the forces (mainly external) that act for the organization of such mechanisms have not been addressed. We observed that this research gap remains without due treatment, which led us to investigate how the governance of sponsored projects in NGOs is configured in the current context of external pressures. Towards that, sponsored projects (that is, the transfer of money from a sponsor to an NGO that may require the performance of specific duties) were understood as temporary organizations. A temporary organization is an organization with an organized course of action aimed at evoking a non-routine process and achieving a goal that has a predetermined point in time or conditional circumstance related to time when the organization and its mission is collectively expected to cease to exist (Ludin & Söderholm, 1995). In addition, empirical studies with a sample of NGOs in a context of low regulation and difficulty to access information about NGO in Latin America in general and in countries such as Brazil in particular are rare. Thus, in environments of low regulation and access to information, relationships can be found between donors and NGOs that complement extant theory.
We use the agency theory approach (Jensen & Meckling, 1976) to examine how the donations market acts to configure the governance rules of sponsored projects in NGOs. In particular, we have developed a study that dialogues with the premises of information asymmetry, different levels of risk aversion, and different planning horizons (Eisenhardt, 1989;Jensen & Meckling, 1976), considering donor companies as the principal and NGOs as the agents in the agency relationship. The research produced data collected through document survey in contracts and call for tenders for project support, as well as semi-structured interviews with executives from donor organizations. The results showed that NGOs could institute governance practices that are limited to project duration (object of the contractual relationship) due to the process of marketization. That is, the project can imply the aspect of the temporality in NGO governance.
In the next section, we present the theoretical framework that supports the discussion. Next, the methodological approach, the results, and the discussion of findings. Finally, the conclusions of this study.

Theoretical Framework
Multiple definitions exist for corporate governance in academic literature, best practice guides, market institutions, and international bodies. Although there is no agreed-upon definition, there is consensus that it refers to the direction and control of an organization. Most conceptual influence comes from the Cadbury Report (Cadbury Committee, 1992) and the principles of accountability, disclosure, compliance and fairness, which are possibly influenced by the definition of the Organization for Economic Co-operation and Development (OECD, 2004) on corporate governance principles. Following the Cadbury Report in 1992 in the UK, other recommendations in the form of codes of best practices were elaborated in many countries (e.g., the UK corporate governance code), including specific ones for the third sector context, like the Standards for Excellence -an ethics and accountability code for the nonprofit sector (Maryland Nonprofits, 2014); and, in Brazil, the Guia das melhores práticas para organizações do terceiro setor: Associações e fundações (Instituto Brasileiro de Governança Corporativa [IBGC], 2016).
We adopt the explanatory basis of agency theory assumed in this study from the classical approach by Jensen and Meckling (1976), from which numerous empirical studies and new theoretical models have been developed (e.g., Eisenhardt, 1989;Glaeser, 2003). In that approach, the firm is defined as a legal fiction that serves as the focus for a complex process in which conflicting individual objectives are brought to equilibrium within a contractual relations framework. The firm, then, represents a way of efficiently integrating the conflicting objectives of the various participants in a legal contractual context. That is, firm behavior is similar to market behavior, resulting from a complex equilibrium process.
In summary, agency theory refers to the relationship between a principal and an agent, in which the principal engages the agent to perform something involving the delegation of decision-making and authority by the principal to the agent; that is, the agency relationship. We assume there are divergences of interests between the principal and the agent because each has different utility functions and there is no perfect contract capable of securing the interests of the principal. This situation is termed the agency problem. As a result, principal and agent incur in agency costs: the principal to align the agent's interests with their own, and the agent to show the principal that their acts are not harmful. Also, the knowledge needed to make different decisions is dispersed among several agents, and there are costs of knowledge transfer between agents. All this implies developing internal and external control systems to reduce agency problems. In this, corporate governance arises with the objective of mitigating agency problems. Pauly and Redisch (1973) were among the first to empirically examine governance in non-profit organizations, more specifically in the context of non-profit hospitals, while Glaeser (2003) was among the pioneers in proposing different utility roles for different actors assumed as principal in agency relationship. For Jensen and Meckling (1976) agency costs arise in any situation involving cooperative effort between two or more persons, even if there is no clear agent-principal relationship. Therefore, we can extend the concept to NGOs, because the problem of agency is not restricted to situations in which there is a hierarchical relationship between principal and agent, rather being present in all activities involving cooperation ties, even if there is no clear agent-principal relationship.
As pointed out by Glaeser (2003), "Non-profit organizations have governance problems that resemble the problems in for-profit firms but are often far more extreme" (p. 39). Fama and Jensen (1983) and Jensen (1994) also advocate the applicability of agency theory to NGOs, for whom the problem of agency is not solved only by the incentive to altruism, typical in NGOs. In this sense, there is a need to institute mechanisms that minimize the problems of agency, observing that such mechanisms overburden agency costs. Nevertheless, it is a recurrent theoretical lens in studies on governance in the third sector (Renz & Andersson, 2014).
From the perspective of agency theory, corporate governance is regarded as a set of internal and external mechanisms aimed at mitigating the agency problem derived from the separation between ownership and management of organizations. Jensen (1993) classifies governance mechanisms as external and internal. External factors refer to factors that are exogenous to the organization, therefore subject to the economic, social and cultural environment, and other factors (such as the capital market, and the political, legal and regulatory system). Internal factors refer to the initiatives of prescribing, monitoring and controlling that are endogenous to the organization, therefore subject to greater control -which Jensen (1993) generally referred to as the internal control system exercised by the board of directors. Glaeser (2003) argues that the donor market in the third sector is equivalent, as an external governance mechanism, to the capital market for publicly held companies. The capital market reflects, by the value of shares in the stock market, the current company performance and the expected future investors performance. This can be understood as an incentive for majority shareholders, in capital concentrated companies, or boards of directors, in dispersed capital, to monitor managers' actions. This acts, therefore, as an incentive on managers to increase the value of shares. Although Glaeser (2003) recognizes that, in donations market, as there are no shareholders, there is no possibility of hostile takeovers, thus (a) donors are free to apply their resources to NGOs that they consider most appropriate, and (b) the risk of NGO reputation losses in the case of inefficient or improper management is an element of regulation of the donations market.
About this issue, Harris, Petrovits, and Yetman (2015) showed a positive relationship between good governance composed of seven dimensions (board, management, policies, access, audit, executive compensation, and minutes) with the volume of donations. In addition, Andrés-Alonso, Cruz, and Romero-Merino (2006), when analyzing the influence of public institutional donor participation in the efficiency of Spanish NGOs, have identified that the participation of an active institutional donor is positively related to greater NGO efficiency. Part of such efficiency is due to the fact that many NGOs have adopted private for-profit market approaches, leading to what Salamon (1997) termed as non-profit marketization. Yetman and Yetman (2012) showed that NGOs with more constraints imposed by their donors report financial information more accurately. Thus, it is evidence for donor's influence in monitoring NGO actions, in addition to the donation market as a governance mechanism.
In Brazil, Tenório's (1999) pioneering studies show that the professionalization of third-sector entities aims at their organizational survival rather than greater effectiveness of actions as social agents. Thus, behind the spectrum of the market, there is rationality that submits the third sector to a mercantile logic.
The discussion on corporate governance demands a specific analysis of its organizational nature. There is a need to align the concepts of corporate governance, generally associated with business organizations, to NGOs. The agent-principal relationship occurs differently as compared to when discussing corporate governance in business organizations, because, in the third sector, there are contractual conformations of intrasectoral and intersectoral partnerships resulting from donations to NGOs. In this sense, according to Jensen and Meckling (1976) and Fama and Jensen (1983), who consider organizations as a complex of contracts (between owners and managers, and between holders of production factors and customers), we understand corporate governance generally as a process that aims to harmonize the different interests between the parties of the contracts in a way that allows for cooperative actions. In the agency relationship, donor organizations assume the role of the principal, while NGOs take the role of the agent. In the present article, we have as a fact the agency relationship in the contract between the NGO (agent) and the donor organization (principal) for the transfer of resources by donation, that is, for the NGO to develop the projects. Therefore, the firm is considered a complex set of contractual relations in which the agent and the principal are present (Jensen & Meckling, 1976).
The literature review shows that theoretical efforts have been undertaken to align the concept of corporate governance with the third sector (e.g., Van Puyvelde, Caers, Du Bois, & Jegers, 2012). Also, empirical studies have been done on the relation of isolated indicators of governance quality and some measure of organizational performance (e.g., Yetman & Yetman, 2012) or a specific dimension of governance, such as the board of directors (e.g., O' Regan & Oster, 2005), as well as on the quality of governance in a more integral way in reference to performance measures (e.g., Harris et al., 2015). In general, the discussion starts with the hypothesis that governance mechanisms influence organizational performance (e.g., Gazley & Nicholson-Crotty, 2018). Most studies we conducted were in relation to the elements of dimension Internal Forces, more specifically the influence of the governance mechanisms present in the NGO structure on variables of interest, as in organizational efficiency.
By not being cautious about the external forces that act for the organization of governance, an important gap of research remains without due treatment, leading us to question what are the donor reflexes in NGO governance. More specifically, how do the external forces, especially the donations market, act over the governance of sponsored projects of private associations and foundations that operate in the environment segment? We assume that (a) the donor organization hires an NGO for the mutual interest of project development, (b) the project is configured as a temporary organization (Ludin & Söderholm, 1995), and (c) an NGO may not act in the best interest of the donor organization. As a consequence, there is a contract of work where the donor organization delegates certain authority level to the NGO for making decisions about the contracted project activities. In the contractual relationship, there is no perfect contract, and it constitutes the foundation of agency theory, called agency relation (Jensen & Meckling, 1976).
Agency theory emphasizes that, in the establishment of the contract, agent and principal act according to rational behavior and self-interests. However, it does not rely exclusively on motivational and objective differences between principal and agent, but also on assumptions of information asymmetry, different levels of risk aversion, and different planning horizon between the parties, implying limited rationality of the economic agents (Eisenhardt, 1989).
The NGO has information about requirements, restrictions and results of its activities that are unknown by the donor organization, configuring the asymmetry of information, which is an agency theory premise. Asymmetry creates conditions for opportunistic action, either by recording inaccurate, incomplete and biased information, including involuntarily or by fraud. Opportunism, in agency theory, may manifest ex-ante or ex-post at the moment of hiring, thus resulting in problems of adverse selection and moral hazard, respectively (Williamson, 1987). This is valid for the third sector, where ex-ante costs (structuring costs, safeguards, etc.) and ex-post costs are incurred (renegotiations, contractual additions, agreements compliance monitoring, evaluation of partial accountability, etc.).
In addition, the donor organization is arguably more risk-tolerant than the NGO, as it is expected to be more able to diversify its grant portfolio and may be indifferent to non-systematic risk, whereas risk-aversion behavior is expected from NGOs, as they have more difficulty to diversify their projects portfolio, and also because the donated resources and NGO credibility are linked to NGO performance in relation to the contract. Thus, the parts of the agency relationship have different levels of risk aversion, which is another agency theory premise.
Finally, time-horizon incompatibilities are also accepted, but inversely in the context of the third sector, considering that it is a general rule in for-profit business organizations. We argue that transience occurs not in relation to the agent, but to the principal. The donor organization allocates efforts in actions whose effects occur within the contract time horizon, while the NGO has a relationship with the firm, in which contract deliverables are sometimes related to longterm objectives. So, the NGO tends to plan projects according to horizons that exceed the contract limits, especially when related to programs (Lacruz & Cunha, 2018). Also, extra-contract performance evaluation tends to rely on different metrics due to different time horizons (Lacruz, Cunha, Moura, & Oliveira, 2019). It is assumed that the firm has a finite life and that, in the third sector, the agent (NGO) considers the firm value as an infinite series of actions, and the main one (donor organization) is limited to great extent by the contract.
Thus, as a consequence of the agency relationship, agency problems arise as well as the due actions (agency costs) to minimize them, such as more complex contracts and more intense monitoring mechanisms.

Method
For this study, we performed a qualitative research and delimited the empirical context of investigation to regular donors. By regular donors, we mean business organizations, corporate foundations, public organizations and NGOs that, between 2012 and 2016, made public calls for support to projects related to the environment, at regular intervals (annual, biennial etc.), in which NGOs operating in Brazil could compete with the submission of project proposals. NGOs generally develop their activities through projects (Diallo & Thuillier, 2004). To do so, they raise funds for projects to carry out the activities defined by their institutional mission, by submitting proposals to national and international bodies. According to the most recent yearbook Environmental Management Analysis 2013/2014 (Análise Editorial, 2013), 97% of the resources of environmental NGOs operating in Brazil were donations from partners, national bodies and international organizations. By delimiting an area of activity in an operational section, we contribute to the homogeneity of the units of analysis among NGOs.
We conducted a survey in search of public calls to support projects in the repository Mapa das Organizações da Sociedade Civil (https://mapaosc.ipea.gov.br/editais.html). We successfully contacted 26 organizations in Brazil that donated resources to NGOs in the environmental sector through regular public calls to support projects. We did not choose the organizations randomly; however, we selected them due to characteristics associated with the investigation, recruiting them in several stages in order to support the development of the study in a flexible way.
In order to obtain several sources of evidence for the same phenomenon, thus creating conditions for data triangulation in the analysis (Casey & Murphy, 2009), data were collected with document search in contracts and public call for tenders, and complementary semi-structured interviews were performed with executives from donor organizations. This research plan is aligned with agency theory, since the contract is the link of the agency relationship. Between September 2016 and March 2017, 52 documents (contracts and public calls for tenders) of 26 organizations were consulted. Complementarily, 16 interviews of approximately 50 minutes were conducted by telephone (nine), in person (three), by Skype (two) and by e-mail (two). Executive activities were related to the donation process, and documents (public calls for tenders and contracts or correlates) were provided by organizations.
For the interviews, it was intended to conduct two interviews per organization and two organizations per sector, for balance and comparison purposes; but by exceeding the number of organizations by sector, we eventually came to saturation by organization rather than by subject, causing us to cut in 2x2, eliminating more interviews and completing the process with missing interviews (Table 1). We initiated data collection with document analysis, in order to identify formally established governance practices, as well as to understand the institutional relations between the parties (donor organizations and NGOs). In order to minimize risks of uncritical bias, we operationalized the sample closure by theoretical saturation by means of observing the objectives defined for the research and the desired depth level (Glaser & Strauss, 1967).
From the document analysis, we elaborated a semi-structured interview protocol (Appendix) according to the suggestions proposed by Spradley (1979), that is, involving introductory, descriptive, structural, contrast, and output issues. All interviewees requested anonymity (including the organization), with the interview protocol being sent by e-mail to them.
In the analysis of data, we did not start with categories defined a priori. In the coding process, we first identified the labels. This step involved three stages: open coding, axial coding, and selective coding (Corbin & Strauss, 1990). In open coding, we navigated through the entire content of the interviews and the documents in order to identify elements whose characteristics would allow for categorization (which we initially performed through a label). When identifying a label, we compared other elements in the texts to look for similarities and differences that allowed for the inclusion of those elements in an existing label or the creation of new labels -with the purpose that passages in texts referring to the same aspect would receive the same label, while different elements would receive more suitable and specific labels. After, we grouped the labels and rearranged them in order to facilitate the hierarchy and the formation of categories. To this end, we created groups of labels linked to a common topic with notes of analysis (memos) that allowed for the illustration of the idea to which each label refers. This grouping and the analysis notes were the baseline for the formation of categories and subcategories that, in short, gathered the labels that we had grouped already. Once the groupings were formed, the next step was the accomplishment of the open codification of categories, which aims to identify categories and their related subcategories.
In axial coding, we established relations between the formed categories and subcategories. We linked the relations through the framework that supports the study, as suggested by G. W. Ryan and Bernard (2003). This resulted in categories and subcategories that formed the basis for the findings of this study. We intended to find uniformities through comparisons and generalizations by seeking a reduced number of high-level concepts. Finally, in selective coding, we unified the categories into a core category by relating previously created categories to a single central idea representing the central phenomenon of the study.

Results and Discussion
From data analysis, three categories emerged: Compliance, Enforcement, and Project. We identified 34 labels, which were grouped and rearranged to facilitate hierarchization. We linked the labels to 15 subcategories (groups of labels linked to a common topic) in order to facilitate the formation of categories and the creation of the central category. Finally, we grouped the subcategories into categories. Table 2 shows the central category, the categories and the subcategories found in the coding process. We observe that the categories relate to the donation market (project and enforcement) and, less obviously, to the political, legal and regulatory system (compliance). The subcategories, in turn, have a connection with the dimensions of governance found in the literature (e.g., compliance, accountability, and audit).
When associating the categories with the process of marketization, Temporary Governance as the central category emerged from the process of codification. Temporary Governance proved to be a consequence of Project, which, in turn, is influenced, within the limits of our investigation, by Compliance and Enforcement (Figure 1).

Figure 1. Relationship between the categories
The Project dimension refers to the object of the contractual relationship between the NGO and the donor organization. In other words, it refers to the business venture executed within a determined time limit, with defined scope, quality and cost, as a counterweight to the donation, which is subject to standardization of the contract signed between the NGO and the donor organization. Thus, the project is a material object of the contract, which in turn is the link of the agency relationship between the NGO (agent) and the donor organization (principal). NGOs, in general, develop their actions, in the fulfillment of their institutional missions, through projects (Diallo & Thuillier, 2004). Furthermore, specifically in the present study, the delimitation of the donations to those linked to projects is aligned to the operational domain.
In turn, the conceptual domain is associated with the definition expressed in PMBOK (Project Management Institute [PMI], 2013), "A project is a temporary endeavor undertaken to create a unique product, service, or result" (p. 9), like in other research (e.g., Muriana & Vizzini, 2017). This concept involves three key features of projects: it is temporary and progressive, and it generates deliverables. Temporality associated to the Project leads to the understanding that governance in NGOs can be largely due to the influence of the project. Each project, with its specificities (many resulting from donor influence), confers to governance practices at the organizational level -albeit temporarily. Temporality is a characteristic that differs projects from processes (Renz & Andersson, 2014) and, in another perspective, the temporary organizations from traditional organizations (Ludin & Söderholm, 1995). By proposing projects as temporary organizations, Ludin and Söderholm (1995) argue that temporality leads to the past and future organizational activities being uncoupled, despite their interactions (Bakker, DeFillippi, Schwab, & Sydow, 2016), and that the activities only happen within the limited life of the project. Turner and Müller (2003) reinforce the aspect of the finite life cycle, highlighting the extinction of the project at a future point.
Decoupling and temporality lead to the proposition that NGOs may, as a result of the process of marketization, incorporate only necessary elements and taxes (transposed) by the donors during the project period into their governance. However, by extinguishing the project, such practices Project Compliance Enforcement Temporary governance could be discontinued in order to be incorporated or modified again into a future project according to the demands of their donors. Thus, organizations become a "portfolio of projects" (Krause, 2012, p. 619) and establish governance that also supports project management. Reflecting on this is strongly associated with the premise of different time horizons of agency theory, which implies the transitoriness of the principal in relation to the NGO, since its efforts are limited to the time of the contractual relationship, that is, to the project, by definition.
We were able to identify the focus on project management and a certain indoctrination, of temporary effect, of NGOs by donors on proprietary project management methodologies.
"At the beginning of the agreement, we deliver extensive training on how to manage projects, on accountability, calculation of results. Throughout the support period, we offer free consultations." (Inf. 5, translated by the authors).
"The [Third Sector Organization -9] has its own structure for the monitoring of projects through manuals, elaborated through its more than 30 years of experience in the area of projects. The NGO should follow these manuals and submit the requested documents." (Inf. 13, translated by the authors).
"As soon as we launch the call for tender [public call for projects], we offer training on the elaboration of proposals. The institutions that have their proposals approved are, at a later moment, skillful in management practices, where we seek to add our expertise to them." (Inf. 10, translated by the authors).
Preceding the Project dimension, we find dimensions Compliance and Enforcement. Compliance refers to legal compliance, and Enforcement manifests beyond the institutional control of the state and relates to the influence of management practices of the donors.
Second Clause -of Obligations 2.1.2 -To the CONTRACTOR g) observe and comply with the rules of Law N° 8.666/93 in concluding contracts necessary for the execution of the object of this agreement, admitted to the adoption of the bidding modality provided for in Law N° 10.520/02; and… (Org. 13, Contract, personal communication, translated by the authors).
"What we have achieved in these years is to give management practices to them [supported NGOs]. The importance of planning, control, measuring and, consequently, managing." (Inf. 7, translated by the authors).
"There is an exchange with the NGOs we serve. They have a lot of openness in the communities served.
In addition, we can help them with the management processes. Many NGOs have this need... and with our support, they can improve their internal processes." (Inf. 6, translated by the authors).
This puts pressure on NGOs. On the one hand, public organizations require that NGOs organize themselves similarly to the State, despite having different legal identities, but with speed in processes that are typically found in business organizations. It should be noted that NGOs are private law organizations, even though they operate public resources. Therefore, even if it favors a public character, given that the beneficiary of actions is the whole society (lato sensu), they are not organizations linked to the State apparatus (Lacruz, 2014). The pressure for NGOs to organize themselves as the State partially contradicts the motivational root of state decentralization actions in Brazil, which demands more agility and flexible processes from NGOs (Pereira, 1998).
Thus, public organizations expect NGOs to speed up the processes of corporate organizations, but with the procedural rites of the State -in an apparent contradiction -, thus imputing NGOs a profile that is often regarded as bipolar: sometimes acting like the State, sometimes as a business organization. On the corporate donor side, NGOs are required to have an organization aligned with their business. Generally, large corporate organizations make social investment either directly or through their corporate foundations. For many NGOs, such management processes go beyond the limits of their organizational structure. There are few large NGOs in Brazil (Instituto de Pesquisa Econômica Aplicada [IPEA], 2018), yet they do not resemble many of their donors.
In summary, the transfer of donor management processes to NGOs produces pressures from public organizations, by the influence of the laws and norms and actions of decentralization of the State, thus demanding that NGOs organize themselves as the State; and on the corporate donor side, by demanding an organization aligned with business. The compliance and regularity dimensions manifested side by side, with the former being mainly during the donation decision process with the call for project support, and the latter during the monitoring process by project donors already running.
The calls for project support are generally public and comprise two stages: legal qualification and merit evaluation. This exerts great force for NGOs to seek to be compliant with the applicable legislation, under the risk of not even being qualified for the litigation of resources in support of projects. Some recurrent practices of governance, although they have no legal provision in Brazil, such as the Board of Directors, are required by donors in the legal qualification stage. Regarding project monitoring related to the Enforcement dimension, this aspect appears to be a concern for donors.
"In addition, the executing agency [supported NGO] must submit a progress report every six months, which can be made available on the Online Monitoring System platform, which helps us to control and analyze documents." (Inf. 14, translated by the authors).
"Every quarter we are presented with accounts, and every month we are sent a form to follow up activities and results." (Inf. 8, translated by the authors).
"Before each new contribution, we are sent partial accountability, both physical and financial. If approved, we will pass on the new contribution. Moreover, annually we make an assessment, with ad hoc consultants, in loco." (Inf. 11, translated by the authors). This is related to accountability, also associated with the regulatory dimension. In the case of NGOs, the accountability process is a specific stage in the life cycle of projects (Lacruz, 2014). It is the responsibility of the NGO to account for its acts and omissions by means of reports of accountability for the physical and financial execution of the project, including processes and purchases and hiring, photographic records, clippings, etc. If the donor disapproves the accountability, the resources intended, as a rule, should be returned. Therefore, NGOs assume the consequences of their acts and omissions in this contractual relationship. "NGOs that are supported by the Call for Tender, that do not meet the goals, terms and conditions established in an agreement signed between the parties and whose justification for such fact does not meet the requirements of the Foundation must return the resource received." (Inf. 9, translated by the authors).
"If the accountability is disapproved, the institution served returns the resource received. Only the value that was interpreted is returned." (Inf. 6, translated by the authors).
We see a clear relationship between these dimensions, in the sense that Enforcement operates in addition to Compliance, since not only legal imputations are required from NGOs, but also donor management practices. Thus, using Jensen's (1993) classification, in the third sector, the donation market acts in complement to the legal and regulatory political system, as an external governance mechanism for NGOs in the delimitation of legal security, given the general legal context and little supervision in Brazil.
In contrast, assuming broader monitoring, the evaluation of the effectiveness of the results achieved with the project fails. We see that, although there is a concern with efficiency and efficacy, there is not necessarily with effectiveness. In other words, donors worry about whether the project was executed within the anticipated limits of time and cost, and whether deliveries OPEN ACCESS are in line with the scope previously planned, but do not verify whether the expected results for beneficiaries have been achieved.
"There is the accountability period in which we need to submit a checklist of technical and financial execution, but, after that, I realize that there is no follow-up, because of other demands/activities of the technicians… If the NGO went ahead with it, we do not follow." (Inf. 1, translated by the authors).
"There is no follow-up. However, before completing the project, the executing agency [supported NGO] should present a document describing the sustainability of the project after the financial support." (Inf. 15, translated by the authors).
"We evaluate the outcome of the project after it is completed, when the NGO requests a new resource. Before having a second project approved, the NGO must demonstrate the current stage of the other approved projects, if applicable." (Inf. 3, translated by the authors).
Of course, this kind of evaluation would require, in many cases, metrics that can be measured before and after the project actions. Moreover, the time for results to be perceived generally exceeds the contractual relationship limit.
The analysis of external forces (e.g., Chokkalingam & Ramachandran, 2015;Desai & Yetman, 2015), especially in the donations market about the configuration of governance in the context of the third sector, still lacks operational and empirical evidence. Our analysis of donor reflexes in NGO governance contributes with evidence to the field addressing agency theory, and it also innovates by highlighting a particular dimension to the third sector environment: project. Renz (2007) proposes that project governance be the answer to the gap between governance (strategy) and project management (operation) in the third sector, reconfiguring the generalities for for-profit organizations. We advance this understanding by suggesting that project management brings a reflection to governance with the presence of temporary governance. Therefore, governance is (re)shaped by the presence of projects in a continuum. The following propositions emerge: Proposition 1: Governance practices of NGOs are impacted by temporary governance derived from projects, which in turn are circumvented by aspects of compliance and enforcement through pressure from donors.

Proposition 2:
Transferring donor management processes to NGOs produces two shadows over NGOs: (a) there is a public organizations' influence of laws, norms and actions of state decentralization that demands NGOs to organize themselves like the State; and, (b) on the corporate donors side, an organization aligned with business is required.
The central argument of these propositions is that project management processes are incorporated into governance practices on a temporary basis and can be incorporated (even after project closure) due to benefits perceived by the NGO, which tends to transfer practices from one project to another, replicating successful actions and seeking alternatives to unsuccessful ones. The propositions involve understanding the project role as a catalyst of management practices in NGOs, which aim, initially, to meet each donor's characteristics, but that can subsequently be captured, stored, developed and shared in other projects, in favor of superior results, and also to business models at the organizational level.
This implies that NGOs operate as project-based organizations, and their projects as temporary organizations act as vehicles that lead to practical benefits and organizational learning as a whole. NGOs absorb good governance practices imposed and driven by various projects over the years that build a foundation of wisdom and know-how in their own governance practices.
In search of understanding the relationship between the different levels of a given dimension of governance in the third sector, it was possible to see the effect of project on governance by means of a temporary governance. Besides, it was possible to see that donor pressure grounded in legal conformity and regulation related to internal processes of donations management -both due to marketization -represented that temporary governance.

Conclusions and Implications
The objective of this study was to investigate the influence of donors in the organization of sponsored project governance in NGOs, using agency theory as the analytical lens. The findings allow us to propose, in addition to studies that associate the quality of governance with the receiving of resources through donation (e.g., Ávila & Bertero, 2016), that not only a good governance can lead to the receipt of more resources in donation, but above all an alignment with the donor management and organization structure. Thus, the influence of temporary governance of the project would form corporate governance, which in a set of projects converge to practices of governance at the organizational level.
This study further revealed donors' emphasis on project governance, considering the project as a temporary organization (Ludin & Söderholm, 1995). This is particularly relevant when discussing the effectiveness of project results. By definition, a project is temporary; with donors being concentrated within the project boundaries, we did not evaluate the outcomes to the beneficiaries, except in cases of continued support through successive projects. Therefore, governance in NGOs, to some extent, may stem from the altruistic goals of the founders and the team, but it also stems from competition in product markets and donation market. Thus, this study seeks to contribute to understanding how donations market acts to discipline the governance rules in NGOs, and to stimulating the governance debate in this segment of recognized social relevance.
Since the Project dimension plays an important antecedent role in our discussion, we should highlight that the cross-sectional nature of our data produced a discussion that might not demonstrate longitudinal evidence; that is, the influence of a phenomenon related to a specific dimension in the occurrence of another phenomenon. However, as the reports of the participants of this study were based on their memories, we accept that temporal relationships can occur. To reinforce the empirical evidence for these relationships, we recommend longitudinal studies.
In addition, investigations with longitudinal sections can focus on assessing the type of relationship between the project and NGO governance. Our findings allowed us to propose that NGOs are able to institutionalize governance practices limited only to the duration of the project (object of the contractual relationship), due to the process of marketization, i.e., that the project can impute temporality to the governance of NGOs. However, there remains an open empirical question regarding the type of relationship between the project and governance in NGOs.
In the same way, it is an open empirical question the possible moderating role of the ownership structure, that is, whether there is a significant difference between NGOs with a predominance of resources from dispersed donors and NGOs with a concentration of resources in a few donors. This is a limitation of this study, because it was not possible to determine the relative participation of donors in the donations received from NGOs, since in this research an NGOs group in particular was not studied, only donors were.
Although the results bring important advances to the area, they should not be interpreted as definitive evidence that all NGOs adopt all governance practices analyzed; for practitioners, it may be useful to guide their efforts towards better fund-raising results with empirical support for donor exerted pressures.
Also, in choosing to allocate efforts to incorporate donor management practices, NGOs move resources from actions more directly related to their mission. Due to a project's temporary nature, and in contrast to the planning perspective, NGOs must reduce their reliance on donors, thus diversifying income sources and increasing unrelated restrictions, in order to keep their actions linked to perennial causes. Therefore, they could incorporate management practices by understanding their importance rather than by the pressure of donors.