Chapter 2: Literature Review
2.2 Agency Theory
國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
can occur during the formation of stronger institutions, which may take a considerable about of time. Many authors have linked the development of social capital with other changes that occur within communities (for example, Gordon, Kayseas, & Moroz, 2017;
Riel, 2017; Urquhart & Wearing, 2017), and these criteria can be used to measure
processes and outcomes during a long-term, ongoing project. The study of the “how” this social capital is built could be useful in assisting with more efficient negotiations and interactions between researchers and communities.
2.2 Agency Theory
At the core of agency theory is the idea of one party acting on behalf of another. Agents are appointed to act on the behalf of principals, with the principal seeking advice or services from a professional, and the agent generally being a person with professional qualifications or particular skills. Principal-agent relations in fields such as economics, management, and law are often expressed through contracts that outline responsibilities and tasks of the agent. Principal-agent relationships also operate and are formalized to limit conflicts of interest, with an expectation that value will be created in the future.
The field of agency theory has roots that are strongly planted in rational choice theory in traditional economics (Braun & Guston, 2003; Shapiro, 2005). In the relationships between two parties, a principal has resources at his or her disposal, but not the resources appropriate to realized their particular goals. An agent can accept the resources to act in the interest of the principal. In the traditional principal-agent theory held in economics, actors are considered to behave according to self-interested rational choice, being constrained against predatory action and moral hazard by certain institutional and contractual designs. Ideally, both principal and agent should profit from this exchange relationship.
‧
In the case of university based action researchers, two main agency roles are played. With the base in cooperation and participatory action with goals informed through community input, research teams act as agents to the community principal. The community principal in this case can be seen as both the community members in general and their
representative leaders. The action researcher’s second main agency relationship is as an agent to the unit that gives access to funding that makes the action research possible.
Agency theory assumes that agents and principals have conflicting goals. Because both parties are expected to act “rationally”, they are likely to maximize their utility which leads agents to not always act in the best interests of the principal. This leads to agency costs due to the effort and resources that must be used in order for the principal to constrain the activities of the agent, for the agent to convince the principal that their best interests are indeed being acted upon, and of the residual losses that occur due to
incomplete utility alignment (Van Puyvelde, Caers, du Bois, & Jegers, 2012).
Agency problems arise when there is imperfect sharing of information and the interests of the principal and agent diverge creating agency costs that, while sometimes difficult to measure precisely, are significant and likely to occur (Bosse, Phillips, & Phillips, 2016).
Much attention has been given to the actions of principals to mitigate negative effects of agents who operate in a classically rational and self-interested way. Along this economic bent are often found detailed calculations that aim of creating contracts that minimize agency costs, such as monitoring and compensation, with the goal of increasing value of organizations and creating higher returns. However, it is unlikely that agents (or
principals) operate without some kind of limits on aggressive or predatory self-interest;
instead they are constrained by psychological and social norms in a way that lead them to act in consideration of fairness.3 While much literature focuses on controlling and
monitoring to reduce moral hazard and shirking on the part of the agent, it is also
3 Not all agents are constrained by psychological or social norms of behavior. Research into corporate psychopaths and CEOs with narcissistic personalities indicates that these people do not operate according to established social norms of fairness (Body 2005).
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
important to consider the possibility of principals to engage in shirking as well (Braun &
Guston, 2003).
While agency theory at its most basic consists of dyadic relationships between actors in the role of principal an agent, in practice complicated interactions at multiple layers of agency relationships occur. Actors play both roles at the same time, they act in groups as multiple principals, and multiple agents compete from different hierarchical levels and with various personal internal interests in a complicated web of interrelated agency relationships. Researchers are confronted with problems such as free-riding, collective action, informational and power asymmetries, operational difficulties in monitoring, and historical context. In some criticisms of the strict economical perspective, taking the individual as a purely self-interested rational actor seeking to maximize benefit has been seen as an oversimplification of an inherently complicated social environment where in reality perfect agency is rare (Shapiro, 2005).
Some critics of agency theory see a reductionist approach to specific tangible variables as risking or ignoring the institutional and social context that influences behaviors and outcomes. For example, Wiseman et al. (2012) bring up the possibly simplistic assumptions of human nature, such as purely self-interested agents who disregard principals in favor of maximizing their own personal benefit. If it is indeed the case that agents (and presumably principals as well) act out of unmitigated self-interest, then there would be little cause to approach agency questions from other non-economic contexts, as there would be considerable difficulties with theoretical validity. On the other hand, it has been argued that motivations of principals and agents have socially derived interests that may not automatically cause pure maximization efforts or complete alignment of interests (Bosse et al., 2016; Shapiro, 2005; Wiseman, Cuevas-Rodríguez, & Gomez-Mejia, 2012).
Put another way, “some agents may genuinely desire to serve the principal, while others may seek to serve a principle even when it conflicts with the wishes of the principal”
(Wiseman et al., 2012).
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Wiseman et al. further note that focusing solely on contracts risks neglecting factors of the broader social, institutional, or cultural context that could affect outcomes, and leaves open opportunities for research to further refine the theory (ibid). They also suggest approaching agency issues by looking at cognitive factors, and though his analysis does not include the influence of social capital, this has been explored by other authors who note that by recognizing the influence of the social environment we can improve our understanding of agency problems. According to Bosse et al., for example,
“understanding positive reciprocity incentive alignments and monitoring mechanisms have the potential to improve aggregate social welfare by creating agency benefits as well as agency costs” (Bosse et al., 2016 p. 290).
With slightly modified theoretical elements agency theory has been successfully extended to other fields, such as management, law, political science, international relations, and sociology. Through reconsideration of the underlying assumptions of unidimensional actors, agency theory can be applied across such various fields. Scholars outside of economics “abandon the assumption of an acontextual, ahistorical, and static relationship between principals and agents (Mitnick 1992)” instead considering the relationships in the broader social context that includes “other agency relationships, competitors, interest groups, regulators, legal rules, and the like” (Shapiro, 2005). By moving away from the traditional mathematical perspective of economics agency theory has expanded to encompass a wide variety of approaches on the quantitative/qualitative spectrum. In this way, agency theory has been able to provide a wide variety of insights to each of these fields.
In a more sociological interpretation of agency relationships, agency costs are still assumed arise if the agent is not acting in the principal’s best interests. It can be difficult for information be credibly conveyed by between the information holder, whether principal or agent, to the other party who would benefit. This credible transference of
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
information can become even more challenging between parties who are separated by language and cultural differences. However, in the traditional economic perspective, information asymmetries can be mitigated through contracts and monitoring, which applies to the sociological interpretation as well. In addition, the sociological
interpretation allows that clear and flexible channels of communication that can make information sharing possible can be improved by building trust, fairness, and reciprocity, i.e. by developing social capital.
Difficulties may arise from a history of previous agency relationships. If these
relationships proceeded badly, resulted in project or social failures, ended before results were achieved, or were concluded too abruptly, locals may come to see relationships with outsider researchers not as repeating interactions, but instead as a series of one-off
engagements. This could be interpreted as a weakness in cooperation or failure to create social capital, or worse, the destruction of social capital that connects principals and agents. The experience of prior difficulties or failed relationships that have occurred in the past may create a context that makes it particularly difficult for new, incoming
researchers to build or rebuild trust. This situation could be made particularly challenging if past agents were not seen to have been acting in the best interests of the principal, whether purposefully or not, and is certainly at play in disenfranchised communities and those with a colonial history.
Agency problems can also present a complicated set of problems for the action researcher. The researcher as an agent acts on behalf of the community to provide knowledge and expertise to benefit the community, while also requiring knowledge, experience, and effort from the community members to be able to create scholarly output that allows access to the funding that makes action research possible in the first place.
This puts the researcher in the role of as an agent acting concomitantly on behalf of the community and the funding institution. Seeking participation from the community and funding from a government institution means that agency problems can operate in both
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
directions, especially when monitoring is made more difficult through physical distance, information asymmetries, or opaque policy.
Some principal-agent difficulties in a post-colonial indigenous community were
presented in a review by Vining and Richards of natural resource extraction development projects by hybrid indigenous non-indigenous organizations in Canada (Vining &
Richards, 2016). These hybrid joint ventures between indigenous and non-indigenous groups provide indigenous communities access to expertise in creating contracts as well as with giving them input into policy decision making processes. For the non-indigenous actors, benefits include access to natural resources, for the indigenous communities other benefits such as financial, management, employment, land use, and social improvement are expected.
Though these programs create links between community insiders and outsiders and a level of benefits are accrued to an extent, they do run into a multitude of issues such as fragmented ownership, divergent goals, difficult to determine ownership statuses due to unclear treaty rights, internal and external constraints on transfer of ownership rights, security from trespass, the long-term survival of indigenous ownership rights (at a collective community level), and conflict between indigenous and non-indigenous principals. Such conflicts are further exacerbated by a pre-existing mistrust of non-indigenous institutions on the part of non-indigenous communities that stems from a long and difficult historical background. Although economic benefits were found in this case, evidence of increased community wellbeing resulting from increased own-source income was less clear. The question then arises of how to mitigate these agency relationship difficulties so as to increase meaningful, though intangible, community benefits.
Parallels to agency issues between indigenous principals and not-indigenous agents in the Canadian context can be seen in Taiwan. Many developments in indigenous areas in Taiwan have suffered for reasons that include past unclear ownership of property
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
resulting from grey or black-market property transactions, restrictive government regulations on the transfer of ownership rights as well as on development of existing properties, land rights issues including past removal from and present limited access to and use of traditional lands, as well a general mistrust of outside institutions. These problems are highly salient to local populations and both result from and lead to difficulties in aligning goals between community insiders and outsiders.
Of particular interest to the present work is found in the more qualitative perspective that can approach issues of “norms, networks, authority, organizations, social control,
regulation, trust, social cognition, and so on” (Shapiro, 2005). In literature related to trust, agents and principals are more likely able to align value and interests when they are closer to each other through kin, community, or associations (ibid). Some authors have presented the concept of bounded self-interest to integrate norms such as of fairness as variables that influence the trajectory of principal-agent relationships. In this approach, trust and reciprocity can generate positive outcomes and benefits for social welfare that are not captured by traditional models that assume pure self-interest (Bosse et al., 2016).
The importance of sociocultural variables can also be seen in cross-cultural studies of agency theory that move outside traditional western population samples, for example between collectivist and individualist cultures or in cultures that place different levels of importance on power hierarchies (Johnson & Droege, 2004).
In addition, an agent can often find him or herself in the difficult position of serving many principals. In the case of the action researcher, creating a successful program often depends on both acting effectively with the community and on producing material that provides access to the necessary funding supports. This means that the action researcher acts as agent to both the community and the funding institution as principals. Yet the outside institution may have goals that conflict with those of the researcher, or for that matter of the target community.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
“Only the rare agent has the luxury of aligning her interests with a single principal. Conflict of interest is hardly about shirking or opportunism with guile; it is about wrenching choices among the legitimate interests of multiple principals by agents who cannot extricate themselves from acting for so many, [these] agents are increasingly buffeted by the conflicting interests of the principals they serve” (Shapiro, 2005)
The agent must then balance goals of both the community principal and the funding institution principal that are not necessarily in alignment. In such cases, conflicting goals make successfully negotiating from their position a delicate proposition, one requiring knowledge of social and cultural environments from principals on either side of the table.
Studies of agency relationships between indigenous and non-indigenous communities are often interested in determining structural conditions that exist for the agent to meet the needs of their principal, but they seem to less often examine the linkages of
communication that are vital to aligning intentions. Some work has been done to investigate the importance of signaling processes in indigenous political policy and regulation in the United States (Worsham & Gatrell, 2005). This work did suggest that communication links are vital in understanding principal-agent relations, though the question was approached from interactions among bureaucracies. When the question is brought into agency relationships at the meso level of the community and micro level of the individual, communication methods remain vital, though it seems that they are also less-well studied.