Political Transactions, the Social Contract, and Administrative Power

Abstract

Although power has frequently been noted as a key to understanding administration, the concept of administrative power has remained ill-defined. In this essay I offer a definition of administrative power based on the social contract, arguing that administrative power is the transactional granting of power by the people in exchange for services and a reduction in uncertainty that agencies provide. I elaborate on the role of communication and persuasion in garnering administrative power and conclude by offering arguments, based on both risk communication and political spin, for how agencies acquire and maintain administrative power in the 21^st Century.

INTRODUCTION

“Power is only one of the considerations that must be weighed in administration, but of all it is the most overlooked in theory and the most dangerous to overlook in practice.” (Long 1949: 257).

Although Long made the observation above in 1949, the critique still largely holds 70 years later. This is not for lack of effort; virtually all efforts to understand the role of the public bureau or the public administrator can be viewed through a lens of defining the legitimate sources and uses of power, and there is an extensive body of research exploring aspects of administrative power such as the study of bureaucratic politics, the use of discretion by street-level bureaucrats, and accountability (Balla 1998; Brower and Abolafia 1997; Carpenter 2001; Meier 1991; Scott 1997). Yet in most of this research, power is left abstract or exists as a concept in the background.

As noted by Durant (2015), Long himself left the term abstract in his foundational essay. He spoke of competition and power between and amongst agencies, and also conferral of power and status within agencies, but never really defined what, exactly, he meant by administrative power. As quoted in Durant (2015: 207), Holden (1966) defined administrative power as “a favorable balance of constituencies” which raises more questions than it answers. Favorable to whom? What is balanced? What constitutes these constituencies? Similarly, Magee and Frasier (2014: 307) defined power as “asymmetric control over valued resources” which raises questions about what symmetry would look like and who values the resources in question. In this essay, I attempt to unpack the concept of administrative power. I argue administrative power is rooted in a transactional exchange of authority and legitimacy that takes place under the auspices of the social contract. I define this exchange as situational, adaptive, and evolving as agencies seek to manage their reputations, communicate meaning about what they do, and maintain or enhance their standing with both political leaders and the public more broadly. Long famously referred to power as the “lifeblood” of administration (Long 1949: 257), and that the attainment and loss of power constitute the context of all public management. Power maintenance and acquisition may not be the focal points of day-to-day management in the public sector, but legitimacy and power are perhaps the key drivers of the environment in which that management takes place. Both scholars and practitioners would do well to acknowledge and embrace the centrality of power arrangements in order to derive deeper understanding of administration and also better outcomes.

DIMENSIONS OF POWER IN ADMINISTRATION

Part of Long’s aim was to open more detailed discussion of power in the context of administration. Tracing the history of thought in public administration, power is often an unstated theme. Wilson (1887) and Goodnow (1900) explicitly sought the removal of political influence (i.e. political power) from the administration of public policy, but implicitly, both were advocating a power structure wherein the administrators were privileged, via their technical expertise, to carry out implementation as they saw fit. The learned administrator, if trained properly, would know how to implement a policy to ensure its effectiveness and thus the proper ordering of power was top-down. It was never stated explicitly in these early writings, but we see hints of depersonalization—the ideal was not that person A had power or authority over person B, but rather that some vague sense of knowledge or expertise would lead both A and B to same conclusion about what was to be done. Long (1949) advocated for a different perspective that recognized that agencies are political actors. However, this time was also the peak of scholarship on the politics-administration dichotomy where power was seen as anathema to administration and politics was seen as a necessary but problematic constraint on administrative rationality (Durant 2015; Kettl 2000). The separate literatures that emerged after Long’s essay illustrate this. One focused on organizational management, conceptualizing power as an internal aspect of leadership and externally important as related to organizational status and reputation (Bustos 2021; Fernandez and Pitts 2007; Salancik and Pfeffer 1974; Vigoda 2000). A second literature dissolved the question of the legitimacy of administrative power in a democratic system to a discussion of performance and accountability, implicitly premised on popular sovereignty and administration’s accountability to the public (Demir, et al. 2019; Hood 1995; Mulgan 2000), while a third literature embraced the political machinations of agencies, but mostly only related to the budgetary process rather than politics more broadly (Balla 1998; Bawn 1995; Calvert 1989; Moe 2006).

The first literature stream depicts power in terms of inter- and intra-organizational relationships—principals and agents, resource dependencies, leadership, and followership—without much consideration of the implications of agencies having power, beyond the inherent value of it (Bustos 2021). The second set of literature starts with acknowledging the fallacy of the politics-administration dichotomy, but still retains a perspective in which politics and administration are fundamentally separate, but overlapping, and this overlap is the realm of accountability (Dunleavy and Hood 1994). The third takes as given that agency survival and thriving is a primary goal of power acquisition but leaves it vague as to why this is a goal in the first place (Bendor and Moe 1985), which was a fundamental concern of Long’s.

Multiple dimensions of administrative power have been noted in these related literatures. First is the bureaucratic politics dimension, in which agencies compete against one another for resources (Moe 1984). In this literature, power is conceptually zero-sum, related to budgetary allocations that agencies receive. The perspective is akin to a principal-agent relationship between elected authorities who have budgetary control and agencies seeking resources. Power is mostly implied as a top-down exercise of budget allocations, with agencies jockeying against one another for these resources. A second dimension relates more to the power of expertise and the conferred legitimacy that expertise provides. Long (1949) specifically noted this dimension, and it is rooted in instrumental rationality and an expectation that expertise will confer the legitimacy to make important decisions, which will in turn result in better governance. This perspective seems rooted in the longstanding discussion of the public interest, traced back to scholars like Wilson (1887), Goodnow (1900), and Friedrich (1940) and their faith in the veracity of expertise and the ethics of administrative professionals to interpret and act according to the public interest. Power itself remains somewhat nebulous here, but it is rooted in popular sovereignty and reflective of the politics-administration dichotomy, with power delegated based on knowledge rather than representation. Power is not in the hands of people in administration but rather is rooted in knowledge and technical expertise.

A third dimension of administrative power relates to the relationship between the public and/or clients and the agencies that serve them. Literature in this area is more diffuse, but perhaps with a common ancestry related to Finer’s (1941) concern with the undemocratic nature of administrative discretion. I discuss Lowi’s (1986) critique in more detail later, but he was concerned with the potential for despotism in administration, and many of the public engagement reforms of the past 60 years relate to a concern of this power imbalance between agencies and the public. Studies along these lines have revolved around the power of police over citizens (Brehm and Gates 1999), heavy-handed regulation and red tape (Bozeman 1993), and fostering of dependency by welfare agencies (Dubois 2014).

Recent work has also considered the ends to which administrative power is attained. Mettler (2011) noted the often-unseen hands of power that guide administrative decision-making and the informal ways that wealthy interests have worked within the bureaucratic process to ensure their interests are paramount. Durant and Ali (2013) situated the current understanding of power within the Progressive movement and explain how the effort to consolidate power may have actually sapped it, while Moynihan and Soss (2014) attempted to understand the relationship between politics and administration in terms of the feedback that occurs between the political process and policy implementation.

Magee and Frasier (2014) note that both status and power in administration are contextual and are distinguished such that “[S]tatus arises from social conferral—a perceptual and judgmental process—whereas power arises from resource control, [with] subordinates play[ing] a more active role in bestowing status on a manager than in granting that manager control over resources” (Magee and Frasier 2014: 309). Status can present an avenue to power, but power can arise in other ways as well. Like much of the literature, these authors do not detail why power is attained, implying that it is attained for its own sake in the name of goal attainment.

Krause (1999) offers additional clarity both in the nature of power relationships and the ends to which power is attained. First, Krause (1999) argues that administrative power is relational and bidirectional, functioning in an adaptive relationship between political institutions and administrative agencies. Administrative power is attained and managed by administrative agencies seeking to influence both policy choices by elected officials and budgets that agencies receive. Carpenter and Krause (2015) extend this idea to describe how the power dynamic between agencies and political institutions is transactional in nature, with entities acting in ways that they perceive as being mutually beneficial. Agencies accept political power over them when they perceive that it is appropriate to follow the lead of public opinion, while at the same time, political leaders recognize the legitimacy of agency expertise to act. Carpenter (2010) notes how agencies do this, via managing their reputation. Describing the power of the Food and Drug Administration, Carpenter (2010: 10) notes that “[r]eputation—understood as a set of symbolic beliefs about an organization, beliefs embedded in multiple audiences—comprises the central response…of American regulatory power in the pharmaceutical world. Reputation built regulatory power in all its facets. And power, once possessed, has been used and managed in ways that maintain reputation, and hence power itself.” In other words, as Long suggested, administrative agencies seek out and use political power, and one of the key ways that they seek and maintain power is by developing and protecting their reputation with key stakeholders and with the general public.

Administrative power is based on legitimacy, expertise, and the managed reputation of agencies in a political environment. It operates between and among political entities and administrative agencies, under the auspices of popular sovereignty towards an end rooted in cognizance of and commitment to the public interest. My argument here builds off the work noted above, particularly Carpenter and Krause (2015), to elaborate on the reputationally based, adaptive, and transactional relationships that form the basis of administrative power in a democratic system. Before specifying my argument, it is important to differentiate concepts, particularly administrative authority, which was the focus of Carpenter and Krause (2015), and administrative power, which was the focus of Long (1949) and my focus here.

POWER AND POLITICAL POWER

Although administrative power may not be well defined, there are conceptual clarifications about the concept of power generally. Power plays out in a relational process—a systematic set of relationships that are constantly evolving, receiving feedback and new information (Krause 1999; Reed 2013). Power exists on an individual/micro level in relationships between individuals, an organizational/meso level in relationships between organizations and between organizations and individuals, and a social/macro level, between and among institutions and the organizations and individuals that live in the society (Moe 2005). Power on each level may be quite distinct from that on another level (or not), but power at each level constrains what power can be on another level (Koliba 2013). Power is intertwined with concepts like status, influence, dominance, dependency (and myriad others), and is often anchored in both self-interest and group interest (Harper 1985).

In short, power is an emergent property of complex social relationships, governed by the specific actors involved and the spatial, temporal, and organizational context (Axelrod 2002). It can be difficult to discern whence power came, because the interactions that form power relationships do not exist in a vacuum. These interactions occur within a social context that is informed by the actors and the setting, including the structure and history of previous power relations, but is also informed by the history of social interactions with institutions and with one another (Alexander 2001). Power is often tacitly accepted (but can also be challenged) and understood rather than overtly discussed or accessed (Simmons 1976). It is a situational characteristic that comes about amongst a particular group of people at a particular point in time, under a particular set of circumstances. In that given situation, rather than any explicit cause of power dynamics, there is a blend of social factors that interact and create power relationships, and this blend is constantly in flux owing to the agency of the actors involved, the rules of social interaction, and the institutional structures that constrain those rules (Clegg 1989; Giddens 1984). The ends and the means are difficult to untangle.

Because of this complexity, power has a wide divergence of definitions. Riker (1964) notes the underlying problem is that taken separately, each definition seems entirely accurate, yet the definitions have little in common. This is partly what makes power at once such an important, yet ambiguous concept. In the early part of the 20th Century, two main schools of thought pervaded study of power in politics. Dahl (2005) exemplifies the pluralist view that power is fragmented and specialized, such that no one group or individual can possess enough power to carry out its entire will. Various groups compete for the power, form coalitions to accomplish short-term goals, and power returns to an equilibrium state with no one consistently in control. If one group obtains too much power, it can be quickly overwhelmed by the combined power resources of other groups.

The second main school is the elite theory of power, most often associated with Mills (1958). In this school of thought, some segment of relatively homogenous elite power brokers determine what issues are on the agenda, and how those issues are decided. The polity has limited power, mostly confined to narrow voting choices, and are compelled to behave as the elite dictate. Bachrach and Baratz (1962) claimed that these studies investigated only exercised power, and that power had a second, potentially more important face in non-decision-making power. Non-decision-making is keeping issues from even being considered. Thus, groups could exclude issues from consideration, maintaining the status quo without having to actually “win” any sort of debate, or exercise any actual power over anyone. Bachrach and Baratz (1962) claimed that to understand who had power, you should investigate who benefits from the status quo; at any point in time, as that is where power must reside. Lukes (1974) built on this idea to add a third “face” of power, the ability of a group to not only exclude issues from consideration, but to convince the non-elite that their interests coincided with the elite group, even when they may actually not.

To Lukes’ point then, power is about a transaction, or more broadly, an ongoing and evolving set of transactions that may or may not be carried out honestly, between two or more parties. It is clearly similar to authority, but authority is usually conceived in formal terms (as opposed to informal power) (Selznick 1953), or from the perspective of a principal-agent relationship as Carpenter and Krause (2015) conceive it. Authority can be viewed as a subset of power, constituting a dyadic relationship between individuals or organizations that cede or exercise that authority vis a vis one another. Power is broader, encompassing multidirectional relationships in a political environment that includes organizations, individuals, and institutions, as well as formal and informal means of influence.

Regardless of these complications, most simply stated, an individual or organization has power over another individual or organization if they/it can get the other to do something. This does not imply that there is no incentivizing or collaboration within that compellation; power over another is not strictly formal or non-negotiable. It is a two-way relationship in that the one that seemingly has power only has that power if the one that seemingly has no power submits to the nature of the relationship. This may happen in formal settings like organizational hierarchies or under a social contract, but it may also happen in situations where one recognizes that another has more information or credibility than they do.

While this may provide some clarity regarding the nature of power, it does not really clarify why power exists in the first place. It is usually taken as given that we want power to ensure that our preferred choices are enacted, but why is power the preferred means of achieving our wants and needs? Why is power always a consideration in the background, even in literature that focuses on collaboration? One might say, in the United States at least, that power and competition in politics arise from the ethos of capitalism, but the acquisition and use of power seem to be consistent across different economic systems and cultures. The roots of power may go deeper.

Although usually termed dominance rather than power in biology, power relationships are a hallmark of understanding animal behavior. In ants, for example, researchers have noted complex social relationships, social groups, and varying levels of status among those groups (Powell 2016). These relationships occur despite lack of evidence that these animals have any level of self-awareness or agency, suggesting that power relationships may be genetic features of adaptation and evolution. Increasingly, research on animal behavior is being influenced by complexity theory, which offers a potential framework for explaining social behavior that is adaptive and interconnected with the environment that the animals find themselves in, but it has been difficult to discern consistent trends in biological dominance (Hemelrijk 2002).

Social dominance appears to be evolutionarily advantageous for both the dominant and dominated within species groups. That is, some level of leadership and followership facilitates a species’ collective goal of surviving and thriving. Studies of our closest primate relatives suggest several reasons for the emergence of social dominance within groups of individual actors. These reasons are consistent with the human rationale for collective governance—ensuring access to food and other resources for survival, efficient utilization of acquired resources to improve group health, evolutionary urges for population propagation, and group safety (Thierry 2008).

Pertinent to understanding administrative power, species have been found to shift dramatically between an egalitarian-oriented culture and a despotic orientation by often subtle shifts in inherited behaviors (Vehrencamp 1983). Thus, many studies of power relationships amongst animal groups focus on the extent to which power is relatively evenly distributed (egalitarian) or concentrated (despotic) (Matsumura 1999; Watts 2010), which sounds much like human conceptualizations of sovereignty (Boehm, et al. 1993). However, sovereignty has tended to refer to social-level and constitutional issues more than to individual or organizational power relationships. Power seems to reside somewhere underneath, but is obviously related to, sovereignty. This egalitarian-despotic continuum is helpful as heuristic; however, because it is easily relatable to discussions of administrative power, where the concerns often focus on the legitimacy of administration holding power in a constitutional system (McSwite 1997).

ADMINISTRATIVE POWER

Administrative power has both despotic and egalitarian aspects, reflected in both formal and informal components, which relate to these legitimacy concerns. Formal components exist within bureaucratic organizations amongst individuals in certain roles, within the vague notions of oversight and accountability to the public, in constitutional notions of executive authority. The legitimacy concern in these formal powers relates to the constitutional role of the bureaucracy (McSwite 1997). Informal components relate to the relationship between the different agencies, the credibility these agencies and the actors within the agencies possess, the extent to which the public views the agency as credible (Carpenter 2010), and the norms and values underlying informal administrative decision-making (Goodsell 2014). Informal legitimacy concerns relate to the well-known tensions between democracy and expertise (Fischer 2009).

The formal powers exist under a Weberian conceptualization of bureaucracy. As Krause (1999) and Carpenter and Krause (2015) note, the expectations of the top-down process of authority colored much of the early literature on administrative power and bureaucratic politics (Rosenbloom 2008). This assumption of the supremacy of political authority motivated much of the discussion of the politics-administration dichotomy, as early public administration scholars sought to define the legitimate scope of administration (McSwite 1997). Later work attempted to distinguish informal power, including Foucaultian notions of power as social control (Farmer 1995), which administrative power simply being another venue for the informal domination of the elite over society.

Formal power, although no longer thought of as unidirectional, is still quite manifest in top-down conceptions of accountability to political authorities (Mulgan 2000), as well as in bottom-up conceptions rooted in rulemaking (DeHart-Davis 2017) and regulatory processes (Eckerd 2014). Working in this regulatory framework, Carpenter (2010) describes how informal power processes play out, as agencies strategically manage their reputations, signaling to both specific actors in political institutions and the public at large. Agencies acquire and use this power both for the ends of achieving organizational goals, and for the commitment to the public interest as described by Goodsell (2014). The venues for these power relations include some top-down processes, like the budget process or oversight (Bendor, et al. 1985; Hill 1995), some bottom-up processes, like petitioning and collaborative governance (Carpenter 2021; McKinley 2018), and the interplay of both of these processes as agencies send signals about their reputations and stakeholders receive those signals and signal their responses in turn (Dubois 2010).

The transactional frame of these interactions is not always explicit, but in describing the use of administrative power, it is usually implied. The bidirectional nature of administrative power is usually framed in some conception of legitimacy, whether as inherent as through expertise, or managed as through reputation, but in just about all discussions of administrative power, the power is conceived as being impersonal—based on ideas and not vested in particular people. Administrators are there to fulfill a role, whether in educating the public or listening to them and translating their concerns into administrative action. However, the considerations discussed here usually reside within a particular political environment. Power is also relevant in the broader environment.

Administrative Power and Sovereignty

Long (1949) situated administration in a role as functional implementer of the “general will” tasked with the power to translate the general will into specific programs and plans. He finds little contemporary use for the likes of Rousseau, Locke, and Smith, noting that with government and the economy irrevocably intertwined, conceiving of a distinction between the concepts of sovereignty and power was as pointless as conceiving of a distinction between politics and administration. He saw public administrators as serving two masters: The clientele (or the mission if you prefer), and the institution. Thus, when Long was talking about power in 1949, he was not referring to the role of public administration in the constitutional system (which did separately in 1952 as discussed below), he was arguing that contemporary administration could not be understood without acknowledging that administrators and administrative agencies sought, held, increased, and lost power. Long recognized administrative power as emergent and adaptive, even if he did not use those exact terms. This is so because administration requires power to implement solutions to social problems yet must also contend with a lack of clarity about exactly who it is responsible to, beyond vague notions of the people or the general will, and it operates within a constantly evolving political context.

He specifically makes no assessment as to whether this is a good or bad thing, but notes that with the intertwined political economy of the 20^th Century, the theoretically representative but increasingly feckless two-party system was steadily becoming dominated in a zero-sum game relative to administration. In 1952, Long elaborated on this latter point while articulating his view of the role of administration in the Constitutional system. He describes several aspects of Congress that render it less representative than it purports to be, which will be quite familiar to us today. A common theme for Long is the intractable partisanship of Congress as unrepresentative of the generally moderate centrism of the American public, and he also notes the overrepresentation of the wealthy and the elite in Congress. From this perspective, administration is much more representative of the American population since most administrators are not wealthy or particularly elite. Like the general public, they are more driven by an ethos of scientific discovery than by partisan or ideological squabbles (which may be less true today than in the 1950s, but it is certainly still debatable). Again, although Long does not really take a position on whether this is a good thing or a bad thing, he seems content with the role of administration in the democratic system, similar to the elaborated arguments along these lines by Rohr and Cook described below.

From a different perspective, Lowi (1986) sees increasing administrative power as a threat to democracy. Relying on science or professionalism and giving extensive discretionary power to administrators is problematic for democracy in that it empowers an unelected individual or favors an epistemology over the preferences of the polity. Unlike earlier scholars who saw power as residing in knowledge, Lowi saw it as being held by people. However, Lowi (1986) is not particularly optimistic about any of the options for introducing more democratic elements to administration either. Opening administrative decision-making to public participation just presents an additional opportunity for the most well-funded and well-organized interests to get their way, and probably more effectively than when trying to influence the legislature. Lowi does not see the representativeness of administration that Long does, arguing that administrators are elites, perhaps not in wealth, but in education and social standing. He says that relying on the moral foundations and administrative commitment to rationalism that Long finds comfort in will concentrate power in the hands of these elites, who will ultimately have the power to decide what is right and wrong.

Long found administrative power well situated within an ethos of representative democracy. To Lowi (1986), the only proper means of administrative power is via a constitutional-legal basis. This constitutional-legal arrangement is premised on the social contract, wherein the polity agrees to two points of delegation—first to an elected representative body, and again to a legal structure based upon the governing of that representative body. This might be conceived as the American ideal that Long found unrealistic: a legislature is elected to represent the constitutional and egalitarian manifestation of power, which necessarily delegates power to administrative authorities via the rule of law, with law (and not individuals) as the manifestation of power. There is not so much a tension between legislative authority and the rule of law, as there is a process of delegation. To Lowi (1986), a duly elected legislature settles the matters of public concern, and the rule of law establishes a consistent order. Law can always be altered by the elected legislature, appointed judges, or constitutional amendments, so the potentially despotic power implied by the rule of law is subservient to the egalitarian power of the constitution, via the legislature. The law can handle the day-to-day operation of sovereignty, always subservient to the popular will which can only be feasibly assessed in elections and served by elected representatives.

Any other manifestation of administrative power besides a constitutional-legal basis would inevitably lead to “patronage...[or] serfdom—the capacity to distribute material resources or privileges on a personalized, individualized basis” (Lowi 1986: 297). In fact, this arbitrariness was Lowi’s chief and an overriding concern, and he argued against enabling other manifestations of administrative power. This includes the sort that Long discusses, which we might call scientific or professional power, but also power that is ceded to particular non-representative publics during administrative decision-making or any manifestation of power that bolsters elites. Perhaps a further subtext is that while Long was writing during a period of time of rapidly expanding power of the presidency, and by extension the executive branch, Lowi was writing during the age of television as the presidency became even more of a power center, but was often seen as competing for power against the rest of the executive branch (i.e. administration) (Livingston 1986).

While Lowi’s (1986) diagnosis of the potential despotism of administrative power resonates, decades earlier Long critiqued this idealized solution of an administration subservient to the legislature and legal basis as not realistic in that public problems are wide-ranging and complex, and neither the legislature nor the people have the time or breadth of knowledge needed to deal with them. This is essentially the basis for Rohr’s (1986) argument about the Constitutionality of the administrative system: legitimacy is beyond legality. There are too many gaps between what law can do and what must be done to govern a society. Rohr (1986) argued the administrative system was thus legitimate in that it existed to govern in the gaps between broad social covenant of the Constitution and the necessarily limited capacity of law. In the grand scheme of governance, the president, Congress, and the courts had very little capacity to govern beyond the most pressing matters, leaving administration in the legitimate position to fill in the day-to-day tasks that could not be managed in the other branches. Although we may not necessarily say that the legitimacy issue is settled amongst the general population, there is at least some consensus amongst scholars (with a few notable exceptions described below) that the administrative system plays some legitimate role in the broad swath of political and governing power filling in the power void of day-to-day governance.

Cook (2014) argued that both Lowi and Rohr were correct: Administrative power could have the propensity to be despotic, but it is not a fait accompli that administrative power will be despotic provided that the bureaucracy does indeed fit within the constitutional structure and its discretion to use its power is balanced by the roles of the presidency, Congress, and the courts. To Cook, some discretion for administrators is warranted given Rohr’s points about the limited capacity of law, but complete discretion is unwarranted because that obliterates the rule of law, which is held up by the three branches. An ethos of administration is adherence to the rule of law, but the rule of law is also a bulwark against a potentially authoritarian administration.

Cook’s argument nicely represented the status quo of thought about administrative power for several decades. However, things shifted early in the Trump Administration where administration and administrative power were held up as the bulwark of the rule of law rather than vice versa. In contrast to Lowi’s (1986), but consistent with Long’s (1949, 1952) arguments, administration was seen as the representative branch that could protect democracy from authoritarianism. Horwitz (2021) offers a good example; in an effort to refute Trumpist accusations of an authoritarian deep state, Horwitz claims that the administrative state’s norms of liberalism and instrumental rationality render it a more coherent protector of democratic values than populist leaders with authoritarian tendencies like Donald Trump. Bauer and Becker (2020) argue that several populist leaders—democratically elected populist leaders—are undermining democracy by undermining the administrative apparatus in their states, while Hatcher (2020) channels much literature on this point during the coronavirus disease 2019 (COVID-19) era, arguing that communications from democratically elected figures can undermine democracy if they contradict expertise. Bannister and Connolly (2018) even go so far as to articulate a rationale for restricting information and communications technology in order to restrict the ability of the public to communicate ideological preferences with one another—essentially an administrative co-optation of free speech, seemingly unironically suggested as means to better foster democracy.

My intent here is not to critique these arguments, as such, but rather to note that our view of administrative power has often hinged on the extent to which the polity elected leaders who governed in ways consistent with the ethos of administration—that is, technical rationality and neutrality—or perhaps more generally a polity that adhered to these norms. When administration was seen to be cloistered, it was opened up to more public engagement, but that engagement still took place within the confines of administrative thinking (Eckerd and Heidelberg 2020). When there was concern about the views of presidential administrations that varied from administration ethos, administration was seen as the bulwark against authoritarianism, even if it operated in ways that could be interpreted as despotic (i.e. mask mandates, restrictions on free speech, and so forth)

Contemporary literature is generally positive about administration entering the power vacuum left by a hyper-partisan, largely inactive Congress, an increasingly inept and populist presidency with eroding legitimacy owing to partisan attacks and the electoral college, and a Supreme Court willfully immolating its legitimacy (Hatcher 2020). Politics has become more akin to a reality television competition that occupies the public’s and the media’s attention, with relatively little discussion of policy or governing. Administration has become the power center in American government—the only entity of government capable of regularly acting because most of the time it exists outside this partisan reality television competition. Administrative power still exists under the same social contract as the other branches but its role within that system is less clear. The basis for the powers of the presidency and legislature are clear—the constitutional system and elections. Even the courts, with the judicial appointment process, are more clearly tethered to the political process. But while administrative power exists in this same realm, it is also different, in both its bases and how power is exercised. We need democracy and the rule of law to protect us from bureaucratic overreach, while at the same time, administration is the protector of the rule of law, and increasingly, democracy as well.

TRANSACTIONAL AND EVOLVING ADMINISTRATIVE POWER

Administrative power is fundamentally about legitimacy under the social contract. Just as the nature of the social contract is vague, we could thus conceive of administrative power as rather amorphously based upon the people submitting to living under the authority of a set of ideas and ideals rather than under the domination of any individual. Administrative power, in the context of a modern democratic state, can be thought of as a set of transactional social contracts, more or less in constant negotiation whether overtly or informally, wherein the governed agree to submit to institutional authority in exchange for the other benefits (e.g., safety, opportunity) that the institutions can provide. That these social contracts are always in flux is illustrated by the tenuousness of power that Long (1949) described. These social contracts are in flux because the benefits that institutions provide are often uncertain—I return to this point below, but first I discuss the framing of power within the social contract.

Power Within the Social Contract

I do not rehash Rousseau’s or others’ conceptions of the social contract here, except to note that the initial conceptualizing of the social contract was an egalitarian response to centuries of despotic domination. In despotic regimes, the question of agency power was mostly a matter of palace intrigue. Administration writ large existed to serve the despot’s interest and had complete domination over the public. Any power variance between different parts of that administrative system was due to jockeying for the despot’s favor. The agency in despotic regime has either one individual or oligarchical group who must be convinced of the value of what the agency does. In a democracy, this power is both more diffuse and it is less clear who has power. The bureaucracy does not have constitutional power over the public, but the people do still agree to acquiesce power over themselves to agencies via the social contract, as reflected in laws that give agencies power. In a democracy, the agencies need to convince the public of the value of that trade-off—a much more difficult proposition than the despot’s bureaucracy which needed only to convince (ultimately) one person of the value of services provided. Administration in a democratic system must accomplish two tasks: Maintain status by convincing the public to not challenge their power and maintain access to resources by convincing the public that the value of what is provided, either in terms of services or in responsiveness to public concerns, is better or more important than what another agency provides.

But who is this public that must be convinced? The American framers realized that the people cannot feasibly rule a complex modern state, any effort to do so would likely be chaotic, and any method through which we attempt to govern introduces a human element that inevitably leads to arbitrariness, despotism, or both—which is exactly counter to the premise of popular sovereignty. This is fundamentally why the Constitution is written as it is, trying to create a delicate balance of ensuring that the popular will informs governance decisions, but that those decisions are not completely subject to the whims of the public.

But even beyond this, the concept of enabling some entity to have power or dominance is uncomfortable under popular sovereignty. Are we comfortable with calling power a “favorable balance of constituencies” when it is unclear as to whose favor that balance tips? The root of the biological concept of power is dominance; accepting the premise of a favorable balance of constituencies means accepting the dominance of one over another, and since governing—solving social problems that require expertise beyond what a representative body can do—must be done, there must be some basis for people to accept domination by administration. I see this as Long’s fundamental point. The American Framers intended a popular sovereignty that was based on reason and an attendant set of institutions that would be guided by reason—in this way shifting the idea of power from dominance of a monarch over subjects to an enlightened society that succumbs to its will to reason. However, this does not fundamentally solve our problem about whose favor the balance tips towards, as clearly the power in a society would be held by those who determine what “reason” means.

The ability of any one agency or institution to dominate is constantly up for challenge—perhaps not in the same way as the fight for dominance within animal species, but with some clear parallels. Dominance functions to better ensure survival of a group by reducing the energy that members of the group needed to expend on unnecessary uncertainties. One of the key functions of a social group is to reduce the complexity that any one individual in the group must deal with. This is ultimately the same rationale for government and for the social contract. I will give up a portion of my right to defend my home and property in exchange for agreeing to live under laws that make me safer to begin with. The complexity of my daily life is reduced because I worry less about being attacked by others, which allows me to concentrate on other matters. But that calculus is always in flux to an extent; if at any point I feel that the institutions that are there to protect me are not effective, then I may revoke that social contract, or at least challenge it. Power requires one party to have some extent of contextualized dominance over others, with the dominated party either accepting that dominance because it provides tangible benefits, like time savings or resolving uncertainty, or challenging that dominance because it does not.

Along these lines, Long’s (1949) article was influential, but perhaps not in a way we might expect. Long was not the first to note the increasing complexity of governance, but he may have been the first to note this complexity—the intertwined nature of government, the economy, and society—as fundamental and endemic to administration, rather than as a particular problem to be solved. One might trace a line from Long to Rittel and Weber’s (1973) conceptualization of wicked problems, or problems that are endemic, that defy easy definition, and that are manageable, but unsolvable. Standing agencies and institutions exist for the management of complex problems, that is, mitigating the effects that these problems have on the members of a society by resolving at least some of the uncertainty associated with them. Long did not elaborate on what dealing with complex problems might mean for administrative power, but he does allude to the role that uncertainly plays in making decisions within the interconnected realms of government and the economy, implying that administrative power arises in the vacuum of this uncertainty.

Heidelberg (2021) addresses this point, arguing that the obvious decisions, that is, those guided by generally accepted knowledge, are immaterial for considerations of sovereignty (or by extension administrative power). We can put rules and processes in place to manage these sorts of commonly occurring problems and not worry about power or sovereignty and just solve the problem. These are the very cases where we have submitted our will to reason and the sovereignty issue has been resolved, as Wilson suggested 150 years ago. Heidelberg (2021) is concerned about who decides in cases of exceptions to the rules—problems for which there is no obvious rule or process or procedure. That is, not only must there be gaps in the laws owing to the complexity of governance as Long described, but there are gaps in knowledge owing to the limitations of human understanding. For such problems, any solution will be partially arbitrary insofar as some human or set of humans will make a decision with, at best, partial information, derived potentially from an interpretation of existing knowledge or conversation with particular publics. An agency will be given the power to dominate where there is no rule to decide, where reason is unclear. If we choose to use less aggressive language, we might say that a street-level bureaucrat is exercising discretion or that an agency is following professional norms in pursuit of its mission, but the point still remains: The decision is an act of power outside the bounds of generalized knowledge or reason, which is the basis for legitimacy in administration. The rules that exist make this palatable not by resolving the uncertainty but rather by providing some measure of consistency to what are fundamentally arbitrary choices.

To keep the system stable, over time we have constructed an administrative state that is a function of rules, processes, and procedures, and the faceless implementers of these rules who have the appearance of dominating, but never see those who they dominate and never see themselves as dominating because they themselves are dominated by knowledge, rules, processes, and procedures. But in a complex society, the rule writing can never end. As Heidelberg (2021: 801) notes:

“[T]he limitations of the decision rules are made increasingly apparent, which leads to a proliferation of rules as the decision system develops increasing levels of comprehensiveness…In a commonplace sense, the goal is for there to be no exceptions to the rule, a radical necessitarianism brought together by that unattainable goal. Unattainability notwithstanding, the goal itself drives the design of systems that constitute the Administrative State.”

Democratic society chafes at the idea of individuals making decisions that affect others without their explicit preferences driving the decision, but we also abhor uncertainty and the resource cost that uncertainty entails. In our efforts to control this uncertainty without ceding our will to other people, we ultimately end up with impersonal institutions and sets of rules. As the democratic social contract has evolved, these rules have become more and more extensive and impersonal. We might conceive of these as formal rules like laws or bureaucratic procedures, but they can also be informal rules like the dictates of science, religion, or experts. The whole process is recursive. Complexity in society leads to more rules, more power given over to rules, and as there are more rules, it reinforces and adds to the complexity, and we are left with an administrative state that has power not because of some fundamental feature of administration within the constitutional system, but rather because we have defined the social contract in terms of reason, and administration has emerged as the province of reason and the attendant rules that governing by reason requires.

The COVID-19 pandemic offers a good example. No matter how the pandemic had been dealt with, it would have been complex and characterized by uncertainty. At the same time, the public consistently supports evidence-based policy decision-making, and it has become a rallying cry during the COVID-19 pandemic that policy be driven by science and evidence. Those of us in the academy likely agree with this premise, but it is also an example of how we attempt to resolve uncertainty by impersonalizing the decision so that it rests not on the power of an individual but rather on an abstract idea. We wear masks not because someone tells us to, but rather because science does. The impersonal nature of this transaction is appealing because relinquishing our will to evidence and knowledge is more comfortable than relinquishing it to a person. We might then expect that the more that an agency is able to convince us that reason and science are the basis for its power, the more power it will have. Or conversely, the more that an agency is reliant on the charisma of its leaders or on unconvincing evidence, the less power it will have.

POWERFUL AGENCIES

We might then expect that the agencies that can most easily rely on settled (or mostly settled) knowledge, and build their reputations from that basis, are in an advantageous position. Gormley (1986) was focused more on legislative or industry control over bureaucratic agencies, then agency power, per se, but the argument is similar for thinking about power. Gormley argued that bureaucratic politics relate to how salient an issue is with the general public and how technically complex it is. Ringquist et al. (2003) note that legislative authority over an agency was more likely to be utilized if the issues the agency deals with are generally highly salient and lacking in complexity—that is, if the public had a strong opinion on an issue and it was understandable to them, then legislative control of an agency was more likely, the inverse of an agency with power. Along similar lines, Bach et al. (2020) found that when agencies had higher status, were more professionalized, and had sufficient resources to be relatively autonomous, they were less subject to external political control, that is, more powerful.

But political control is only one venue of administrative power. Long was speaking more broadly than this, and administrative power is more than a continuum between legislative control versus autonomy. Part of Long’s point was that agencies have power vis a vis one another, and often power over the other branches and the public more generally. Drawing on the literature noted above, it seems that salience of the work the agency does is important, as is the complexity of the issues with which the agency deals. However, it could be argued that most of what administrative agencies do is complex, and salience is likely dependent on a host of external factors over which an agency has relatively little control. Besides this, both salience and complexity are tricky in a political environment. While legislative authorities may have historically attempted to assert control over salient issues, it has also been well documented that agencies manage their public relations to ensure that public support for the agency’s work helps keep that legislative authority at bay (Carpenter 2010; Durant and Wilson 1993). Beyond this, while it seems likely that salience rooted in negative perceptions of any agency likely limits an agency’s power, positive perceptions could do the opposite. Salience is also situational; issues may be highly salient with a very narrow set of stakeholders but not the broader public. Complexity is similarly nuanced in that complexity comes in many different varieties—not only technical complexity, but also complexity rooted in large and diverse stakeholder groups, complexity related to interacting social systems, and the adaptiveness present in the policy environment.

Salience and complexity are perhaps better thought of as situationally endemic factors of an organizational environment that agencies seek to politically manage within (Pandey and Wright 2006). That political management often takes the form of reputation management along the lines described by Carpenter and Krause (2015). They may do so by a) ensuring that the public is aware of what the agency does, focusing on the positive aspects for the society; b) providing clear and consistent messaging about the issues the agency deals with in order to make the complexity of these issues appear to be manageable; and c) project confidence about the agency’s ability to manage these issues in a way that communicates the social value of their work to the public. In short, when an agency can convince the public that the transactional nature of the power trade-off is worth it, particularly if they do so in ways that engage the public, they have power both relative to other agencies and with their ability to act on decisions that affect the public. They can then exercise this power without exposing themselves to political risk; and this is ultimately what agencies and the public want. Agencies want to avoid political risk as it puts budgets and discretion in jeopardy, and the general public wants to avoid risks to their safety and welfare.

Risk Communication: Confidence, Clarity, and Consistency

In recent decades it has been increasingly clear that complexity is a context in which policy and management take place rather than a problem to resolve—complexity can be managed within, but it cannot be eliminated (Morçöl 2013). Salience functions similarly. The attention span in the policy cycle is variable and subject to external events that may have little to do with anything that an agency is actually doing (Jones 1994), although of course, it could be because of what an agency is doing. Agencies can do little about increasing complexity or increasing salience, but that does not mean they are completely at the whim of these effects.

The problem this poses for an agency is that people do not like the uncertainty complexity fosters as it taxes resources and that is why we cede power to begin with. What we might expect is that administrative power is concentrated in those agencies and institutions that are best able to convey certainty about decisions, a certainty that rests upon some impersonal, but widely accepted social constructs, like technical clarity, evidence, or scientific reasoning. The certainty might be relatively realistic in the sense that we have good information, or that the problem is well structured, or it might be conveyed to those ceding power that there is certainty where there actually is little, focusing on the reputation of an agency as a whole in a way that is uncoupled from a particular issue with which the agency deals. Agencies can choose to communicate uncertainty or certainty, somewhat regardless of the actual level of certainty in large part because the general public is not in a position to know how much uncertainty there is.

There is a fine line in communicating information to the general public that acknowledges that there is uncertainty but also stipulates that there is actionable information that an agency can use to make decisions. Further, it is a challenge to communicate valid information in a way that is approachable to non-experts (Matheny and Williams 1985). Agencies themselves work within uncertainty (Potoski 1999), operating with an understanding of risk based upon uncertainty that is aggregate in nature, impossible to eliminate but mitigatable. They assess risk as something to minimize but not eliminate, recognizing that, in the aggregate reduction in risk is a key goal. The public, however, does not view risk similarly (Eckerd 2014). For individual members of the public, risk is viewed more personally—the risk to the society is a secondary consideration to the risk to oneself. To this end, people tend to limit consideration of risk information to that which is identifiable to themselves, both because they are not experts and can only process certain information, but also because they are (quite reasonably) focused on the individual risk to themselves (Zinn 2005). As noted by Alaszewski (2005), members of the public are not passive recipients of information, but rather active seekers of it—but seekers of only relevant information to their context.

Effective risk communication is about bridging this divide. As Sterman (2008) notes, to effectively communicate risk, experts need to recognize the mental models of recipients, aligning communication in a way that both conveys the expert understanding of the problem and fits within the mental model that recipients have of the problem. It is difficult to do this without some level of engagement with the affected public. Effective communication conveys what is known about a problem in understandable language couched in positive, actionable recommendations about what individuals can do. This is not to say that information is obscured or miscommunicated but rather that it is presented in a digestible way that frames action to be taken in a positive way.

However, this is generally not how administrative agencies communicate risk information. Administrators tend to use technical language and convey aggregate risk in a way that reflects their own understanding of risk rather than the community’s understanding of risk (Eckerd 2014). Government officials emphasize the top-down communication of “robust scientific and expert knowledge” (Boholm 2019: 1695) which is often inaccessible to the public and, at best, fosters uncertainty in the general public rather than resolving it to the extent practicable unless the agency is already trusted by the public (Holmes et al. 2009). This is not to say that people cannot handle difficult information, but rather that the information needs to be provided in way that makes sense to them (Cope et al. 2010).

Risk communication literature, most of which exists outside of public administration literature, offers a range of different ways to think about informing the public about policy issues. There are several consistent themes. The information provided to the public should be appropriate for their concerns, but it should also be honest about the uncertainty that is being conveyed (Leiss 2004). However, the public does not necessarily have the knowledge to distinguish between levels of risk or the particularities of uncertainty (Holmes et al. 2009). Therefore, agencies that communicate confidently likely establish more trust with the public that they are trying to inform (Cope et al. 2010). This should not be misplaced confidence, but rather confidence that indicates the expert understanding of an issue and extent to which there is expert consensus. Rather than highlighting the uncertainty of the issue, agencies would likely foster more trust, and thus legitimacy and power by highlighting the aspects of the issue that they are confident about, and while acknowledging areas of uncertainty, do not dwell on them but rather offer positive steps that can be taken. Even under the most complex situations, there are bound to be things that an agency is confident about, and a focus on these will likely resonate more with an audience.

The information provided should not be overly complicated. There is a process of simplification that must occur any time that complex information needs to be conveyed and this does not mean that communication needs to be dumbed down, but rather simplified in a digestible way that fits within the understanding of the audience receiving the message (Eckerd 2017). Obviously, there are different audiences, and those audiences have different levels of education and knowledge so the information may be more or less technical in nature depending upon the audience, but even sophisticated audiences do better with clear, concise depictions of issues than with opaque or complicated explanations. The clarity of communication can enable a message to be received and understood, which allows an agency to be seen as being helpful, thus also fostering trust and legitimacy.

Finally, few things erode trust in an institution, and thus legitimacy and power more than messaging that is inconsistent (Galford and Drapeau 2003; Lee 2009). Agencies that offer consistent information that becomes increasingly familiar to the public likely also fosters trust, legitimacy, and power. This is not to say that communication should not change as information changes, but rather that the messaging should convey that information in a consistent way that makes sense to an audience. An agency with power is unlikely to recommend one set of behaviors and a short time later recommend an entirely different approach.

FROM CLIENTELE SUPPORT TO SPIN

In this article, I have argued that administrative power is fundamentally about a social transaction. Agencies with power are able to effectively convince key audiences that the work they do is worth accepting power over them. I expect that agencies that communicate clearly, confidently, and consistently to their stakeholders will have more power than agencies whose messages are more muddled. In 1949, Long noted the importance of a powerful clientele and in 1980 Meier noted the importance of clientele support for agency power. Today, agencies are seen to be more affected by the public more generally than just clientele (Denhardt and Denhardt 2000), but throughout the growth of the administrative state, support from both directly affected publics and the public at large has remained a key determinant of the power that an agency has. While others have noted the importance of agency performance in securing this support (Brewer and Selden 2000; Rainey and Steinbauer 1999), past performance is just one factor and the framing and communication of that past performance might be more important to interested stakeholders than the performance itself (Behn 2003; Melkers and Willoughby 2005; Yang and Holzer 2006).

This may be particularly acute in more recent times as administrative decisions have become more politicized. Miller and Mooney (2010) argue that “spin” and public relations, or the orientation of information to fit the political environment, from corporate actors, has increasingly affected the policy environment, and the same could be said of public agencies themselves. Spin has a negative connotation (Roberts 2005), with a supposition of blame shifting being the primary motive (Hood 2010), but the difference between the recommendations of risk communication and the spinning of communication to be favorable in a political environment are not that different. While the 20^th Century focus of communication may have been focused on specific clientele (Meier 1980), the 21^st-century focus may fit more in the context of spin. Bell et al. (2010) note the ways that government agencies shift messaging depending on the audience and the intention of the communication. Gilad et al. (2015) identify blame-shifting (i.e. spin) as an important strategy for agencies in maintaining their legitimacy, while Figenschou et al. (2022) note the different roles that agencies take on as a communicator depending upon the framing need (i.e. task orientations versus value orientations). Spin is the strategic framing of information to put the agency in the best light politically and improving its reputation (Carpenter 2010), while good risk communication is the strategic framing of information to build trust to ensure that people act in ways that minimize risk. In our current times, communication that achieves both of these goals may be optimal for garnering power.

This is because the goal now is the same as the goal then: persuasion (Bell et al. 2010) and reputation building (Carpenter and Krause 2015). Power is acquired by administrative agencies that are successful in persuading the right people that the agency deserves power. In the past, that persuasion may have been focused on relatively small or cloistered clientele groups that were part of the iron triangle that agencies used to maintain power, and in the present that might be more focused on the media (traditional and social) and communications more directly with the public, but in either case, powerful agencies likely communicate in a way that the intended audience can understand and relate.

Long wrote in 1949 that the presiding view of administrative power—of power in the executive branch flowing through the chain of command evenly and cleanly from the president—was an “idealized distortion of reality” (Long 1949: 258), noting instead that power flowed up and down, from the outside in, and from the inside out. No level of a public agency is immune from considerations of power, both decision-making power within the organization, and the political standing of an agency externally. Long viewed the acquisition and utilization of power by agencies not as a cynical ploy to acquire resources but rather as a realistic response to the nature of the work they do and the political environment in which they work. Power enables agencies to have the political and monetary resources they need to meet their missions, and as I argue here, constitutes a central theme of agency work in ensuring that the agency’s reputation makes the power conferred to it by the public seem worthwhile.

REFERENCES