Skip to main content

Principles and Practices

Constructions of Public Office: An Analysis
Mark Philp

This is a slightly complicated paper (for which I apologize) in which I try and map the conceptual dimensions of understandings of public office. A truncated powerpoint presentation that summarises the central elements of the paper can be found here.The thrust of the paper is simple – there are many dimensions to office, they pull in different directions, and are potentially conflicting, which means that they give rise in different historically specific locations to distinct understandings and sets of practices. As such, this is best understood as challenging the view that there is only one right way of understanding public office.

King Lear: Thou hast seen a farmer's dog bark at a beggar?
Gloucester: Ay, sir.
King Lear: And the creature run from the cur? There thou mightst behold the great image of authority: a dog's obeyed in office.

Office – a position or post to which certain duties are attached (OED).
A duty attaching to a person's station, position, or employment; a duty, service, or charge falling or assigned to one; a service or task to be performed; a person's business, function, or part. (OED)

Office seems intimately related to command and authority. On one set of claims, what conveys authority on the office is not the occupant but the office – ‘a dog’s obeyed in office‘ – that is, a set of formal elements of office confer rights on the office holder (whom I will refer to as the agent) and reciprocal duties on the part of others (whom I refer to as members of the office’s domain) towards the office holder. But this is only one form of office and only one basis for authority.

We might ask from where the authority of the office derives – in some cases one office derives its rights from another office – for example: ‘in virtue of my powers as Lord Chamberlain I create an office for the execution of a subset of my powers and duties and commission a person to assume that office and its functions.’
Alternatively, an office may have its authority on other bases –
i. where an agent commands in virtue of their person (rather than the office conferring authority on the agent, the agent confers authority on the office);
ii. or where the personal authority of the principal confers authority on a sub-ordinate office. When I task someone with a responsibility, we might say ‘this is my officer’ – which emphasises my authority – in contrast saying ‘This is the/his/her office’ which opens space for a degree of independence for the office from the will of the occupant of the superior office (the Principal).

This also raises questions about the basis on which the occupant of the office – the agent - recognises a right on the part of the principal to command him – and the extent to which he/she accepts/attributes/confers/acknowledges the authority of the principal to accord a set of duties and rights – and tasks – to the agent. (Although I will frame things using the language of Principal and Agent, as should be clear from the following, this is not a single type of relationship but may take many forms. The Principal may be a single individual – as with a king – but it is also common to think of office holders as acting in the interests of the public and at their behest, even if that relationship is mediated by other actors and processes – such as elections, referenda, etc.).

An office is task (which may be more of less defined, more or less open-ended, and more or less open to interpretation and discretion) or set of responsibilities/duties/tasks. It is often framed by one (the principal – or in the name of the principal, as when we talk about the sovereign people) to whose command we submit (on various grounds)
The office might make no claims on others (over no domain) – but merely fulfil the desire/commands of the principal. Or it might make claims upon itself as a claim to a distinct authority and to the corresponding compliance or subordination of others. P commanding A to spy of Q gives no authority to A over Q to which Q need respond. But P may be in a position to accord authority to A in a situation where Q has a duty to comply with the orders of P and his/her subordinates.

'Office' then involves (in broad terms) the delineation of a task. The task is a thing in itself. It is not the task of the agent qua individual directly; rather it is a task that the agent is accorded – by grant directly or by acceptance of the office, or sometimes by self-stipulation (I make it my office to….), and in the execution of which certain responsibilities are entailed, which take the agent beyond their personal and private actions and preferences.

That delegation may be informal – asking for the office/service of a friend; or formal, in the delegation of certain rights and powers. But my sense is that this has to be seen as a continuum – from the tasking of A being tied up with a wholly personal relationship – to the tasking of A being a concomitant of a wholly formally defined office. In contemporary politics we still see this played out at the intersection between the strictly political and the strictly administrative – where Ministers want people they can command on the basis of loyalty and ideological affinity (SPADS), but where the formal civil service sees itself as providing an administrative service which is political neutral (and which they tend to see as sufficient for ministerial advice). In part, though, they seem to be disagreeing over what the task is – and how far it falls within a personal/political service, rather than an administrative service.

There is a potential (and often actual) separation of the person from the office; and from the agent’s interests qua individual, and the concerns and responsibilities of the office. Weber has very interesting things to say about the way in which the conception of an office that was to various degrees independent from the principal comes about, in the shift from the household to patrimonial order where the Lord has an interest in not adding further to the household, and so uses free men to conduct his major offices, so that the office develops a degree of independence – and in doing so might become something that could be sold or transferred. He’s also interesting about the way in which there is a movement in the Western Church to resist the practices of the Eastern Church, where the priesthood ceases to be part of a monastic community and becomes a prebendary or independent benefice, which then placed restrictions on the degree of control Bishop could exercise – partly because benefices could be linked to parishes under secular control – a practice the legacy of which might be seen in the control of parish appointments by institutions such as Oxford and Cambridge colleges. But at a more abstract level (if that is possible) we can see this as a struggle between different components and degrees of precision and discretion:
The principal and the functions s/he wants the office to serve
The office and the formal tasks that are assigned to it
The arrangements for financing of the office
The control or accountability of the office given that it acts increasingly at a distance from the ultimate source of delegated power
And the interests that the office holder has – as a person, as a holder of the office, as a subordinate to the source of office, and as an agent in relation to the intended human domain that the office commands.
In different offices, within different structures there may be more or less interest in some or all of these dimensions. In farming out taxes, where the agent’s interest is to maximize their independence, their income, and to resist the withdrawal of their office (see Weber, Economy and Society University of California Press, p.1033-4) – so that the office becomes a property that the agent is able to sell (after 1567 (in France) with a fixed fee paid by the successor). The principal’s interest might be to optimize tax returns while minimizing costs incurred to do so.

The above comments emphasise the relation between the agent, the office, and the structure of authority/principle’s commands, within which the office derives.
Not all offices serve the public (for example, in firms the role of the agent is determined by those who control the definition and pursuit of the interests of the firm); but most office involves something (a set of claims to rights and a set of responsibilities) that goes beyond the immediate personal interests of the person who holds that office, and that implicitly raises a potential for a conflict of interests between the agent qua agent and the demands of the office the agent occupies. This gives rise to the classical Principal-Agent (P-A) problem: It is in the agent’s interest to ally their responsibilities as closely as possible with their personal interest; it is in the interests of the delegator of authority that the office and its responsibilities serve their designated ends. But, as in much contemporary analysis of P-A problems, the degree of independence ascribed to the office, which on the one hand makes the delegation more perfect, simultaneously renders the delegator dependent on the delegate for an account of how the office is being exercised (and if you appoint a third party, you face the problem of regulatory capture and ‘who guards the guardians').

How then does the principal ensure that the agent performs his/her office in ways that are compatible with/optimise Ps interests. Note, that in the case that the office derives from the personal relationship between the principal and the agent, where the grounds for A’s accepting the office are motives such as friendship, loyalty, fealty etc., the interests of the agents are more closely aligned with the interests of the office – I do this for you, because I do it for you. Where the office is undertaken on the basis of other motives – a position, an income, etc, then the P-A issue raises its head more powerfully – since interest drives the taking of office, how do we sequester interest off in the conduct of the office?
This suggests a potential tension in all office: that the more personal and personally motivated the undertaking of office is, the more aligned are the interests of the agent and the principal; the more impersonal the relationship to office is – the less the agent’s motive is to serve the principal – the more the principal needs mechanisms to ensure that the agent’s actions align with the responsibilities of office.

One dimension of the ‘task’ that office often (but does not necessarily) confers concerns the authority of the office over a domain. For A to have authority, A must be authorized by the office – but that authorization has two distinct dimensions – the authorisation by the P of the A with respect to some action or domain; and the authority of A with respect to the rights, claims, persons in the domain (D). For those in D to recognize A’s authority in some relevant respect, involves attributing to A a right to command in a certain respect in relation to the D.

All office then raises questions about the extent to which the office is one that arises from the personal relationship between the P and the A; how far the office confers rights or claims in respect to a particular domain which may or may not be formally under the control of the P, where different P’s may create different relationships (for example, the commands of the sovereign qua Crown are likely to have a different character to those where the P is identified broadly in terms of the people and their interests, interpreted through a range of institutions); and how far the office is institutionalised so that it derives its authority from the rules and formal structures of an institution or set of institutions (the state) and how far the domain (over which the A attempts to exercise that authority) recognises A’s authority, and whether they associate it with the formal structure (i.e., the fact that P has delegated/entrusted/contracted to A), or whether in fact they resist and deny that the A has any such authority over them by denying that A derives authority from P (is acting ultra vires); or denies that the system of rules legitimates As claim to authority in this respect; or denies that P has a right to claim to confer authority over D on A.
Finally, there is an issue about ‘nesting’ – that is, how the various claims to authority and commands are stacked relative to each other: As the boss of a firm I can task you with various responsibilities that serve my interests (as a family firm), or those of my shareholders (in a corporation), but what it in my power to require is limited by the law – both in terms of what it is legal to do, and in terms of workers’ rights. Moreover, when I contract with the state to provide a service my capacity to insist that the interests of the firm come first is further limited, although by how much it is limited is contested.

What is a Public Office and what does the dimension of 'Public' add?
Sir George Cornewall Lewis wrote:
Public, as opposed to private. Is that which has no immediate relation to any specified person or persons, but which may directly concern any member or members of the community, without distinction…a theatre, or a place of amusement, is said to be public, not because it is actually visited by every member of the community, but because it is open to all indifferently. In the language of our law, public appear to be distinguished from private acts of parliament, on the grounds that the one class directly affects the whole community, the other some definite person or persons. (Brian Barry, Political Argument (RKP, 1965), 190-191.

On this basis we can distinguish very tentatively between public and private office (but we should bear in mind Raymond Geuss’s comment on public vs private: ‘there is no such thing as the public/private distinction’ – on the grounds that the lines between them are drawn in different ways, and serve different purposes on every occasion, and it is essential that we should be attentive to that variation).
The office one does for a friend is a private office; the office one performs within a wider public institution may be public. The office that identifies particular persons as its beneficiary (the executor of a will or trust) is private; whereas an office that serves members of the public indiscriminately is a public office. Thereby giving two dimensions and four possibilities:

1. An office may be within a private domain and serve private ends – servant, friend, member of the household. Here loyalty or fealty is central.
2. Be in a private domain and serve public ends – my servant distributes Maundy money; a private company in a PPI? SPADS – who serve the minister in the capacity of a personal office, but where the office has a public orientation. This office relies on loyalty, but there is an equivocation in that it is loyalty owed to a specific individual, but unclear whether it is to that individual as the holder of an office whose powers and responsibilities are justified by reference to the public interest (where the justification for the appointment of a SPAD is that they assist the minister in their capacity as a political office holder – where political office and public office can be seen as in tension); or whether – as is often claimed in relation to developing countries - the responsibility is seen as deriving from the personal relationship. In many cases the factor of time-frame is crucial here – the less confidence in the future the more personal the interpretation of responsibilities; the more confidence, the more abstract and formal the basis of one’s conduct can be – where that comes in part to influence the extent to which authority is seen as personal in form rather than public orientated. (so time frames are a component in the way people construct/interpret office). Similarly, space and location matter a lot, with different spaces being structured by different expectations and traditions. Consider for example, the way that authority in English factories in the Far East were affected by the other authorities exercised in those locations.
3. In the public domain and serve public ends – western bureaucrats – whose office is constructed within a structure of bureaucratic offices with designated public ends.
4. And In the public domain and serve private ends – here there seems to a double feature – depending on whether the serving of the private interest has a public interest justification. The tax official may be delegated to investigate specific citizens’ tax returns – so that his/her activities are not indiscriminate. At the same time how he gets those is more indiscriminate; and the functions performed by investigating those have a wider public significance. The bodyguard to the PM serves a particular interest – but does so in the wider public interest. A tax farmer looks as if she holds a public office in the public domain, but the ends she serves are complex – partly private, but legitimate because seen as functional to public ends (or to the private ends of managing the king’s estate). And, as in 2 above the significance of private ends rather than public ones may well be sensitive to time-horizons.

Four further complications arise:
1. From the separation of politics and public office – and the mixed messages the former gives – it this a formal office, is it wholly partisan? – it is clear that MPs are not bound by standard rules and norms of public office. But exactly what their status and responsibilities are is deeply contested.
2. From a historical tradition in which powers and responsibilities are delegated from above, but that then evolve in the direction of the emergence of a language of ‘the public’ as the intended beneficiary of the office and who are assigned ultimate sovereignty over the assessment and allocation of office (that is shifting from one form of legitimation that appeals upward, to another that appeals downward).
3. From the emergence of a discourse of public service in which the administrative order (and judiciary) sees itself as the prime point for the assessment of the public interest and thereby of the responsibilities of those in public office (so that the interests of those in office themselves, qua office holders, may lead to conflict with both the Principal and the Domain.
4. And from a discourse of political legitimacy that asserts the primacy of political control and political authority in the determination of the interests and goods that should be served by the public service. (where much depends on the practices of and claim about legitimacy that operate in the public domain).
From what I have said about each of these categories, it is clear that the relationships between the different components can be very complex. And that the different elements can be in tension with each other. This suggests that different traditions will give rise to different conceptions of office in which the different components are differentially negotiated, and in which different factors will be identified as salient in the filling of public office – for example, the differences between states in their requirements for and training of senior public officials – with the UK picking gifted amateurs, Germany Lawyers, France, technocratic graduates from the Grande Ecoles, etc.
In comparing patrimonial prebendal decentralization of the administration with rationalization and with institutions such as judicial independence, Weber draws a distinction between constructions of office in which the aim is the independence of office as designed to achieve and protect impartiality in the exercise of office, and cases where the aim is to protect the official’s right to his/her office. But we might also multiply these dimensions and examine the extent to which there are components of each in any particular construction of office, for example:
Protection against partial influence – from political pressure; from popular pressure; from sectional interests; from institutional and administrative interests; from personal interests of the office holder.
Protection of the right to office, from political usurpation, from sectional interests, from bureaucratic infighting.
Protection of the sphere of delegated judgment, from political influence, from competing institutional interests, from sectional interests among the regulated domain etc.
Which is given priority depends on a mix of path dependency, perceived threats, institutional sensitivities and aspirations, etc. And, of course, such issues will be contested between different interested agents, so that the outcome will depend heavily on the particular forces and alliances that are dominant when the system is progressively institutionalised.

There is no archetypal public office: there are a series of dynamics which encourage the formation of types of delegated or authorised powers, whose exact form is historically contingent, but where we can recognise a series of tensions that have to be negotiated. How they are negotiated and contested will have an impact on the forms of public office that emerge, the kinds of requirements they make of those in office, and the way in which relationships between office, its principles, and those under its domain are constructed.
If that is true, then there is no one right way of establishing and structuring public office. But there are going to be better and worse ones. Some favour the interests of the person in authority or of the agent more than the intended beneficiary of the office – for example the ordinary member of the public. Some give more weight to the intended beneficiary than is compatible with establishing a level of professional competence and trust between the principal and the agent on the one hand, and the beneficiary on the other. Asking whether the relationship is working – and working for whom on which dimensions, is always a good starting point. So we can begin to build a more normative account of ideas of public office on the basis of giving greater weight to certain dimensions of the relationships involved than to others. But, such considerations need to be sensitive to location, traditions, time frames, the complex expectations map within which they operate, and considerations of practicality and feasibility of objectives - not so as to be overwhelmed by them, but so as to develop both a sensitivity to local understandings and to feasibility constraints in the formation of a sense of what kinds of assessments and expectations are appropriate for this context.