Corruption is considered one of the biggest threats to government trust across the world. However, there is growing awareness of the limitations of addressing corruption. These limitations are conceptual: overcoming corruption itself only entails meeting the very minimum standards of public office (Heywood and Rose Reference Heywood, Rose, Hardi, Heywood and Torsello2015, 102). They are also empirical: fighting corruption has often been ineffective, sometimes counterproductive, and is generally insufficient to restore trust in government institutions (Rothstein and Tannenberg Reference Rothstein and Tannenberg2015, 9).
This has led to calls for a more praiseworthy governance goal: one that not only addresses corruption but entails institutions truly worthy of trust. There has been a surge of interest in defining a concept of ‘public integrity’ to play this role, led by researchers (Heywood et al. Reference Heywood2018; Huberts Reference Huberts2018; Menzel Reference Menzel2015; Montefiore Reference Montefiore, Montefiore and Vines1999; Mungiu-Pippidi, Dadašov and Fazekas Reference Mungiu-Pippidi, Dadašov and Fazekas2015), international organizations (OECD 2018) and non-governmental organizations (Integrity Action 2018; Pope Reference Pope2000).
However, most authors have adopted what we might call an ‘officer-first’ approach to defining public integrity. They have sought to define public integrity primarily as a quality of individual public officers; and only derivatively, if at all, as a quality of public institutions themselves. This article argues that this approach is flawed. Analysing the current debate about public officer integrity, we discover a need to define a role-specific sense of ‘praiseworthy’ behavior for public officers. In turn, we can only define this role-specific sense by reference to a public officer's contribution to the overall moral ideal of her institution. Assuming that this ideal itself is a form of public integrity, it then follows that we must define such institutional integrity ‘first’, in order to then define a public officer's praiseworthy contribution to it second.
Substantively, I argue that ‘public institutional integrity’ is the robust disposition of an institution to pursue its purpose efficiently, within the constraints of legitimacy, consistent with its commitments. ‘Public officer integrity’ is the robust disposition to support the integrity of her institution, within the course of her duties, to the best of her abilities.
The article is structured as follows. First, I clarify the methodology employed and certain assumptions about institutional virtue. Secondly, I assess the current ‘officer-first’ strategy, its failures and the case for a new ‘institution-first’ strategy. I then execute this strategy by defining public institutional integrity and then public officer integrity. Finally, I note two implications for the cognate debate about institutional corruption.
Methodology
Defining a concept of ‘public integrity’ entails encompassing the logical, cognitive and affective associations of personal integrity, but in the public realm with public actors. It is an exercise in analogy. The method, therefore, is to identify a concept in the public realm with a similar set of qualities to those defining ‘integrity’ in the personal realm.
What are those qualities? The conceptual core of personal integrity is coherence and consistency (McFall Reference McFall1987; Monaghan Reference Monaghan2017; Montefiore Reference Montefiore, Montefiore and Vines1999; Williams Reference Williams, Smart and Williams1973). It involves consistency between one's normative beliefs and their coherence with actions and motivations (Calhoun Reference Calhoun1995; Davion Reference Davion and Card1991; Taylor Reference Taylor1981). However, this is insufficient for integrity (contra Bigelow and Pargetter Reference Bigelow and Pargetter2007). It requires some additional condition(s), but those condition(s) are disputed.
We do not need to resolve this dispute about personal integrity in order to begin assessing conceptions of public integrity. We only need to adopt the criteria philosophers have used to assess competing accounts of what those additional condition(s) for personal integrity might be. Just as competing plausible accounts of personal integrity have sought to best fit these criteria, so will any competing plausible accounts of public integrity.Footnote 1
There are at least four such criteria. First, the additional condition(s) for integrity must explain why, under normal circumstances, having or acting with integrity is morally permissible. Normally, it runs against the logic of integrity to say that someone has impermissibly high levels of integrity, or that they are under a duty to compromise their integrity for some other end.Footnote 2 Secondly, the additional condition(s) for integrity must explain why, under normal circumstances, having or acting with integrity is not merely permissible but is morally praiseworthy (Cox, LaCaze and Levine Reference Cox, LaCaze and Levine1999).. This entails that the beliefs, actions and/or motivations brought into coherence and consistency must be in some sense important (Calhoun Reference Calhoun1995; McFall Reference McFall1987); that the act of coherence must be an act of will for which the person is responsible; and that there must be some difficulty, or at least substantial risk of difficulty, in so cohering (Taylor Reference Taylor1981). Thirdly, the additional condition(s) must explain why integrity is a virtue: a robust disposition to act in a praiseworthy manner.Footnote 3 Finally, these additional condition(s) must explain why integrity serves as a rational basis for trust. A person who has integrity is trustworthy (Calhoun Reference Calhoun1995, 237; Heywood and Rose Reference Heywood, Rose, Hardi, Heywood and Torsello2015, 112; Philp Reference Philp2007, 152; Zimmerman Reference Zimmerman2009). ‘Trustworthiness’, I shall assume, means being reliable with respect to one's commitments (Hawley Reference Hawley2014).
In sum, just as philosophers have judged competing conceptions of personal integrity using these criteria (consistency, coherence, permissibility, praiseworthiness, virtue and trustworthiness), I shall take them, by analogy, as the criteria upon which to judge competing conceptions of public integrity.
The Possibility of Institutional Virtue
Given these requirements for public integrity, we must resolve one further preliminary question: are public institutions the kind of thing that can have ‘integrity? Can they possibly be coherent and consistent, praiseworthy, trustworthy, virtuous and act justifiably? Or is our search for a quality that brings all of these elements together a conceptual non-starter? Justifying the metaphysics of collective agency and virtue are beyond the scope of this article. Instead, I rely upon the following four assumptions.
First, I adopt a narrow definition of ‘institution’. Sometimes we use the term in a broad sense to include any complex of human action and norms that sustains itself over time (Harré Reference Harré1993). By contrast, I shall refer to institutions in the narrower ‘organizational’ sense, as the sub-set of such complexes with individuals playing defined roles, bound by internal rules, authority structures and decision-making procedures (Isaacs Reference Isaacs2011, 4).
Secondly, I assume that such institutions are capable of collective intention, purpose, action and disposition. I associate collective intention with collective policy apt for implementation (rather than premising intention on the existence of some ‘joint commitment’) (French Reference French1979; Isaacs Reference Isaacs2011, contra Fricker Reference Fricker, Gendler T and Hawthorne2010; Gilbert Reference Gilbert1992). So long as an institution's collective rules, authority structures and decision-making procedures are capable of making such policy, then it is capable of collective intention. An institution's purpose(s) is a policy that sets the end for which other policies and actions should aim. So long as an institution is capable of collectively acting on such intentions qua policies and purposes, then it is capable of action. Further, institutions can have ‘dispositions’ to act in certain ways, that is, given their internal composition they will tend to act in one way or another given relevant circumstances (Byerly and Byerly Reference Byerly and Byerly2016, 36).
Thirdly, I assume that since they are capable of collective intention and action, institutions are collective agents. This agency cannot necessarily be reduced to the agency of its members (Byerly and Byerly Reference Byerly and Byerly2016; French Reference French1979; Gilbert Reference Gilbert1992; Isaacs Reference Isaacs2011; List and Pettit Reference List and Pettit2011). The institution may F, but none of its members need F. As agents, the institution may have collective normative powers, such as certain rights and permissions. These are exercised by its internally empowered decision-making bodies, or by appropriately delegated members acting on its behalf. In this vein, it is able to make commitments. Thus, it can be trustworthy (Kirby, Kirton and Crean Reference Kirby, Kirton and Crean2018, contra Hawley Reference Hawley, Faulkner and Simpson2017). It is liable to fall under duties, and it can be held responsible for its (in)actions. Finally, its (in)actions may be more or less praiseworthy as a function of whether it exercises its agency, importantly, and with possible difficulty.
Finally, being capable of dispositions to act in more or less praiseworthy ways, institutions are capable of virtue (Byerly and Byerly Reference Byerly and Byerly2016; Fricker Reference Fricker, Gendler T and Hawthorne2010, contra Cordell Reference Cordell2016; Konzelmann Ziv Reference Konzelmann Ziv A2012). Following Byerly and Byerly (Reference Byerly and Byerly2016, 43), I hold that: ‘A collective C has a virtue V to the extent that C is disposed to behave in ways characteristic of V under appropriate circumstances’, where characteristic V is in some sense praiseworthy.
Strategy: Public Officer or Institution ‘First’?
As stated in the introduction, the implicit strategy used in the current literature is to define public officer integrity first – as the primary moral concept at play in our normative public ethics – and to worry about institutional integrity, if at all, second. The problem, however, is that current attempts to execute this strategy continue to fail. In this section, I shall diagnose these failures, and explain why they suggest the opposite strategy: defining institutional integrity independently, first, and public officer integrity, secondly, as a function of it.
The Officer-First Approach: A Critique
The current debate about public officer integrity is the story of two approaches: what I call the ‘minimum standards’ and ‘abstract ideal’ approaches. In order to explain these two approaches, however, we must start with a legal (not moral) distinction: between legal duties and discretionary powers.
In administrative law, public officers are bound by legal duties. Such duties oblige a public officer to act in certain ways under certain conditions. Others can demand that the public officer act in such a way by legal right, and are legally wronged if they do not. Thus, a failure to perform one's duty is generally liable to judicial review and sanction. Appropriately, therefore, such duties must be relatively ‘black and white’ (Cane Reference Cane1996, 32). They must satisfy requirements such as generality, clarity and publicity so that public officers can be reasonably expected to know, prospectively, how they are obliged to act. Of course, many decisions that public officers must make are not black and white: they are context dependent, interpretative, ambiguous, controversial, new and unexpected, and ultimately require judgment. Typically, therefore, the law does not (and cannot) regulate such decisions through legal duties. Instead, they fall within the public officer's domain of ‘discretionary power’. Unlike a legal duty, a discretionary power legally permits choice between options. Thus no one else can demand that the officer choose in any particular way by legal right and cannot be legally wronged by any particular choice. Yet, just because such a power is not regulated by a legal duty does not mean that its exercise cannot be assessed according to other standards. A discretionary power may still be exercised, correctly or incorrectly, well or poorly, optimally or otherwise. The standards underpinning these assessments, however, are likely to be more abstract ideals, the application of which is liable to require as much judgment as the exercise of discretionary power itself. For this reason, something like judicial review is inappropriate for assessing the exercise of discretionary power. But forms of ‘administrative’ (or in some cases political) accountability are appropriate, such as performance review, public critique, recommendation, promotion and non-renewal of contract.
With this legal distinction in place, we can characterize the first approach to defining public integrity: the ‘minimum standards approach.’ This first approach is to define public integrity by reference to certain minimum standards of morally permissible behaviour that could (and, generally, should) be legislated as legal duties of public office. In this way, others can demand public integrity, as a moral value, by legal right from public officers, and public officers can be held to account, judicially (or quasi-judicially) for failures of public integrity. In this vein, some authors – at least by implication – identify public integrity with the most minimum of standards, that is, the absence of corruption (Anechiarico and Jacobs Reference Anechiarico and Jacobs1996, Rose Reference Rose2014, 151). Others, looking for a higher minimum standard, add impartiality (Mungiu-Pippidi, Dadašov and Fazekas Reference Mungiu-Pippidi, Dadašov and Fazekas2015, 8). Others add transparency qua ‘not actively attempting to conceal information’ (Heywood and Rose Reference Heywood, Rose, Hardi, Heywood and Torsello2015). Still others, issue a concrete list of ‘integrity violations’ to be avoided that include a range of public (and private) forms of misconduct (Lasthuizen, Huberts and Heres Reference Lasthuizen, Huberts and Heres2011). And, official codes of conduct tend to emphasize legality and obedience (Rothstein and Sorak Reference Rothstein and Sorak2017).
Regardless of the specific standards, however, three generic problems arise for this approach. First, there is often little justification offered for the particular minimum standards listed. Authors merely appeal to intuition in a rather ad hoc vein. There is certainly no attempt to draw upon a deeper set of principles that explain why these standards are coherent, exclusive and exhaustive of the requirements of public integrity.
Secondly, each standard tends to be vulnerable to counterexamples. For example, surely, avoiding corruption cannot be a necessary condition of integrity because some corruption can be done permissibly for a noble cause? Surely, transparency cannot be a necessary condition of integrity because intelligence services permissibly keep operations secret? Surely, legality cannot be a necessary condition of integrity because sometimes it is permissible to break the law in order to leak gross government misconduct to the press? Without a strong theoretical justification, authors are unable to push back on such counterexamples, reinterpret their standards in light of them or justify the limits of their standards so as to characterize them as principled exceptions.
Finally, compliance with such minimum standards of public office is insufficient for praiseworthiness. While one might think that this follows by definition, since complying with the minimum standards can never be praiseworthy, this is not the case. Depending on the context, fulfilling such standards may be deeply important, difficult and take great acts of will. For example, in a systemically corrupt police force, older officers may place coercive pressure on a new recruit to engage in corruption to secure her loyalty and minimize the chance of her reporting their behaviour. Such pressure might be social or physical, or connected to their prospects of advancement. In light of such threats, it is praiseworthy to rebuff corruption, while at the same time it remains a moral (and presumably legal) duty to do so.Footnote 4
However, mere compliance with the minimum standards of public office is still insufficient for praiseworthiness. This is because one can still be unpraiseworthy using the discretion that lies within those minimum standards. Our police officer, for example, might be praiseworthy in the particular act of rebuffing corruption, but unpraiseworthy in the way in which she carries out other duties of office: she might be lazy, unhelpful, obstructive, mendacious, uncooperative or unreliable without contravening the stated minimum standards. So long as public officers have such discretion they can be (un)praiseworthy within it.
Of course, one possible response is to identify public integrity with the elimination of discretion: to try and define minimum moral standards capable of being turned into legal duties, so to speak, ‘all the way up’ (Huberts Reference Huberts2014; Mungiu-Pippidi, Dadašov and Fazekas Reference Mungiu-Pippidi, Dadašov and Fazekas2015). However, this is both implausible and undesirable. It is implausible to have public rules that prospectively determine what to do, clearly, in every possible scenario for most public officers. Further, prior research has demonstrated the negative impacts that such rule- and compliance-based cultures can have on an institution's efficiency and effectiveness (Mayer Reference Mayer2013, 60). It would seem counterintuitive, therefore, that ideal public officer integrity depends on both an impossible and undesirable lack of discretion (Philp Reference Philp2007).
On this basis, many authors conclude that we cannot define public officer integrity merely by reference to moral duties that are all sufficiently general, clear and public enough to be legally enshrined as minimum standards of office, and admitting of judicial review. Instead, unavoidably, public integrity must be defined – at least in part – by reference to some abstract ideal(s). Such ideals may be telic in nature (goals to be promoted and balanced), but they may also include moral duties. However, these duties will be too abstract to be justiciable minimum standards of office. For example, a ‘duty to be ethical’ is still a moral duty, but it is an implausible legal duty. It might plausibly form the basis of administrative or political, but not judicial, accountability.
Searching for such a comprehensive abstract ideal, some authors define public integrity as ‘staying true’ to one's own deep normative beliefs, in both the private and public spheres (Montefiore Reference Montefiore, Montefiore and Vines1999). Others note that this view would implausibly credit any internally consistent but evil public officer with ‘public integrity’ (Heywood et al. Reference Heywood2018, Lasthuizen, Huberts, and Heres Reference Lasthuizen, Huberts and Heres2011, n3). These others turn to so-called objective values to secure praiseworthiness: integrity is always ‘doing the right thing, in the right way’ (Heywood and Rose Reference Heywood, Rose, Hardi, Heywood and Torsello2015, 112); it is ‘a characteristic or quality that refers to accordance with the relevant moral values and norms’ (Huberts Reference Huberts2014, 4); ‘[p]ublic integrity refers to the consistent alignment of, and adherence to, shared ethical values, principles and norms for upholding and prioritising the public interest over private interests in the public sector’ (OECD 2018); ‘simply put, “integrity” means having a genuine, wholehearted disposition to do the right and just thing in all circumstances, and to shape one's actions accordingly’ (Fleishman Reference Fleishman, Fleishman, Liebman and Moore1981, 53); or public integrity means just simply ‘a general way of acting morally’ (Brenkert Reference Brenkert2004, 5).
The important thing to note, however, is that all the relevant abstract ideals cited (acting morally/ethically/justly/rightly/in the public interest), at least without further clarification, are not role specific. By this I mean that one could plausibly recommend that anyone, public officer or not, ‘do the right thing in the right way’, ‘act morally’ and in ‘accordance with relevant moral values’. Even the distinction between ‘private’ and ‘public interest’ is salutary in ordinary life: I should prioritize the public over my private interests by decreasing my personal carbon emissions. Given the generality of such moral concepts, therefore, two problems follow.
The first problem is that while always acting ‘morally/ethically/justly/rightly/ in the public interest’ might be plausibly praiseworthy in one's personal life, it is not always morally permissible in public office. For example, a public officer may disagree with her minister as to what the ‘public interest’ is in relation to a discretionary tax exemption that the public officer must administer in accordance with the minister's directions. The public officer may think it will increase inequality, and that this is unjust. The minister might agree that the exemption will increase inequality. But she might reasonably believe, contrary to the public officer, that such inequality is just. Let us assume that the public officer is correct. This would still be an insufficient basis for her to disobey the valid directions of her minister in order to promote the true public interest. Pursuing the true public interest, even if otherwise praiseworthy, might still be impermissible.
One response to this first problem is to reintroduce the minimum standards, set out by the first approach, as side constraints to acting praiseworthily in the public interest (for example, Heywood and Rose Reference Heywood, Rose, Hardi, Heywood and Torsello2015). According to this view, an officer with public integrity should pursue the public interest, unless it conflicts with the minimum standards of office, which ‘trump’ the public interest: the duty to obey one's superior, for example. A dialectical cost of this strategy, however, is that such a combined view inherits the first two problems of the minimum standards view. It still owes a background theory that justifies such minimum standards, and their exceptions, and now, further, also justifies their priority over other values like the public interest. However, even assuming that such a theory can be provided, a second, fatal problem arises. Pursuing a general abstract moral ideal like ‘acting in the public interest’, even if only conducted within one's discretionary power, is not necessarily praiseworthy in public office.
Let us return to our tax officer to illustrate. Let us assume that she concedes our point: while she is correct about what constitutes the public interest in the case, disobeying her minister breaches a minimum standard. However, within the bounds of her legal discretion she can still frustrate the minister's intention by ‘slow-walking’ the policy. She can interpret it, plausibly, but inconsistently with another department commitment, inviting a public rebuke, and hopefully a policy change. She can allocate resources away from its implementation. She might see a mistake being made by a colleague that will undermine the policy's effectiveness and decide not to act proactively to resolve it. Now, ex hypothesi, since the relevant tax policy is against the public interest, ceteris paribus each of these actions will promote the public interest. Furthermore, ex hypothesi these are within the minimum standards. However, intuitively, no matter how praiseworthy it might be for others outside the institution to frustrate the policy by whatever legally permissible means, it is not praiseworthy for this particular public officer to do so. It is inconsistent with her role as a member of her institution. She is not being praiseworthy in her particular role as a particular type of public officer in a particular institution, even though she might be promoting the public interest in a legally permissible fashion. Praiseworthiness, and thus public integrity, must in some sense be specific to her role. It cannot be defined by a general abstract ideal – that is, something that anyone, anywhere might plausibly have reason to pursue. How can we define such a role-specific ideal? This is the key problem that needs to be resolved in order to plausibly define public officer integrity. In order to do so, we need to undertake a new institution-first strategy.
The Institution-First Approach: A New Strategy
Public officers, collectively, embody public institutions. By ‘embody,' I mean nothing particularly metaphysical. I simply mean that what a public institution is and what it does at any point in time – its intentions, dispositions, purposes, virtues, character and actions (all the concepts we defined above) – are a function of how its public officers qua constituent members, themselves are and what they do, at any point in time. This function is rarely total or simple. The former is not necessarily reducible to the latter. But regardless, this partial, complex function holds as much for Premiers, Prime Ministers and Presidents, as it holds for ‘street-level bureaucrats.' Public officers always have ‘two bodies,’ so to speak: their own, and the public institution that they each, partially, represent.
This embodiment relation entails that we cannot hope to define what we should ideally want from public officers in their roles as parts of their institution until we first define what we should ideally want from their institution as a whole. Just as we cannot comprehensively define what we want from individual football players until we define what we want from their team as a whole. The collective ideal must be defined first, the individual member's ideal role-specific contribution to that collective ideal defined second.
So what is the collective ideal for public institutions? I shall assume that it is the same as at the individual level: public integrity, whatever that might mean upon further investigation. This is to say, intuitively, we should want institutions that are coherent, consistent, act permissibly, are trustworthy, praiseworthy and virtuous. Justifying this claim in full is outside the scope of this article. However, my assumption is that if public integrity is an ideal for public officers qua public agents, then it will also be an ideal for public institutions qua public agents. We rightly value public integrity in public officers because they are morally responsible public agents, not because they are individual persons. As such, since institutions are also morally responsible public agents (although they are collectives), ceteris paribus we should have the same reason to value public integrity in them. Upon further investigation it may turn out that, mutatis mutandis, this assumption does not hold. However, setting out a conception of institutional integrity, and an intuitively satisfying relationship with public officer integrity in this article, is a sensible preliminary to that debate.
Thus, assuming in this article at least, that the moral ideal of public institutions is a form of ‘public institutional integrity’, then we must resolve the debate regarding its definition first, in order to then second justify and give concrete content to the praiseworthy contribution that individual public officers can make to this collective ideal. Let us execute this strategy.
Public Institutional Integrity
A number of philosophers refer to institutional integrity (Ceva Reference Ceva2018, 119; Ferretti Reference Ferretti2018; Philp Reference Philp2007, 229ff; Thompson Reference Thompson2013, 19; Thompson Reference Thompson2018, 505; Waldron Reference Waldron2016, 35, 36, 65ff), but to my knowledge only three offer a conception. Gillian Brock (Reference Brock and Safra2014, 5–6) defines ‘institutional integrity’ as when an institution (1) achieves its purposes effectively and equitably, (2) does not improperly depend on or promote the interests of other parties, and hence (3) public confidence in it is appropriate and (4) public confidence in its practices, operations, and policies can survive appropriate transparency and accountability tests. Daniel Wueste (Reference Wueste2005, 21) holds that institutional integrity is maintained by ‘the effective realization of an institutional purpose’ and compliance ‘with the normative constraints intrinsic to [that] enterprise’. Allen Buchanan (along with Robert Keohane), do not offer a definition as such, but they do explicitly hold that a lack of ‘institutional integrity’ entails a lack of ‘legitimacy’, arising from ‘a pattern of egregious disparity between [the institution's] actual performance, on the one hand, and its self-proclaimed procedures or major goals’, and that ‘if an institutions fails to satisfy the integrity criterion, we have reason to believe that the key institutional agents are either untrustworthy or grossly incompetent, that the institution lacks correctives for these deficiencies, and that the institution is therefore unlikely to be effective in providing the goods that would give it a claim to our support’ (Buchanan and Keohane Reference Buchanan and Keohane2006, 422–4).
In other disciplines, sociologist Philip Selznick argues that institutional integrity is the structuring of an institution's elements – including its members’ attitudes, decisions and forms – around its ‘distinctive competence’. This distinctive competence gives the institution its identity, and its integrity is a ‘value in itself’. As such, ‘the protection of integrity is more than an aesthetic or expressive exercise, more than an attempt to preserve a comforting, familiar environment’ (Selznick Reference Selznick1983, 138–9; see also Terry Reference Terry1995). And, in public administration, Breakey, Cadman and Sampford (Reference Breakey, Cadman, Sampford, Schwarts, Harris and Comer2015) hold that institutional integrity involves ‘an institution asking hard questions about its value and values, giving honest and public answers, and living by them’. This includes its raison d'etre. These answers have to satisfy what they call a ‘public institutional justification’ – that is, a justification ‘wholly chosen by the institution-members/representatives’ to ‘show the public it deserves their support or at least tolerance’; it must reflect ‘community values’ (Breakey, Cadman, and Sampford Reference Breakey, Cadman, Sampford, Schwarts, Harris and Comer2015).
I define public institutional integrity as the robust disposition of a public institution to pursue its purpose efficiently, within the constraints of legitimacy, consistent with its commitments. I shall address each aspect of this definition, identifying its analogical equivalence with personal integrity, and justifying it, where relevant, in contradistinction to the other conceptions.
Purpose
Each definition of institutional integrity above assumes, as necessary, some overall institutional purpose (or major goals, distinctive competence, raison d'etre). However, having a purpose is not a necessary feature of personal integrity. Some paradigmatic integrous individuals are indeed single-mindedly purposive (for example, the committed artist). But, conversely, an integrous individual might lack an overall purpose, doing what pleases her from time to time, but having consistent moral commitments about how she lives and treats others in the process.
Why is purpose, then, a necessary condition of institutional but not personal integrity? Only Selznick, above, offers any justification for the centrality of purpose to institutional integrity. According to his view, an institution's particular purpose (or ‘distinctive competence’) is essential to its identity. Hence, assuming that institutional integrity involves some sense of being ‘true to oneself’, then institutional integrity must involve some kind of faithfulness to purpose.
The problem with this view is twofold. First, it entails an essentialist logic. If an institution is defined by a particular purpose, then a ‘true purpose’ distinct from other possible ‘false purposes’ must be attributable to it. Hence, we can never have a clear idea of an institution's integrity unless we have a clear and authoritative account of its true purpose. But most institutions’ purposes are always to some extent vague, ambiguous and controversial among their members and other stakeholders. As such, their integrity will be equally vague, ambiguous and controversial. The second problem is that such a view struggles to explain how an institution can change its purpose over time while maintaining its integrity, which institutions can surely do. But if integrity for an institution involves a commitment to a particular purpose, then it must be a failure of integrity to drop that purpose in favour of another purpose.
A better argument for the necessity of institutional purpose is as follows. Institutions, unlike people, are constituted by multiple different agents, applying many different rules, exercising different powers, sharing common resources, and partaking in collective responsibility for the actions and future of the institution as a whole. Within this structure, each agent is constrained by rules and localized purposes. However, such rules and purposes always need to be interpreted, and interpretations are liable to conflict. Further, the activities of agents are liable to conflict with respect to resources and collective responsibilities.
A relatively clear common purpose offers the only robust means of avoiding and/or resolving these conflicts in a way that will make activities across the whole institution consistent and coherent. It is no coincidence that organizational psychologists define the difference between a mere group and a ‘team’ to be the sharing of a common goal or purpose (Kozlowski and Bell Reference Kozlowski, Bell, Borman, Ilgen and Klimoski2003). It is, of course, theoretically possible that all activity within an institution can be disposed to remain consistent and coherent without a common purpose. However, it is impossible for such a non-purposive disposition to be robust, and a robust disposition is necessary for the relevant quality to ground a virtue. Finally, note that the purpose need only be clear enough to ensure a robust disposition towards coherence and consistency. This requirement is consistent with a level of vagueness, ambiguity and incompleteness that is inevitable in any real-world interpretative instrument such as an institutional purpose. It also permits integrity to survive a change of purpose, so long as that change is consistent and coherent with the institution's procedures for setting its purposes, and any other relevant conditions for integrity (for example, keeping commitments to other stakeholders). In this way, having a relatively clear, institutional purpose is a necessary condition of public institutional integrity.
Legitimacy
‘Legitimacy’, as I shall define it, turns upon an agent having the ‘right to rule’ over others in some domain. By this I mean that, regardless of pre-existing rights and duties that those other agents have, the ‘ruling agent’ has the power to change some of those rights and duties in that domain because it says so (Raz Reference Raz2009, Ch 5). Depending upon one's substantive theory of legitimacy, an agent will only be legitimate because it satisfies certain moral preconditions (for example, democratic election, delegation, being the best available means to serve the interests of the governed). Further, such a right will be restricted in scope by other preconditions to a certain domain (for example, consistent with the constitution, certain basic human rights, directions of its higher ruling agent). An institution has a ‘legitimate purpose’ if its purpose is consistent with those preconditions. An institution pursues that purpose ‘legitimately’ if it acts intra vires, that is, within the scope of its legitimate powers.
Some public institutions do not obviously have any ‘right to rule’ over anyone, or if they do then they do not spend all their time creating new rules. A public institution (say a library) might simply manage a public resource: pursuing a purpose (promoting public access to books) by exercising certain rights to use and expend that resource (a budget, land and property) within a certain domain of discretion. However, such rights of institutions like libraries are invariably sourced in the prior exercise of some ‘right to rule’ that is either held by itself or another, higher agent. Thus the library qua ‘public institution’ only has its rights as a function of the exercise of the right to rule by the local authority (creating duties for money to be handed over, and its purposes respected, and so forth). Thus, the library or any other such institution can be said to be legitimate insofar as it continues to satisfy the conditions of having those rights. And its purposes and actions can be said to have legitimacy insofar as they fall intra vires, that is, within the domain of power granted by such rights.
The right to rule to which I am referring is a moral (or de jure) right as distinct from the mere de facto or legal power to direct others (Raz Reference Raz2009, 128–9). According to most substantive theories of legitimacy, these three concepts will be closely related, but they will sometimes come apart. For example, a law might permit torture, but arguably both the law – and thus any action purportedly in accordance with it – would be illegitimate. However, what is perhaps most important is that according to almost all theories of legitimacy, the right to rule is rarely a simple function of what is otherwise just, equitable, ethical or moral. In fact, typically, a central aim of such theories is to explain why an agent might have the right to direct us to do things that we might consider unjust, inquitable, unethical or immoral; and we might even be correct; yet we are still obliged to obey. Extremes of injustice, for example, may define the limits of that power (arguably the case in the example of torture), but there is always some domain in which our normal moral reasons are liable to be overridden.
By including legitimacy as a condition of public institutional integrity we ensure that it satisfies the permissibility condition for integrity, and it does so by definition. Since something done by right is permitted, then a legitimate purpose is a permitted purpose, and a legitimate mode of pursuit is a permitted mode of pursuit. Conversely, if a purpose or mode of pursuit is not legitimate, then it is not permitted.
In this way, the norm of legitimacy maps the contours of permissible institutional behaviour in a more accurate fashion than other suggestions. It is an obvious advance on Selznick, who has no condition that could plausibly ensure permissibility in the first place. The same holds for Buchanan and Keohane, since they only see legitimacy as the result of integrity, not as a requirement. Wueste does refer to operating within the ‘normative constraints intrinsic to the enterprise’. However, this simply names rather than helps us identify those constraints. Arguably, Breakey, Cadman and Sampford's idea of public institutional justification is best interpreted as a substantive conception of legitimacy: in short, so long as the public agree or tolerate the justification put forward by the institution, then it is legitimate. As a theory of legitimacy, it is open to a range of critiques (for example, what if the majority tolerate grievously unjust, unconstitutional institutions that oppress a minority?). However, regardless, it is more accurate to define institutional integrity by reference to whatever substantive theory of legitimacy is true, not any particular controversial theory like their own. Finally, ‘legitimacy’ possibly gives content to Brock's reference to ‘(im)proper’ dependence. But unlike Brock's definition, it ensures that the institution's purpose is permissible. And finally, for the reasons discussed above, it also provides a reason to dismiss her claim that an institution of integrity must achieve its purposes ‘equitably’.
Consistency with Commitments
Brock, as well as Buchanan and Keohane, rightly associate institutional integrity with being worthy of (or ‘making appropriate’) trust (or ‘public confidence’). However, they both identify institutional trustworthiness only with the reliable fulfilment of institutional purpose and the reliable fulfilment of some cognate of legitimacy. Assuming a ‘commitment view’ of trustworthiness, however, these conceptions are too narrow.
It is true that institutions make commitments, explicit or implicit, to pursue their purpose(s) and operate within the constraints of legitimacy. However, they also make a range of other commitments to particular stakeholders. They make representations and promises to the community, or particular communities (to build a road, deliver a new hospital, keep citizens' data secure). They enter into contracts and institute policies with respect to individual members qua employees (to pay them on time and in full, protect against workplace harassment, provide training). They enter into a range of contracts, agreements and fiduciary relationships (with businesses, other public institutions and even foreign entities). In order to be trustworthy, an institution needs to have a robust disposition to meet all such commitments.
A difficulty with this condition is that sometimes fulfilling a commitment to another party may conflict with a public institution's commitment to pursue its purpose or to act within the constraints of legitimacy. For example, an institution may commit to deliver certain public services that later turn out to be either not the best way to achieve its purpose or beyond its legitimate powers. The immediate implication of such conflicts is that they compromise the integrity of the institution. The institution is not internally consistent with respect to its purpose, legitimacy and commitments; nor can it cohere to all of them.
An institution with high levels of integrity simply avoids, or at least minimizes, such conflicts. However, what if such inconsistency is not avoided? With respect to a conflict between its other commitments and the constraints of legitimacy, an institution must adhere to the latter. This is because compliance with such other commitments would be ultra vires. The institution does not have permission to fulfil commitments in ways that exceed that power. By contrast, with respect to a conflict between its commitments to other parties and its legitimate purpose, an institution should adhere to the former.Footnote 5 This is implicit within such commitments. Other actors can reasonably expect the institution to be responsible for ensuring that its own commitments are consistent with its own purpose. Such information, power and accountability will most often be outside the reach of the other actors. And the institution cannot reasonably expect these other actors to bear the cost of its own failure to make commitments consistent with its purpose. Further, such costs can generally be far more easily, and justly, borne by the former than the latter.
Efficient Pursuit
The efficient pursuit of purpose, within the constraints of such legitimacy and consistent with commitments, is necessary for praiseworthiness. Brock and Wueste make the subtle but significant mistake of identifying institutional integrity with the achievement of purpose. Praiseworthiness (and integrity) require responsibility, and institutions are not necessarily responsible for failing in their purposes. The achievement of some purposes may be impossible (to cure a disease that is ultimately found to be incurable), or unlikely (to defend the nation valiantly against an overwhelming enemy), or never ending (to secure the proper administration of law), or frustrated by exogenous shocks or externals constraints such as being under-resourced.
Instead, an institution is responsible for pursuing its purpose. This demands intention as demonstrated in its policies and action. Pursuit by itself, however, is not sufficient for praiseworthiness. This is because one can still pursue one's purpose poorly, or well but easily. In order to ensure praiseworthiness, an institution must pursue its purpose efficiently. Ceteris paribus, an administrative task is performed better expending fewer resources than more. Doing so is more difficult, and importantly, those resources can be used for other public purposes. Thus, ‘efficient pursuit’ secures the responsibility, importance and difficulty necessary to ensure institutional praiseworthiness as a condition of institutional integrity.
Robustness
None of the definitions of institutional integrity mentioned above include ‘robustness’. However, the mere disposition under current circumstances to pursue purpose efficiently, within the constraints of legitimacy and consistent with commitments is not sufficient for an institution to have integrity. This disposition must also be robust. Otherwise, it does not constitute a true virtue, a trait that in some sense ‘goes deep’ to the character of the institution, a trait that others can rely upon the institution to have, come thick or thin.
In some sense, the absence of robustness from current definitions of institutional integrity is ironic, since so much of what might broadly be called the integrity ‘agenda’ focuses on mechanisms designed to secure robustness: public accountability (Integrity Action 2018), transparency (Brock Reference Brock and Safra2014, 202; Philp Reference Philp2007, 217, 229), limiting forms of discretion (Mungiu-Pippidi, Dadašov and Fazekas Reference Mungiu-Pippidi, Dadašov and Fazekas2015), institutional competition, and so on. However, none of these more particular mechanisms should, in themselves, be considered necessary elements of institutional integrity. This is because none of them is unconditionally important for robustness (Philp Reference Philp2007; Rothstein and Teorell Reference Rothstein and Teorell2008, 202). Promoting robustness is (generally) the relevant aspect of integrity against which they should be measured. For example, processes of accountability – to citizens, other institutions and higher authorities – may be generally integrity promoting. However, sometimes too much (or certain processes of) accountability dispose the institution to red tape, risk aversion and an environment in which no one publicly says what they really think. Such an excess of accountability does not further institutional robustness, and thus ideally should be eliminated for the sake of institutional integrity. It is only justified insofar as it promotes overall institutional integrity.
With our final condition of ‘robustness’ in place, we have our definition of a public institution's integrity: the robust disposition to pursue its purpose efficiently, within the constraints of legitimacy, consistent with its commitments. I now return to public officer integrity to complete the institution-first strategy.
Public Officer Integrity
Our original task was to capture the role-specific sense in which public officers with integrity are praiseworthy. To this end, I now argue that public officer integrity should be understood as: the robust disposition of a public officer to support the integrity of her institution, within the course of her duties, to the best of her abilities. Let me first illustrate the meaning of such a disposition, and then justify this claim.
A public officer can support an institution's integrity in four main ways. First, assuming the legitimacy of the institution's purpose(s), actions, and internal and external commitments, a public officer generally supports the integrity of her institution by performing her everyday responsibilities, pursuing her immediate purpose(s) and complying with the institutional rules that apply to her.
Secondly, insofar as the public officer has discretion, either within those rules, outside her immediate purposes or when interpreting both, she should act in ways that support the integrity of the institution. She should interpret and pursue her own immediate purpose(s) in the way that best promotes the overall purpose(s) of the institution, not merely the interests of her sub-section; she should not merely ‘stay in her lane’ but also reinforce the work of others; she should not merely avoid pathologies, like corruption, herself, but also actively seek to eradicate the risk of them occurring elsewhere in the institution.
Thirdly, if and when the institution falls into, or there is a risk of, inconsistency or incoherence between purpose, procedure and commitments, the public officer should proactively take appropriate responsibility to ameliorate such conflicts.
Finally, and in the most difficult of scenarios, the public officer must ameliorate institutional illegitimacy, whether through internal protest, public statements, lobbying for accountability and intervention by other appropriate actors, resignation, whistleblowing or even subversion.
This conception of integrity is closer to the abstract ideal approach to defining public officer integrity than the minimum standards approach. However, the relevant ideal – the integrity of one's own institution – is not general, but role specific. It is not an ideal that many other people could have (like being ethical). Instead, one can only have it because one is a member of a public institution. Its content is specific to that institution. Furthermore, it is praiseworthy for a public officer qua member of this institution to support this ideal to the best of her abilities. It requires willing choices by public officers for which they can be held morally responsible. It is important insofar as the integrity of the institution as a whole is important. Furthermore, it is difficult to maintain to the best of one's abilities, constantly in everything one does, across all kinds of pressures. It requires an ‘ethic of responsibility’ not just for one's narrow set of defined duties, but for the behaviour of others, the rules, culture and norms that define institutional culture, and the constant risks to institutional integrity, and the opportunities to improve it.
To illustrate, let us return to the example above of the public officer who disagrees with her minister about whether a mandated policy serves the public interest. We can now explain why her actions, despite being within the minimum standards and in the ‘public interest’, would be contrary to public officer integrity. First, attempts to frustrate the policy, like ‘slow-walking’ it and allocating resources away from its implementation, all manifest a diversion from institutional purpose, assuming the purpose of her institution includes the implementation of government tax policy. Further, by interpreting the policy in a way that leads to overall inconsistency in the institution's commitments, she undermines its trustworthiness. Finally, by letting a colleague outside of her own mandated area of responsibility make a significant mistake that she could have prevented by intervening, she displays a lack of responsibility for the institution as a whole.
In this way, as a role-specific but abstract ideal, supporting the integrity of one's institution to the best one's abilities provides a standard by which to assess the exercise of one's discretion within public office. However, it also provides a way of justifying the typical ‘minimum standards’ of public office that might circumscribe that discretion when enshrined as legal duties: bringing them into theoretical coherence and explaining principled exceptions. Generally, compliance with such minimum standards by public officers will be required to support the legitimacy element of institutional integrity. It is illegitimate for an institution to breach such minimum standards; thus it is illegitimate for a public officer (as its embodiment) to do so on its behalf. However, the norm of legitimacy also explains exceptions. For example, it is arguably legitimate to break the law and disobey orders when doing so is necessary to address gross illegitimacy elsewhere in the organization (for example, by whistleblowing).
In this way, our conception of public integrity defines a role-specific sense in which public officers can be praiseworthy in their use of legal discretion, but also justifies the limits of permissible action in a manner that can generally be enshrined in legal duties, but with due recognition and justification for exceptions.
Corruption: An Opposite of Institutional Integrity
While the literature on institutional integrity remains underdeveloped, its apparent opposite – institutional corruption – has gathered attention. Before concluding, it is worth considering how our conceptions of public integrity – both institutional and public officer – interact with and further this concurrent debate in two ways.
First, our conception of institutional integrity situates ‘institutional corruption’ as but one of many possible opposites. Take Lawrence Lessig's influential definition: ‘Institutional corruption is manifest when there is a systemic and strategic influence which is legal, or even currently ethical, that undermines the institution's effectiveness by diverting it from its purpose or weakening its ability to achieve its purpose, including, to the extent relevant to its purpose, weakening either the public's trust in that institution or the institution's inherent trustworthiness’ (Lessig Reference Lessig2013, 553).
This definition permits a range of institutional states that are not necessarily corrupt, but certainly not ideal either. Deploying our conception of institutional integrity, we can diagnose these ‘pathologies’. We might define ‘institutional confusion’ as when an institution loses clarity over purpose, rather than being strategically and systematically diverted from purpose, or ‘institutional mission drift’, when it becomes diverted from its purpose, but not necessarily for any private interest. We might define, as a limit case, ‘institutional cipher’, where an institution cannot be diverted from its purpose because it lacks any collective purpose at all. We might define ‘institutional inconsistency’ as a situation in which an institution's constitutive parts begin to act in inconsistent ways, ‘institutional duplicity’ as when an institution makes commitments it has no intention of keeping, ‘institutional exploitation’ when an institutional fails to live up to its commitments to its own staff, and ‘institutional capture’ as when the authority dictating the organization's purpose is corrupted and changes its purpose to serve private ends, and so on. Providing a full taxonomy of the pathologies of institutional integrity, and an account of how they interact, is a task for another article. The only point I would stress here is that ‘institutional corruption’ clearly does not cover the field, and that institutional integrity, as an opposite, ties all these pathologies together.
Secondly, our conception of public officer integrity usefully informs public officers’ responsibility to address institutional corruption. There is an anxiety that focusing on institutional corruption will absolve or obscure individuals’ responsibilities within institutions (Ferretti Reference Ferretti2018). In response, as Dennis Thompson notes, ‘most … institutionalists, [take] care to show that even when corruption is purely institutional, individuals can still be held responsible for trying to eliminate it, or at least moderate its harmful consequences’ (Thompson Reference Thompson2018, 498). This is true, but their care does not extend as far as describing the limits of such responsibility, which is surely not unlimited. Let us take the case of US campaign finance, favoured by Thompson and others. Members of Congress might recognize its current corruption, and they might even recognize their duty to ‘eliminate’ or ‘at least moderate its harmful consequences’. But how far does their duty extend? Should members of Congress divert all their own political capital from other important causes to resolve this one? Should they divert the legislative agenda from, say, impeaching a corrupt president to focus on campaign finance instead? Should such members be willing to lose their own seats in protest against laws? These are hard questions. But they are answerable, or at least can be productively debated, when we see any duty to address institutional corruption as derivative of broader requirements of public officer integrity. Within this framework, a member need only ask: what would best support the overall institutional integrity of Congress? Within this framework, one can balance the competing institutional values (purpose, legitimacy, keeping commitments, efficient pursuit, robustness) to discern both the justification and limits of such a responsibility. Hence, for example, one might be obliged to divert some political capital and some legislative time, but not so much as to compromise the proper execution of the impeachment proceedings; perhaps, one should be willing to lose an election to promote the cause, but not if it would mean the serious risk of an opponent gaining power who aims to cut funding to key oversight institutions.
Conclusion
In this article, I have adopted an institution-first approach to defining public integrity. I have defined public institutional integrity first, and derivatively, public officer integrity.
In order to justify this approach, I assumed that we have an over-riding moral imperative to promote the coherence, consistency, permissibility, praiseworthiness, virtue and trustworthiness – that is, the integrity – of public institutions. I have argued that, upon reflection, we can define the quality that delivers up these qualities in a public institution as the robust disposition to pursue its purpose efficiently, within the constraints of legitimacy, consistency with commitments.
I have argued that with this institutional ideal in place, we can define integrity for public officers as a function of their praiseworthy contribution to that ideal. Public officer integrity is the robust disposition to support the integrity of their institution within the course of their duties, to the best of their abilities.
This set of inter-related conceptions of public integrity – institutional and public officer – defines a new governance goal. It is necessarily opposed to corruption – both institutional and public officer – and implies that efforts should be taken to minimize it. However, it also puts such efforts into context. Corruption is just one of many pathologies, many ‘opposites’ of integrity. Thus efforts to address corruption should not come at the expense of overall public integrity. An institution and its public officers must balance such efforts against other imperatives implied by integrity: robustness, the efficient pursuit of purpose, internal consistency in commitments and so on. Only in this way can institutions not only address corruption but aspire to be truly worthy of praise and public trust.
Acknowledgements
I wish to acknowledge the financial support of the Leverhulme Trust, Amersi Foundation, and Economic and Social Research Council. Many thanks to Anna Petherick, Jo Wolff, Jody Laporte, Lucinda Cadzow, Thomas Robinson, Ngaire Woods, Bo Rothstein, Mark Philp, Emanuela Ceva, Maria Paola Ferretti, A J Brown, Charles Sampford, Marie Newhouse, Gillian Brock, Ken Smith, Christoph Demmke, Peter Collins, Robert Klitgaard and three anonymous reviewers for their comments on this work. I owe especial thanks to Paul Heywood for his tireless efforts to help me improve it. Many thanks also to the practitioners and organizations who engaged critically with this work, particularly Julio Bacio Terracino (OECD), Robert Barrington (Transparency International), Fredrick Galtung (Integrity Action), Francesca Recantini (World Bank), Andrew Preston (UK Cabinet Office) and Sean Hagan (IMF). Finally, thanks to academic and practitioner audiences in Paris, Oxford, Hamburg, Prague, Riyadh, Rio de Janeiro, Brasilia, Sao Paolo, Melbourne and Singapore.
