Free and Frank Advice: The Bounded Plurality Finding

02/05/2026

I bring the six postures, practical statecraft and the four fault lines together as the bounded plurality finding, explain why I treat it as a sensitising concept rather than a definition, and set out why Aotearoa’s constitutional culture makes an undefined convention workable.

This is the fifteenth of twenty posts drawn from my doctoral thesis on free and frank advice in Aotearoa. The last four posts set out the parts of the empirical structure one at a time; this post reads them together as the central finding of the thesis. It covers what the bounded plurality finding claims, why I hold it as a sensitising concept, and how the constitutional culture of Aotearoa helps explain why a convention named as essential can remain undefined and still carry force.

Free and frank advice is undefined in statute and in scholarship, and the work of saying what it means has fallen, in practice, to the ministers and officials who give and receive it, which is why I went to practitioners to find out what it is. What they told me, through the interviews, the Q-sort and the validation, is that free and frank advice reaches past both a single agreed definition and a scatter of private views, and in my thesis I call what lies between those two a bounded plurality. I do three things in this post: first, I set out what the bounded plurality finding claims when its parts are read together; second, I explain why I treat it as a sensitising concept rather than a definition; and third, I describe how Aotearoa’s constitutional culture makes that kind of open convention intelligible, and where its reach stops.

What the finding claims

The bounded plurality finding has three parts, and the finding is the three read together: a shared boundary, which is practical statecraft; six postures inside that boundary; and four fault lines, over mandate, challenge, occasion and timing, along which the postures part company. Each part depends on the other two for its meaning, because the boundary read alone would look like consensus, the postures read alone would look like fragmentation, and the two read without the fault lines would show difference with no pattern to it, which is why I hold all three inside the one finding.

The plurality lies in the six postures, which show practitioners sorting the same 34 statements as though free and frank advice carried more than one shared meaning. For the Institutional Pragmatist it is workability under pressure, for the Democratic Constitutionalist a statutory duty attached to role and office, for the Collaborative Sense-Maker collective sense-making across knowledge holders, for the Moment Chooser judgement exercised in time, for the Political Servant service to the Government of the day and protective warning for a minister, and for the Westminster Protector senior, protective role-work inside Westminster propriety, and each of those gives a recognisable account of what free and frank advice is for, how it should be performed and where its legitimacy sits.

The boundedness lies in what those postures share, which is practical statecraft, the commitment to advice that brings evidence, expertise and practical knowledge of consequence together, and that is made between ministers and officials out of their different knowledge, authority and responsibility. The postures part company over mandate, challenge, occasion and timing, yet they continue to treat advice as something that has to be of use to the Government of the day, and for that reason I read the finding as saying more than that free and frank advice is fragmented. Practical statecraft is the empirical boundary around the plurality, and the fault lines show how the plurality is organised inside it.

The four fault lines belong to the finding itself, because they record where and why the postures disagree inside the boundary they share. The first runs through the ministerial mandate, and how far advice must fit a minister’s democratic mandate; the second through challenge, and where challenge belongs in the advisory relationship; the third through occasion, and whether free and frank advice is the ordinary work of governing or a protective warning kept for serious risk; and the fourth through timing, and when the obligation to speak arises. In my reading, the fault lines are what give the plurality its structure, since they show that practitioners disagree along a small number of recurring lines rather than at random, and that the same institutional condition can support one posture while constraining another.

The shape of the solution gives the plurality a recognisable form, which I take as evidence that the differences are structured. Four postures carry 48 of the 50 defining sorts and most of the shared pattern, two narrower postures sit at the outer edge where free and frank advice turns toward protection, political service, senior craft and Westminster propriety, and the eight hyperplane sorts sit between postures, inside the solution, so the postures are neither equal-sized population types nor isolated opinions, and in my reading they are best described as shared advisory orientations within a patterned field.

The finding also changes how practitioner disagreement can be read, and the Stage Three validation bore that reading out. The data show practitioners holding more than one considered view of free and frank advice, with each orientation fitted to different institutional pressures, some weighted toward duty, some toward usefulness, some toward relationship and collective knowledge, and some toward timing, protection or seniority. In Stage Three, practitioners said the postures gave them words for differences they had lived but had until then put down to personality, style, courage or political judgement, and I take that as the clearest sign that the plurality is real.

A sensitising concept rather than a definition

I hold bounded plurality as a sensitising concept, in the sense Blumer (1954) gave that term, and the distinction from a definition matters for what the finding can be asked to do. A definition would state in advance the conditions advice must meet before it counts as free and frank, and would hold those conditions for every case, whereas a sensitising concept gives a direction in which to look, at postures, at fault lines and at the threshold practitioners treated as the limit, and leaves what is found there open to the next study.

The threshold in my thesis was recovered from practice rather than stipulated, and that is why I am careful about how far it reaches. It is what 58 practitioners placed in common when they sorted the same 34 statements, sharpened by what they said when the map was returned to them, and because the unit of analysis throughout is the shared viewpoint, the finding reports how practitioners hold the obligation and stops short of stating what the obligation is. A different group of practitioners, in another jurisdiction or at another moment, could place the threshold elsewhere, and the design allows for that.

Reading the finding this way redirects the problem the thesis began with. The under-definition of free and frank advice first looked like an absence to be corrected through a better definition, but the findings suggest that the work of definition has been going on within practice all along, through the judgements ministers, officials and political advisers make about mandate and independence, the record and protected counsel, routine obligation and protective failsafe, and timing and the duty to speak. The more useful question, in my view, is how practitioners sustain a recognisable obligation without a rule, and how they manage to do it as coherently as the finding suggests they do.

Why the silence is workable in Aotearoa

The answer I give in the thesis is that Aotearoa’s constitutional culture makes the definitional silence intelligible, and in most respects allows it to work. Constitutional culture, on Palmer’s account, is how a constitution actually operates in the beliefs and behaviour of those who run it, and in his phrase the constitution is “not a thing but a way of doing things” (M. S. R. Palmer, 2007, p. 565), which leaves room for the executive, as much as the courts, to give working form to the unwritten parts of constitutional practice through what ministers and officials do. Read through the state as cultural practice (Bevir & Rhodes, 2010), free and frank advice is one of the practices through which that happens.

That culture is plural, carrying four inherited traditions at once: the Westminster and Whitehall inheritance, the contested shift within and beyond Westminster, Te Tiriti o Waitangi, and the settler constitutional culture Palmer describes. Te Tiriti places the advisory relationship within a constitutional order that holds more than one account of authority, of whose knowledge counts and of what good government requires, and Māori authority endures despite the settler state’s attempt to absorb it by claiming a sovereignty beyond what Te Tiriti conferred. I return to Te Tiriti as the limit of the field I mapped in a later post, and record here only that free and frank advice that reckons with the Westminster model alone meets part of what the advisory system faces.

The settler constitutional culture carries three dispositions, egalitarianism, authoritarianism and pragmatism, and the fault lines my thesis found fall where the claims of those dispositions cross. Egalitarianism supports both deference to an elected minister’s mandate and plain speaking between people who regard one another as equals, which are the two poles of the first fault line, and it reaches the second and third as well; authoritarianism wants government strong and decisive, deepening the mandate pole, and also wants a capable service that stewards institutions across governments, which meets the protective pole of the third. Pragmatism declines to resolve those crossed claims, preferring convention to code, and it reaches the fourth fault line, where timing and judgement in the moment are at stake, so that mandate and occasion are each reached by two dispositions, challenge by egalitarianism alone and timing by pragmatism alone.

The claim I make here concerns intelligibility rather than causation, and I hold it at that strength deliberately. My study was designed to map practitioner subjectivity, and its evidence leaves open whether the constitutional culture produced the fault lines or whether culture and practice are responding to the same conditions independently, and it also leaves open whether a pragmatic culture can keep absorbing a convention that Parliament has increasingly put into statutory form. The obligation itself, on McLean’s reading, remains a norm observed “as if no one is looking, because in fact no one is looking” (McLean, 2024, p. 285), and in my view the bounded plurality is what practitioners have made of free and frank advice in the space that observation describes.

In the next post I read the six postures as situated enactments of the traditions practitioners inherit.

References

Bevir, M., & Rhodes, R. A. W. (2010). The state as cultural practice. Oxford University Press.

Blumer, H. (1954). What is wrong with social theory? American Sociological Review, 19(1), 3–10.

McLean, J. (2024). Conventional norms and the comparative constitutional law project. Comparative Constitutional Studies, 2(2), 272–291.

Palmer, M. S. R. (2007). New Zealand constitutional culture. New Zealand Universities Law Review, 22, 565–608.