Free and Frank Advice: A convention everyone invokes and nobody defines

04/10/2025

I open the series by setting out the puzzle my thesis began with, that free and frank advice is written into statute and invoked across the public service while nobody has said what it means, and I outline what I found.

This is the first of twenty posts drawn from my doctoral thesis on free and frank advice in Aotearoa. In this post I set out the puzzle the thesis began with, the public episode that made it concrete and the questions I asked of it, and, in outline only, the finding that the rest of the series will take apart one piece at a time.

Free and frank advice is one of the phrases people in and around government in Aotearoa reach for most often, and in my experience one of the hardest to explain when someone asks what it requires, in front of a particular minister, on a matter going badly. In this first post I do three things: first, I set out the puzzle my thesis began with, a convention and statutory principle that is constantly invoked and nowhere defined; second, I describe the public episode that made the puzzle concrete for me and the questions I asked of it; and third, I sketch the central finding in outline.

A convention without a definition

The Public Service Act 2020 makes free and frank advice one of the public service principles, placing on officials a duty, “when giving advice to Ministers, to do so in a free and frank manner” (Public Service Act 2020, s. 12(1)(b)), yet nothing in that Act, or in the statutes that came before it, says what advising in a free and frank manner actually requires. Before that elevation, free and frank advice had long operated as a convention of executive government, treated as binding in practice while remaining under-identified and vague in its content (Marshall, 1984; McLean, 2024). In my view the interesting thing is that a practice so central to the advisory relationship has gone on working for so long without anyone having to say what it is.

That absence has a history, because Parliament has returned to free and frank advice over four decades, each time giving it more standing while leaving its substance open. It first appeared in the Official Information Act 1982 as a basis for withholding information, designed to shield internal deliberations from disclosure; it became a responsibility and duty of chief executives under the State Sector Act 1988; and it was raised to one of the public service principles under the Public Service Act 2020. I read that repetition as a pattern of elaboration without definition, drafted with care each time by people who understood the convention well.

The scholarship has followed much the same course, approaching free and frank advice for the most part from the outside in, and seldom from the meaning practitioners make of it inside the relationship. One strand sounds the alarm about pressure on the convention, a second seeks to measure its erosion, and a third maps the systems and statutory settings that carry and protect the advisory relationship (Eichbaum & Shaw, 2008; Mulgan, 2012; Shaw & Eichbaum, 2020). In my reading this work explains a great deal about the pressures around free and frank advice, and leaves open the question I most wanted to ask: what the ministers and officials who give and receive it understand it to mean, and how they act on that meaning.

The 2013 episode and the questions it raised

The puzzle became concrete for me in 2013, when confidential Cabinet papers on restructuring the Ministry of Foreign Affairs and Trade were leaked to the Opposition and the State Services Commissioner commissioned an inquiry into the leak. The inquiry found no evidence that the leak came from the two senior diplomats later criticised in its report, and it nonetheless concluded, from their private emails and phone records, that they had created a perception that opposition to the restructuring could acceptably be aired outside the ministry (Rebstock, 2013). The episode has stayed with me because it showed how quickly the conduct of senior officials can be recast once nobody has an agreed account of what they owe.

The criticism was substantially undone in 2016, when the Ombudsman found that the inquiry had been flawed and unfair, having exceeded its terms of reference, denied the officials fair notice that their own conduct would be examined, and reached some criticisms the evidence did not support (Paterson, 2016). What I take from the Ombudsman’s findings is that they corrected the process and the evidence, and left for someone else the further question of how the conduct had come to be read as it was in the first place.

That further question is the one my thesis takes up, because the inquiry assessed the two officials through political neutrality, through loyalty and compliance, and through the instruction to give feedback by an approved channel, and it did not assess their conduct as candour among and between colleagues. The report itself accepted that contemplating and debating public policy through widely shared formal messages “was not out of the ordinary” for a policy agency (Rebstock, 2013, p. 41), and the same conduct could therefore be read either as improper opposition to a lawful instruction or as senior officials debating the direction of a major institutional change. In the absence of any authoritative account of free and frank advice, the officials, the inquiry and the Ombudsman were each left to construct their own answer, and in my view that ambiguity, and the doubt it creates inside the advisory relationship itself, is the real subject of the case.

I came to the giving of advice before I came to its study, having learnt my practice at Te Puni Kōkiri and worked since at the point where advice meets decision, and I have watched free and frank advice done well and done badly, and seen how costly the difference is. That experience shaped three questions, asked in a deliberate order: what free and frank advice is, and how practitioners understand and construct it; why it matters to them; and what conditions they identify as supporting or constraining its practice. The first is the question the thesis is built to answer, and I treat the second and third as serving it, because a convention can bind in practice while remaining under-identified in its content.

To answer those questions I used a three-stage interpretive design, in which 32 elite interviews in Stage One generated the field of things practitioners say about free and frank advice, 58 practitioners sorted 34 statements drawn from that field in a Q-methodology study in Stage Two, and 88 participants tested and sharpened the interpretation through wānanga, workshops and one-on-one interviews in Stage Three. Of the 141 people approached, 126 took part substantively in at least one stage, and I count the method among the contributions of the thesis.

What I found, in outline

The central finding of my thesis is that free and frank advice is a bounded plurality, a term I use as a sensitising concept that organises the evidence and marks distinctions within it, carrying only the weight the evidence can sustain and stopping short of a definition. It has three elements: a shared boundary, six postures inside it, and four fault lines between the postures.

The boundary comes first, because practitioner understandings reach a limit, and free and frank advice stays recognisable as such when it supports what I call practical statecraft. I state practical statecraft as a two-part test in a fixed order, in which the content test asks whether advice brings together evidence, expertise and practical knowledge of consequence, especially knowledge of how a policy will work out once it meets the world, and the collective test then asks whether it is brought together between ministers and officials out of their different knowledge, authority and responsibility. In my reading the test marks the edge of what practitioners recognise as free and frank advice, and it is only once advice passes it that the differences between them come into view.

Inside that boundary the plurality becomes visible, in six practitioner postures, the recurring positions practitioners take up when they practise free and frank advice: the Institutional Pragmatist, the Democratic Constitutionalist, the Collaborative Sense-Maker, the Moment Chooser, the Political Servant and the Westminster Protector. Each posture is read configurationally, from the whole array of 34 statements as its practitioners arranged them, and each is a coherent orientation in its own right, which is why I resist treating any single posture as the convention itself.

Four fault lines run between the postures and organise the disagreement among them: mandate alignment against principled independence, challenge as public reason against protected counsel, routine practice against protective failsafe, and the strategic craft of timing against the duty to speak. Read through the idea of the remainder, free and frank advice appears as an unresolved part of the advisory relationship, a site where tensions between ministers and officials over mandate, challenge, candour, timing and trust are contained without being settled (Honig, 1993). In my view the absence of a definition is doing something, because it places judgement with officials in the one relationship where candour, responsiveness, neutrality and political authority all meet.

Te Tiriti o Waitangi sits at the edge of this picture, and my claim about it is deliberately careful, because Te Tiriti stayed outside the first two stages of the study, even when participants were prompted, and emerged most clearly in Stage Three validation, which suggests to me that it presses on the advisory system in Aotearoa while remaining only partly integrated into it. I treat Te Tiriti as marking the constitutional limit of the mapped field, a pressure at the edge of the advisory relationship that this study could only partly explain, and I think the next account of free and frank advice will need to begin from that wider constitutional footing.

Read through the bounded plurality, the 2013 episode looks different, because the failure the Ombudsman corrected was also a failure of recognition, in a system without the vocabulary to distinguish disloyalty from principled candour. The two officials can be understood as acting from a posture concerned with institutional integrity and with what a major restructuring would do to the capability of their ministry, the kind of practical statecraft a permanent public service is expected to contribute, and the damage was done when one posture was treated as the whole convention and another was read as a breach. I began the thesis expecting to find a definition, or at least a contest over one, and found practitioners who were clear-minded about free and frank advice and handling real tensions with skill and care, inside a convention that democratic government requires and leaves finally open.

In the next post I go beyond Aotearoa, to look at how Australia, the United Kingdom and Canada carry a kindred obligation of candour in different institutional forms, and at the silence the four systems share.

References

Eichbaum, C., & Shaw, R. (2008). Revisiting politicization: Political advisers and public servants in Westminster systems. Governance, 21(3), 337–363.

Honig, B. (1993). Political theory and the displacement of politics. Cornell University Press.

Marshall, G. (1984). Constitutional conventions: The rules and forms of political accountability. Oxford University Press.

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

Mulgan, R. (2012, May 30). What future for free and frank advice? [Address]. Institute of Public Administration New Zealand Centennial Lecture, Wellington, New Zealand.

Paterson, R. (2016). Investigation into SSC conduct of MFAT leaks inquiry. Office of the Ombudsman. https://www.ombudsman.parliament.nz/resources/investigation-ssc-conduct-mfat-leaks-inquiry

Public Service Act 2020 (NZ).

Rebstock, P. (2013). Report to the State Services Commissioner on the investigation into the possible unauthorised disclosure of information relating to the Ministry of Foreign Affairs and Trade. State Services Commission.

Shaw, R., & Eichbaum, C. (2020). Bubbling up or cascading down? Public servants, political advisers and politicization. Public Administration, 98(4), 840–855.