Free and Frank Advice: Alarms, Measurers and Mappers, and What the Literature Leaves Out

03/01/2026

I read four decades of scholarship on free and frank advice as three dispositions, set out the three things the map leaves out, and explain why the conditions of the convention are mapped while its meaning is not.

This is the seventh of twenty posts drawn from my doctoral thesis on free and frank advice in Aotearoa. Having shown that the statutes elevate the convention without defining it, I ask here whether the scholarship supplies what the statutes leave out, and I find that for the most part it does not.

The scholarship on free and frank advice is careful, cumulative and often urgent, and when I read it as a whole for my thesis I found that it leaves undone the same task the statutes leave undone, which is the empirical mapping of how practitioners themselves understand and enact the obligation.

In this post I do three things: first, I describe the literature as a single discourse that has passed through three dispositions, the alarm sounders, the measurers and the system mappers, and I show where each stopped short of the question of meaning; second, I set out three things the map of the advisory system leaves out, concerning evidence, authority and delivery; and third, I explain why the literature, read together, turns an old constitutional observation that conventions are slippery into a problem for public administration, and why that problem calls for a different kind of inquiry.

Three dispositions, one gap

In my reading the literature is one evolving discourse in which each generation of scholars has built on the one before, asking three questions that still remain open. The alarm sounders asked what the New Public Management reforms would cost the advisory relationship, the measurers asked whether that cost could be observed, and the system mappers asked what the advisory world had become once the reforms had run their course. None of the three asked what practitioners took the obligation itself to be, and that common omission is the gap my thesis sets out to fill.

The alarm sounders wrote as the reforms were enacted, and their warning was specific: a purchasing model would strain to specify a product whose value lay in the exercise of judgement, and what escaped specification would in time cease to be bought. The reforms renegotiated the public service bargain, the exchange by which officials trade loyalty and competence for tenure and the standing to reach an independent judgement, with free and frank advice counted among the things owed back across it (Hood & Lodge, 2006). A review of the new model for the State Services Commission warned that a regime of specified outputs “narrows responsibility to simple compliance with what is on the list, and prompts Chief Executives to disregard responsibility for items not specified” (Schick, 1996, p. 1), and the collapse of the viewing platform at Cave Creek in 1995, which killed 14 people, later gave that warning a human form, since the knowledge that might have prevented it existed in pieces across the department and stayed below the point of decision (Gallen, 1995). What strikes me about this first literature is that it knew what was at risk without saying what the thing at risk consisted of.

The measurers then set out to establish whether the damage the alarm sounders feared had in fact occurred, and the record they produced is careful and granular, although it studied the conditions around the convention more fully than the understandings within it. The first survey of politicisation by ministerial advisers in Aotearoa put a 68-item questionnaire to 546 senior officials and drew 188 responses (Eichbaum & Shaw, 2007a); the State Services Commission’s own project on the quality of policy advice recorded, among the reasons quality went unevaluated, “the seeming lack of awareness on the part of the Public Service of what constitutes free and frank advice” (State Services Commission, 1999, p. 27); my own small study in 2014 asked eight policy professionals, each with thirty years of giving and receiving free and frank advice, how they understood the practice (Te Kawa, 2014); and the Public Service Census of 2025 found that 71% of those who prepare advice for ministers were confident that their organisation gave free and frank advice (Te Kawa Mataaho Public Service Commission, 2025). Even the direct measure asks how officials regard the convention, and in my view the more important question, what they take the convention to be, is left unasked by every one of these instruments.

The system mappers stepped back far enough to take in everyone now in the room, and the field they drew is a crowded one, with ministers taking counsel from consultants, lobbyists, think tanks and ministerial staff as well as from the public service (Craft & Halligan, 2020). They named the purple zone, the common space between politicians and administrators that the reforms opened (Matheson, Scanlan, & Tanner, 1997), and they traced a functional politicisation that bubbles up from within the service as officials fit their advice to anticipated preference without any instruction being given (Shaw & Eichbaum, 2020). The closest the mapping comes to my question is the account of free and frank advice as the confidential counsel of a trusted insider (Mulgan, 2012), together with the argument that sound advice fuses research evidence, political judgement and the practical wisdom of delivery (Head, 2008), a fusion I return to later in this series under the name the thesis gives it, practical statecraft. The mappers placed the convention on the map of public administration, and in my reading still treated what practitioners mean by it as a background assumption.

What the map leaves out

The map the system mappers drew is careful, and it leaves out three things that are, in my view, three sides of one inheritance: evidence, authority and delivery. The first omission is evidence, which the advisory map treats as the firm base on which candour rests, although the more recent literature is much less sure, writing of a widely shared sense of a crisis of expertise in liberal democracies in which evidence-based ideas have come to be distrusted (Head, 2024). If evidence is one form of knowledge among several, contested and revisable, then advice resting on it is fallible, and those who treat it as the settled base of free and frank advice assume a stability the knowledge does not have.

The second omission is authority, and it follows once evidence is admitted to be one knowledge among several, because the next question is whose knowledge and whose authority, and in Aotearoa that question carries both a name and a constitutional standing. Work on wairua in the public service shows a way of knowing that a bureaucratic order built on impersonal and calculative reason sets aside (Maynard, 2025), and Critical Tiriti Analysis reads policy directly against the articles of Te Tiriti o Waitangi, in the face of constitutional arrangements that have tended to maintain structural racism against Indigenous people (Came, Baker, & McCreanor, 2020). None of this excuses the order the reforms replaced, since the closed, elite service that preceded them was itself paternalistic, and was the instrument of the settler state’s confiscations and assimilation. My reading is that the reforms and the order they replaced are flawed in different ways, and that the managerial map leaves the authority of Māori and the knowledge of Te Ao Māori out of view.

The third omission is the most ordinary and, I think, the most consequential, because the managerial map pictures evidence running downhill from analysts into policy and then into advice, and misses that knowledge also rises from delivery. Front-line discretion in schools, clinics and offices does more than carry policy out, since in practice it becomes the policy (Lipsky, 2010), and good policy rests on feedback thinking, the deliberate readmission of what delivery and experience teach (English, 2021). The split between policy and delivery cut that path, and it has yet to be rebuilt. These three omissions each press on free and frank advice from outside the managerial frame, and I treat them in my thesis as conditions that bear on the relationship between permanent officials and the ministers they advise, which is the relationship the statutes govern and the one the inquiry keeps in view.

From a slippery convention to a public administration problem

The statutes, the inheritance and the scholarship arrive at the same finding by different routes, which is that the working meaning of free and frank advice stays elusive, and I want to be careful about what follows. Constitutional scholarship has long accepted that conventions resist exact statement, and I treat that general proposition as settled. My claim is a public administration claim, set in a deeply political context: free and frank advice is a slippery convention that has been put to work as a professional duty, a statutory principle, a quality standard, a withholding interest, a leadership obligation and a measure of system health, and public administration has asked it to do that daily work while leaving unmapped the meanings practitioners give it.

Four decades of scholarship did produce something of real value, which is a precise account of the conditions under which free and frank advice is expected to survive. It requires an administrative settlement that gives officials standing and security, working knowledge drawn from delivery, a relationship an official can read and move through, institutions capable of hearing unwelcome information, and a professional culture that carries the duty to speak across generations of officials. Each of those five conditions has been studied, and the cost of their failure counted in inquiry after inquiry, while the public service’s own census still finds fewer than three-quarters of those who advise ministers confident that their organisation is free and frank. In my view the distance between a statute that names free and frank advice a principle of the public service and a service in which many feel unable to give it is itself a measure of what the gap in our knowledge costs.

What the literature leaves open is an account of what the officials and ministers working inside those conditions understand themselves to be doing when they call advice free and frank. If the content of the convention sits outside the statutes and the scholarship, and is nevertheless supplied every day by those who give and receive advice, then the questions worth asking are questions about them: what they take the obligation to be, what significance they attach to it, and which conditions they identify as supporting or constraining it. Those are empirical questions, and they are not answered by another survey that asks whether free and frank advice is occurring before asking what practitioners take it to be. They call, in my view, for a method that can study subjective understanding, situated judgement and patterned disagreement together.

In the next post I set out that method: the decentred approach to interpretive inquiry, Q-methodology, and the three-stage design that carried the thesis from interviews to shared patterns of meaning.

References

Came, H., Baker, M., & McCreanor, T. (2020). Addressing structural racism through constitutional transformation and decolonization: Insights for the New Zealand health sector. Journal of Bioethical Inquiry, 17(4), 583–595.

Craft, J., & Halligan, J. (2020). Advising governments in the Westminster tradition: Policy advisory systems in Australia, Canada, New Zealand and the United Kingdom. Cambridge University Press.

Eichbaum, C., & Shaw, R. (2007a). Minding the minister? Ministerial advisers in New Zealand government. Kotuitui, 2(2), 95–113.

English, B. (2021). Better public policy via feedback thinking. In S. Mazey & J. Richardson (Eds.), Policy-making under pressure: Rethinking the policy process in Aotearoa New Zealand (pp. 134–146). Canterbury University Press.

Gallen, T. W. (1995). Commission of Inquiry into the collapse of a viewing platform at Cave Creek (Part one). New Zealand Government.

Head, B. W. (2008). Three lenses of evidence-based policy. Australian Journal of Public Administration, 67(1), 1–11.

Head, B. W. (2024). Reconsidering expertise for public policymaking: The challenges of contestability. Australian Journal of Public Administration, 83(2), 156–172.

Hood, C., & Lodge, M. (2006). The politics of public service bargains: Reward, competency, loyalty—and blame. Oxford University Press.

Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the individual in public services (30th anniv. ed.). Russell Sage Foundation.

Matheson, A., Scanlan, G., & Tanner, R. (1997). Strategic management in government: Extending the reform model in New Zealand. State Services Commission.

Maynard, K. T. (2025). He kakano no Rangiatea: Indigenising and spiritualising public administration [Doctoral thesis, Victoria University of Wellington].

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

Schick, A. (1996). The spirit of reform: Managing the New Zealand state sector in a time of change. State Services Commission and Treasury.

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

State Services Commission. (1999). Essential ingredients: Improving the quality of policy advice (Occasional Paper No. 9).

Te Kawa, D. (2014). Free, frank and fearless: A small-scale investigation [Unpublished paper]. State Services Commission & Victoria University of Wellington. Te Kawa Mataaho Public Service Commission. (2025). Public Service Census 2025: System report.