Executive Power and What Cabinet Committees Reveal About How Prime Ministers Actually Govern

04/02/2026

This post draws on research I am preparing for a journal submission, examining thirty-three years of Cabinet Office circulars to trace how successive prime ministers have configured the cabinet committee system in Aotearoa. The hypothesis I am advancing is perhaps a little too simple: that cabinet committees constitute the primary mechanism through which governments furnish what Alex Matheson and others termed the “purple zone”: that institutional marae-ātea where political authority and administrative expertise must integrate rather than separate: the space where political will becomes administrative action (or, not).

There is a particular kind of political analysis that focuses on what governments say: the speeches, the manifestos, the carefully calibrated announcements, often with “horse-race” tones and pace. And there is another kind, rather less glamorous, that attends to how governments organise themselves to do the actual work of governing. Analysis of the cabinet committee system belongs firmly to this second category, and it deserves considerably more attention than it typically receives.

By the end of this post, I hope to have shown you three things. First, that the apparent stability of our cabinet committee system is largely illusory: beyond a small procedural core, each prime minister rebuilds the architecture according to their own governing logic. Second, that this rebuilding is not merely administrative housekeeping but a political act with material consequences: one that shapes what advice officials must provide to be heard, and what questions the purple zone makes mandatory. Third, that the configuration of this space matters for performance in ways that my earlier analysis of Performance Improvement Framework data helps to illuminate+. If, as that analysis suggested, leaders have the most impact when ministers ensure a clear and coherent strategic framework, then cabinet committees are where that coherence is either built or left to chance.

The implications extend beyond constitutional curiosity. If free and frank advice operates within the purple zone, and if successive governments furnish that zone differently through their committee structures, then what counts as complete, relevant, or even sayable advice shifts with each reconfiguration. Officials navigating the advisory relationship must learn to distinguish why each piece of furniture has been chosen.

When Christopher Luxon assumed the premiership from Chris Hipkins in late 2023, the transition involved something rather more profound than the customary changing of the guard. Of the ten cabinet committees that had structured Hipkins’ administration, Luxon retained precisely three. Seventy percent of the purple zone’s furniture was dismantled and rebuilt according to an entirely different logic. The two governments shared only the Appointments and Honours Committee, the Cabinet Business Committee, and the Legislation Committee: the essential plumbing of the executive, one might say, but hardly the rooms where the substantive work of integrating political will and administrative capability actually occurs*.

The Instability Beneath the Veneer

This is not, I would suggest, administrative trivia. It is a window into how prime ministerial authority is actually constituted and exercised within our Westminster-derived system.

A surface reading of the Cabinet Office circulars spanning the past thirty-three years might suggest a degree of institutional stability. Two committees appear with near-universal consistency: the Legislation Committee, present in every administration since 1990, and the Appointments and Honours Committee, absent from only two. Yet these bodies perform functions that are essentially procedural: ensuring bills are legally sound, reflect previously taken policy decisions and sign off on public appointments. Their persistence creates what one might call a veneer of sameness, behind which everything of genuine political consequence remains entirely contestable.

Once one moves beyond this administrative core, the committee architecture becomes something rather more revealing. It becomes, in effect, a materialisation of how each prime minister chooses to configure the workspace where democracy meets bureaucracy: and, crucially, of what strategic framework officials will be expected to operate within.

The purple zone, as Matheson and others observed in their foundational work, is where “a critical element in coherent government policy was the understanding amongst the key players of the intended direction of the whole.” Cabinet committees are the primary mechanism through which that understanding is given institutional form. They determine which questions must be answered together, which ministers must coordinate, and therefore what analysis officials must provide to be considered adequate to the task.

My earlier analysis of the Performance Improvement Framework reviews offers some empirical grounding for why this matters. Examining fifty-three published PIF reviews, I found that even when controlling for how well an agency was run operationally, strategic clarity from ministers had an independent effect on agency performance. Leadership influences government priorities, but that relationship weakens considerably when strategy is taken out of the equation. The implication is worth stating plainly: without a clear and coherent strategic framework from ministers, even the most capable senior officials struggle to keep agencies aligned with government goals. Incoherent political strategy, the analysis suggests, will eventually degrade outcomes.

Cabinet committees are where that strategic coherence is either created or left to emerge haphazardly. When a government establishes an Expenditure Control committee, as National-led governments have done repeatedly, under Bolger in 1990, under Key in 2008, under Luxon in 2023, it privileges certain questions and makes them procedurally mandatory: what does this cost, what efficiency gains are possible, where can spending be reduced? When a government establishes a Social Wellbeing committee, as Labour-led administrations have favoured, different questions become central: what are the distributional consequences, who benefits, who bears the costs, what is fair?

The structure is not merely a set of meeting rooms. It is, rather, a set of assumptions about what matters, given institutional form. And that form shapes what free and frank advice must contain to be heard.

What the Furniture Demands

This matters for how officials navigate the advisory relationship. The PIF analysis revealed that efficiency and effectiveness are closely tied even when leadership is taken out of the equation: suggesting that investing in tools, processes, and accountability structures can improve performance directly. But those structures do not exist in a vacuum. They operate within the purple zone configuration that ministers establish, and they must answer the questions that configuration makes salient.

Under Bolger and Shipley, as I have traced elsewhere in the Free and Frank series, Strategic Result Areas crossed portfolio boundaries with explicit targets for reducing Māori disparities: embedding Article Three Te Tiriti obligations into the purple zone’s integrative machinery. For officials offering advice on any significant initiative, this infrastructure made certain questions mandatory. Advice that failed to address whether a proposal would improve or worsen Māori–non-Māori disparities was not merely incomplete; it was inadequate to the purple zone’s requirements. The committee structure gave Te Tiriti and equity a procedural voice.

When Clark’s government replaced this framework with lifecycle approaches and the reducing inequalities and then the opportunity-for-all framing, that procedural voice disappeared. The constitutional obligation remained, of course, but the infrastructure that made honouring it procedurally expected rather than personally risky was removed. Under Key and English, elaborate performance frameworks and Better Public Services results privileged evidence-based professional authority: officials had systematic channels for frank advice about what was working. Under Ardern, dramatic simplification around wellbeing and child poverty left political will somewhat unmoored from the administrative capability infrastructure that might have supported it: and, as the PIF analysis would predict, created conditions where strategic alignment became harder to maintain.

The essential quality of free and frank advice may remain constant across these configurations, but the form it must take, the marae-ātea it has, the questions it must answer, the framing it must use to engage ministers, all shift as prime ministers reconfigure the purple zone through their committee choices.

In the context of our MMP environment, committee chairs become a form of political currency: a mechanism through which coalition partners are granted genuine authority within the purple zone rather than merely the appearance of it. Under the present government, ACT’s David Seymour chairs the Expenditure and Regulatory Review Committee, while Winston Peters presides over Foreign Policy and National Security. The pattern is not new; back in 1996, Peters chaired the Cabinet Committee on Industry and Environment as part of his arrangement with Jim Bolger.

The committee structure, in other words, is where coalition deals are given operational reality. It is where the promises made during formation negotiations either acquire substance or quietly dissipate. The purple zone must accommodate not merely the relationship between ministers and officials, but the relationships between coalition partners with different priorities and different bases of political authority.

This creates distinctive challenges for officials navigating the advisory relationship. Advice that satisfies one coalition partner may concern another. The purple zone furniture must somehow enable coordination across these tensions whilst maintaining the integration function that Matheson identified as essential to coherent governance. And as the PIF analysis suggests, when that coherence fractures, because the strategic frameworks become unclear or contested or filled with buzz-words, performance eventually suffers, regardless of how capable individual leaders or agencies might be.

Where Coherence Is Built or Left to Chance

There is something instructive, too, in how prime ministers adjust their structures over time. The data suggests that most learn on the job, consolidating and streamlining as they become more confident in the role: and, I would argue, as they discover which purple zone configurations actually enable the strategic coherence that performance requires.

Helen Clark began with a sprawling thirteen committees in 1999: a structure that perhaps reflected the complexity of coalition management in the early MMP era, but reduced this to a more manageable seven by her second term. John Key proved particularly adept at this kind of refinement, progressively condensing his cabinet from seven committees in 2008 to five in 2011, and finally to a remarkably centralised four by 2014.

This trajectory suggests something important about the relationship between purple zone infrastructure and governing effectiveness. Too much structure creates what Derek Gill characterised as an “iron cage” of performance management: elaborate frameworks that constrain rather than enable integration, formal documents that become irrelevant as soon as they were written yet must still be produced. Too little leaves political will unmoored from administrative capability, and officials without the structured channels that might make frank advice procedurally expected rather than personally risky. Prime ministers appear to discover, through the accumulated experience of governing, where the balance lies for their particular style and coalition configuration.

The question of what happens when a prime minister is replaced by someone from their own party proves unexpectedly revealing. One might anticipate continuity, yet the pattern is inconsistent in ways that illuminate different theories of prime ministerial authority.

When Jenny Shipley succeeded Jim Bolger in 1997, she retained his committee structure entirely: a signal, perhaps, of stability during an already turbulent period. The purple zone furniture remained in place; officials could continue navigating familiar configurations, answering familiar questions. Bill English, taking over from John Key in 2016, did precisely the opposite: he expanded Key’s streamlined four-committee system into nine, asserting his own authority through structural transformation. Chris Hipkins followed a similar path when he succeeded Jacinda Ardern in 2023, restructuring the cabinet to signal his shift toward what was characterised as “bread and butter” concerns.

Each of these choices carried implications for the advisory relationship. When Shipley maintained Bolger’s structure, officials faced continuity in what questions their advice needed to address. When English reconfigured the purple zone, the requirements shifted: new committees meant new coordination demands, new questions that advice must answer to be considered adequate. The strategic framework changed, and with it, as the PIF analysis would predict, the conditions under which officials could effectively align their work with government priorities.

The most striking finding across three decades of data concerns Jacinda Ardern’s first term. When she formed her government in 2017, she did something no other incoming prime minister had done across party lines: she retained her predecessor’s committee structure in its entirety. The nine committees Bill English had used were the same nine Ardern employed. The only difference being the addition of the Crown-Māori Committee.

In spite of the addition of the Crown-Māori Committee, which signalled a distinct political and policy priority, the choice was a remarkable moment of institutional continuity, almost certainly explained by the extraordinary complexity of the three-party coalition negotiations that brought her government into being. Indeed one might say the purple zone furniture was inherited rather than chosen: perhaps a constraint imposed by the political economy of coalition formation rather than a deliberate design choice. One might speculate that the strategic coherence the PIF analysis identifies as essential to performance was, in this configuration, somewhat accidental: a framework that none of the parties would independently have designed, yet one that provided structure nonetheless. Certainly an area for more in-depth study.

The arrangement did not survive her majority victory in 2020. With the constraints of coalition removed, Ardern restructured the cabinet to better reflect her own priorities: simplifying dramatically around wellbeing and child poverty. The anomaly of her first term illuminates something important about the difference between governing within the constraints of a fragile coalition and governing with the relative freedom of a parliamentary majority. It also raises questions, which I am not yet in a position to answer definitively, about whether inherited structures can provide the same strategic coherence as deliberately designed ones, or whether the absence of intentional configuration carries costs that only become visible over time.

Cabinet committees are not, then, the boring machinery they might first appear. They are the architecture through which executive power is organised, distributed, and constrained. They reveal how prime ministers see the world, what they believe requires dedicated attention, and how they navigate the political relationships upon which their authority depends.

More fundamentally, they constitute the furniture of the purple zone: the workspace where political authority and administrative expertise must integrate rather than separate. How that workspace is configured shapes what questions officials must answer, what coordination is procedurally expected, and therefore what free and frank advice must contain to be considered adequate. The PIF analysis suggests this configuration has material consequences: that strategic coherence from ministers independently affects performance, and that without clear frameworks, even capable leaders and agencies struggle to maintain alignment with government goals.

The next time a government is formed, or new ministers are welcomed to the top table, I would suggest we also attend to not only to who receives which portfolio, but what committees are created, who chairs them, and what they are charged with considering. The purple zone is being furnished. The configuration will shape what can be said, what goes without saying, and perhaps what cannot be said at all without challenging the arrangements that hold the governing coalition together.

The ghost in the machine, as I have termed free and frank advice elsewhere, must inhabit whatever space it is given. Understanding how that space is constructed may be the first step toward understanding how advice actually operates in practice: and why the disagreement about what it means may be less confusion than constitutional design.

References

Gill, D. (Ed.). (2011). The iron cage recreated: The performance management of state organisations in New Zealand. Institute of Policy Studies.

Matheson, A. (1998). Governing strategically: The New Zealand experience. Public Administration and Development, 18(4), 349–363. https://doi.org/10.1002/(SICI )1099-162X(199810)18:4<349::AID-PAD22>3.0.CO;2-5

Matheson, A., & Fancy, H. (1995, November). Future directions in public sector management in New Zealand: Towards strategic management [Paper presentation]. Public Sector Convention, Wellington, New Zealand.

Matheson, A., Scanlan, G., & Tanner, R. (1997). Strategic management in government: Extending the reform model in New Zealand. In Benchmarking, evaluation and strategic management in the public sector (OECD Working Papers, Vol. V, No. 67). OECD.

*The primary data for this analysis comprises Cabinet Office circulars issued between 1990 and 2023, which specify the cabinet committee structure, membership, and terms of reference for each administration. These circulars are issued by the Cabinet Office within the Department of the Prime Minister and Cabinet at the commencement of each new government and, on occasion, during mid-term restructuring. They are publicly available through the Cabinet Office’s published records. The dataset encompasses fifteen distinct committee configurations across nine prime ministers: Bolger (1990, 1996), Shipley (1997), Clark (1999, 2002, 2005), Key (2008, 2011, 2014), English (2016), Ardern (2017, 2020), Hipkins (2023), and Luxon (2023). I have them as pdf’s: email me if you want a copy.

+For the full analysis of the fifty-three published Performance Improvement Framework reviews, including the Spearman’s correlation and partial correlation methodology, see Te Kawa, D. (2025). What Drives Government Performance? A Look at the Numbers. https://debtekawa.com/what-drives-government-performance-a-look-at-the-numbers/