Proposed Amendments Public Service Act (2020): Joining the Threads or Tightening the Grip?

09/06/2025

Paper Three in the reform suite promises to break down silos and deliver a more integrated public service. On the surface, it revives familiar tools, for example, shared services, system leaders, and cross-agency levers, and stitches them into law. That could bring needed coherence. But the deeper risk is brittleness: integration without legitimacy, central control without discretion, and national security powers exercised with little contestability. The question is whether these joins will hold the pattern, or pull it too tight to flex.

If Paper One in the public sector reform suite of proposals strengthened the frame of the tukutuku panel, and Paper Two pulled the weave tight at the centre, then Paper Three turns to the joins: the messy threads between panels where the public service operating model frays, service delivery models double up, and coherence quietly unravels.

In language that promises simplicity, the third and final paper proposes ways to break down silos and coordinate more efficiently across government. But its real ambition is broader: a more focused, more integrated, and arguably more effective public service.

There’s nothing especially new here. Much of what’s proposed echoes the Better Public Services agenda from over a decade ago: shared back-office functions, system leadership, and centralised levers to drive alignment. But this time, the reforms are wrapped in statute. That matters.

What was once a strategic preference has now become a legislative expectation. And with that shift, we should be asking not just what’s gained in efficiency, but what might be lost in discretion, legitimacy, and trust.

What it restores: tools, levers, and lines of sight

Paper Three reactivates some underused tools in the 2020 Act: interdepartmental ventures, joint operational agreements, and system leader standards. These were designed to enable joined-up delivery but have largely sat idle.

The paper also proposes considering requirements for smaller agencies to use shared services and using central procurement powers to better coordinate large-scale digital investments. This work is underway, and is a good thing.

Perhaps most notably, it reintroduces the concept of ‘key positions’ across the system. These roles, like Chief Information Officers and Chief Financial Officers, require dual sign-off from both the departmental chief executive and the Public Service Commissioner (or a delegated system leader).

It’s a structural echo of the 2013 State Sector Amendment Act, when Treasury sat in on most agency CFO appointments under its delegated authority. I believe that provision was removed in 2020, as it was deemed too burdensome and too centralised. Now it’s back, but in a more tailored form.

The Commissioner is also proposed to gain new powers to restrict or prohibit specific vendors, services, or technologies when they pose risks to national security or the national interest. This is framed as a risk management measure, and it may well be needed in an increasingly contested digital environment. However, it represents a significant shift: a statutory direction-making power based on classified information, with limited routes for public contestability or challenge.

Before issuing such directions, the Commissioner must consult the Minister for the Public Service and seek advice from an extensive network including the Directors-General of the intelligence and security agencies, the Department of the Prime Minister and Cabinet as the National Security Advisor, and the Secretary of Foreign Affairs and Trade regarding New Zealand’s international obligations and interests.

These reforms, taken together, strengthen the state’s capacity to act in concert, making shared choices, coordinating delivery, and managing whole-of-system risk. That’s no small achievement. But it’s only half the picture.

What it risks: integration without legitimacy

The danger with central coordination is that it can start to behave like a command.

Dual-key appointments, for all their logic, may reduce space for agencies to choose leaders who understand their communities, cultures, and the context in which delivery happens.

Shared services may create efficiencies but risk flattening an institution’s character.

Centralised ICT governance may bring down costs, but if over-applied, it could displace the discretion needed to serve different populations in different ways.

And the security-based vendor restrictions, while serious in intention, raise difficult questions. We will wait until the Bill is tabled, but I think this means the Commissioner would be authorised to issue binding directions, informed by classified intelligence, and applied across the Public Service and agencies mandated in the Protective Security Requirements.

That’s a large power to sit with a single statutory office. Yes, there are safeguards: proportionality, international obligations, and inter-agency consultation. But we will wait to see what the Bill says about public hearings, appeal mechanisms, and transparency. In the name of national interest, a wide range of technologies could be removed from service, without public visibility into why.

Let’s be honest: the desire to break down silos is real. But silos aren’t only structural, they’re cultural. They emerge not just from poor wiring diagrams, but from mistrust, misalignment, and fear of failure. They also emerge when the centre pulls too tight and the frontlines are left with no room to engage with one another.

This paper treats fragmentation as a design problem for systems. It might be more accurate to call it a problem of institutional confidence, relational capability, and uneven incentives. You can’t fix that with a better wire diagram or Meccano set.

So, yes, the pattern may become more cohesive. However, the risk is that it becomes brittle, a tight weave with no room to flex.

What Parliament should ask

If this paper is to restore not just coordination but democratic accountability, then a few questions need asking:

What ensures public legitimacy when national security decisions are made behind closed doors? The paper proposes binding vendor bans based on classified information. But democratic legitimacy depends on the ability to contest, explain, and review decisions. Where will those protections live?

Who holds system leaders accountable in dual-key appointments? If system leaders now control appointments and reviews for key roles across departments, how are they themselves assessed? What prevents the emergence of an unelected inner circle of influence?

How are capability differences accounted for in shared service mandates? Smaller agencies will likely bear the brunt of centralisation. What support is being offered to lift their capability, not just consolidate it?

Where is the everyday legitimacy of the system protected? These proposals restore ministerial control. But legitimacy isn’t just handed down from the ballot box; it is earned daily through trustworthy institutions, respectful relationships, and service that feels fair. Where is that protected in this pattern?

The deeper risk

Let’s be clear: these reforms do restore a version of democratic accountability, if we define that narrowly, as alignment with elected government priorities. They sharpen levers. They clarify roles. They give ministers and central agencies greater visibility across the system.

But real democratic legitimacy is a thicker weave. It’s not just representative, it’s relational, deliberative, and participatory. It lives in the integrity of everyday decisions, the discretion to adapt, the ability to challenge, and the humility to listen. Our democracy only works when elected governments act in good faith, and when coordination serves communities, not just control.

None of these tools can substitute for that.

In fact, these reforms look a lot like Better Public Services, stitched into law, only this time, with more security language, tighter levers, and less trust in the system’s ability to work things out through relationships.

Like all tukutuku, the joins matter. Not because they are tidy, but because they carry the tension. They are where strands cross, patterns are held in balance, and meaning emerges through contrast, not uniformity. Strong joins don’t just connect, they absorb strain, allow flex, and preserve the integrity of the whole.

These reforms risk mistaking the frame for the full design. They tighten the outer edges, standardise the verticals, and fix the grid. But they leave little space for the sideways moves, for discretion, dialogue, resistance, and repair, let alone for the contextual, interdependent and inexplicable nature of problems, and therefore the things that make the pattern live.

And that’s what’s at stake. Not just better coordination, but a public service that can move with complexity and contextuality. That knows when to hold firm and when to yield. That treats relational work not as inefficiency, but as essential infrastructure. That understands its constitutional role is not to be nice to Ministers, but to offer free and frank advice and implement the policies of elected officials.

In tukutuku, the join is where the story deepens. We would do well to remember that.

Closing Reflection on Papers One to Three

Across the three Cabinet papers, a clear pattern emerges. The outer frame has been redrawn. The internal threads have been pulled tight. And the joins between panels, once loose, fraying, and uneven, are now being restitched with firmer tools.

This reform suite offers something the system hasn’t had in years: a coherent, deliberate theory of public sector governance and performance grounded in institutional theory, not just managerial language or progressive nods.

That deserves recognition.

But we should be honest about what’s happening here.

These are not just tidying amendments. They are constitutional reforms by another name. They shift where authority lies, how discretion is exercised, and what legitimacy looks like in the everyday work of governing. They strengthen accountability, if we mean accountability of unelected officials to elected ones. But democracy is more than representative democracy, in the same way that legitimacy is more than having a warrant issued by the Governor General.

These reforms assume good-faith ministers, transparent institutions, and a political culture that respects limits and serves and governs for the benefit of all.

If those assumptions fail, the tighter pattern won’t hold. It will crack.

A well-made tukutuku does not rely on control.

It holds because of balance, tension, yes, but also comfort with conflict and give. That balance is what holds the weave through crisis, change, and time.