Te Rā Whakamana: The Coherence Paradox

19/08/2025

I take Paul Cairney’s argument that perfect policy coherence is unattainable, set it beside the Performance Improvement Framework evidence and the Land and Water Forum, and argue that what matters is practical outcome coherence between advice and delivery.

Today I turn to Paul Cairney’s (2025) provocation that perfect policy coherence is unattainable and may even be counterproductive. Cairney reminds us that the gap between ideal and reality is a permanent feature of governing rather than a mistake, and at times even a strength. Te Rā Whakamana is a regular series that examines what happens after policy is announced. It draws on peer-reviewed research and writing about the slow, often invisible work of delivery, where systems hold or fracture and where legitimacy is earned (or lost) in context.

Too many policy advisory tools still begin from the fantasy that policy is a box to be packed neatly and closed tight, with priorities stacked and racked, funding instruments matched, capability ready, willing and able, and coherence sealed in.

Tools of this kind are useful as starting points, because they force officials and ministers to think harder about what they are trying to achieve, to expose tensions, and to confront the trade-offs that public sector delivery demands.

They are not end points, however, and treating them as though they resolve complexity and make a good idea ready for delivery is a category mistake. As I have argued throughout this series, what looks tidy on the page rarely survives contact with the contested, relational, and often messy reality of front-line practice, and an announcement is not the same thing as delivery.

Cairney’s (2025) latest provocation is aimed squarely at this preoccupation with coherence in advisory systems. He uses an ideal type to show what perfect advisory coherence would demand: priorities set in order from Cabinet down, evidence integrated across agencies, advice consistent at every turn, and the machinery of government singing from the same score.

He then carefully punctures the illusion. In his account the coherence gap, the distance between aspiration and reality, is a permanent condition of governing rather than a failure. Silos exist for good reasons, to hold expertise and relationships and to secure accountability; conflicts over goals are the substance of politics rather than a distraction from it; and discretion at the edges will never be removed by central strategy. On this reading, chasing coherence across policy advisory systems is impossible and also risks making things worse (Cairney, 2025).

If advisory coherence is a mirage, the question is what holds systems together in practice, and here the empirical evidence from the Performance Improvement Framework (PIF) is instructive.

Across 53 agency reviews, a consistent picture emerges in which performance is neither random nor explained by resources or rhetoric. What matters is internal outcome coherence, the alignment between strategy, outcomes, leadership, financial management, and the operating and delivery models, and it is this internal coherence that tends to drive results.

The correlations are telling. Strategy and effectiveness are significantly linked (0.449), and the relationship holds when financial resources are taken into account. Efficiency and effectiveness move together (0.504), even when leadership effects are removed. Leadership shapes strategy directly and strongly (0.578), and not only through operational channels, while financial management is associated with efficiency (0.508) and service quality independently with efficiency (0.475). I read these as real patterns in how high-performing agencies operate: when strategy and institutional capability line up, performance improves in a systematic way.

The Land and Water Forum is a telling case. Established in 2009 under Minister Nick Smith, it brought together nearly 60 stakeholders from farming, Iwi, science, industry, and environmental groups, in a structured ecosystem of advisory coherence that was far more than a token consultation.

Its independent chair, Alastair Bisley, required adversaries to face each other directly. Over its active period the Forum produced more than 150 recommendations, building what Elizabeth Eppel (2013) described as a form of collaborative governance that removes the middleman and places the onus of finding a way through complexity on the stakeholders themselves. It was advisory coherence at scale, designed and hard-won through repeated negotiation.

It nonetheless failed at the wall of implementation. Fewer than half of the recommendations were taken up (Cheyne, 2017), and those that were adopted often lacked traction because the delivery system had been left outside the room. The regional councils, compliance officers, and community organisations expected to give effect to new water management regimes had played no part in the design. Readers have also reminded me that the government cherry-picked the consensus outputs, which undermined the legitimacy of the process, and that key stakeholders walked away as a result.

The advisory ecosystem was coherent, but it was disconnected from the delivery system, and without that connection its coherence did not survive contact with practice.

This, I think, is the crux of the matter. Cairney shows that perfect advisory coherence is an illusion; the PIF evidence suggests that coherence within outcome systems, linking strategy and outcomes to leadership, finance, efficiency, and service delivery, is what drives performance; and the Land and Water Forum shows that coherence in advice, without coherence with delivery, tends to collapse. Advisory systems can be as sophisticated as you like, but unless they are joined to capacity and ownership in practice they remain brittle.

The lesson is to redefine coherence rather than abandon it, and this matters particularly for elected officials. Accepting that perfect policy coherence is impossible does not let the authorising environment off the hook; if anything, it asks more of it.

The task of the authorising environment shifts from imposing uniformity to something more demanding, which is curating coherence where it counts. That means providing narrative coherence (a clear account of direction), coherence of permission (explicit authority to vary the approach), and coherence of accountability (clarity about outcomes, with flexibility about means).

Most importantly, elected officials need to resist creating incoherence through contradictory directives or by cherry-picking advice, because even the most capable officials cannot compensate for ministers who work against the alignment of the system.

What effective ministers do, in my experience, is champion what I am calling practical outcome coherence: the alignment that emerges when strategy meets delivery and advice connects with ownership, and when that connection is strong enough to hold under pressure. These ministers make sure delivery agents are in the room from the first day, asking “what can we actually make work?” as well as “what should we do?” The Land and Water Forum failed in large part because the authorising environment lacked that discipline; that stakeholders reached agreement in isolation from those who would have to implement it is a failure of the authorising environment, not of the stakeholders.

Practical outcome coherence goes further than advisory coherence. It is about building systems in which the capacity to advise and the capacity to act are joined from the start, and in which those who must deliver take part in shaping what they will deliver.

Pragmatism may glue things together for a time, as Palmer (2007) argued, but the practical state requires something more deliberate: advisory ecosystems connected to implementation systems, where legitimacy is earned in practice rather than assumed at the centre. Ministers are not short of blueprints; what they want, as my doctoral research has found, is free and frank advice that tells them whether those blueprints will hold once they reach the ground.

The task, then, is to join advice and delivery in ways that keep trust, context, and ownership together, rather than to eliminate fragmentation or force advisory systems into an artificial unity, because coherence that exists only on paper is not coherence at all.

References

Cairney, P. (2025). Why perfect policy coherence is unattainable (and may be ill-advised). Policy Sciences, 58(3), 619–642. https://doi.org/10.1007/s11077-025-09582-9

Cheyne, C. (2017, June 28). Can collaboration trump adversarial environmental planning? Insights from New Zealand’s Land and Water Forum and proposed statutory collaborative planning process [Paper presentation]. International Conference on Public Policy, Panel T02P09, Collaborative Governance and Deliberative Policymaking in Comparative Perspective, Singapore.

Eppel, E. (2013). Collaborative governance case studies: The Land and Water Forum (Working Paper No. 13/05). Institute for Governance and Policy Studies, Victoria University of Wellington.

Palmer, M. (2007). New Zealand’s constitutional culture. New Zealand Universities Law Review, 22, 565–597.