Te Rā Whakamana: The Story So Far

07/04/2026

I recap the series so far, from fifty years of implementation research through the politics of evidence, the fault line between top-down and bottom-up, and street-level policy entrepreneurs to the interpretive hinge, and I name what the hinge cannot yet do.

This is a recap. If you are new to Te Rā Whakamana, it will bring you up to speed, and if you have been here from the beginning, it will remind you where we have been and why; by the end, you will have the shape of the argument so far and a clear sense of where we are going. The series has covered five core ideas: fifty years of implementation research and why we ignore most of it; what “evidence-based” actually means when you press it; the fault line between top-down control and bottom-up reality; the frontline workers who are making policy, whether we admit it or not; and why an interpretive approach to all of this is a necessary act rather than an academic indulgence, with a working concept I have been calling the interpretive hinge, which I consolidated in a recent side quest. The argument running beneath it all has stayed the same. In my view our problem lies less in our politics than in the way our policy advisory systems and our delivery systems have come apart, and we have, for the most part, forgotten how to put them back together.

For the past few months I have been working through one argument, which is that we keep designing policies that look brilliant on paper and fall apart in practice, and we keep acting surprised. We restructure, we write better instructions and we demand tighter compliance, and then we are surprised again.

My argument is that this is mainly a systems problem, and only secondarily a political one. Our policy advice and our delivery have become disconnected, and somewhere along the way we forgot how to implement effectively, not all the time, but often enough for everyone to notice. When delivery does work, it tends to do so because someone on the front line is carrying more than they should have to, quietly bending rules and solving problems the system never authorised them to solve.

The series is a wander through the implementation and policy advisory systems literature to see what it reveals, what it confirms and where it falls short, particularly in Aotearoa, where Te Tiriti o Waitangi brings a constitutional dimension that the institutional and managerial scholarship does not address.

We have been arguing about this for fifty years

We started with fifty years of implementation research, and Sætren’s (2024) intellectual history set out three generations of thinking. The rationalist, top-down first generation of the 1970s treated failure as a problem of insufficient control, while the bottom-up second generation of the 1980s gave us Lipsky’s (1980) street-level bureaucrats and the insight that policy is what is enacted in thousands of daily micro-decisions, whatever is written in law. The hybrid third generation, from the 1990s onward, accepted that implementation sits somewhere between pure control and pure chaos, as a matter of continuous negotiation.

My reflection on that history was that public discourse in Aotearoa remains largely caught in first-generation assumptions, reaching for levers, demanding alignment, and expressing ministerial surprise when elegant plans meet messy reality, despite five decades of scholarship showing why the control model does not work. I do not offer that as a counsel of despair; it is an observation about where the opportunity lies.

What Sætren’s fifty years teach us, in my view, is that there are no simple answers to why policies fail or succeed, and that implementation is a messy, contextual and often inexplicable craft. The challenge is to build the institutional capacity to work with that messiness instead of trying to resolve it, and that capacity is what I was beginning to call the interpretive hinge: the capacity to translate between the world of policy advice and the reality of delivery, an idea the series has been developing ever since.

Evidence-based according to whom?

The next post pressed on the politics of evidence through a deceptively simple question from Gade (2024): when is it actually justified to claim that a policy is evidence-based? The question matters because “evidence-based” has become one of those phrases that does more work as a rhetorical move than as a description of epistemic practice. It is invoked to close down debate, to signal seriousness, and to distinguish oneself from those who proceed on mere ideology or intuition. Gade’s paper slows that claim down and asks what, precisely, everyone is claiming when they reach for the authority that their advice is evidence-based.

Gade’s answer is that a claim is justified only if three conditions are met. First, the claimant must possess comparative evidence, that is, evidence about the effects of the proposed policy relative to at least one alternative; “This works” is not an evidence-based claim, while “This works better than that, on this dimension” might be. Second, the policy must be supported by that comparative evidence according to at least one of the claimant’s preferences. This is where the argument turns quietly radical, because whether evidence supports a policy stops being a purely empirical question and comes to depend on what you are trying to achieve, which dimensions of effect you care about, and what you are willing to trade off against what. Third, anyone claiming their option is evidence-based must be able to show their workings, explaining both the evidence and the preferences that together support the claim, and “The research shows” without further elaboration does not meet the threshold.

Those three questions (compared to what alternatives, according to which preferences, and can you show your workings) matter enormously for policy advisory systems. When an official tells a minister that a policy is evidence-based, Gade’s framework suggests they should be able to answer all three, and if they cannot, the claim is unjustified because its epistemic warrant has not been established, whether or not the policy itself is wrong. If that is the test for an official, it ought to be the same test for a think-tank, a consultancy, a lobbyist or an academic.

The Wai 4 Kaituna River case illustrated what happens when those conditions are never applied. The Ministry of Works spent twenty years insisting its technical solution was neutral, and its engineers had flow rates, nutrient loads and cost-benefit analyses, but they lacked any framework for weighing spiritual contamination, the severance of whakapapa relationships, or the loss of traditional gathering sites. When Ngāti Pikiao witnesses brought mātauranga Māori to the Tribunal, they were providing evidence rather than alternative opinions, read through different preferences about which dimensions of harm mattered. The Ministry’s claim to technical neutrality collapsed once its preference set was made visible, and when the Tribunal required consideration of the whole picture, the culturally appropriate solution proved cheaper and more effective across every metric. The evidence-based pipeline turned out to be expensive, outdated and destructive, and as I read the case the reason was that the Ministry had never been required to show its working on whose evidence counted and why, since it had no shortage of evidence.

That question, whose evidence counts, pointed toward the structural fault line that the next post took up.

The view from the top versus the view from the ground

Where Gade pressed on the politics of evidence, Buzogány and Pülzl (2024) helped us understand the structural fault line that runs through the implementation literature itself. In my view the reason we remain stuck in first-generation thinking has less to do with ignorance than with the fact that the centre and the front line start from fundamentally different places, as well as operating at different levels of the same system. Where one begins one’s analysis determines what one sees, what one values and what solutions one proposes. The centre sees a set of ministerial instructions to be followed, while the front line sees a problem to be solved in a community that does not care much for politics or political rhetoric and wants the system to work. Both are right in their own terms, and that is precisely what makes the tension so persistent, because it cannot be resolved by pulling harder on the same lever and has to be understood first.

Two Aotearoa cases illustrated the difference in practice. The Treasury’s Living Standards Framework, rational, comprehensive and conceived entirely from the centre, produced the performative compliance that bottom-up theorists would predict (Treasury, 2019): agencies learned to speak the language of living standards in reporting documents while continuing to operate according to their own professional and agency logics. The Māori Communities COVID-19 Fund did the opposite, devolving trust, resources and discretion to Iwi, Hapū and Māori community providers, and it succeeded precisely because it abandoned the top-down impulse (Te Kawa & MacDonald, 2023). The fund trusted communities to know what to do instead of telling them.

That post went further than the international literature is prepared to go. The classic debate assumes the bottom consists of public servants operating within a single sovereign order, and in Aotearoa that assumption fails, because Iwi me Hapū stand beside the Crown’s hierarchy rather than at its bottom, as parties to a constitutional compact that the implementation literature was never designed to comprehend. Nor is it only Iwi me Hapū, since a significant proportion of public outcomes in Aotearoa are delivered by community organisations, third parties, private contractors and arm’s-length entities, none of whom fit neatly within the sovereign hierarchy assumed by the classic top-down discourse. The bottom of that hierarchy and the front line are much more complex and much more contested than the literature recognises.

This raised the question the series has been sitting with: if our policy advisory systems have little understanding of how this arm’s-length implementation actually works across these varied delivery arrangements, who is doing the actual work?

The person behind the counter is also making policy

The next post took up that question directly through Cohen (2024), who builds on Lipsky’s (1980) original insight and takes it further. The street-level bureaucrat in Lipsky’s account is the frontline worker facing impossible demands (overwhelming caseloads, ambiguous rules, inadequate resources) who develops informal routines and shortcuts to get through the day. The argument Cohen makes is that these workers are innovating as well as coping: they identify problems invisible to policymakers, devise solutions that work in their specific contexts, and build informal networks to spread new practices to their colleagues, and in Cohen’s framing they are policy entrepreneurs.

The examples are familiar: the Work and Income case manager who finds an obscure provision to help a struggling family, the hospital social worker who builds an informal network of community organisations for patients being discharged without support, and the social worker who uses a GoFundMe so that no child she comes across goes without nappies. I read these as acts of street-level policy entrepreneurship that the system simultaneously depends on and refuses to legitimise, more than as charming anecdotes about pragmatism.

That is the problem Cohen puts his finger on: we celebrate the outcomes when they succeed but implicitly condemn the rule-bending that was necessary to achieve them, and we rely on frontline innovation to function without authorising it, investing in it, or giving frontline workers explicit permission to experiment. Our best frontline innovators are left to operate in the shadows, taking personal risks and developing workarounds that run parallel to the official system, and they are vulnerable to being shut down by managers who put compliance ahead of outcomes.

The COVID-19 response made this visible in ways the system found uncomfortable. Public health nurses developed new ways to reach vulnerable communities, social workers created virtual support networks for isolated clients, and teachers pioneered remote learning approaches that were later adopted system-wide, none of it planned or mandated from the centre. The response of the system was telling, because some innovations were celebrated and scaled while many others were ignored or actively discouraged, and the default assumption was that any deviation from official procedures was potentially problematic, even when it was clearly producing better results.

This is the accountability problem at the heart of the series. We elect ministers and hold them responsible for outcomes they cannot directly control, delivered by people they have never met, in contexts they do not understand, and the traditional response is to eliminate discretion through more detailed rules and tighter supervision. Cohen shows why that is both impossible and counterproductive: no set of rules can anticipate every situation a frontline worker will encounter, and trying to eliminate discretion prevents exactly the kind of adaptive, responsive delivery that citizens need and that makes government effective.

The post closed by signalling that frontline workers may already be doing some of the work of the interpretive hinge without being the hinge themselves. That distinction needed to be drawn carefully, and it pointed to the next post in the series.

The interpretive hinge unpacked

The next post took the working concept and gave it a clearer shape. Drawing on Bevir and Rhodes (2003), it argued that there are no brute facts, because every piece of evidence is seen through somebody’s lens, and the same water quality data tells completely different stories depending on whether you are a farmer, an environmental scientist, an Iwi kaitiaki or a recreational user.

That insight led me to reread two cases the series had been carrying. The Land and Water Forum spent seven years building sophisticated consensus across nearly sixty stakeholders and produced over 150 recommendations, fewer than half of which were adopted, and the Forum collapsed. The advice was not the problem, in my reading; the delivery systems had never been in the room, and when the recommendations arrived they made no sense given the frames of interpretation that regional councils, compliance officers and enforcement agencies were operating from (Eppel, 2013, 2014). The leaky homes crisis followed the same logic, because the performance-based building code had no feedback loops to detect how councils and builders were interpreting the new regime until homes began to rot (Gill, 2016).

The Māori Communities COVID-19 Fund showed what the alternative looked like. Instead of imposing an interpretive framework, it authorised different actors to act on their own understanding of what their communities needed. Our evaluation found that the dominant narrative was about manaakitanga more than about hitting targets, though the fund exceeded them: getting a foot in the door of some whare and rebuilding trust where mainstream systems had failed, as a precursor to a kōrero about vaccination (Te Kawa & MacDonald, 2023). The interpretive diversity was the strength of the fund rather than its problem.

The post then brought the working concept of the interpretive hinge into clearer view. I describe the hinge as the space in the machinery of government where policy advice and delivery meet and where the work of connecting actually happens, which is why I resist treating it as a stage, a process or a person. Three strands of work happen inside it: translation between the strategic intent of the centre and the operational reality of the front line; sense-making in the face of contested evidence and competing preferences; and the relational work that keeps the system honest with the people it is meant to serve. When the hinge is built well and kept in working order, the connection holds, and when it is poorly built, or left to wear out, the connection fails.

The post anchored the concept in the scholarship the series has been drawing on, taking its interpretive footing from Bevir and Rhodes (2003), its language of meaning-making from Yanow (1996) and its account of situated agency from Aprimadya (2025). It also drew on a recent application of the Policy Capacity Framework to primary health care in Aotearoa, which found that operational capacity, understood as a creative force generating the ideas and practices that make implementation work rather than narrowly as implementing directives, is the most important driver of progress and the one most often missing (Tenbensel et al., 2026). Operational capacity is, I think, part of what the hinge depends on.

The post also left one matter deliberately open. The people doing the work of the hinge do not fit neatly into the categories the literature provides, since they are not street-level bureaucrats in the traditional sense, nor policy entrepreneurs, nor boundary spanners, nor ministerial-facing policy advisors. In a forthcoming article I argue that they are something else, and I call them the weavers. The image is deliberate: weaving brings three strands together without dissolving any of them, and it is patient, relational work that does not assume the strands begin in the same place, only that they can be brought together by hands that know how to hold all three. The article deals with the people, and the series will develop the concept further.

What the hinge cannot yet do

The working definition has its limits, and they are worth naming before the next post in the series. The hinge concept, as I have developed it so far, assumes that the parties doing the translating, sense-making and relating share enough of a frame for the work to be possible, and the Land and Water Forum case shows the limit of that assumption, as does the Kaituna case. The frame the centre carries is not always continuous with the frame the community holds. When the frames are commensurable, weaving can be done; when they are not, the weaver is left holding strands that do not belong to the same cloth.

Critical theory has been making this point for decades, and decolonial, Indigenous and feminist scholars have been making it for longer than the implementation literature has existed. The next post in the series steps back from the practical question of who is doing the work and asks what those traditions have to tell us about implementation itself. Implementation is never neutral, because the frame the system carries with it has politics built in, and the questions follow from that: whose knowledge counts, whose stories are heard, whose authority is recognised, and whose power is protected by claims to neutrality? Recognising that is, I think, the first step toward understanding what the weavers actually have to weave.

References

Aprimadya, M. H. (2025). Rethinking situated agency: An interpretive framework to policy implementation. Policy Studies. Advance online publication. https://doi.org/10.1080/01442872.2025.2547859

Bevir, M., & Rhodes, R. A. W. (2003). Interpreting British governance. Routledge.

Buzogány, A., & Pülzl, H. (2024). Top-down and bottom-up implementation. In F. Sager, C. Mavrot, & L. R. Keiser (Eds.), Handbook of public policy implementation (pp. 116–126). Edward Elgar.

Cohen, N. (2024). Moving beyond traditional policy implementation: Street-level bureaucrats’ policy entrepreneurship. In F. Sager, C. Mavrot, & L. R. Keiser (Eds.), Handbook of public policy implementation (pp. 148–155). Edward Elgar.

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.

Eppel, E. (2014). Improving New Zealand water governance: Challenges and recommendations. Policy Quarterly, 10(3), 66–73. https://doi.org/10.26686/pq.v10i3.4501

Gade, C. (2024). When is it justified to claim that a practice or policy is evidence-based? Reflections on evidence and preferences. Evidence & Policy, 20(2), 244–253. https://doi.org/10.1332/174426421X16905606522863

Gill, D. (2016). Regulatory coherence: The case of New Zealand. In D. Gill & P. Intal Jr. (Eds.), The development of regulatory management systems in East Asia: Country studies (pp. 175–227). Economic Research Institute for ASEAN and East Asia.

Lipsky, M. (1980). Street-level bureaucracy: Dilemmas of the individual in public services. Russell Sage Foundation.

Sætren, H. (2024). An intellectual history and state-of-the-art assessment of policy implementation research and theory. In F. Sager, C. Mavrot, & L. R. Keiser (Eds.), Handbook of public policy implementation (pp. 12–31). Edward Elgar.

Te Kawa, D., & MacDonald, L. (2023). Independent evaluation: Māori Communities COVID-19 Fund. DTK and Associates for Te Puni Kōkiri.

Tenbensel, T., Middleton, L., Silwal, P., & Cumming, J. (2026). Using the policy capacity framework to evaluate New Zealand’s primary health care policy implementation. Public Administration and Policy, 29(1), 36–49. https://doi.org/10.1108/PAP-11-2024-0183

Treasury. (2019). The Living Standards Framework 2019. New Zealand Treasury. Yanow, D. (1996). How does a policy mean? Interpreting policy and organizational actions. Georgetown University Press.