Administrative Burden: The Foundational Text

20/02/2026

I go back to the founding 2015 article by Moynihan, Herd, and Harvey, set out its three costs and its argument that burden can be a political choice, and propose a fourth cost, the burden of impermanence.

This is the second post in a series examining the real costs of dealing with the state in Aotearoa. Last month, I introduced the administrative burden framework developed over the past decade by American scholars Pamela Herd and Donald Moynihan, and asked whether the Ministry for Regulation’s focus on “cutting compliance costs” is the right frame for the problem of compliance costs. This week, we go back to the founding text: the 2015 article by Moynihan, Herd, and Hope Harvey that started the field. I sit with its most unsettling argument, that the friction citizens experience when dealing with the state is sometimes a deliberate choice.

In this post, I want to do three things. First, I want to unpack the three-part typology at the heart of the 2015 article (learning costs, compliance costs, and psychological costs) and, in doing so, show why distinguishing between them matters. Second, I want to sit with the article’s most challenging argument: that administrative burden, which is sometimes a design failure, can also function as a deliberate political strategy and policy choice. Third, I want to propose a fourth cost that the framework does not name, one I am hearing about constantly from clients in Aotearoa right now: the burden of impermanence, of a state that keeps forgetting who you are because it keeps restructuring itself.

Three costs, carefully distinguished

The argument of the 2015 article is that when citizens interact with the state, they bear three distinct types of cost, and that failing to see all three produces incomplete analysis and, ultimately, bad policy.

The first is learning costs. Before any compliance can occur, a citizen must establish that a service or obligation even exists, whether they are eligible, and what the rules of engagement require.

The state may offer a benefit in good faith, but if the eligibility criteria are so complex that working out whether you qualify requires hours of research, or money for professional advice, then the benefit is, for a significant portion of its intended population, practically inaccessible.

In Aotearoa, one might think of the supplementary benefits available through Work and Income. The information is technically public, but for a person in crisis, without reliable internet access or the energy to navigate the website, the effort of working out which of the dozens of available benefits applies represents a cost that never appears in any impact assessment.

The second category, compliance costs, covers what we more traditionally call “red tape”: the form-filling, the document-gathering, the transport to a government office, the half-day of unpaid leave to attend an appointment. These costs tend to be regressive in their structure. A fifty-dollar fee for a certified document is an inconvenience for a professional household with savings and flexible employment, but for a family living from one pay to the next it can be an insurmountable barrier. The cost functions as a gatekeeper, quietly determining who gets access and who does not.

The third category is where the framework becomes genuinely illuminating. Psychological costs are the stress, the frustration, the stigma, and the erosion of dignity that accompany these processes: the anxiety that a single error might have serious consequences, the experience of being asked, repeatedly, to prove that one is deserving, and the feeling of a system that treats you as a number, or that seems, in its very design, to be inviting you to give up. The effect is more than unpleasant, because psychological costs deter eligible people from applying for help to which they are entitled, and they erode trust in government over the long run.

What the three-part typology reveals is that a process might have relatively low compliance costs (a short online form, for example) while simultaneously creating serious barriers through high learning costs and significant psychological burden. Burden is multi-dimensional, and each dimension has its own logic.

Burden as strategy

The second major contribution of the article is its argument that burden, which is often treated as a technical accident, can function as a deliberate political and policy tool.

The authors illustrate this with Medicaid policy across three governorships in Wisconsin. What makes the case striking is that it is more than a simple partisan back-and-forth. Both Tommy Thompson, a Republican, and Jim Doyle, a Democrat, reduced administrative burdens and expanded access to the programme. When Scott Walker took office that changed; the eligibility rules stayed the same, while the administrative environment around them shifted. Walker’s Commission on Waste, Fraud and Abuse systematically reframed the burden-reduction work of his predecessors as a threat to programme integrity, and set about reversing it. The changes were deliberate, and they were effective.

The authors called the changes “hidden politics.” A legislative proposal to cut a benefit would be visible, contested, and politically costly, whereas a quiet administrative decision to add a form, require an additional document, or close a local office can achieve substantially the same result with far less scrutiny. In effect, the burden becomes the policy, because access has been made difficult.

A fourth cost? The burden of impermanence

The framework was developed in the United States context, and it describes something real. But sitting with it over recent months, and listening to local corporate clients, I find myself wondering whether there is a fourth cost it does not fully name, one that is particularly acute right now.

I call it the burden of impermanence, or the cost of restructures to citizens.

The case manager who understood your situation is no longer in that role. The account manager who spent eighteen months learning the technical architecture of your regulatory environment has taken a redundancy. The policy analyst who had genuinely absorbed the science underpinning your consenting framework has moved to another sector, taking with her two years of accumulated context that was never written down anywhere. And you, the citizen, the small business, the regulated party, must now begin again. You must re-establish the shared understanding that made the relationship functional, and re-educate the state about itself and its role.

This is unlike a learning cost in Moynihan, Herd, and Harvey’s sense, because the knowledge already existed; it was simply attached to a person, and that person has gone. It differs from a compliance cost, because no new rule has been introduced. And it is only partly a psychological cost, though the exhaustion and quiet despair it produces are real enough. In my view it is the cost of the state’s own institutional amnesia, transferred wholesale onto the person who had the misfortune of investing in a working relationship with it.

The scale of restructuring across the public service over the past few decades has been substantial. When an organisation loses experienced staff in large numbers over a short period, it loses more than capacity: it loses the informal knowledge that never made it into a briefing note, the detailed understanding of how a sector operates and the accumulated professional trust that reduces friction on both sides of the counter. That loss is borne, in the first instance, by remaining staff. It is also borne, repeatedly and invisibly, by the people and organisations who had built their engagement strategies around relationships and knowledge that no longer exist.

The Herd and Moynihan framework asks us to see burden as a political choice. This fourth dimension asks us to see institutional continuity as a public good, one with a real cost when it is depleted, even if that cost never appears in a redundancy calculation. I will return to it later in this series.

What it all means for us

The Ministry for Regulation’s focus on “compliance costs” is right as far as it goes, but it is incomplete. If we only measure the time it takes to fill in a form, we miss the time it took to find the form, we miss entirely the psychological toll of the process, and we miss how long it takes for the form to make a difference because the account manager has taken redundancy. The Treasury framework, from which so much of our regulatory impact analysis still draws, has the same blind spot: it is designed to make the costs to agencies visible, and it leaves the costs to citizens out of view.

And if we do not see burden as a political variable and a policy choice, we risk being naïve about what is actually happening when a process becomes more difficult. Is it really just poor design, or is someone, somewhere, making a quiet choice about who should have easy access and who should not?

The 2015 article gives us the tools to ask those questions: to look at a government process and ask “how much does it cost the agency to run?” and also “who bears the learning, compliance, and psychological costs, and is that equitable?” Most importantly, it lets us ask whether the level of burden we see is the result of neglect or the result of a choice.

These questions are uncomfortable, but they are the right ones.

Next month, we turn to the evidence on distribution. Administrative burdens fall unequally, and the research is unequivocal about where they fall hardest and why.

References

Moynihan, D., Herd, P., & Harvey, H. (2015). Administrative burden: Learning, psychological, and compliance costs in citizen-state interactions. Journal of Public Administration Research and Theory, 25(1), 43–69. https://doi.org/10.1093/jopart/muu009