Institutions can become remarkably skilled at recognising what is wrong with their present arrangements while remaining surprisingly unable to imagine themselves organised in any fundamentally different way. Officials may describe recurring coordination failures, acknowledge that an inherited classification no longer fits the problems they encounter or recognise that a particular distribution of authority reflects conditions that have long since changed, yet the alternatives they consider remain variations of the architecture they already possess. Processes can be accelerated, reporting lines adjusted, technologies replaced and responsibilities redistributed at the margins, while the deeper assumptions about what the institution is and how it must work remain largely untouched. The difficulty is not necessarily resistance to change. Sometimes the institution has simply lost the cognitive ability to represent itself otherwise.

Recognising that the present could have been different is an important beginning, but it is not the same as being able to imagine a different present. Historical self-knowledge can reveal that a structure once treated as inevitable was actually produced by particular choices and circumstances, thereby moving it from the category of necessity into the category of contingency. Yet contingency creates possibility only in principle. An institution may understand perfectly well that its current arrangement was not historically predetermined and still possess no credible mental representation of what another arrangement might look like. Between recognising that something need not be as it is and conceiving how it might be otherwise lies a distinct cognitive capability: institutional imagination.

The word imagination can sound uncomfortably abstract in administrative contexts, where decisions are expected to rest on evidence, legal authority, operational knowledge and practical constraints. Yet every meaningful institutional change contains an imaginative element because institutions must be able to represent a configuration that does not yet exist before they can evaluate whether moving towards it would be desirable. This does not require fantasy or unconstrained speculation. It can begin with disciplined questions about familiar arrangements: what if authority sat somewhere else, what if this category did not organise the problem, what if these two functions were connected differently, what if information moved directly between actors who currently communicate through several intermediary layers, or what if a responsibility now concentrated in one unit were distributed across a network?

Such questions matter because existing institutions shape not only what people do but what they find easy to imagine. Someone who has spent years inside a particular organisational structure learns its categories, dependencies and practical realities so thoroughly that they can begin to function as the background conditions of thought. The organisational chart becomes more than a description of responsibilities; it becomes a map of plausible action. Existing data classifications become more than technical conventions; they become ways of perceiving the population or problem being governed. Established divisions between policy, operations, technology, analysis and service delivery can gradually become assumptions about how institutional work must necessarily be separated. In this way an institution’s architecture can reproduce itself cognitively before any formal decision to preserve it has been made.

Expertise can reinforce this effect. Deep institutional experience provides indispensable knowledge about constraints, previous failures and the practical consequences of apparently simple reforms, but the same experience can narrow the imaginable future if knowledge of why change is difficult becomes indistinguishable from knowledge that change is impossible. The experienced official often understands precisely why a proposed arrangement would collide with legislation, budgets, information systems, professional boundaries or political realities, and such knowledge protects institutions from naïve reform. Yet if every alternative is evaluated exclusively through the assumptions of the current architecture, institutional knowledge can become an efficient mechanism for reproducing that architecture. The institution knows its present so well that the present begins to define the boundaries of the possible.

Institutional imagination therefore requires a temporary separation between two questions that administrative reasoning often combines too quickly: could the institution be different, and should it become different? The first asks whether another configuration can be represented coherently enough to become thinkable; the second asks whether that configuration would be desirable, legitimate and feasible. Collapsing the questions means that unfamiliar possibilities can be rejected before they have been sufficiently articulated to be examined. Separating them does not imply that feasibility should be ignored, only that feasibility analysis becomes more meaningful once there is an alternative substantial enough to analyse.

Comparative experience can help expand this imaginative range because other institutions provide evidence that functions considered inseparable in one context may be organised differently elsewhere. Another jurisdiction may distribute authority differently, structure relationships between central and local government in another way or use categories that expose dimensions of a problem hidden by domestic classifications. Historical comparison can perform a similar function by revealing that the institution itself has previously operated through arrangements that current members regard as unusual. Neither comparison provides a model to copy automatically. Its value lies partly in demonstrating that familiar arrangements occupy only one region of a larger institutional possibility space.

Counterfactual reasoning can perform the same function without requiring an existing model. Asking what the institution would build if a particular inherited constraint disappeared, or how a capability might be organised if existing departmental boundaries were not taken as the starting point, can expose assumptions that ordinary reform processes leave untouched. The purpose is not to pretend that inherited constraints really have disappeared; they have not. The counterfactual creates analytical distance from them long enough to reveal how much of the proposed future is being derived from necessity and how much is simply being reproduced from familiarity.

This becomes particularly important during technological transformation. New digital systems and artificial intelligence can create genuine possibilities for organising information, expertise and decision support differently, yet institutions often use new technologies to reproduce old architectures more efficiently. A paper workflow becomes a digital workflow, an inherited classification becomes a data schema and an existing chain of approval becomes an automated sequence. These changes may improve performance considerably, but technological novelty should not be confused with institutional imagination. An institution has exercised the latter only when it becomes capable of asking whether the underlying configuration itself remains the most appropriate way to organise cognition and action.

Artificial intelligence may also become an instrument for exploring alternatives, for example by helping analyse scenarios, compare institutional arrangements or reveal patterns that challenge existing categories. Yet generative capacity is not equivalent to institutional imagination either. A system can produce hundreds of organisational possibilities without the institution acquiring any greater ability to understand what those possibilities mean in its legal, political and operational context. Institutional imagination remains a collective cognitive capability because alternatives must become intelligible to actors who understand the institution sufficiently well to recognise what would actually be different.

There is consequently a discipline to imagining institutional difference. Alternatives must be different enough to expose assumptions embedded in the present, but coherent enough to support serious examination. They should make it possible to ask which capabilities would improve, which might be lost, which new dependencies would appear, whose authority would change and what forms of coordination the alternative would require. At this stage, however, the objective is still not to select or implement the future institution. Imagining institutional difference is not designing the future institution. Design requires choices about legitimacy, feasibility, transition, resources, governance and implementation that belong to a later stage of institutional agency.

The distinction is important because imagination loses much of its cognitive value when it is forced prematurely to become a proposal. People become understandably cautious once an idea appears to carry an implicit commitment to reform, and institutional politics can begin before the possibility itself has been properly explored. Creating room to imagine alternatives without treating every alternative as an advocacy position allows institutions to examine their own assumptions with less immediate pressure to defend or overturn them. Some possibilities will prove unrealistic, undesirable or inferior to the existing arrangement, and discovering that is part of the process rather than evidence that imagining them was wasted effort.

An institution capable of imagining itself differently therefore gains something more fundamental than a catalogue of reform ideas. It gains distance from its own present architecture. Historical self-knowledge may have revealed that today’s institution was made rather than given, but institutional imagination allows that recognition to become cognitively productive by giving form to the proposition that it could, under different choices and conditions, be made otherwise. The result is not freedom from constraint and still less a blueprint for redesign. It is an expanded field within which institutional judgement can eventually operate.

Institutional evolution becomes possible when an institution can do more than recognise that its present form was not inevitable. It must also become capable of representing credible forms of difference without immediately reducing them to the assumptions of the system it already knows. To imagine being different is not yet to decide what to become, but without that capacity the future remains cognitively trapped inside the architecture of the present, and institutional change can amount to little more than rearranging a structure whose deeper alternatives have ceased to be visible.