Governments are expected to make decisions about societies they can never observe completely from inside their own organisations. They collect administrative data, commission research, employ specialists, maintain statistical systems and accumulate enormous quantities of institutional experience. Modern public administrations may consequently possess forms of knowledge that few other organisations can reproduce. Yet the existence of this capacity can encourage a misleading assumption: that the knowledge required for governing is, or eventually could be, contained within government itself. In reality, much of what public institutions need to understand is generated elsewhere.
This is partly a consequence of scale. Governments act across economies, communities, infrastructures, ecosystems and social systems whose complexity exceeds the observational capacity of any single organisation. Relevant knowledge is produced continuously by businesses, universities, professional communities, civil-society organisations, local associations, service providers and citizens themselves. Some of it is formal and highly codified; some exists as professional expertise, operational experience or situated understanding. The important point is not merely that government lacks information. It is that GOVERNANCE-RELEVANT KNOWLEDGE CAN EXIST OUTSIDE GOVERNMENT.
That distinction changes how we understand institutional cognition. If the problem were simply incomplete governmental information, the obvious response would be to collect more data until the gap became smaller. But knowledge outside government is not always waiting to be extracted and transferred into an administrative database. Some knowledge exists because external actors occupy positions government does not occupy, experience consequences government does not experience or participate in systems from perspectives that public institutions cannot reproduce internally.
A transport authority, for example, may possess detailed data about routes, capacity and service performance while passengers know where formal reliability statistics diverge from the practical experience of making a journey. A health ministry may understand national epidemiological patterns while local organisations observe barriers to care that have not yet appeared in administrative records. A regulator may possess extensive industry reporting while engineers, workers, consumers or independent researchers recognise emerging risks through forms of experience that the reporting architecture does not capture. None of these perspectives provides a complete account, but each may contain knowledge relevant to public judgement.
This does not mean that external knowledge is inherently better than governmental knowledge. OUTSIDE GOVERNMENT ≠ EPISTEMICALLY SUPERIOR. External actors have their own interests, blind spots, incentives and limited perspectives. Firms may possess valuable technical information while also having reasons to present it selectively. Citizens may report genuine experiences without being able to observe system-wide patterns. Local organisations may understand particular communities exceptionally well while lacking comparative evidence. Academic research can offer analytical depth while remaining distant from implementation conditions. The cognitive value of external knowledge therefore depends on what it reveals, how it was produced and how it relates to other evidence.
The argument is instead about cognitive position. Different actors see different parts of reality because they occupy different relationships to it. Government’s institutional position gives it unique forms of visibility: access to administrative records, authority across jurisdictions, long-term continuity and the capacity to observe patterns at population scale. Other actors possess different forms of visibility precisely because they stand elsewhere. The challenge of governance is not to decide which position owns the truth, but to build sufficient cognitive architecture for relevant perspectives to become available to public judgement.
This matters because organisational boundaries can easily become cognitive boundaries. Institutions naturally invest in systems for circulating knowledge internally: reporting structures, databases, analytical teams, meetings, briefings and decision processes. Over time, what travels through those structures can come to define what the organisation treats as knowable. Information produced outside them may appear anecdotal, irregular or difficult to verify, even when it captures aspects of reality that internal systems systematically miss.
Administrative knowledge is especially vulnerable to this problem because public institutions often know society through the categories required to administer it. People become cases, taxpayers, patients, students, claimants, businesses or service users according to the function being performed. These categories are necessary for government to operate, but they do not exhaust the reality of the people and systems being governed. Knowledge generated outside government can therefore matter not simply because it adds more observations, but because it can reveal aspects of a problem that administrative categories make difficult to see.
The distinction between ACCESSING EXTERNAL KNOWLEDGE and ABSORBING EXTERNAL KNOWLEDGE is important here. Government does not need to internalise every relevant source of expertise or experience. Attempting to do so would be impossible and, in some cases, counterproductive. Knowledge may depend upon the independence, professional position or local embeddedness of the actor that produces it. Bringing every source inside the administrative hierarchy could alter the conditions that made the knowledge valuable.
What government needs is cognitive reach rather than cognitive ownership. It must be able to encounter, evaluate and use relevant knowledge without assuming that all valuable knowledge must first become governmental knowledge. This requires a different image of institutional intelligence: not a self-contained repository accumulating everything it needs to know, but a system capable of reasoning across a wider epistemic environment.
The challenge becomes particularly visible during periods of rapid change. Emerging technologies, new economic practices, social transformations and novel risks often develop faster than formal governmental knowledge systems can adapt. Expertise may initially concentrate in research communities, technical firms, professional networks or affected populations. Waiting until all relevant understanding has been translated into established administrative structures can leave public institutions cognitively behind the systems they are expected to govern.
Artificial intelligence offers a clear example. Governments increasingly need to make decisions about technologies whose technical development, operational deployment and social consequences are distributed across companies, universities, standards bodies, public agencies, professional communities and users. Public institutions can and should build internal expertise, but internal capability alone cannot reproduce the entire knowledge environment surrounding such a fast-moving field. Governing effectively requires the ability to reach beyond institutional boundaries while retaining independent public judgement.
That final qualification is essential. Recognising external knowledge does not mean transferring public authority to whoever possesses expertise. KNOWLEDGE DISTRIBUTION ≠ AUTHORITY DISTRIBUTION. Democratic institutions remain responsible for decisions that involve legitimacy, rights, public values and collective priorities. A technology company may know more about the architecture of a system it has built; that does not give it the authority to determine acceptable social risk. A local group may understand community consequences better than a central administration; that does not automatically resolve conflicts among communities or national obligations.
Knowledge and authority therefore need not occupy the same place. This is not a defect to be eliminated but a condition to be governed. Public authority can remain public while the cognition supporting it draws on knowledge distributed across society.
The same principle protects against another mistake: equating consultation with cognition. An institution may invite external actors to meetings, publish calls for evidence or conduct public consultations without allowing what it learns to modify its understanding of the problem. External participation can become procedural while the underlying cognitive model remains unchanged. The relevant question is not simply whether outside voices were heard, but whether relevant external knowledge was able to become consequential within public reasoning.
This does not imply that every claim should alter policy. Institutional cognition requires evaluation as well as openness. External evidence may be unreliable, strategically presented, unrepresentative or incompatible with stronger evidence elsewhere. Cognitive reach is not epistemic surrender. The capability lies in maintaining permeable boundaries through which knowledge can enter while preserving the institutional capacity to assess its relevance and validity.
There is also a temporal dimension. Governments may initially know less about an emerging problem than actors closer to it, but their comparative and longitudinal capacities can become increasingly valuable as evidence accumulates. Conversely, institutional knowledge that was once adequate can become outdated while external actors encounter new conditions first. The boundary between internal and external knowledge therefore shifts over time. A cognitively capable government must be able to recognise when important understanding is developing beyond its existing observational architecture.
This suggests that the appropriate unit of analysis for public-sector intelligence cannot always be the individual organisation. A ministry may make a decision, but the knowledge on which good judgement depends may originate across an ecosystem of actors. The institutional question then becomes how government remains capable of public reasoning when cognition relevant to that reasoning is distributed beyond its formal boundaries.
The answer cannot simply be “collect everything”. No government can contain the full knowledge of the society it governs, nor should it attempt to. The more important capability is knowing that its own cognitive boundary is incomplete and designing public decision processes accordingly. This means recognising where external knowledge may exist, understanding why certain actors can see things government cannot, creating routes through which relevant knowledge can become visible, and retaining the capacity to evaluate that knowledge without confusing expertise with authority.
Government does not hold all the relevant knowledge because the world it governs is cognitively larger than the organisations through which it governs. Public institutions possess indispensable forms of knowledge, but so do actors whose positions, experiences and expertise lie beyond governmental boundaries. Institutional intelligence therefore cannot depend on cognitive self-sufficiency. It depends on a government’s ability to reach beyond itself for knowledge while preserving the public authority, critical judgement and accountability through which that knowledge is ultimately used.