top of page
  • Geoff Mulgan


The traditional government department or ministry is organised as a pyramid made up of blocks.  At the top sit the ministers and the top officials. Below them are a series of divisions, and units, cascading downwards and outwards.  Multiple organograms show this structure clearly.

What then flows down and out from the departments are emanations of power: rules, laws, directives, funding programmes and regulations of all kinds.  These then land on the people doing the action in the world: teachers, police, doctors, social workers, businesses big and small, as well as citizens.

This model is so common it is rarely questioned.   I’ve been involved in many versions of it – sometimes guided by top-down strategies, and once again by the language of missions (popular back in the 1990s and now in the 2020s), all of which set the managerial task of ensuring all parts of the system are aligned with the priorities of the top.

But there are alternatives.  An opposite model imagines the department as sitting in the middle of a circle.  Instead of emanating authority, it aims to support the work of the doers.

It regularly asks them – the schools, hospitals, police forces, charities - what they help need with to do their jobs better and tries to provide it.  The things the doers need may include money, knowledge, training, infrastructure and many other things (or just being left alone).  The quality of the department is judged by how useful the doers find it.

These models don’t preclude an interest in performance – but they tend to focus on whether the end-users are happy, or outcomes being achieved, rather than only compliance with rules.

Such models may appear fanciful to anyone brought up in a traditional bureaucracy, or the ideas of new public management.

But they are not unknown.  The Finnish educational service is one example of trying to flip the model– though when they first asked schools how they could be helpful most schools struggled to believe they were serious.  Amazon is in some ways a very tightly controlled hierarchy, but it also constantly asks units what they need from the centre. More networked models like Buurtzorg operate in a cellular way, less like a pyramid, and so more naturally ensure that the central teams are useful to the doers (and a fascinating recent book by Mark Considine, 'The Careless State', showing the disasters of contracting out in Australia, confirms that these more cellular models, which make the most of the commitment at the bottom rather than the targets from the top, often work much better in achieving outcomes).

My hunch is that we need many more inside out departments in the future. Our public services are overburdened with bureaucracy, audit, metrics and the stresses of compliance.  Large central bureaucracies weigh down on them rather than lifting them up.

We regularly see scandals which result from the over-reach of powerful centres – the Post Office which ignored the needs and interests of users and sub-post masters.  The headteachers struggling with Ofsted inspections.

Democratically elected ministers will often have a mandate to shift or reform the system, and the interests of the public won't always align with the interests of the professionals and the doers. So there will always need to be a balance between the top down and the bottom up.

But in many governments there is too much centralisation and too much interference, often unproductive. At a minimum any departments responsible for a large public service should be regularly asking them what an ‘inside out’, or ‘upside down’ perspective would bring.    If they surveyed their users would they report that the centres are useful, provide them vital information and make it easier for them to do their jobs?  How would they far with ‘net promoter’ scores?   And what would the world look like if the telescope was turned around?


bottom of page