5th September 2026
John Swinney has set out perhaps the most ambitious programme of public-sector restructuring Scotland has seen for many years.
The direction of travel is clear.
Fewer organisations. Fewer layers of management. Fewer duplicated functions. A smaller workforce. More shared services. Greater use of technology and artificial intelligence. And, ultimately, a public sector that costs less to run while delivering better services.
On paper, it is difficult to argue with the basic objective.
Scotland has more than 130 public bodies. The Scottish Government itself says that some have overlapping or duplicated functions and that the system can be difficult for the public to navigate.
But there is an important question which needs to be asked before anyone starts counting the projected savings.
How much will it cost to make all these changes?
Because public-sector restructuring is rarely free.
Indeed, the biggest danger with major reorganisations is that governments concentrate on the eventual saving while underestimating the cost of getting there.
And those costs can arrive long before the savings.
The scale of what is being proposed
Swinney's Programme for Government is not simply about trimming a few departments.
The Government is looking at the structure of Scotland's public services from top to bottom.
The NHS is perhaps the most dramatic example.
The existing 14 territorial health boards are to be replaced by two strategic health boards, alongside wider changes intended to alter how healthcare is planned and delivered.
Local government is also being reconsidered.
The Government has announced a three-month dialogue about the future shape of local governance, including whether resources could be pooled regionally while strengthening community-level decision-making. It is also looking at how social care can be reorganised to achieve economies of scale.
Then there is the wider public-body landscape.
More than 130 organisations are potentially within the Government's sights for simplification, merger, closer working or other changes.
The environmental sector is another example. Plans have emerged to bring Scotland's four national environmental bodies together into a single overarching organisation.
Transport is also being reorganised, with proposals involving Transport Scotland and the creation of a centralised ferries body.
This is not one reform.
It is a programme of reforms taking place across government and public services at roughly the same time.
That makes the potential long-term savings significant.
It also makes the potential short-term bill significant.
The first problem is people
The Government says the Scottish Government itself is already becoming smaller, with reductions achieved largely through natural attrition and tight recruitment controls.
Its five-year civil service plan says further options could include a voluntary exit scheme, subject to ministerial approval and engagement with trade unions.
That sounds relatively painless.
But reducing a workforce still has costs.
Voluntary redundancy payments cost money.
Early retirement packages cost money.
Pension arrangements have to be dealt with.
Employment lawyers may be required.
HR departments need additional resources.
Managers have to spend time reorganising teams.
People remaining in the organisation may need retraining.
And during a major reorganisation there can be a temporary fall in productivity as employees learn new responsibilities and systems.
None of that necessarily means the reform is a bad idea.
It simply means that the saving does not arrive on day one.
If a department eventually saves £20 million a year but has to spend £15 million getting there, the economics are rather different from announcing a £20 million saving.
And that £15 million is only worthwhile if the £20 million annual saving actually materialises.
Then there is IT
This could become one of the biggest hidden costs.
Governments increasingly talk about digital transformation and artificial intelligence as ways of reducing administrative costs.
Swinney's own civil service five-year plan says the Scottish Government intends to expand the use of AI and automation and have these capabilities embedded in its operations by 2030. It also plans new systems for correspondence, parliamentary questions, Freedom of Information, HR and IT support.
There is considerable potential here.
A well-designed digital system can eliminate repetitive work, reduce paperwork and allow staff to concentrate on tasks that genuinely require human judgement.
But replacing an old computer system is rarely cheap.
Now imagine replacing systems across organisations that are being merged.
Different databases.
Different software.
Different procurement contracts.
Different security arrangements.
Different payroll systems.
Different records.
Different working practices.
All of them have to be made to work together.
The NHS provides an obvious example of the potential complexity.
Merging organisations isn't simply a matter of changing the name above the door.
It means integrating thousands of employees, budgets, contracts, IT systems, buildings, procurement arrangements and management structures.
Buildings are another cost
A smaller public sector should eventually require fewer buildings.
That could produce substantial savings.
But getting there may involve significant expenditure.
Leases have to be terminated.
Buildings may have to be sold.
Properties may need refurbishment before staff can be moved into them.
IT and communications systems have to be installed.
Offices may have to be redesigned.
Some buildings may have little market value if a large public-sector employer withdraws from a town.
This is particularly relevant in rural Scotland.
Central government can look at a spreadsheet and see a building as an asset or a cost.
The community may see it as a major employer and part of the local economy.
If a public body disappears from a town, the consequences can extend well beyond the public-sector payroll.
That is why the Government's promise that devolution should bring government closer to the people needs to be tested against what actually happens on the ground. Its own civil-service plan says that civil servants working face-to-face with communities and partners is important, while also proposing more shared buildings and more flexible working.
The NHS question could be particularly difficult
Reducing 14 territorial health boards to two strategic boards is an enormous structural change.
The argument is that too much management capacity is tied up in separate organisations and that a simpler structure can free resources for treatment.
That may ultimately prove correct.
But there is a difference between reducing the number of boards and reducing the cost of running the NHS.
The new organisations still have to perform the work previously undertaken by the old ones.
Indeed, the creation of larger organisations can sometimes produce another problem.
The bigger an organisation becomes, the greater the distance can become between the people making decisions and the communities affected by them.
That is particularly important in a country such as Scotland, where the healthcare requirements of a densely populated city are very different from those of Caithness, the Highlands and the Islands.
A large strategic organisation may produce economies of scale in procurement, administration and specialist services.
But there can also be diseconomies of scale.
Travel becomes more difficult.
Local knowledge can be diluted.
Decision-making can become more remote.
And if the restructuring goes wrong, fixing the problem can itself be expensive.
Local government could be even more sensitive
The proposed discussions around local government could eventually prove to be one of the biggest changes.
There are 32 Scottish councils.
The Government is now considering how local governance should operate, including regional collaboration, pooling resources and stronger community-level decision-making.
There is a perfectly reasonable argument that councils could share more services.
Procurement could be combined.
IT could be standardised.
Back-office functions could be shared.
Specialist expertise could be pooled.
But a merger is not automatically a saving.
Two councils may each have finance departments, HR departments, IT systems and management structures.
Put them together and those functions may eventually be reduced.
But somebody has to undertake the work of merging them.
Contracts have to be reviewed.
Staff have to be reorganised.
Systems have to be integrated.
Buildings have to be rationalised.
Political structures have to be changed.
And elected members have to agree on how the new organisation operates.
That can take years.
There is a danger of counting the same saving twice
This is something taxpayers should watch particularly carefully.
Suppose three public bodies each claim they can save £5 million by sharing procurement.
The theoretical saving is £15 million.
But if they all begin sharing procurement, the £15 million cannot then be counted again as part of another central-government efficiency programme.
It sounds obvious.
Yet large public-sector transformation programmes can become extremely complicated, particularly when several reforms overlap.
That is why the Scottish Parliament, Audit Scotland and the public need transparent figures showing not simply gross savings, but the actual net savings after implementation costs.
The crucial number is not:
"We will save £100 million."
It is:
"We will spend £X million implementing the reform and then save £Y million a year, producing a net saving of £Z over five or ten years."
That is the calculation taxpayers need to see.
What about the staff who disappear?
Swinney has been clear that the Government's approach is not based on compulsory redundancies in the immediate plans, with natural attrition and voluntary departures playing a major role.
That may reduce some of the social cost.
But it could make the process slower.
If one person retires and their post disappears, that is relatively straightforward.
If ten people retire but the organisation still needs eight of their functions, the remaining employees may simply end up carrying more work.
That creates another important question.
Are we reducing bureaucracy or simply reducing the number of people doing it?
There is a big difference.
If unnecessary work disappears, excellent.
If the same amount of paperwork and regulation remains but fewer people have to process it, the public may eventually experience slower services rather than greater efficiency.
The Government's stated ambition is to use technology and automation to remove low-value administrative work and allow staff to concentrate on higher-value activity. That is the right principle.
But it needs to be demonstrated rather than assumed.
The upfront cost could be the political problem
This is where Swinney's plans become particularly interesting.
Scotland faces tight public finances.
The Government is simultaneously trying to increase spending in some areas while reducing the underlying cost of delivering public services.
That makes restructuring attractive.
But it creates a cash-flow problem.
Imagine that a reform costs £500 million over several years but eventually saves £200 million a year.
It could be an excellent investment.
But somebody still has to find the first £500 million.
If the Government does not have that money available, it may have to borrow, divert money from existing services or delay other spending.
The public therefore needs to know not only the eventual saving but how the transformation will be financed.
There is a lesson from previous reforms
Scotland has considerable experience of public-sector mergers.
Police Scotland was created by merging eight regional police forces into one national force in 2013. The intention included protecting frontline services, delivering efficiencies and strengthening specialist capabilities.
The lesson from such reforms is not that mergers are necessarily wrong.
It is that the promised efficiency does not automatically appear simply because organisations are put together.
The savings have to be identified, delivered and sustained.
And the quality of the service has to be monitored at the same time.
Otherwise the Government can end up with a bigger organisation rather than a more efficient one.
There is potentially a very good outcome
None of this means Swinney's reform programme should be dismissed.
Quite the opposite.
There is a strong case for asking whether Scotland has accumulated too many organisations, too many layers of administration and too many duplicated functions.
The Scottish Government itself now makes that case.
The opportunity is to redirect money from administration towards frontline services.
That is something most taxpayers would support.
The difficulty is making sure that "frontline" really means frontline.
A saving in headquarters does not help a patient if it results in more pressure on local services.
A council merger does not automatically improve bin collections, roads or social care.
A new computer system does not automatically make government more efficient.
And reducing staff numbers does not automatically reduce the amount of work government has to do.
The test should be simple
There should be three numbers attached to every major reform.
How much will it cost to implement?
How much will it save each year once fully operational?
How long will it take before the savings exceed the implementation costs?
Those three questions would transform the debate.
They would also allow taxpayers to distinguish between a genuine efficiency programme and a reorganisation that merely moves costs from one part of government to another.
Swinney has given Scotland a five-year programme.
That is long enough to make substantial changes.
It is also long enough for taxpayers to demand evidence that the promised savings are actually appearing.
There is nothing inherently wrong with spending money to save money.
Businesses do it all the time.
A company may spend millions replacing an inefficient factory because it knows that the investment will reduce costs for decades.
Government should be capable of doing the same.
But the business case has to add up.
And that is where the scrutiny of Swinney's plans should begin.
Scotland does not simply need a smaller public sector.
It needs a more productive public sector.
If ten people can do the work previously done by fifteen because technology has removed unnecessary administration, that is a genuine efficiency.
If two organisations can provide the same service for less money without reducing its quality, that is a genuine efficiency.
But if an organisation is simply renamed, its headquarters moved and its management structure changed while the underlying cost remains much the same, taxpayers have gained very little.
The coming consultations will therefore be important.
They should not simply ask what should be merged?
They should ask what will it cost, what will it save, when will the saving arrive, and what will happen to the service in between?
Because there is one certainty about public-sector reform.
The bill for changing the system arrives before the savings from the new system.
And with Scotland's finances already under considerable pressure, taxpayers deserve to know just how large that first bill could be.