At a recent seminar, we decided to take a look under the hood of the Swedish and Canadian consolidations in the 1990s and find out what their administrations actually did. Senior civil servants from both countries gave us much food for thought.
A tale of two countries
The event quickly brought out the very different approaches to reducing public spending taken by Sweden and Canada.
In Sweden, the focus was on reducing the cost of transfers (such as unemployment insurance), but leaving departments and agencies largely untouched in terms of staff numbers and service delivery.
The Canadians reduced some transfers and subsidies too. However they took a more fundamental look at what the government should no longer be doing or could do differently. This led to a radical re-think of how several federal departments operated.
The government here appears closer to the Canadian approach than the Swedish one. With cuts for most departments likely to average around 25% (and some departments could be much higher), it is clear business as usual is unsustainable. Transforming Whitehall and the wider network of service delivery it oversees will be vital.
The Swedish rationale
It is worth dwelling briefly on the rationale for the Swedish approach before moving on to the lessons from Canada. In the midst of an urgent crisis (interest rates temporarily spiked at 500%), the Swedes prioritised cutting back forms of spending that would yield an immediate, large and certain cash saving.
They felt cutting back - or even reshaping - the capacity and capabilities of government risked a difficult and uncertain journey which may well have harmed service delivery, and, more importantly increased the risk that their consolidation would fail.
It was this approach that allowed the Swedish government to achieve nearly half of its consolidation within the first financial year.
Lessons from Canada
The challenge the UK now faces is to avoid the risks of decimating services for no real gain.
Transport Canada cut its headcount by over 75% and its budget by 61%, whilst also being recognised by the Office of the Auditor General as a model for other departments. Specific policies (such as selling off assets and commercialising large areas of delivery) are inevitably unique to the particular context but several broad lessons emerged:
- Plan ahead and settle early – Transport Canada went into Program Review with a clear plan of its own making; other departments that were more reluctant to engage in the process had not only the size but also the shape of the cuts imposed upon them from the centre
- A coherent story to tell – having this helped to marshal efforts within the department and gave a broad direction of change for outside interest groups. 'Vision' was felt to be too grand a word given the complex and messy nature of any reorganisation. Communicating the story was said to consume about half of the deputy minister’s (equivalent to permanent secretary) time while implementing the cuts. Importantly, this was a case of communicating decisions that had already been taken by the department rather than consulting
- Pace – plans will never be perfect and mistakes will be made, but it is vital to move quickly on the fundamental changes that need to take place and the loose ends can be tied up later
- Greasing the wheels – Transport Canada realised far greater revenues from the sale of assets than expected and was allowed to keep a small proportion of the funds to help sweeten the pill for adversely affected interest groups. Similarly, offering generous terms was key to the process of making staff redundant or moving them out of the department
Bringing it home
A couple of notes of caution emerged.
First, the current context in the UK appears to be much more challenging than for Sweden and Canada in the 1990s. There are tougher economic conditions both at home and abroad and the depth of public acceptance of the need for cuts is more uncertain.
Second, Canada did not succeed at the first attempt. It was not until the sense of crisis was shared nationally that the cuts could be implemented.