Why is public service reform so difficult?
Government is not easy. There are no perfect solutions to entrenched poverty, poor public health, or high levels of inequality. There are also practical limits on what the state can do to act on them given both the need to secure fiscal sustainability and the open and connected nature of the modern economy. Constant Parliamentary and media scrutiny about what the Government is doing also makes it very difficult to act with the same ruthless efficiency or appetite for risk that you might expect in a business or start-up.
But all that being said, you can’t help but feel that the current system of Government – a top-down and heavily centralised bureaucracy – makes it harder than it ought to be for a political party to deliver effective change. One big problem that seems to be at the root of many failed reforms or wasteful projects is that public sector leaders struggle to manage the complex dynamics that exist when introducing change.
The complexity puzzle exists at two levels: one is cultural and the other is more straightforwardly a management issue.
The first element of complexity is the types of problems that Government is trying to solve. Perhaps with the exception of big infrastructure projects (e.g. Crossrail and the Olympics), the problems we try and act on are not the complicated science, engineering or construction problems, they are complex social ones that involve multiple independent actors and contain lots of unknowns. Mark Foden, the management thinker, was the first to draw this distinction to my attention and has applied it to the public sector. Mark cites improving public health as a good example of a complex problem. Poor health has a mixture of known and unknown causes, there are a number of known and unknown cures and the actions of any one single actor (doctors, nurses, families, neighbours and ourselves) will not necessarily solve the problem.
However, the bureaucratic response to trying to solve a complex problem tends to be to approach it like an engineering puzzle.
New reforms are almost always approached in a top-down and additive way – new funding, new rules, new managers, or new technology is introduced, rather than allowing solutions to emerge from the bottom-up. In other words, reforms work on the assumption that structure and process can have a direct impact on performance. However while these reforms can be successful on their own terms they very rarely lead to the sorts of sustainable outcomes that actually reduce dependency on the state over the long-term. What we tend to end up with is a random hotchpotch of initiatives that creates new obstacles to improving productivity or innovation.
Under the Coalition Government, for example, there were 30 separate funding pots available for local change programmes provided by different departments, with different criteria and methods for allocating money. And according to the Institute for Government, Central Government has tried to encourage local service integration no less than 59 times over the last 19 years without delivering a substantial improvement in services.
The second element of complexity is in bureaucracy itself. It is well known that big organisations take on a life of their own and a Government bureaucracy is perhaps the best illustration of this. More often than not, bureaucracies tend to: resist changes in direction; find unnecessary work to do to fill time; soak up all available funds by thinking of new initiatives; and mindlessly imitate the behaviour of peer organisations. This is what Warren Buffett calls the ‘institutional imperative’ and it makes controlling an organisation through top-down planning very difficult.
The public sector is particularly vulnerable to mismanagement because it often lacks the necessary in-house capability to manage projects effectively as well as the financial resources to compete with the private sector for the best talent (more on that in a previous Policy Exchange paper here). This lack of expertise means that consultants are often drafted in to work on projects. But consultants are heavily biased to claim value in the short rather than long-term – you often hear about consultancies agreeing to do whatever their client wants, regardless of how stupid it is. Government is also more vulnerable to capture by technologists who tend to build complex systems or processes into a project. In the end even simple projects can balloon in size very quickly.
Universal Credit (UC), the project which aims to streamline six existing benefits into one payment, has been plagued by these types of management issues. It started with the assumption that streamlining benefits was the right solution for the whole system without testing it small scale. Success depended heavily on the introduction of new technologies that tried to imitate smart automation systems used in the private sector without fully understanding their replicability. Timings and resources were all determined upfront in a detailed business plan that made adaptation and improvisation much harder further down the line.
So how can we change the system to make public sector reform more effective? This is something Policy Exchange will examine in a forthcoming paper on delivering effective change in the public sector. In the interim, here are three suggestions.
First, we should recognise that managing things at scale can create major problems in complex contexts. Small is better as when you multiply the number of people involved, the nature of the problem can change as well. Skunkworks projects that operate outside of conventional Civil Service rules and conventions are a useful tool for trying new things. The Government Digital Service, for example, used these tactics to deliver new digital services at great speed and with fewer costs.
Second, rather than imposing top-down reforms we should encourage bottom-up solutions instead. The Government’s devolution programme should help with this, allowing for the simultaneous development of different policy solutions by competing Local Authorities. For example, the idea of combining health and social care with a single budget and joint commissioning has been long advocated but it was the Greater Manchester City Deal that has been the vehicle for implement it. Future City Deals are therefore an opportunity to experiment with service reform. They are also less likely to be burdened with the same levels of expectation, media scrutiny and unionisation that can undermine radical public service reform at the national level.
Third, experimentation should be built into the mainstream policy making process. The Government could, for example, establish a new legal basis for experimentation, creating an operational environment that is more accommodating for testing, learning and subsequent changes in direction if something does not work as originally planned. The Green Book could also require departments to provide evidence of prototyping or ethnography as part of a business case submission. This would allow the Government to find out what works by actually testing it small-scale and with an open-mind.