Evidence Session 2: Best Practice
International best practice in public administration
Evidence session with Sir Bill English, John Micklethwait and Adrian Wooldridge
The Commission for Smart Government held an evidence session on Thursday 19 November 2020, where Commissioners put questions to Sir Bill English, Prime Minister of New Zealand 2016-17 and Minister of Finance 2008-16; and John Micklethwait & Adrian Wooldridge, authors of The Wake Up Call: Why the Pandemic Has Exposed the Weakness of the West - and How to Fix It.
The session focused on international examples of best practice in public administration, including comment on the Commission’s case study New Zealand’s Innovative Government Reforms, and discussion about the systemic problems in governance exposed by the COVID-19 pandemic.
Commissioners present at the session included Lord Bichard, Sir Suma Chakrabarti, Sir Ian Cheshire, General Sir Chris Deverell, Dame Jayne-Anne Gadhia, Baroness Finn, Verity Harding, Lord Herbert of South Downs (Chair), Husayn Kassai, Sir Paul Marshall, Lord Nash, and Sir Mark Rowley.
Highlights from Sir Bill English’s evidence
On New Zealand’s performance improvement framework
In putting New Zealand’s performance improvement framework together we started with a more political and business view, rather than a state-service view. It was a very engaged, practical way of dealing with the senior management of a department. The model was very successful in flushing out issues and re-setting direction. Expenditure reviews and restructuring tend to be a waste of time. But if you focus on the performance of half a dozen people who run large entities [then it has much more impact].
We picked about 12 targets that included some pretty tough social targets – like reducing the number of substantiated child abuse cases. There were three elements about these targets that mattered: they were published regularly; they had direct personal accountability; and they were unambiguous goods.
As an indication of effectiveness of the targets, over four or five years with a bit of focus, and quite a lot of money, we got the rate of Rheumatic Fever down by about 40 per cent, which no one had thought would be possible.
On the need to focus on the particularity of customers: which people, where?
I can’t emphasise too much, that if you want to change anything in public service, you have to motivate [the system] by what happens to real people. What actually happens to a young child who comes from social housing into the A&E department with bronchial problems, and has to wait six hours, his mother has to go off to a shift job, so his big brother comes in, he gets treated, and then they go back to their social house that’s cold and got mould? That’s what should drive your policy.
So we spent a lot of time seeing what happens to real people. If you put those stories, those real people, in front of those who run your big agencies, it can’t help but affect them. Particularity matters, so for any policy the first question is which people, where?
And the use of public service data
New Zealand has the best linked public service data in the world. We can look at anonymised data of public service use of every citizen, some of it going back 30 years. If you rank everyone by public service use, the first 85 per cent use half of all public services. They’re reasonably well served by monopoly or commodity supplied services - standard hospitals and schools.
The problem is that the other 15 per cent use the other half of all public services. So the need curve is logarithmic. But the models are all built around normal distribution. You really need to understand those [15 per cent high use] customers. Those are your big users, your frequent fliers. They get small bits of commodity services that don’t work, and they exhaust themselves going around looking for it.
On longitudinal investment calculations
We calculated the net present value of the return on investment for a person or cohort. The key element is to look at a population not at departments. Departments can’t tell you about a real person, they can only tell you about a pupil, a student, a prisoner, not the family that’s got all three of those in it. They only know about the feedstock for their institutions.
For example, it turned out that it was worth it for every sole parent under 20 in New Zealand to be individually supervised, and we went through every single one of them. Because their kids are the most vulnerable people in our community.
And the use of evidence to inform policy and service design
The data and technology are now so cheap that even in public service we can know anything about anyone for practically nothing. Digitalisation crushes transaction costs. The tools that help you run global supply chains can help public services. Knowing this evidence used to be a professionalised, institutionalised specialised function, requiring statisticians and policy analysts. Now you can get a bunch of 23-year olds to get you evidence on a whole range of interventions and show you the return on investment in an hour.
On the new legal frameworks for cross-government working
Cross agency collaborations can fail because they lack a sense of common purpose, and they have mismatched accountability and authority structures. You need to understand the trail of financial and operational accountabilities that lead to decisions. This is a problem of internal governance. How do you get authoritative decision-making across a range of agencies? In the market you do this as a joint-stock company. The new legislation provides for internal subsidiaries with proscribed governance processes. This is being trialled with family violence, and I hope it works. It could be used for a whole lot of tough social problems. You need accountability over time. Agencies struggle to sustain a relationship with each other over time. So you need to lock in a statutory entity that holds them together for the life of the project.
On tackling tough social issues and the predictability of need
When you look at spatial economics of cities, what’s surprising is how predictable these phenomena are. Really good big data gives really simple answers. In this suburb, there are going to be 463 juvenile delinquents, who are going to commit this many burglaries. It’s not an overwhelming tide of need, it’s totally predictable phenomena. The data can paint that picture. The way to deal with this is one by one. But our institutions were never designed to do that. Our public services cannot produce the institutional relationships capable of holding long term problems with people with chronic problems. You have to change your models to services that focused on particular families, and that are trust based and long term. The social investment calculations will show you there’s a big pay-off.
Our welfare system has been recast to calculate a bottom up liability. This showed in vivid technicolour the heterogeneity of groups. A thousand people who need 1 per cent solutions, not a thousand people for which there’s one solution. And of course the agency’s not organised like that. It showed the gap between what’s actually needed and what’s delivered.
On data analysis of populations
Clever use of data can show the stories of populations. You need internally an agency that has the protection and the power to focus on [analysis of] actual populations, independent of the main agencies – such as New Zealand’s Social Investment Agency. It proved to be a fantastic catalyst. [But] you have to really believe in that because it tells a truth that’s inconvenient. The more holistic view you have of a family or a community the more it’s obvious that any agency is not designed to deal with that need, only a part of it. And in order to [meet that need] you’ll have to work with someone else or change the model.
On the ‘Beehive model’ of co-locating ministers
All our ministers are in one building, separate from their ministries. A weakness in our system is that ministers don’t get as well trained as they don’t get as much exposure to public servants as in other jurisdictions. But the decision-making process is a lot better. The cabinet process is much more collective. We have four or five cabinet committees which meet at least every week. They generally work on having to get agreement before they go to cabinet and that agreement can involve ten ministers.
Because the ministers are all in the building, you can also have an informal stream of activity. If you get ministers together and present data around how people experience public services, you can solve problems more quickly. The problem with politicians is you have to reduce the amount of space that politics takes up in their head, and expand the space taken up with thinking about actual problems.
Highlights from John Micklethwait and Adrian Wooldridge’s evidence
On what the pandemic has shown about why government matters
The pandemic was like an exam for governments. Around the world, government was the difference between living and dying. In Singapore they have a death rate [from Covid-19] of 5 per million; we have a death rate of 650 per million in the UK. Germany is six times better than us [at protecting its people] at around 150 per million. Look at London, New York and Seoul, which are all big bustling mega-cities. New York has lost 25,000 people, London has lost 6,000, Seoul has lost about 50. It’s not about culture; Seoul is a big chaotic city, but it had better government. If you look at the countries at the top of international health and education league tables, the same countries did well at Covid.
On the weaknesses in the UK’s system revealed by the pandemic
There was a lack of preparedness, flexibility and resilience in the system. Over-centralisation is a huge problem. We have one of the most centralised government systems in the world. In the UK, local governments spend less than 10 per cent of GDP, which within the OECD is comparable only to Luxemburg and Portugal, both much smaller countries. The relationship between local and central government is dysfunctional. Another problem is that the public sector instinctively avoids working with the private sector. In addition, the cabinet structure was not equal to the task. A small number of cabinet ministers were massively over-stretched, and others were left with only a limited role. The process was dominated by a handful of figures who had to deal with political, presentational and operational issues. We needed a clearer distinction between the political and operational functions of ministers.
On the strengths in Germany’s federal system when it came to responding to the pandemic
The federal system didn’t have the same tension between the centre and local government. Their decentralised health system, under local public health directors, which are well resourced with large staffs, was able to respond very well.
On Singapore’s theory of government and culture of public service excellence
Singapore is the first country in a long time to produce a new, comprehensive and effective theory of government. Countries in the West innovated with a succession of theories of government over the late eighteenth and nineteenth centuries, up until the establishment of the welfare state in the middle of the twentieth century. From the 1960s to 1990s, [Singapore’s first Prime Minister] Lee Kuan Yew advanced the notion of a meritocratic system with a limited democracy and an austere welfare state. The Singaporean state is activist and developmental, it encourages the development of a knowledge economy, and the role of the government in pushing private sector companies up the value chain. It is a much smaller state which provides better outcomes for its customers.
Singapore’s success is largely not inherent to authoritarianism and is founded on lots of things that Western democracies could adopt, namely the creation of an ethos of excellence in the public sector. In Singapore, people enter the civil service as a career path associated with honour and status. It is a dynamic state constantly trying to improve, with a strong capacity to learn and innovate. It is much more results-orientated, with performance related rewards. Singapore’s state is not successful because it’s more authoritarian than our state, it succeeds because it’s more like our private sector.