Moonshot

The death of public service began in the late 20th Century.

First came a political disillusion that more money did not necessarily make for better outcomes. Then systems thinking confirmed what we already knew, that services for people are complex and our sense of control a fallacy. The economic crash of 2008 led to a decade of cuts and increasing demand, and we often saw the unkindness and inequality in our communities. At the same time the world exploded in a digital revolution and we began to redefine the relationship between the state and local citizens.

So we need to talk about about rebirth — navigating from traditional public service to a new public system. A more compassionate public system that is cost effective, but also kind and supportive for our communities, creating resilience and growth as well as being the safety net.

In short there are three coordinates that point to the future public system. A compass that helps to guide our innovation and ambition for change: a moonshot for better outcomes.

• Understand and anticipate — new data sources and big data analytics are creating a step change in our understanding of population needs and outcomes. Better understanding increases the accuracy of predicted need, making it economic to target individuals and offer help much earlier.
• A compassionate system — professionals working in public services are compassionate, but our systems are not. Families are often pushed away until their lives get worse, experience long waiting times to manage demand, and receive services that don’t fit their need. There is now an economically viable alternative to reach out and support citizens earlier – a compassionate public sector system.
• Integrate to control demand — the more we understand the local public system the more it is seen as indivisible. Integration is critical to control where demand is met, to build trust between providers and manage how the statutory, community, universal and digital resources are used to support citizens. Inevitably funding will move from expensive statutory services to real early help, increasing the volume of demand but reducing overall costs.
  

Our current model of public service was set up in the 1940s. There have been enough changes in practice, data and technology, community relationships and systems thinking to usher in a new approach. Statutory minimum is a bust mode, so collectively we must reach with ambition to the next paradigm…

Understand and anticipate

The first challenge is to better understand and then anticipate the needs of local citizens — enabling much more responsive and cost-effective forms of support. Our intelligence will be integrated to create a single view of the citizen and family unit, including a wide variety of partner data (for example education, care, health, benefits, housing, marketing and social media).

But this intelligence is insufficient — for instance we barely know our population until their needs escalate — which gives us little time to be pro-active. So, we can expect new measures such as annual surveys of children’s outcomes, and real-time measures of patients’ health using wearable tech, so people can stay at home rather than in a precautionary hospital bed.

Predictive analytics has been used in business since the 1990s, for instance to check credit risk, detect fraud, or target product marketing. At its heart is the analysis of trends, and comparing the needs of individuals we understand, with others to make assumptions about what they might need in the future.

In public services, we want to find out the factors that protect residents, the factors that expose them to more risk, and a list of individuals or families whose need is likely to escalate over the coming years. As the data we collect increases, we can analyse more inter-related factors to spot trends such as housing, benefits, education, health data and social media. When data-sets become too big for a person or team to understand, we use artificial intelligence learning algorithms to develop hypotheses and spot data-relationships. For example:

Behavioural Insight Team analysis of social workers’ case notes can detect which children are at a high risk of being re-referred.

Los Angeles uses decades of crime data to identify areas and times with high probabilities for certain crimes, and deploys officers more efficiently. Police forces in the UK have used the same software to predict drug crime and robbery.

The New York City Fire Department service analyses building inspection records to predict the risk of fire. They have now added another 7500 factors to include behavioural characteristics in predictions.

Research in New Zealand combines local authority and health data to predict, at the age of two years, which children will be on the child protection register in the future. Similar approaches are being trialled in the UK.

If we get the intelligence and predictive analytics right, we can do three things:

• Build the science behind our services, based on local evaluation of the longitudinal impact of interventions, so we can invest in what works. We might also find it’s not traditional services that deliver the most sustainable outcomes, but family and friends, the community, local economy and environment.
• Commission for outcomes and not services — giving responsibility to providers for example for the mental health of a population, rather paying for entry to services or number of sessions (in lean terms, paying for failure).
• Anticipate the needs, including hidden needs, of every citizen and offer targeted and tailored early help. (Which might feel different to the current experience of rationed services.)
  

A compassionate public system

Public service failure is often institutional, obscured by layers of history and culture. The older person whose complex needs don’t fit box A, B or C, but that’s the only offer. Or the hidden children where 24 per cent of a class has a mental health need but we only have services for three per cent. The new parents who can’t read but are given a pack of information about local services. Waiting time targets that are gamed by institutions and used to manage demand. Or the thresholds that say we can only help if you’re really ill.

Professionals working in the public sector are compassionate — but the public service model is let down by a conspiracy of institutional blindness. For example, in children’s safeguarding there is a process of triage with a quick assessment against thresholds. If the need is below the threshold then there is no further action, reducing demand to expensive social care services. Logically this makes sense to ration resources, but it also pushes families away until their needs escalate and lives become worse. Now some local areas are removing this stage and offering appropriate help for all referrals — using the wide range of community and public services to help families and intervene early.

What a public sector system might encourage is move away from cuts and statutory minimum services to a compassionate design — that understands citizens and reaches out to help them.

It’s a simple choice: demand is either managed by ignoring it, or because we help people earlier.

But belief in early help has been shaken. Business cases that promised a return on investment haven’t delivered. This is often because the business case assumes that presenting need would eventually cost society, but ignores hidden need that has the potential to escalate. We have now found that many traditional case-holding models of early help are too expensive and have increased thresholds in order to be more financially viable, leaving a gap that we might call real early help.

So what we define as a service is changing. Intelligence moves our interventions from reactive to pro-active giving time for different types of support to be more effective. However, there is currently a vacuum of interventions in this new tier, so we need innovation. And at the same time we are seeing traditional services blending with community resource, blending with universal services and blending with digital delivery. So, we cannot be sure what this new tier of support will look like, but it is likely to grow rapidly, and it won’t feel like a normal public service.

For example: volunteers shoulder to shoulder with professionals; service users recognised for their assets and asked to support others; self-help and a culture of local compassion seeded by the state; teachers doing a little more with target families; pharmacies, taxi drivers and other local businesses supporting resilience; online training and guidance to democratise professional knowledge; or digitally pushing a tailored list of local community groups to residents through social media based on analysis of their future needs.

So, in summary we can expect to see a radical expansion of real early help in the tier before traditional service models are viable. The effectiveness of this support will be aided by better understanding and anticipation of citizens’ needs, and targeting individuals to offer help. And there is room for innovation and investment in these new models — to reduce demand into statutory acute services, and to form an important layer of the new public service system.

Have councils and commissioners fallen into a rationing trap? It’s easy to assume that a parent coming into the housing service is looking for social housing, and the officer across the counter knows there isn’t much social housing. The relationship that has been created is already in conflict: the client will fight for a house, the service will fight to ration supply.

But if people coming into housing services are asked what they want, 70 per cent just want some help, not housing. Thinking about behavioural insights, people could be asked to write down what they are looking for, and maybe use that as the starting point for the conversation, replacing conflict with co-production. Many local areas are recognising the importance of social capital and deliberately creating a new relationship with the community — to encourage connections, self-help, and neighbours and families to help each other a little more.

We talk about communities but I sometimes wonder if we’ve created a geographic convenience rather than a helpful construct to public service design. I was struck a few weeks ago by my four-year-old daughter talking about her friends at school, how she develops, makes connections and becomes more resilient is already personal to her. My daughter’s community is not mine — so how can we think of geographically coincident people as homogenous?

Perhaps each resident has a different community. I’m tending towards a revised model which thinks of individual resilience connected through four webs:

Local geographic connections

Family and friends who support each other

Communities of interest such as work, football club, etc

And digital connections which whilst un-personal are good at some aspects of resilience such as advice and guidance and bringing together strangers with similar interests across a wider sphere.

Perhaps we should be designing a strategy to support all these individual communities in England, all 65,648,000? As the convenor of community support, I’d suggest local government should be aiming for the following objectives:

Connections between individuals and groups

People know where to get help, both locally and from digital help

Individuals and families develop their emotional resilience, this is key and curiously there aren’t a lot of activities or services in this box

And we want a culture of reciprocity and lowering inhibitions so people are more willing to reach out and help each other.

Integrate to control demand

It feels uneasy to talk about control in the same breath as compassion. But control will be a defining characteristic of the public sector system — so that citizens get the help they need from the many different services and types of innovative support. And demand is met in the most cost-effective part of the system instead of the de facto maze of pathways.

Single point(s) of access will make the public system easier to navigate and show us where more support is needed to reduce demand to expensive statutory services. For example, increasing capacity in primary care through digital consultations — which in-turn reduce demand to acute hospital services.

So, the new public system is dependent on trust between leaders — to stop cost-shunting, share risks and use the total resource in the system. Integration will help to improve efficiency and create the trust — through vehicles such as Alliances that give a legal basis to risk sharing for providers. Or Combined Authorities which give a similar basis for risk sharing between providers. New providers will be incentivised to grow innovative services blended with community, universal and digital resource that promote resilience and reach out to citizens.

The new structures are likely to operate in layers. National governance and policy, regional combined authorities, hospital level alliances, local service integration around 50,000 populations, and voluntary and community based support at street level. Too much focus on one layer to the exclusion of another will be sub-optimal.

And not only will we create larger public sector delivery vehicles in each area, but commissioning will also integrate and scale up to manage these competing providers. A Combined Authority model gives the right sort of integrated governance so we can move to outcome based commissioning and making best use of all public services. Pursuing population outcomes, but also pushing inclusive growth and resilience, and shifting funding from acute to early help.

Democracy exists by the will of the people: a delicate balance between basic rights, managed expectations and hope, and capitalism that drives growth. Some communities however experience a sense of inequality from post-recession recovery, a lack of control and little hope for improvement. We know that inequality is linked to high cortisol levels during pregnancy, which in turn leads to children who are less trusting and mature more quickly. Research such as The Spirit Level has shown the link between inequality and poor outcomes for the whole population – affecting both the haves and the have-nots.

So do our services help or hinder inequality? Do we design for a master-slave relationship to ration services, or co-production that encourages resilience? Can we develop a more compassionate and human public system?

Moonshot

Moonshot is a term to describe the great leaps the human race can take when we are collectively ambitious and bloody-minded. We’ve been playing with systems thinking and pulling the threads of the 1940s social reform for long enough. This is the moment, this is our moonshot to a compassionate public system.

First published in the PSTA State of Transformation 2018: http://www.publicservicetransformation.org/wp-content/uploads/2018/07/PSTA-public-service-state-of-transformation-report-think-pieces-v2.1NS-e-version.pdf