The public sector can’t deliver everything – so how governments partner and outsource is a vital question

Governments have a long list of responsibilities – from educating their citizens, to looking after people’s health, to maintaining a transport network, to dealing with crime, to helping the most vulnerable. Delivering these not only requires hundreds of billions in public money, but a vast body of tools, systems, people, expertise and capital – much of which does not exist within the public sector. 

Consequently, governments spend many billions buying goods and services from non-government providers. There are advantages to this – more on that below – but there are also challenges and risks, not least the chance that an organisation goes under in the middle of providing an essential public service. 

That risk was infamously realised in 2018 when the company Carillion, which held contracts for 420 UK public sector projects (from serving school dinners, to cleaning prisons, to building railways and hospitals), collapsed. As just one example of the many fallouts, a new hospital Carillion was meant to be building in Liverpool was delayed by five years and ended up costing the taxpayer £1bn rather than the originally budgeted £350m. Local people had desperately needed a new facility at the start of the project, let alone the much-delayed end: one resident bluntly described the old hospital as “a craphole”. 

As a research communicator, I try to explain academic findings to general audiences. But when it comes to research findings around how governments partner and outsource – which didn’t go so well with Carillion, and which is what the Blavatnik School’s Government Outcomes Lab works on – this has never been easy. Commissioning and contract management are topics that demand lots of contextual understanding and specialised vocabulary. 

The reasons for doing the research are easier to convey, though. Whether people flourish or struggle, especially vulnerable people, depends in large part on their government’s partnerships with other sectors. 

Why governments outsource and partner – and why it’s hard 

The public sector is large – in the UK, for example, over one in six employed people work in it – but it’s not large enough to deliver everything governments pay for and provide. And most people (and most countries) agree it shouldn’t be. Not only does outsourcing to other sectors allow everything to be done, it offers governments many advantages: harnessing outside innovation, accessing large systems and facilities the private sector has built, deploying expertise held in specialised organisations, working through charities that have trusted contact with vulnerable groups, and leveraging private capital. 

The challenges, though, include not only risk mitigation to avoid another Carillion experience, but the headache of managing contracts and the complexity of bringing together multiple partners. 

Governments face wickedly difficult questions in these areas. How to structure contracts so that providers deliver the right outcome? How to harness the creativity and expertise that exists across sectors while still ensuring certain results? How to allow for responsive, real-time innovation during a project without making a contract dangerously vague? And even when a specific product or service has been provided, how to measure whether it changed individual lives? 

Pioneering experimentation in the UK 

Some committed UK policymakers have been innovating on these questions over many years. Back in 2010, the UK’s Ministry of Justice made a world first in how it financed a programme at Peterborough prison aimed at reducing reoffending. The programme would provide prisoners with support in areas linked to post-release reoffending, like substance abuse, mental health issues, struggling to find accommodation, or not being able to navigate benefits. The support would be delivered by a mix of public, charity and private providers (from prison staff to Mind to YMCA), and it would be financed in a brand-new way, through the world’s first ‘Social Impact Bond Payment by Results’.

See the tricky vocabulary creeping in? Let's break it down. 'Payment by results' is as it sounds: paying for the outcome, not the service. It's a bit like paying a gardener only when you're actually eating apples, rather than when they've planted the apple tree.

What about the 'social impact bond' bit? Bonds help government finance activity. In a normal bond, investors lend money to a government for a set period, after which they are repaid with a fixed rate of interest. But in this social impact bond, repayment was contingent not on time elapsed, nor even on delivering the agreed support, but on achieving the ultimate outcome: reducing reoffending. In other words, investors paid for the programme upfront, and they wouldn’t be refunded by the taxpayer unless it succeeded. The investors took that risk because they wren't just moneymakers, they were people and organisations who cared about reducing reoffending.

It did succeed: by 2017, the programme had reduced relevant reoffending by 9% compared to a national control group, exceeding the 7.5% target. Since then, social outcomes partnerships (as they are now known) have been set up in 35 countries. 

Ongoing UK innovation 

In 2016, the government launched the Life Chances Fund, a nine-year £70m outcomes fund that was at the time the biggest of its kind in the world. The Government Outcomes Lab, or ‘GO Lab’, was established at the Blavatnik School at the same time, as a fully independent research lab that could be the fund’s evaluation and learning partner. The GO Lab has since expanded its funders and remit internationally, while remaining a source of expertise for the UK. 

At its launch in 2016, the Life Chances Fund was pioneering in several ways – not only was it designed to propagate the adoption of social outcomes partnerships, it was also hugely ambitious in its size (the largest of its type at the time, spanning nine years and multiple spending review cycles); the breadth of themes considered (which enabled considerable experimentation and the application of social outcomes partnerships to social challenges where they had not been deployed before); and the way it connected local and central government. It helped fund 29 projects across England with over 50,000 beneficiaries in areas like youth unemployment, mental health and homelessness. 

The GO Lab has just published its final evaluations of the fund, which go into detail about what did or did not work and why, for the future benefit of policymakers worldwide setting up similar partnerships. The Life Chances Fund Final Report evaluates the fund overall and each of the 29 funded projects, and the accompanying Evaluation Synthesis Report distils nearly a decade of data and research on the fund into a single document. 

These cap a large body of GO Lab evidence on outcomes-based cross-sector partnership: twelve evaluation reports, eight case studies, seven comprehensive technical guides, thirteen peer-reviewed publications, and an open-access dataset with granular outcomes data. They aren’t easy reading for a layperson like me, but they have been extremely useful for policymakers and have been downloaded over 7,000 times. 

The Life Chances Fund evaluations come at a pivotal time in the UK: the fund’s much larger successor, the Better Futures Fund, is under development, and will be up to £1bn strong, with £500m invested by the government and a goal of matched investment from partners. Again, it is the biggest of its kind in the world. It aims to “support up to 200,000 children and their families over the next ten years by bringing together government, local communities, charities, social enterprises, investors, and philanthropists to work together to give children a brighter future.” The government says it could fund projects like “providing support in schools to improve attendance, behaviour and overall achievement of pupils; intervening to free children from a life of crime; and offering employment support to secure their futures.” 

Wider impact 

As well as evaluating schemes governments are trying, the GO Lab’s international team of researchers, policy specialists and data experts undertake academic research and policy engagement focussed on outcomes-based partnerships. The GO Lab also hosts a global knowledge database where lessons from different partnerships are collected, pulling together experiments in better ways to help people from all over the world. ‘INDIGO’ covers over 40 countries and holds data on over 1,000 organisations and projects. 

The GO Lab’s research has made important scholarly contributions. Its academics’ recently published OUP book (Contracting for Public Value) is described by Nobel Laureate in Economics Oliver Hart as “a terrific analysis of how to procure public services”, while a 2024 paper by three of its researchers won the Bleddyn Davies Prize (acknowledging scholarship of the very highest standard by early career academics). A programme of policy engagement and capacity-building ensures the knowledge generated through such research is actively used by policymakers and practitioners across the globe, and includes an executive programme on leading cross-sector partnerships. 

The GO Lab has hosted international workshops and exchanges with governments from Denmark, Canada, Singapore, South Africa, India, Pakistan, Malaysia, Australia, and Brazil, and contributed to global platforms including high-level discussions with the UBS Optimus Foundation and a UNESCO and UEFA workshop on social impact investment in sport. A partnership with the ShikshaLokam Foundation in India will research how governments can enable grassroots social movements, with a focus on education. 

And back in the UK, building on its track record of working with the government as they experiment with different approaches, the GO Lab will next be the learning and evaluation partner for the Test, Learn and Grow programme, a £100m initiative designed to reform public services by using localised experimentation. 

Intricate insights, crucial topic 

The GO Lab presents me a challenge as a research communicator. I struggle to illuminate the lab's detailed findings, or make them compelling for anyone not working in the area. The team dives deeply into questions around payment structures, contract management, data collection, and other topics that require considerable background knowledge to understand. Furthermore, their insights are context-specific: what works well for people at risk of homelessness in a particular area of Yorkshire may not translate to a classroom in India – or at least, not easily. 

Still, that need for learning between one context and another is what makes the GO Lab’s scholarship, outreach, and global database of shared knowledge so important. And while generalists might find the details boring, the consequences are anything but. It’s our money as taxpayers that is being spent on providers of public services, in its billions. And it’s our lives that are affected if schools, prisons, housing and health go wrong – or if big issues like climate change aren’t tackled collectively by all sectors. I, for one, am very glad the GO Lab team, and policymakers around the world, are delving deeply into the details of how governments can partner best with other sectors. All our wellbeing depends on it.