The Covid-19 pandemic has shown how important a strong state is. Around the world, governments have had to take steps to protect people, their health, their livelihoods and their futures.
But during this time of great responsibility, the gap between expectations and capacity has never been clearer. Governments have struggled to scale their response to the pandemic, while deep structural inequalities have been highlighted across government institutions and public services.
This is why the primary challenge for governments in the next decade is how to reconfigure the state using technology to make better decisions, better resolve crises, and improve people’s lives. This paper sets out the progressive case for reforming the operational model of the state so that it looks more like a modern, tech-enabled institution and less like a 20th-century bureaucracy. We offer some practical suggestions about how to begin that process and where to look in the world for inspiration, but this is only a starting point. We want to provoke a wider discussion about how to deliver radical change rather than merely incremental reform.
Today’s model of government is insulated from change. State silos block wholesale reform, and top-down hierarchies can no longer deliver in an era of overwhelming complexity. Progress relies on those in power overcoming immense inertia to help small bits of the system catch up, instead of the underlying conditions encouraging continuous improvement.
The costs of this are stark. Structural inequalities go unaddressed, public services fail to deliver the best possible outcomes for citizens, and scarce resource is wasted instead of being used to support those most in need. But proper implementation of technology should break the constraints inherent to this industrial model of government.
Technology reduces the barriers to scale, personalisation, iteration, feedback, experimentation and collaboration. Platforms open up service delivery to a wider range of actors to compete on quality. Digital marketplaces can reduce procurement costs and raise standards. Common functions can be expressed in infinitely scalable software, built once, and reused elsewhere to save precious resource.
Most importantly, this generates material benefits for people’s lives. Some of the best governments and organisations already illustrate the possibilities. Estonia has built proactive child benefit payments, delivered automatically and without administrative burden, all underpinned by a flexible, modern data infrastructure and secure, user-centric digital identity. Amazon consistently provides an excellent service, with all the background complexity of the technology stack, logistics and organisational processes oriented entirely around users. Taiwan uses novel online and offline tools to promote participatory governance, refocusing the policymaking process around public deliberation rather than treating it as an afterthought. While no country or organisation has mastered the tech revolution, and none of these cases provide a ready-made blueprint for others to simply follow, they share some common traits: investing in foundational capabilities made possible by technology and organising around the needs of people, not historical bureaucracies.
Several obstacles stand in the way of industrial-era governments, like the UK, making the necessary change. Siloed organisations mean far-reaching reform requires extraordinary, long-lasting consensus that is unavoidably elusive. A poor delivery record for government IT projects and public concerns about power and privacy mean change is high-risk but low-reward. Reform is seen merely as technocratic rather than a key doorstep issue, even if a new approach might deliver a step-change in the quality of public services. In turn, leaders must spend both financial and political capital, but the benefits are often diffuse and rarely accrue within a single political cycle. These are not conditions that encourage the reform necessary for government to meet its ever-evolving responsibilities.
What conditions will allow governments to reach escape velocity and reconfigure public institutions to deliver effectively, creatively and sustainably going forward? This must be the principal question in relation to government for policymakers today. Merely trying to reinforce 20th-century government with technology isn't enough. Better websites, e-petitions and remote working are positive but fall far short of what’s required.
For some, the answer is to break the whole system (including the constitution) and rebuild from the ground up under the direction of a small group of authority figures. In the process, they are prepared to trample norms, people and experience in order to assert full control. Although this approach seeks to improve the ability of the system to cope with an increasingly complex environment, it is extremely risky. Ruthless confrontation may get the system moving in the short-term but has little resilience once political will runs out.
The alternative, progressive approach asserts that accountability and trust still matter, but knows this must be coupled with optimism. This vision must start with a radically more open approach to government that puts users first. This means government and public services must be subject to the same pressures and consumer demands that push other organisations to evolve continuously; that in the internet era, no single actor can succeed alone; that authority and credibility are earned from delivery not conferred on arrival; and that resilience relies more on agility than artificial order. It should also involve far greater engagement with the public: Government is well placed to aggregate broader societal needs and will need this insight to assess how others are meeting them.
In practice, governments must shift from delivering what they always have to ensuring people’s needs are met in the best possible way. This should include opening up delivery to a greater range of partners from both the private and charity sectors where they can provide a better service. Users should have the right to choose from an array of modern services, and those services should be able to integrate fully with a publicly controlled data infrastructure with public-interest guidelines that govern access. This means extending the model of CityMapper, Xero and many others – which each focus resolutely on the varied needs of users and can often execute faster or better than government itself – and allowing for tighter integrations that unlock new possibilities. Consider Universal Credit as a platform, with Citizens Advice providing an improved interface for people to access support. There will always be parts of government that don’t fit this approach, but the point of this shift is that a bias towards empowering users over bureaucracies should provide the necessary push for states to keep pace with fast-evolving needs.
Radical models of innovation that disrupt established ways of working are not always comfortable. But long-term stagnation has an unacceptable impact on the state’s ability to meet people’s needs. The challenge is therefore to expose the key trade-offs and resolve them as best as possible. Three should take priority:
1. How states can give up control to encourage innovation while protecting quality and in-house capacity
A more open approach to government can bring real benefits in terms of innovation and service delivery, but also sharpens the requirement to ensure equitable outcomes. Similarly, the pandemic has shown the value of protecting resilient, responsive in-house state capacity instead of letting this degrade over decades.
Resolving this trade-off requires a framework to assess where and how to encourage a more open model of policymaking and delivery. Services where the stakes are high or where real-time iteration is hard may require more work to be front-loaded, but technology creates new levers for quality control for the remaining class. Governments can:
Control software services’ access to government platforms and datasets via public-interest guidelines which govern activity, e.g., App Store guidelines.
Evaluate and iterate services in real-time by setting transparency requirements and leveraging analytics and user feedback, instead of relying solely on ex ante permission, e.g., eBay and other digital marketplaces.
Focus contracts on outcomes rather than outputs, measured over time using data, to promote experimentation while upholding standards, e.g. Social Impact Bonds.
Robust quality control should mitigate the risks of opening up services to competition. But this is not a licence to relegate in-house capacity. This pandemic has shown the tension between expectations that states act as insurers of last resort, financially and logistically, and “effective governance [that] cannot be conjured up at will”. The lesson is that governments must double-down on some competencies where they have a special responsibility. At minimum this would include tail-risk planning and providing foundational technology infrastructure for others to build on. Government is also well placed to engage citizens and aggregate their needs – it should use that insight to evaluate how a more open ecosystem of service providers is meeting them.
2. How to reorganise the state around scale economies underpinned by technology while moving delivery closer to people’s lives
Leaders should recognise that it is consistent to seek economies of scale in some functions while also unbundling the services that are built on top. Indeed, selecting from a suite of well-maintained, constantly improving, and cheap and easy-to-use platforms, rather than needing to procure or build the same common components for each new service, should free up capacity and resource to improve each locally tailored service. Instead, the risk is about misallocating power, so for this model to be sustainable, any technology stack should truly be open to everyone on the right terms, not just those that the centre of government favours. For example, GOV.UK Notify or Ordnance Survey’s Open Address System are maintained centrally but are either not available to those outside government or only with significant cost and friction.
Radical organisational change will also be necessary but will look different depending on where responsibility lies in government. For central government, policy should be organised around portfolios of multi-disciplinary teams that operate closer to users. But an alternative model will be needed for delivery, which accounts for 85 per cent of civil servants in the five largest departments. Local government is different once again.
The starting point must be recognising that top-down, centralised departmental hierarchies lost their ability to control the external operating environment long ago. For frontline services, operations vary greatly, so instead of seeking artificial order with one-size-fits-all reorganisation, the focus should again be on setting the right conditions for optimal services to come to the fore. This requires recognising that government’s monopoly on service delivery has more to do with history than principle. States should organise around users and provide the foundations that allow new services to operate, while letting go of controlling the last mile of service delivery. A better way forward is a more collaborative approach that encourages communities, charities and companies to design more tailored services, and enables people to choose those which best meet their needs.
3. How to better listen, engage with and adapt to peoples’ views without undermining the basic tenets of representative democracy
A core part of product and service design both in business and in the public sector is scoping based on user needs. With modern software, constantly updated and refreshed with improvements and features, this process is continuous. Users generate data which is fed back into the product to refine and iterate again. The product launched on day one will be unrecognisable from the product delivered two years later.
The challenge for governments is how to make better decisions, and more specifically in the modern operating environment, how to continuously make good decisions by setting the right parameters and design for the systems put in place. This is difficult for governments who launch with great fanfare and are held to account immediately. This naturally makes the way governments operate risk averse.
Service design in government already shares many of these feedback loops by virtue of digital delivery on apps and websites, but the policymaking process in government looks like the policymaking process of the last 20 years, only with a new digital interface.
Departments formulate policy based on manifesto commitments, ministerial direction or necessity and then consult before going through something similar to a green paper, white paper, draft bill process. Iteration, the cornerstone and opportunity of the internet era, is naturally limited by a closed process.
Governments of course conduct quantitative and qualitative research to inform their decisions and consult, but it is a front-loaded process where the delivery may be separated from the core insights by a number of years. Existing consultation approaches also remain out of step. They have high barriers for engagement, do not track views as they evolve, miss crucial segmentations and are routinely ignored. But modern technologies should allow governments to use their unique position to engage the public and aggregate views. Just as with user research in service design, these insights should also only inform, rather than substitute, the policymaking process.
A closer relationship between the citizen and state, properly scoped, should help leaders understand the plurality of public opinions on thorny policy issues, in turn potentially addressing the growing disconnect between public institutions and those they represent. Internet technologies have amplified the voice of the networked public, creating political pressure that’s here to stay. It’s far better to channel this constructively, using technology to be more responsive to people’s needs and underrepresented voices, than allow it to take on a destructive approach and be externalised elsewhere.
Getting from the status quo to a more open model will be challenging, given the internal constraints of public institutions. The following ideas are a starting point for how governments could break some of these constraints and should be considered a foundation for wider debate:
Ensure every new service publishes and documents an Application Programming Interface (API) to enable platform-based innovation and deliver choice to users. This should be based on a micro-services architecture to promote smaller, modular contracting which can improve value.
Provide open, policy-agnostic platforms for common needs, such as payments, identity, notifications, split testing and simulation, that private- and third-sector innovators can integrate into services.
Focus procurement around a single, user-centric digital marketplace to simplify access to public tenders for small providers and establish common standards for issues like subcontracting transparency.
Replace hierarchical policy departments with portfolios of multi-disciplinary teams, organised around user needs and able to move around easily.
Decouple data functions from policy teams and create a network of Data Registrars independent of departments who are empowered to make data registers findable, accessible, usable and trustworthy – this will help to improve leaders’ situational awareness.
Launch a two-way fellowship programme between the tech and public sectors, based on the US Presidential Innovation Fellows programme, to accelerate learning from novel operating models.
Establish a presumption of choice for service providers. Just as app stores enable users to choose from a wide range of vetted but competing services to meet their needs, government should seek to dramatically increase the choice available to citizens seeking to get their needs met. The state should provide the platforms and other enabling conditions to facilitate this.
Aggregate purchasing through a single procurement marketplace to incentivise competition and bring data from various pre-existing procurement portals into one place to improve government monitoring of value and quality.
Create a more scalable funding model for services that disrupt industrial-era assumptions. For example, virtual GP services shouldn’t be stalled just because old funding formulae link delivery and geography.
Create a new Digital Citizen Service built on the vTaiwan model to deliver scalable feedback and citizen engagement for all services, modernising democratic engagement as GDS did for service delivery.
Provide insights openly as a public good: Aggregate public sentiment and real-time data about policy challenges and make them available both within government and beyond, e.g., Mind or Headspace can improve their understanding of loneliness to iterate mental-health support.
Use public engagement to assess in real-time the outcomes of partner services to ensure that people’s needs are being met and that users have avenues for redress where necessary.
It’s clear that government needs to change. The structural stasis of recent years constrains the state’s ability to deliver for citizens already, and the challenges of the tech revolution will only amplify this responsibility. But the debate remains stale because it is focused on the wrong questions. Politics is stuck arguing over who will spend more money or which needs matter more, rather than rethinking the system which forces these trade-offs in the first place. No one is interrogating the question that really matters: In today’s world, what is the best way for governments to meet their responsibilities?
The way forward is to focus on establishing the conditions which force continuous improvement over time: Provide platforms to open up delivery to a range of actors competing on quality, empower users to push organisations to organise around their needs, articulate public-interest guidelines to regulate this more open approach to service delivery, and put understanding people’s needs and views at the heart of governance to restore faith in public institutions.
The real test will be how this can improve the lives of people in a material way and restore trust in the ability of public institutions to help meet people’s needs. There remain some outstanding challenges and trade-offs, and the ideas set out here are only a starting point. But to make progress, we must recognise the fundamental need to escape the limits of the status quo. Only a radically new approach can put us back on the right path.