Part 3: A Working Definition of Government as a Platform (2019)

By Richard Pope, Formerly of GDS and Fellow at Harvard Kennedy School

“Government as a Platform” (GaaP) has come to mean many things to many people since the term was introduced by Tim O’Reilly in 2011. It has been called a route to better public services, a way to break down organizational silos, a toolkit for civil servants, an open platform to build upon, a new public infrastructure, a shorthand for coproduction of policy, and a harbinger of new institutions fit for the digital age.

In much the same way that the term “smart cities” is used to cover everything from parking apps to delivery robots and invasive digital advertising, there is a risk in using the ever-more important concept of GaaP to cover such a broad range of definitions that sometimes become too vague or may be seen as contradictory. In addition, important considerations, such as public safety and democratic accountability, often appear to be missing from today’s debate.

With more governments around the world adopting a platform approach, the lack of a clear, actionable definition of GaaP could become a problem. An accepted “working definition” of GaaP could act as a useful decision-making tool for civil servants. This article presents such a definition by examining the ways the term has been used, emphasizing considerations of safety and democratic accountability and aiming to resolve some of the apparent contradictions in usage.

Seven Framings of GaaP

The seven framings of GaaP that follow are based on examples from the limited published literature on the subject from government programs around the world that have adopted a platform-like approach and from more general commentary on the subject. It is important to note that the quotes included almost certainly don’t represent the totality of what the people quoted think about the subject. The aim here is to illustrate the variety of definitions that exist rather than present an exhaustive review of the literature.

Better Public Services

GaaP has been described by several governments as a route to better public services. The U.K.’s GDS, for example, adopted the GaaP label in early 2015 for the next phase of digital transformation work that began after the launch of gov.uk. From the early days of the U.K. GaaP program, there was a clear articulation that it could provide a route to developing “brilliant, user-centric government services” that were “more closely targeted at user needs.”

Writing more recently about U.K. government efforts around GaaP, technologist Jerry Fishenden has discussed the latent opportunity for technology platforms to have a “Researching in depth how and where technology could play a genuinely significant role in improving the quality of life for citizens and public employees alike — not focusing on websites but applying the best lessons learned from modern organizational structures and processes, particularly their ability to respond to real-time feedback and make continuous service improvements.”

The Estonian government has talked about using its X-Road platform to provide “invisible services” for 15 significant life events like learning to drive, starting a business, and having a child, where the government proactively provides services — such as automatically sending pension checks when one hits retirement age — without a citizen needing to register for the benefit or service. Singapore’s GovTech agency is following a similar approach and has developed a “Moments of Life” service, which is built on shared government platforms.

In the United States, Aneesh Chopra and Nick Sinai, chief technology officer and deputy chief technology officer, respectively, in the Obama administration, have written about the promise of application programming interfaces (APIs) to enable the federal government to “partner with local government, nonprofits, and businesses to better serve the American people — extending its reach and helping solve our nation’s challenges.”

Removing Silos, Making Things More Efficient

There is a lot of duplication in government. Generally, departments and agencies are vertically integrated, with each running its own version of a system (a public website, a payment gateway, a printing service, an address lookup, etc.) that’s similar but not identical to all the others. Platforms have been cited as a way to remove that duplication and break down organizational silos.

“Siloed approaches to transformation don’t work,” Mike Bracken wrote in 2015. “Reinventing the wheel every single time we build a service has led to far too much duplication and waste. That’s not good enough.”

In the U.S., the Obama administration’s digital government white paper Building a 21st Century Platform to Better Serve the American People makes a case for government to “build for multiple use cases at once.” Similarly, the Australian Digital Transformation agency explicitly uses the term “whole-of-government digital platforms.”

A Toolkit for Government

Another way to explain the potential for platforms to remove duplication is to view cross-government shared components, such as login.gov in the U.S., GOV.UK Pay in the UK and Il Cloud della Pubblica Amministrazione in Italy, as part of a toolkit for designing services.

For example, the Scottish government describes its payment platform project as “building something centrally that is easy for service teams to plug into and re-use, without additional procurement. That saves them time, money and hassle.”

The U.K. GaaP team came to see their work in this light and the focus of their program became making tools for digital teams elsewhere in government. Today, its platforms are part of a “service toolkit” that, along with official guidance, standards, and a design system, help users to build and run government services. This has become a point of criticism. For example, Mark Thompson, Professor of Digital Economy at Exeter Business School, asked in 2015 if the GaaP team was developing “government as a platform, or a platform for government.”

Open Platforms for Anyone to Build Upon

Similar to the “government toolkit” concept, but with an explicitly broader intent, is the idea of government platforms as a set of tools for government and nongovernment organizations alike.

The Indian government’s IndiaStack is a set of components and APIs for, among other things, identity verification, digital payments, and storing official documents. These components are explained in terms that make it clear they are aimed at wider societal transformation rather than government transformation alone. IndiaStack is described as “a set of APIs that allows governments, businesses, startups and developers to utilise an unique digital Infrastructure to solve India’s hard problems.”

In the U.S. too, there is a strong narrative around building open APIs to allow the private sector and others to develop services that meet the “long-tail” of user needs. For example, the U.S. Veterans Administration has begun to build its services around APIs, which it discusses in terms of “Empowering our partners to build innovative, Veteran-centered solutions.”

Digital Public Infrastructure

There is also a narrative of APIs, data, and cross-government components as infrastructure, analogous to government projects in the physical world.

According to the book Rebooting India, which details the development of Aadhaar and other platforms, the former Indian President Pranab Mukherjee made the following analogy: “See, it is just like a railway platform. Different trains pull up at a railway platform, each with a different destination, and people get on and off depending on where they are headed. In the same way, the technology platform is a central location where various state governments, institutions, and citizens can gather. All government services are offered on the same platform, and citizens can enrol for all eligible services in one place.”

The Societal Platform project, also in India, imagines platforms for cities, healthcare, and education as “foundational infrastructure,” and the Government of India’s proposed National Health Stack uses similar terms. As part of its work to encourage better use of data, the Open Data Institute also spurs governments to think about “data infrastructure,” comparing it with a country’s road networks.

New Institutions for the Digital Age

Others have talked about GaaP as providing a route to new institutions fit for the digital age. This definition sees digital government entwined with institutional reform and requiring a reorganization of the very work of government.

“If you were to create government today, you would not build it around large, free standing Departments of State,” writes former U.K. cabinet minister for the Cabinet Office Francis Maude in the foreword to a white paper from the think tank Policy Exchange. “Instead of a series of siloed hierarchies, you would structure it as a platform responding to the needs of the end user. . . . Government as a Platform will not happen without clear direction from the top.”

In a blog post entitled “Making government as a platform real,” Tom Loosemore, a consultant formerly with the U.K.’s GDS, writes,

If you want a natively digital nation, or a state, or a city, or whatever, my message today is you actually need to be bold enough to create some new institutions; institutions that are of the internet, not on the internet.

In India, this narrative has become a reality, with new institutions set up to operate digital identity and taxation platforms. The Goods and Services Tax Network oversees APIs and other infrastructure needed to operate the national harmonized sales tax (which itself required a change to India’s constitution). GSTN is a nonprofit company owned partly by the national and state-level governments. Aadhaar, the identity verification platform, is operated by the Unique Identification Authority of India.

Coproduction of Policy and Services

Some descriptions of GaaP are less focused on the mechanics of government, or the component pieces of platform government. These tend to relate to a general aspiration or opportunity to use digital to engage more citizens in the running of their government through “coproducing” policies or services by having citizens write applications or manage services that leverage underlying government services.

The international Organisation for Economic Co-operation and Development’s 2018 review of Sweden’s digital government defines GaaP as using data to harness people’s creativity and jointly address policy challenges. David Bartlett, the former Premier of Tasmania, has used similar terms. “[W]hile there is a role for experts in government as a platform,” he writes in an essay in Opening Government: Transparency and Engagement in the Information Age, “there is a much more significant role for the non-expert population in co-creating solutions.”

In their comparison of private- and public-sector platforms, academics Lydia Ottlewski and Johanna Gollnhofer define public-sector platforms in terms of of systems that help citizens to connect and to exchange services to solve societal problems.

What’s Missing?

The framings above show a lot of breadth. That should not be a surprise. Tim O’Reilly’s original 2011 article was itself a broad call for government to make the most of the opportunities digitization provides, and we can see the germ of many of the examples above in that article.

However, a key concept that O’Reilly covered is missing from many of today’s conversations about GaaP, and that is how to make sure these new systems are accountable and safe. O’Reilly talked about the importance of “rules of the road” to ensure governments operate in a way that generates the best outcomes without creating unintended harm. He also notes that choices about how platforms operate must be revisited periodically to ensure that this remains the case.

With all we have learned about the impact of commercial digital platforms on society in the past decade, trust, accountability, and safety issues demand more space in the debate around digital government. Though platform government creates new opportunities for better public services and saving money, it also raises questions about the risks of data centralization and the privacy rights of citizens. And platforms bring questions of accountability and democracy: Which government agency should be held accountable when a shared platform fails? If a central and a local government both rely on the same platform, what does it mean for local democracy and devolved power?

The debate around the Aadhaar identity platform in India — whether it was created via legitimate vehicles and whether it violated Indians’ right to privacy — provides insight into these types of issues, which other countries will likely face as they adopt a platform approach to government. These are fundamental questions about how the mechanics of government should operate and be held accountable in the digital age. Until they are answered, aspirations for “coproducing” policies and services feel like a distraction when it comes to examining the role of citizens in these systems. Instead, it is important to focus on ensuring positive outcomes and providing clear routes for accountability.

Why Create a “Working” Definition of Platform Government?

As the above examples show, there are many ways in which the term “Government as a Platform” and the more general concept of digital platforms in government have been deployed since 2011. The question is, is there a working definition that encompasses some of the above, while remaining useful? To answer, we first ask more questions: “Useful to whom?” and “Useful how?”

The intended audience here is digital-services units; that is, organizations set up within governments to design, build, and operate digital services. Such groups are being created in governments around the world, and increasingly, they are developing shared platforms.

A good working definition of GaaP should give such units a framework for their work, helping them to present it as a whole — including the aspects that may be politically difficult or superficially less important. It should name the constituent parts of GaaP that digital-services units will need to consider, while avoiding generalities and vague language. It should also aim to resolve any apparent (but false) contradictions, such as to what degree GaaP is for use inside or outside government, is primarily about technology or institutional reform, or is meeting the goal of efficiency or accountability. It is all of these things and all are necessary.

Ultimately, the working definition should help teams to make decisions about which work to prioritize by guiding them in asking and answering questions like, “Does this project we’re starting fit as part of government as a platform?” and “Is that smart-city-blockchain-widget Consultancy X is trying to sell going to help?”

Proposing a Working Definition

With all this in mind I propose the following definition:

GaaP should reorganize the work of government around a network of shared APIs and components, open-standards and canonical datasets, so that civil servants, businesses, and others can deliver radically better services to the public more safely, efficiently, and accountably.

This definition aims to clarify several things:

1. That the goal of platform government must be to enable radically better services for the public and that everything else is subservient to that aim.
2. That services for the public can be delivered by a range of different actors — government, charities, businesses, etc. — all built on common foundations.
3. That these common foundations are made up of canonical datasets (definitive lists of things like tax rates, licenses, or addresses), open-standards (common ways of modelling data), shared APIs that expose the business logic of government, such as tracking the status of an application, and shared components (things like shared hosting or common design systems).
4. That considerations of safety, accountability, and democracy must at all times be viewed as equal to considerations of efficiency.
5. Finally, that the adoption of platform approaches in government will result in changes to how the work of government is organized. Platform government is, in part, about institutional reform, not just making the way things are done today more efficient.