What is Digital Public Infrastructure and why does it matter?

Last year, India made Digital Public Infrastructure (DPI) a central theme of its G20 presidency, championing how it uses technology to facilitate economic development, improve access to banking and reduce the costs of government services.

DPI in some ways represents a sharp departure from previous conversations around e-government, digital government or even digital transformation. Often those conversations focused on the digitization of specific services. DPI advocates seek to address something both broader and more structural.

DPI seeks to transform not a specific service but provide a digitized version of critical societal functions. This includes enabling individuals to digitally authenticate themselves (have a digital identity or a credential), transact (participate in digital payments), and share data between services.

In our research, we scanned definitional documents in DPI from funders, on-ground implementers, technology experts and development leaders, and found consensus on two key attributes.

Firstly, DPI has an infrastructure capability, meaning it performs a foundational and common role across public and private sectors. Organically, this means DPI needs to be interoperable (i.e. communicate with multiple, diverse modules of technical architecture).

Secondly, DPI serves a public interest – specifically prioritizing access and inclusion – and thus has a governance structure that supports the public interest.

There’s consensus around increasing state capacity and doing so in a fiscally responsible way. The successes of DPI would typically include financial inclusion examples from Brazil and India, as facilitated by access to digital identity and digital payments.

It is true, however, that its true nature remains elusive, and the picture changes radically from one country to another. As governments, multilaterals and technical advisors convene to discuss such infrastructure pieces, we ask: what are effective global benchmarks to build this infrastructure? How can their design reflect variations in state capacity and dimensions of DPI?

The DPI Map offers a global view of the state of DPI. It effectively maps how the national or regional deployment of digital identity, digital payment and data exchange systems (treated as foundational components of DPI) embody technical and non-technical characteristics of DPI.

We’ve gathered meaningful insights on how these infrastructure pieces look on the ground, and where there are gaps in the collective intelligence. For one, we know that DPI is far more widespread around the world than we initially thought. As many as 100 countries have systems to exchange data and information in an interoperable way. Half as many have an identity system that does so.

In addition, because DPIs often build on existing state functions, it is unsurprising that several proto-deployments of DPI are emerging. These systems resemble most, but not all, features of an aspirational DPI deployment. This means they embody the most essential features of a DPI-like system – like interoperability or exchanging data in real-time – but adopt alternative ways to do so. Data exchange systems, for example, can be created through different models: an X-Road adoption (a Nordic open-source solution) is the most sophisticated of them. There are also API database and manager models, as well as enterprise service buses (ESBs).

Worth highlighting is how heterogeneous the governance of these systems can look. DPI doesn’t require being wholly state-owned and operated. We find more evidence for public-interest governance structures, recognizing the role of the market in shaping and scaling DPI. Illustrating this, all payment systems in low-income countries are operated by private operators. Almost half of the systems in middle-income countries are privately operated. All of these are under the regulatory purview of the central bank or financial system regulators. These operate across sectors and include systems that facilitate real-time payments for P2P (person-to-person), and P2M (person-to-merchant) transactions.

Evidence on how heterogeneous DPI systems can look is only just emerging. At least partially, this heterogeneity can be explained through the different motivations that incentivise the development of digital infrastructures in governments.

Other than economic growth, there is an increasing need to establish and extend sovereign control over digital systems. This directly addresses concerns around cyber risks and emphasises a preference for being open source. Relatedly, digital public infrastructure can allow states to shape markets by promoting competition for services built on a public-interest technology stack. This reduces abusive rent-extractions prevalent among monopolies.

Whatever the motivation, momentum around DPI appears to be building. The 50-in-5 initiative, a multilateral effort to aid 50 countries in designing, launching and scaling their digital identity, digital payment and data exchange systems, is a testament to this.

The World Bank, through its ID4D initiative, is on track to aid 60 countries in building their digital identity programmes by 2025 while concurrently facilitating G2P (government-to-person payments) capacities in 35 of these.

DPI is also finding support delivery across outcomes. Recent efforts to compile evidence on its impact make its potential clear: from facilitating financial inclusion for those on the margins to economic growth for broader society. This picture will soon be even clearer as more heterogeneous stories of DPI adoption and clearer ways to measure its impact emerge.