The concept of digital public infrastructure (DPI) – a system of state coordinated shared, multi-use digital services – has gained significant traction as a distinct approach to government-led digital transformation since emerging as a priority area with India’s G20 Presidency in 2023. But both the definition and the achievements of DPI remain uncertain. In this blog, Sumedha Deshmukh and Stephanie Diepeveen discuss the policy questions involved in implementing DPI systems.
Since 2023, global initiatives and funding have enabled country-level digital public infrastructure (DPI) exploration and adoption worldwide (especially in small and low income states). DPI has continued to be featured in G20 agendas, in high-level government strategies, like the UK’s “Blueprint for Modern Digital Government”, and global events focused on sustainable development.
Though the policy momentum for investing in DPI is growing, conceptual and definitional ambiguities remain. Most discussions anchor on the example of the “India Stack”, which is based on digital identity, digital payments infrastructure, and a data exchange platform. However, other services like digital wallets and open networks for digital commerce, and use case-focused services within healthcare and education, are also considered within the DPI approach. DPI can also be connected to related approaches like the platformisation of government.
Further, what is being achieved through DPI is couched in some uncertainty. Based on early examples and the promise of de-duplicating intra-government efforts, proponents argue these foundational systems can reduce costs, increase access to services, and enable new forms of market activity. Yet, developments in digital platforms and marketplaces and evidence from early implementations raise questions about governance, market dynamics, and state capacity.
Governance
First is the fundamental question of governance: How can public interests be protected while enabling innovation? For now, there may be a gap between DPI’s democratic promise and its practice.
While presented as a transformation responding to public demand, many initiatives are in fact shaped by networks of philanthropic funding and technical expertise. Citizens often have to use these systems to access essential services, even amidst concerns about distributional gaps, surveillance and repression potential, and security vulnerabilities.
Moreover, several core concepts central to DPI’s appeal to governments – like “sovereignty” and “openness” – remain ill-defined. In some conceptualisations, sovereignty is equated with the power of domestic versus foreign firms, or the power of any actor to use technology for its empowerment. Both definitions differ from academic debates about political sovereignty. And while “openness” is emphasised as if it is an obviously good feature, the term’s meaning has evolved. The distinction between open source, open standards, and open APIs often becomes blurred, allowing systems with limited potential for scrutiny or modification to claim openness. For instance, the source code released for one DPI-branded contact tracing app was not the code for the actual app – compromising meaningful oversight of privacy protections. Some authors have highlighted that the language of “openness” obfuscates commercialisation or that many parts of the system are proprietary (referred to as “open washing”). This lack of clarity extends to newer sovereignty-centric digital industrial policies.
Market structure and competition
A second critical consideration is market structure. DPI promises “public rails for private innovation,” but early evidence suggests it may contrarily on the contrary enable new forms of market concentration.
Despite visions of an “hourglass model” whereby government enables diverse private sector participation above (e.g. applications) and below (e.g. networks, devices), there are often limited providers in practice. Even with open protocols, multinational corporations control approximately 85% of the market share at the service layer of one payment system, while the public app has much lower usage numbers. Major cloud providers have also hired “specialist DPI teams” and released offerings like “DPI in a Box” to capitalise on the population-scale market opportunity.
The question of data governance adds complexity. While DPI systems generate enormous amounts of transaction and interaction data, frameworks for managing this public resource remain underdeveloped. With concentrated digital payments, valuable transaction data remains in private hands. Another DPI-based application generates significant mobility data, but its open data portal only reports the app’s usage and not socially useful metrics such as congestion – limiting governments’ ability to use open data to make meaningful social interventions.
State capacity
Countries must also consider what capabilities they need to oversee and maintain digital infrastructure effectively – and, relatedly, what public accountability is required.
Smaller nations, in particular, face implementation challenges. For example, one country’s national ID system has registered only 6.4% of the population after three years. ID enrolment elsewhere has faced repeated postponements due to technical and financial constraints. Hundreds of thousands have faced suspension of DPI-dependent services due to technical glitches or not registering biometrics with the government. In some cases, the privatised nature of funding and service provision means that the channels for public accountability are limited.
Unlocking the promise of DPI
Nonetheless, there are also promising aspects of, and approaches to, these various challenges. Publicly runPublicly-run digital payment systems allow localisation amidst increased scrutiny over card networks’ global rent-seeking. DPI projects have provided innovative examples of government coordination and engagement of startup communities. As DPI systems mature, they are exploring pathways to sustainability. One country, for instance, strategically earmarked 1% of GDP commitment for ICT and parliamentary oversight over the systems (which is possible for wealthy countries but may not be for others where there are other pulls on budgets).
As governments worldwide consider their approaches to digital transformation, these policy choices become increasingly important. The next phase of DPI must then move beyond general claims about DPI’s potential toward concrete frameworks for ensuring DPI serves public interests.
Looking ahead, several priorities emerge for those involved in government and multilateral policy discussions about DPI:
- A demand-based approach to government-led digital transformation, which addresses real needs and capacity issues by country, and is tailored to each context. This entails creating space in global forums to discuss the unique properties of states and specific opportunities, like the potential of DPI data as a public resource.
- Design of rigorous evaluation criteria that are aligned with demand in specific contexts. This likely requires opening up to different frameworks, e.g. considering inequality and competition alongside efficiency.
- Shifting how DPI interventions are designed, with scope to explore diverse interventions that meet concerns about openness, accountability, and so on.
- Ensuring funding models and priorities are diverse enough to create space for diversity in demand-led DPI interventions.
This blog is based on a working paper written by the authors along with anonymous collaborators, available at: https://bit.ly/Global-DPI-Agenda
The views and opinions expressed in this post are those of the author(s) and not necessarily those of the Bennett Institute for Public Policy.
