Learn about our approach to mapping DPI as a concept and real-world implementation here.
As governments, multilateral actors, and technology implementers collaborate towards making DPI a reality across the globe, the DPI Map asks: how prevalent is DPI across the planet? This project seeks to answer this question.
To do so, we had to wrestle with some fundamental issues. How does one define DPI as a concept? How do you translate any definition into foundational qualities that can be objectively measured across different national contexts around the world? This piece seeks to explain our approach in addressing these questions, to enable researchers, technical advisors, policymakers and advocates to understand the context of this data set and its limits, so they can effectively use it in their work.
The resulting ‘Measuring DPI Framework’, offers a first take on translating the normative concept of DPI into a measurable framework to evaluate real-world implementations. The framework has two layers.
The first layer is inductive. It proposes a set of normative attributes, a set of common technological and governance qualities that describe how DPIs should be designed.
The second layer is deductive. It is composed of a set of outcome indicators that describe how attributes can be realised in real-world deployments.
We arrived at both attributes and outcome indicators through desk investigations and expert consultations (see illustration below).
Our framework proposes four attributes identified through a literature review: interoperability and extensibility; transparency, accountability and oversight; privacy, protection and security; and inclusion and non-discrimination. These attributes describe the qualities that define artefacts - like digital identity systems - as public and not private infrastructures. In addition to the four attributes, an essential condition to implementing DPI is the supportive ecosystems capacity and coordination capability. A final component for assessing DPI’s attributes is through the adoption of DPI, or the extent to which actors other than the DPI operator leverage the DPI.
To assess the attributes of real-world deployments, the DPI Mapping team had to determine: how do we know if DPI is interoperable? Do we look at its technical standards, or whether the architecture is made accessible through documentation? Or both?
The resulting set of outcome indicators sought to elaborate on the attributes by describing effective norms, rules, technologies and processes observed in real world deployments and suggested by expert practitioners and area specialists (e.g. privacy and security experts).
Each foundational DPI (i.e. ID, payments, data exchange) has both unique and shared indicators. (38 for ID, 20 for payments, 20 for data exchange).
It’s worth mentioning that not all indicators are equal.
In version 1.0 of this framework, we present an exhaustive version of indicators for all attributes. These indicators however can be divided into subjective and objectively measurable (based on our data scoping methods, accessible here). We further divided indicators into two other categories: ‘must haves’ referred to indicators that more directly assessed the essence of an attribute, while ‘good to haves’ reflected those that did so more indirectly or partially. In an effort to both be more objective and make effective use of our limited resources, version 1.0 of our map collects data on only measurable must-have indicators.
We purposely share all the indicators we derived for several reasons. First, we hope users and contributors might suggest edits that would shift non-measurable indicators into being measurable. Second, that our list of indicators might informs others research. And third, just because an indicator is non-measurable does not mean it is not important. We hope that policy makers and DPI developers will find all our indicators useful when developing or deploying DPIs.
Using the framework
We hope that researchers, policymakers, technical advisors, and advocates will use this framework to advise and inform the measurement of DPI as a concept.
A complete list of works consulted directly or indirectly in developing this framework will be accessible in our upcoming whitepaper. Feedback is welcome and will be integrated with subsequent versions.
Download the framework (.xls)
Citation guidelines
Eaves, D. & Rao, K. (2024). Measuring digital public infrastructure framework. Unpublished manuscript.
Digital public infrastructure is understood as ‘society-wide, digital capabilities that are essential to formal participation in society and markets as a citizen, entrepreneur, and consumer in a digital era’.
The following DPI components are treated as fundamental to a country’s DPI infrastructure, and form the primary focus of this project. Nevertheless, we acknowledge that this list may expand as countries progress in their respective journeys towards sophisticated DPI. Digital ID: A secure, interoperable system that includes digital authentication (not just enrollment) to access services provided by both public and private sector actors. Digital Payment: An interoperable, open-loop system that securely connects banks (money custodians), payment rails (routes the payment messages) and front-end payment applications (apps and fintechs) for end-users. It is ideally instant (i.e. instant retail payment system). Data Exchange: An interoperable system that enables individuals and public and private sector entities to exchange and digitally verify data with one another securely.
The Map of DPI deployments puts the ‘Measuring DPI’ framework to test, combining a mix of data points from publicly accessible data sources. Typically, these sources include government-reported updates through their websites, press releases, as well as data from other third-party reporters (World Bank, UN, regional development banks, regional technical capacity-builders).