Abstract: 

As digital transformation accelerates globally, Digital Public Infrastructure (DPI) is emerging as a key framework for building interoperable public systems. In an effort to better understand which states are best suited for DPI adoption, this study investigates which country-level conditions, specifically population size, enabling infrastructure, and state capacity, are associated with the successful adoption of DPI. To examine this, a dataset of 80 - 120 countries will be constructed using publicly available global indicators, including both DPI adopters and non-adopters. The analysis will combine descriptive cross-country comparisons, linear regressions for each key variable, and a grouped median-split approach to assess both the individual and joint effects of population size, enabling infrastructure, and state capacity on DPI adoption. This will be supplemented by case studies of countries that deviate from expected patterns. This research is particularly pertinent as digital infrastructure is increasingly dominated by conglomerates, concentrating data in private hands and raising concerns around privacy. DPI offers a public alternative to privately controlled digital infrastructure, but its viability is not universal. By identifying the conditions under which DPI adoption is most likely to succeed, this study provides a framework for governments to assess whether they are well-suited to implement DPI and likely to benefit from such investments.

Research Objective: 

This project will test the hypothesis that DPI adoption is driven by the interaction of three types of factors: scale pressures, enabling infrastructure, and state capacity.

Scale pressures are expected to shape the cost-benefit analysis for DPI adoption. Countries with larger populations face higher administrative complexity and higher costs of delivering public services through traditional systems. At the same time, DPI systems exhibit strong economies of scale due to high fixed costs and low marginal costs. As a result, DPI becomes more economically attractive in larger countries because it improves service delivery efficiency and reduces per-user administrative costs. 

Second, enabling digital infrastructure, such as internet penetration and mobile connectivity, determines whether DPI systems are technically feasible and widely usable. Even if DPI systems are implemented, low connectivity limits uptake and reduces effectiveness. Therefore, stronger existing digital infrastructure is expected to be a necessary precondition for effective DPI adoption. 

Finally, state capacity constraints, measured through government effectiveness and administrative capacity, are expected to be the determinants of whether DPI systems can be successfully designed, implemented, and maintained. DPI requires coordination across agencies, reliable execution capacity, and institutional stability. Countries with stronger bureaucratic systems are therefore more likely to move from feasibility to actual implementation.

Background: 

Over the course of nearly two decades, India has constructed one of the most ambitious digital governance projects in the world. Through a set of interoperable platforms for digital identity, payments, and data sharing, known as the “India Stack,” the government has enabled more than a billion people to prove their identity, move money instantly, and access public and private services online. Today, over 97% of Indians possess a digital identity, and digital payments account for nearly 90% of all transactions by volume, signaling a fundamental shift in how citizens interact with markets and the state (Srivastava et al., 2025). As governments search for scalable strategies to promote financial inclusion and administrative capacity, India’s digital public infrastructure (DPI) is increasingly presented as a model for the Global South. Digital public infrastructure refers to foundational digital systems, such as identity frameworks, payment rails, and data-sharing mechanisms, that allow individuals, firms, and governments to interact securely in a digital environment (Reim & Vega, 2025). Rather than delivering services directly, DPI operates as an enabling layer between physical infrastructure and sector-specific applications including social protection, healthcare, and e-commerce (Marskell et al., 2024). Outside of India, successful cases appear in other populous middle-income countries, such as Brazil, where fixed costs can be spread across hundreds of millions of users, and in small but high- income states like Estonia and the United Kingdom, where higher fiscal capacity substitutes for scale (Reim & Vega, 2025). This suggests that population scale, state capacity, and existing enabling infrastructure may be necessary preconditions to DPI adoption to justify large upfront investments. By contrast, many low-income, sparsely populated countries face a dual constraint that may render DPI economically unviable.

Methodology:

The research will begin by constructing a cross-country dataset of 80 - 120 countries using publicly available global indicators. The sample will be designed to include a mix of low, middle, and high-income countries, as well as DPI adopters and non-adopters, to ensure variation in outcomes and underlying conditions. 

Once the dataset is assembled, the first stage of analysis will focus on descriptive cross-country comparisons, examining how average levels of key indicators differ between DPI and non-DPI countries. This will help identify whether countries with DPI systems systematically differ in terms of infrastructure, governance quality, and state capacity.

The next step is to test the relationship between DPI adoption and the main country characteristics using linear regression. DPI adoption will be treated as the outcome variable (coded 0 = absent, 1 = partial or work in progress, 2 = mass adoption), and each key factor will be tested separately. For each regression, I will compare how DPI adoption changes as one variable increases. For example, whether countries with higher internet penetration are more likely to have higher DPI adoption levels. This approach allows me to directly estimate the strength and direction of the relationship between each factor and DPI adoption independently.

Then, I will examine how the key variables appear together rather than in isolation. This will involve grouping countries based on whether they fall above or below the median for each of the three core variables: population size, enabling infrastructure, and state capacity. This will create eight possible country groups (e.g., high population-high infrastructure-high state capacity or low population-high infrastructure-low state capacity). I will then compare the prevalence of DPI adoption across these groups to determine whether countries with strong performance across all three dimensions are more likely to implement DPI, as the theoretical framework predicts, or whether other combinations of conditions are also associated with successful adoption.

I will also include a small set of case studies based on my findings. These will focus on countries that do not fit the expected patterns, such as countries with strong enabling conditions but limited DPI adoption, to better understand contextual factors that are not fully captured by the tested variables.

Potential Impact: 

As digital infrastructure becomes more concentrated among a small number of private companies, concerns over data privacy and the erosion of public oversight have grown increasingly urgent. The privatization of core digital systems places enormous control over citizen data in corporate hands, creating structural vulnerabilities that governments and civil society are ill-equipped to address. DPI presents a counterweight to this trend, offering states a pathway toward reclaiming sovereignty over foundational digital systems. Understanding when and how DPI adoption is most likely to take hold equips policymakers with a practical lens through which to evaluate their own readiness and the potential returns of such investments.