NextBillion - A WDI Publication

“[Digital public infrastructure] (DPI) is not an issue of needing deep pockets: It’s about having deep conviction.” With this statement, Nandan Nilekani, the architect of India’s DPI stack, captured a central truth from India’s experience: While the financial cost of building DPI may be modest, implementation requires political will.

The World Bank defines DPI as “an approach to digitalization focused on creating ‘foundational, digital building blocks designed for the public benefit.’” Within this broad umbrella term, our research has focused on two of the three broadly accepted categories of DPI: instant payment systems (IPS) and open finance (the third category is digital identity). IPS are networks which consumers can use to send payments in close to real time across financial institutions, while open finance networks enable consumers to share their transactional financial data with other firms in a secure way that’s under their control, in order to access other services such as credit.

Our conclusion from researching IPS and open finance schemes around the world suggests that Nilekani’s dictum is incomplete. DPI may not require deep pockets — but it does require long ones. In other words, sustainability at scale depends less on upfront funding than on how ongoing costs are allocated over time. This makes DPI sustainability fundamentally a governance issue, centered on how pricing policies are set and adapted to distribute long-term costs across an ecosystem in a credible way.

Three levers of DPI sustainability

Across IPS and open finance, three policy levers determine whether DPI can be sustained over time:

• Pricing rules — who pays, how much and under what conditions
• Participation mandates — who must join and contribute
• Governance and ownership — who sets and adjusts the rules

These levers operate together to allocate costs across three groups: participants, end users and government. Sustainability arises when this allocation remains credible over time — especially as systems scale, risks evolve and new use cases emerge.

From deep pockets to long pockets

The expansion of IPS and open finance has been rapid: More than 120 instant payment systems are live globally as of 2025, and close to 100 jurisdictions have made commitments to open finance.

The upfront costs of these deployments, while not trivial, are typically manageable, and usually involve mainly the capital expenditures to design the initial system and develop related software and governance systems. IPS infrastructure can cost upwards of a few million dollars, and early open finance implementations have ranged from single-digit millions to tens of millions annually. These amounts are relatively modest: To put this in perspective, a single large bank in India such as ICICI or SBI spends hundreds of millions of dollars each year on IT, and large tech companies globally are investing trillions in new data center capacity. In this context, Nilekani’s observation still holds: DPI indeed does not demand deep pockets at the outset.

But focusing on upfront costs obscures the real challenge. DPI systems generate ongoing expenditures that grow with scale and usage. These include:

• Continuous investment in system capacity and resilience
• Rising costs of fraud detection and mitigation
• Costs of regulatory oversight and dispute resolution
• Expansion to new use cases and sectors

Experience from DPI systems such as Brazil’s Pix shows that fraud management alone can become a major and rapidly increasing cost for the ecosystem. These expenditures are distributed across operators, participants and regulators — and they persist indefinitely.

This is what we mean by long pockets: not the capacity needed to finance large initial funding, but credible mechanisms to finance a long tail of costs over time, while maintaining incentives for participants and trust among users.

Who pays for DPI in the long-term, and how?

If sustainability depends on long-term cost allocation, the key question becomes: Who pays, and how?

Historically, retail payment systems — the infrastructure that enables the transfer of electronic funds — operated on a “club good” model: Banks funded them as a collaborative initiative from which they also extracted benefits. Banks funded the shared infrastructure — the central IT switching software, and the hardware and interfaces used by participants — collectively, and they expected to recover costs through fees charged to users. This model aligned incentives: Banks had an incentive to keep the costs of the infrastructure low, but also not too low, since they were funding it. It also relied on a relatively closed ecosystem.

Over the past decade, that model has shifted. The entry of non-bank players — fintechs and mobile network operators — has made voluntary cost-sharing more complex, while incumbent banks have often been slow to collaborate. In response, central banks in many countries have taken a more active role: mandating participation among certain types or sizes of financial institution, shaping pricing rules, and in some cases owning or operating IPS infrastructure directly.

At the same time, in many markets outside of India, development partners have supported early-stage DPI through concessional funding, and by providing open-source solutions covering the suite of interoperable services involved in DPI — both of which help to lower the initial barriers to setting up this infrastructure.

This combination of public leadership and expanded participation has often been accompanied by explicit pricing mandates, especially for end users. In a sample of IPS launched since 2016, more than half include such mandates, with a majority of those setting consumer prices at zero.

Open finance has followed a similar trajectory. Early regulated open finance or open banking regimes typically prohibited charging for data exchange. Instead, governments and large financial institutions funded the initial implementation. Over time, however, regulators have begun allowing more diverse pricing approaches — especially as usage scales, industry and consumer benefits are demonstrated, and costs become more significant.

Pricing as the core allocation mechanism

Because no DPI system is truly “free,” pricing policy is ultimately about allocating costs across stakeholders. This is the central governance challenge for DPI.

Three broad pricing approaches are emerging:

• Zero-price models, often supported by government subsidies or cross-subsidization by financial institutions with other income streams
• Cost-recovery models, where participants or users pay directly
• Hybrid models, combining free and paid elements across use cases or thresholds

Each of these approaches reflects different policy priorities and trade-offs.

In IPS, zero-price models can accelerate adoption, but they may favor large incumbents able to cross-subsidize, or well-funded entrants. Cost-recovery models may improve sustainability, but they risk excluding smaller players. Hybrid approaches — such as offering person-to-person payments for free if the amount falls below a certain threshold, while charging merchant fees — attempt to balance these objectives.

In open finance, similar patterns are emerging. Basic data access may remain free to promote innovation, while premium APIs offering high-volume usage or additional data or data analytics services may be priced at fixed rates or at open-market pricing. These high-volume users include financial service providers (e.g., lenders) that buy greater access to the open finance scheme, which lets them access a borrower’s full transactional data across other providers in the scheme as part of a credit scoring process authorized by the customer: They have both a business case for this data, and enough revenue to pay for it. These models may present challenges to larger institutions due to set-up costs, but they increasingly aim to distribute costs more equitably as ecosystems expand.

Convergence of IPS and open finance

IPS and open finance are often discussed separately, but in practice they are converging. Payment initiation is now a core feature of many open finance systems, while data exchange underpins advanced payment use cases.

This convergence creates new possibilities for cross-subsidization in DPI:

• Revenues from payments can support data infrastructure
• Charges for data access can help fund payment systems

Different jurisdictions around the world are experimenting with different approaches to this cross-subsidization. In some cases, payment initiation is free while data exchange may be monetized; in others, the reverse applies. Hybrid models are increasingly common, reflecting local policy priorities and market conditions.

The key insight is that financial sustainability is no longer confined to a single system: It is a property of the combined DPI ecosystem.

Sustainability over time: adapting the pricing model

A critical lesson from open finance is that pricing models must evolve as systems mature. In the early stages, concentrating costs on governments and large incumbent financial institutions or tech platforms may be the simplest way to launch a system. But as participation broadens and transaction volumes increase, this approach becomes less viable — and less equitable. Over time, costs must be distributed more widely across participants and use cases.

This requires governments to build adaptive mechanisms into their DPI that allow pricing and funding arrangements to change. Therefore, a sustainable DPI typically combines:

• A credible, long-term (five year +) funding horizon, giving participants confidence in system continuity
• Diversified funding sources, reducing reliance on any single group
• Adaptive pricing rules, capable of responding to growth and new risks

These features emphasize adaptability: Static models — whether fully subsidized or rigidly cost-recovery based — are unlikely to remain effective as systems scale.

A mindset shift for regulators

If pricing is central to the sustainability of this infrastructure, then DPI requires a shift in regulatory thinking.

In networked sectors such as electricity and telecommunications, regulators routinely set tariffs to balance investment, efficiency and access. In contrast, financial regulators have traditionally focused on risk management and consumer protection, often avoiding direct involvement in pricing.

DPI challenges this approach. Because pricing rules shape adoption, competition and long-term viability, they cannot be left entirely to operators or market forces. Nor can they rely indefinitely on implicit or explicit subsidies.

This raises institutional questions. In some contexts, financial regulators may need to develop new capabilities to oversee pricing and market structure. In others, there may be a case for involving competition authorities or establishing specialized digital economic regulators.

The appropriate model will vary by country, but the underlying requirement for DPI regulation is clear: Credible, legitimate governance must underpin pricing decisions over time.

Conclusion

The distinction between deep pockets and long pockets in DPI is important to understand, as these systems continue to gain traction globally. The challenge is not simply to fund initial deployment, but to establish durable mechanisms for allocating costs as systems grow and evolve.

This makes sustainability a core question of governance: how pricing rules are set, who participates in funding, and how these arrangements adapt over time. The answers will differ across countries, reflecting policy priorities and market conditions. But the question itself can no longer be avoided.

Without long pockets — and the governance structures to manage them — DPI cannot sustain the scale, trust and innovation on which its promise depends.

David Porteous is the founder and CEO of Integral: Governance Solutions; Rafe Mazer is the Director of Fair Finance Consulting.

Photo credit: metamorworks