Menu
Burke Index
RESEARCH
11.11.2026, 08:52
The Paradox of Digital Scale: Estonia and Kenya

Abstract 

A state with a small domestic market may enjoy greater freedom of action in the digital domain if it retains control over interoperability rules, has replaceable infrastructure components, and can restore critical functions. The comparison between Estonia and Kenya, however, supports this proposition only conditionally: a distributed architecture does not eliminate shared intermediaries, just as a mass-market commercial platform does not imply an absence of state control.

A Note on the Index Scores

In the data provided, Estonia and Kenya have identical scores, not only in the aggregate but across all seven dimensions.

Dimension Estonia Kenya
Political 61 61
Economic 54 54
Technological 70 70
Informational 55 55
Cultural 59 59
Cognitive 57 57
Military 33 33
Total out of 700 389 389

The contrast in scale is substantial: Estonia had approximately 1.36 million inhabitants as of January 1, 2026, whereas Kenya’s regulator recorded approximately 54 million mobile-money subscriptions by the end of June 2026. The latter figure counts subscriptions rather than unique individuals; it therefore cannot be directly compared with population figures or the number of citizens using public services.

The size of a user base primarily indicates the extent of infrastructure adoption. For the study of sovereign capacity, the more consequential questions are who can change the rules governing that infrastructure, replace a critical provider, and maintain essential functions following its failure.

The dependent variable is thus not the “level of digitalization” but the state’s operational freedom under infrastructural constraints. Scale of adoption should be examined separately—as a potential source of resources and bargaining power, but also of systemic dependence.

Two Architectures of Digitalization

The Estonian model organizes interaction around shared rules for data exchange. X-tee employs X-Road technology, enabling public and private organizations to interact through a common interoperability infrastructure. Within this architecture, data are exchanged directly between participants, while the ecosystem operator establishes participation rules and maintains central components.

The Kenyan model exhibits a different configuration: at the end of June 2026, Safaricom accounted for 88.8% of mobile-money subscriptions. This indicates a high degree of concentration in that particular segment, not the company’s share of all payments nationwide or the magnitude of potential losses in the event of a disruption.

The contrast between “public Estonia” and “private Kenya” is overly simplistic. A more precise distinction is between an architecture of distributed functions governed by common rules and one characterized by a highly concentrated mass-market service. Both combine public and commercial elements.

Four Mechanisms of Sovereignty

The combination of Boundary + Second-Order Sovereignty + Dependence Architecture + Reproductive Sovereignty is best treated as a set of analytical mechanisms rather than four additional rankings.

Boundary refers to the capacity to establish and enforce limits on access to data and critical functions. The relevant questions are who authorizes access, who restricts it, and who monitors compliance.

Second-Order Sovereignty refers to the capacity to change the rules under which a digital system operates. It is assessed by identifying who determines standards, interfaces, connection requirements, and transition procedures.

Dependence Architecture refers to the capacity to structure dependencies so that no single intermediary becomes irreplaceable. The assessment examines which functions have independent alternatives and which dependencies are shared.

Reproductive Sovereignty refers to the capacity to sustain and renew digital competencies over time. The assessment asks who can maintain the system, remedy vulnerabilities, modify its architecture, and train successors.

These mechanisms must be distinguished: a state may possess the legal authority to require a change without having the technical capacity to implement it promptly; likewise, access to source code does not necessarily mean that a qualified team is available to maintain the system.

Boundary: The Limits of Control

In the Estonian architecture, the ecosystem operator establishes participation rules, while individual organizations retain control over their own data and services. The state defines the general conditions of interaction, whereas each participant manages the service it provides.

Kenya also has legal instruments that constrain platform lock-in. Section 38 of the Data Protection Act establishes a right to receive personal data in a machine-readable format and to transmit them to another controller or processor. Yet personal-data portability is not equivalent to payment-function portability. The transferability of service provision, integrations, settlement procedures, and access to funds must be assessed separately.

In assessing Boundary, the key consideration is not the state’s maximum discretion over data, but the existence of enforceable and verifiable limits on access. Otherwise, an expansion of administrative control may be mistakenly interpreted as an improvement in digital-system resilience.

Second-Order Sovereignty: Authority over the Rules

Estonia illustrates how a small market can participate in the evolution of an infrastructure model through institutionalized cooperation. NIIS develops and maintains X-Road, and the product evolves openly in collaboration with participants and the wider community. This does not confer unilateral control over every decision on Estonia, but it provides an institutional channel for influencing the architecture itself.

Kenya, too, can change the rules governing a commercial ecosystem. Interoperability for person-to-person transfers across operators was introduced in 2018, and merchant-payment interoperability among three providers was launched in April 2022 with the involvement of the central bank. Kenya therefore cannot be portrayed merely as a state compelled to accept the rules of a dominant platform.

The distinction lies in the depth of intervention: can the state merely negotiate discrete adjustments, or can it ensure a transition to a different infrastructural configuration? Answering this question requires examining technical specifications, contractual rights, and the actual enforcement of requirements.

Dependence Architecture

X-Road’s openness and standardized exchange provide grounds for reducing dependence on particular implementations, but do not in themselves guarantee complete provider substitutability. A transition also requires compatible data models, documentation, access rights, and qualified personnel.

The availability of ID-card, Mobile-ID, and Smart-ID should not automatically be counted as three independent layers of resilience. In 2022, RIA both emphasized the value of alternative identification methods and noted the role of SK ID Solutions in the digital-signature ecosystem.

In Kenya, cross-operator interoperability reduces network isolation, but Safaricom’s large subscription share leaves open the question of critical-function concentration. Interoperability among functioning platforms and the capacity to withstand the failure of one of them are distinct properties: the ability to transfer funds between networks does not establish that customers can continue to be served without their original operator.

Reproduction of Competencies

Estonia’s infrastructure has an institutional mechanism for developing a shared product through NIIS. Yet Reproductive Sovereignty cannot be assessed solely by the availability of open-source code. It also requires examining personnel succession, maintenance funding, and the ability to modify critical components independently.

Nor should the Kenyan model be construed as lacking domestic technological expertise. Safaricom’s 2025 report describes M-PESA modernization, platform upgrades, and investment in technological infrastructure. The research question concerns the distribution of these competencies: do they remain confined to a particular company, or are they accessible to the broader national ecosystem?

Commercial expertise can enhance national capabilities without automatically enabling the state to reproduce the corresponding function when a provider is replaced.

Testing Critical Failure

To test the argument, this study proposes a qualitative N−1 stress test: one critical intermediary becomes unavailable while the remaining components continue to operate. This is a proposed analytical scenario, not an account of a comprehensive test already conducted.

It is essential to compare equivalent functional levels. One cannot contrast the failure of a single identification application in Estonia with the failure of Kenya’s leading payment platform as a whole and then interpret the difference as an established national outcome.

In the event of an identification-provider failure, the assessment examines the availability of alternative login methods, the independence of certificate verification, and the ability to issue new credentials.

In the event of a leading payment-operator failure, the assessment examines alternative banks and payment channels, shared infrastructural dependencies, access to payments and funds without M-PESA, the capacities of other networks, and the continued use of cash.

If a contractor ceases to provide support, the assessment examines access to code, documentation, keys, and maintenance personnel, as well as service-migration rights and the feasibility of switching technology providers.

In the event of a prolonged outage, the assessment examines the restoration of services, queued transactions, settlements, agent-network liquidity, and services dependent on the affected infrastructure.

For each scenario, the analysis should record recovery time, the share of critical functions that remain operational, switching costs, the volume of unrecovered data, and the extent to which assistance from the failed intermediary is required. Technical failure, termination of cooperation, and financial insolvency must also be distinguished.

In Kenya, it is important not to conflate a payment platform with state-issued identity. The National Registration Bureau is responsible for civil registration and the issuance of identity documents, with its services accessible through the eCitizen platform. An M-PESA outage cannot automatically be equated with a loss of state identification capacity.

Estonia provides concrete evidence of limited payment redundancy. In June 2026, connectivity was disabled for one hour at six stores to test payments using physical bank cards and PINs. The exercise demonstrated that the solution worked under the tested conditions, but did not establish the resilience of the entire payment system to a prolonged failure of a major bank or infrastructure provider. 

Conditions for Confirming the Hypothesis

The hypothesis of an Estonian advantage would gain support if, under a comparable failure, Estonia could restore critical functions more quickly, preserve more independent service channels, and complete a transition without the authorization of the unavailable intermediary. Multiple interfaces or a legal right to demand portability would not suffice.

The hypothesis would be weakened if Estonia’s apparent alternatives proved dependent on a single shared provider, while Kenyan arrangements enabled switching that was operationally feasible. Such a finding would not negate the importance of architecture, but would demonstrate that architecture cannot be inferred from public or private ownership alone.

The comparison must account for alternative explanations, including the quality of public administration, available financing, infrastructure coverage, institutional support from partners, and differences in the original objectives of digitalization. Without these controls, a two-country comparison can identify a plausible mechanism but cannot establish that this mechanism accounts for all observed differences.

The Study’s Central Argument

Digital sovereign capacity is determined not by the number of users, but by whether a socially essential function remains available when the rules, provider, or conditions of access change. Scale is a resource; dependence architecture is the mechanism through which that resource is converted into capacity; and resilience following failure provides a means of testing the outcome.

The Estonian model offers ground for expecting an advantage in interoperability governance and the development of shared infrastructure, although common trust-service providers require separate scrutiny. The Kenyan model combines a high concentration of mobile-money services with a demonstrable capacity on the part of the state to promote cross-operator interoperability.

The paradox can therefore be stated most precisely as follows: a small market may afford greater digital freedom when the state can organize substitutability and reproduce critical functions; a large market may generate substantial opportunities while simultaneously increasing the cost of abandoning a dominant platform. This is a testable hypothesis about the structure of dependence, not a predetermined judgment that one country is superior to the other.