Data governance and open data: West Africa confronts its lag

Can a country steer its development without knowing precisely its population, its births, its harvests or its pupils? In West Africa, the question is not rhetorical. Open, reliable and usable data has become an asset of sovereignty as much as an instrument of accountability, and Africa has just posted the strongest improvement of any region in the world in the Open Data Inventory (ODIN) 2024/25 assessment, with an average jump of 23%. Yet this momentum masks a very low starting point and a persistent paradox: political intent is advancing faster than the real capacity to produce, host and release usable data. It is in this gap, between stated ambition and execution, that the quality of the region's public policies is now being decided.

Real progress, but a very low starting point

The 2024/25 signal is unambiguously positive. According to Open Data Watch, ODIN now covers 197 countries, the highest number ever assessed, across 22 categories of official statistics, and Africa's average 23% rise in score is the highest of any region in the world that year. West Africa is explicitly among the sub-regions that improved the most. But this catch-up starts from a low base and remains uneven. In the ODIN 2024/25 ranking, Nigeria, the region's largest economy, scores only 56 overall (96th worldwide), the result of a coverage subscore of 48 and an openness subscore of 63. Togo follows closely at 55 (101st worldwide), with a coverage of 55 and an openness of 54. The other assessed West African countries fall below this level. None comes close to leading open data standards, and the region's largest economy stays below the world median.

Catching up is not a fated outcome: the lesson of trajectories

It is tempting to read these scores as a structural, almost immutable, state. Trajectories show the opposite. The open data movement only truly emerged on the continent in the early 2010s, and the recent pace of catch-up proves that an ODIN score is built, or lost, over a few assessment cycles. The continent-wide average rise of 23% in the single 2024/25 exercise (Open Data Watch) illustrates a fast catch-up effect: countries starting from a low base can gain many points once the fundamentals, national portal, open formats, regular publication, are in place. Conversely, West Africa's digital economy has paid dearly for the lack of redundancy: the rupture of several submarine cables in March 2024 cut off internet access for a large part of the West African coast, a reminder that opening data rests first on resilient infrastructure.

This dynamic is also visible in the digital economy, of which open data is an input. Across the continent, the share of the digital sector in GDP rose from around 1.1% in 2012 to a projected 5.2% by the end of 2025, with a projection of 8.5% by 2050. In other words, the value created by data is growing faster than the economy as a whole. The chart below sets this rise of the digital economy against the ongoing open data catch-up: two curves pointing the same way, tracing a window of opportunity, provided infrastructure and governance are not allowed to fall behind usage.

The mechanics of the lag: three locks, not a cultural fate

Why does West Africa remain behind despite the momentum? The lag is not a matter of political will, now broadly asserted, but of a data production chain that breaks at three precise links. Decomposing these mechanisms is the condition for acting in the right place rather than funding empty portals.

First lock: usability

The main obstacle is not the absence of data, but its format. According to the Open Knowledge Foundation's Africa Open Data Index, only 14.5% of the public datasets assessed across the continent (61 out of 420, in 30 countries) are released in at least one machine-readable format; most remain locked in printed reports, PDFs or HTML pages. In practice, data trapped in a PDF can be neither aggregated, nor cross-referenced, nor reused at scale: it exists without serving. The same reality is visible in the Open Data Barometer (4th edition, 2016), where Sub-Saharan Africa shows a readiness of 27, but an implementation that drops to 12 and an impact of 10. The intent is there; operational execution does not follow.

Second lock: sovereign infrastructure

Governing one's data requires the ability to host it locally. Yet, according to sector estimates, West Africa has only around twenty data centers, a small share of a continental fleet concentrated in Southern Africa (South Africa alone accounts for several dozen). This deficit weighs on digital sovereignty and the resilience of public services. It is compounded by a usage deficit: according to the ITU, only 38% of Africans used the internet in 2024, against a global average of 68%, with one of the world's widest urban-rural gaps (57% in cities versus 23% in rural areas). Without regional hosting capacity or mass connectivity, open data remains structurally dependent on the outside and out of reach for some of its intended beneficiaries.

Third lock: the underfunded statistical base

Before being opened, data must be produced. This is the most neglected link. According to the Mo Ibrahim Foundation (IIAG 2023), statistics receive only 0.34% of total official development assistance, and the aid Africa receives for data and statistics nearly halved between 2018 and 2021. When the production base dries up, opening data becomes a facade exercise. West Africa's lag is therefore first a problem of funding the upstream, not only of publishing the downstream.

• Usability: only 14.5% of the public datasets assessed in Africa are machine-readable (Open Knowledge Foundation). The rest is trapped in PDFs, hence unusable at scale.
• Infrastructure: around twenty data centers in West Africa (estimates), and only 38% internet users on the continent in 2024 (ITU), versus 68% worldwide.
• Upstream funding: 0.34% of official development assistance goes to statistics, and aid for African data nearly halved between 2018 and 2021 (Mo Ibrahim, IIAG 2023).

Data trapped in a PDF exists, but does not serve. As long as 86% of public data remains non-machine-readable, evidence-based governance will stay a promise.

Sizing the reservoir: how much is unopened data worth?

The gap to close is not just a ranking, it is a reservoir of dormant value. If 14.5% of the assessed public datasets are usable, nearly 86% are not: this is the share of public information already produced but feeding neither analysis, nor decision, nor private innovation. Set against a digital economy that rose from 1.1% to 5.2% of continental GDP in thirteen years, this reservoir represents growth potential directly throttled by a problem of format and governance, not by a lack of raw material. The data already exists; it is simply inaccessible.

The cost of inaction: steering blind

What happens if nothing changes? The cost is measured first in decisions taken without data. The Mo Ibrahim Foundation documents a worrying situation: in 14 African countries, the last population census predates 2010, meaning that all recent planning rests on estimates. As of April 2023, only 10 African countries, representing 19.6% of the continent's population, had a system registering at least 90% of births; and only 3 countries (7.8% of the population) registered at least 90% of deaths. An unregistered child is invisible to school, health and social protection; a poorly counted population distorts budget allocation, electoral boundaries and crisis response. Inaction is not neutral: it translates into mistargeted services, misallocated funds and policies evaluated on fragile bases.

To this steering cost is added a sovereignty cost. Without regional hosting or a common framework, sensitive data transits and resides off the continent, and the analytical value is captured elsewhere. The lack of infrastructure redundancy, illustrated by the submarine cable cut of March 2024, turns every technical incident into a public-service interruption. Postponing investment in data means accepting that West African public decision-making keeps relying on aging projections.

What averages hide, and why fine-grained measurement changes the decision

An aggregate national score is a starting point, never a destination. Behind an average ODIN score lie considerable gaps between categories: a country may publish its national accounts well and ignore its health, environment or gender statistics. Behind a birth registration rate lie disparities between urban and rural areas, between regions, between girls and boys. And behind a flattering continental average, +23% in 2024/25, lie sub-regions that are slipping. This is precisely where rigorous data work adds value: not reproducing the average, but breaking it down.

This is CRAD's approach. Measuring fine-grained, disaggregating by sex, age and territory, and geolocating, surfaces what the aggregate conceals and thus makes the decision actionable. Knowing that only 23% of rural dwellers use the internet (ITU) is useful only if one knows where these 23% are. Knowing that births are registered for only 19.6% of the population guides action only if one identifies the municipalities where registration collapses. Open but aggregated data informs; open, disaggregated and geolocated data decides. West Africa's lag will not be closed by publishing more averages, but by producing data fine-grained enough to target intervention.

The institutional link: a regional framework under construction

The good news is that the governance framework, long absent, is taking shape. The African Union adopted its Data Policy Framework on 28 July 2022, a roadmap for an African single digital market. Above all, the Malabo Convention (the African Union Convention on Cyber Security and Personal Data Protection), adopted in 2014, finally entered into force on 8 June 2023 after reaching the required threshold of 15 ratifications, nearly nine years after its adoption. That delay says it all: adopting a framework is far quicker than ratifying it, and quicker still than actually implementing it. Several West African countries, including Niger and Senegal, are among the parties, but the published lists of ratifiers still diverge from one source to another, which speaks volumes about the transparency gap in this area. ECOWAS, for its part, has had a Supplementary Act on personal data protection since 16 February 2010. Across the continent, 39 of the 55 African countries had adopted a data protection law by the end of 2024, of which 34 had a regulatory authority.

This legal architecture is necessary but not sufficient. Having a data protection law does not guarantee that it is applied, nor that an authority has the means to enforce it. The risk now is no longer the absence of a framework, but its symbolic implementation. It is the same gap as the one measured by the Open Data Barometer between readiness (27) and impact (10): the region excels at adopting texts and remains fragile at operationalizing them.

Regional champions showing the way

The picture is not uniform. According to ICTWorks' analysis, of the fifteen West African nations, eight have launched open data projects and five have effective national initiatives: Burkina Faso, Ghana, Nigeria, Senegal and Sierra Leone. Ghana has shown that rapid progress is possible, ranking among the African countries that gained the most positions across the editions of the Open Data Barometer. On the regulatory front, several countries, including Benin, Ghana and Senegal, are among the African states that have formalized a national artificial intelligence strategy. These advances prove that the catch-up trajectory is attainable, provided strategies are turned into execution mechanisms, since none of these countries has yet put in place a formal framework for regulating AI.

The gender blind spot: counting those who go uncounted

Among the most deficient categories are gender statistics. The Mo Ibrahim Foundation notes that data on violence against women, child labor and the informal economy, sectors with a strong female presence, are among the most lacking on the continent. Yet what is not measured is not funded. A birth registration system that reaches 90% only for 19.6% of the population leaves millions of women and children in the shadows, with direct consequences for access to rights, inheritance and services. Opening data will be transformative only if it is paired with a requirement to disaggregate by sex; otherwise it will reproduce, more visibly, the existing blind spots. This is a dimension CRAD integrates systematically, from survey design to reporting.

← All insights