Excerpt
Development reporting is shifting fast from narrative accounts to quantified dashboards, and donors are demanding more precise, more granular metrics than ever before. But the sector’s statistical literacy hasn’t kept pace, producing volumes of data that are harder to interpret, not easier, and that often measure what’s easy to count rather than what actually matters. The problem was never a data shortage. It’s an interpretation shortage.
The Rise of Data-Driven Development
There used to be a time when a development project could close with a narrative report, a written account of what happened, what changed, and what was learned, told largely through stories behind a few highlights of key performance indicators (KPIs).
That era is ending quickly. We’re seeing again a rapid shift after a short few years of pushing for better and more holistic, dignified human stories of impact. Across the sector, narrative reporting is being replaced by something harder, more quantified, and increasingly carbon-focused.
Donors want precise metrics. They want detailed and perfectly precise KPIs. They want numbers that can be aggregated, compared, and presented aesthetically on a dashboard.
A Shift That Isn’t Inherently a Problem
This shift is not inherently a problem. Why not? Better measurement has genuinely improved monitoring and evaluation practice over the decades, giving organisations sharper tools to track what is actually happening and move closer to holding projects accountable for results rather than good intentions. Donors, rightly increasingly suspicious of vague claims, are asking harder questions before they fund and during implementation. That scrutiny is a healthy development in many ways.
But there is a gap opening up beneath the surface of this trend, and it is rarely discussed openly: a growing distance between the numbers being reported and the realities they are meant to represent.
The Technical Problem: Statistical Literacy Hasn’t Kept Pace
Part of the issue is technical. Statistical literacy has not kept pace with the demand for data. There are now well-documented cases of donors requesting data on every single participant in a programme, rejecting the idea that a well-designed sample of a hundred people could represent a much larger population just as validly (and far more efficiently). The logic of sampling, a basic tool of rigorous research for decades, is being abandoned in favour of an assumption that more raw data is automatically more credible. It often is not. It frequently produces volumes of inconsistent information that are harder to interpret, not easier, and that obscure rather than reveal what is actually happening on the ground and is costly in both time and effort to collect.
The Cultural Problem: Counting What’s Easy, Not What Matters
The other part of the issue is more cultural than technical. Leadership reporting increasingly focuses on metrics that are easy to count rather than metrics that are meaningful: toilets built, trainings delivered, hypothetical tonnes of carbon avoided.
These numbers are clean, comparable, and good for a slide deck but they often say very little about whether a system is actually functioning, whether people are using what was built, or whether change will outlast the project duration and funding cycle.
Carbon accounting illustrates this tension well. Translating local environmental realities into globally legible metrics is valuable. This accounting allows comparison across contexts and unlocks climate finance. But the translation is skewed. A carbon figure can make a fragmented, context-specific intervention look standardised and comparable in ways that flatten the complexity that determines whether it will work or not.
From a Data Shortage to an Interpretation Shortage
None of this is an argument against data. It is an argument for understanding what data can and cannot tell you, and for resisting the comfortable illusion that more numbers automatically mean more insight.
The sector does not have a data shortage but an interpretation shortage.
The next phase of credible measurement will not be defined by who collects the most data, but by who asks the most honest questions about what it actually means.
Comments are closed