What IATI’s own publisher registry reveals about apparent fragmentation growth
Part 2 of this series argued that IATI data cannot distinguish real fragmentation growth from reporting expansion. This analysis tests that argument empirically using IATI’s own publisher registry — and finds that roughly half of organizations entering each country-sector after 2015 were new IATI publishers.
The IATI publisher base grew from 79 organizations in 2012 to 1,699 in 2024. Growth peaked during 2016–2020, with 2018 alone adding 274 new publishers. The publisher base is dominated by OECD-country organizations (UK: 487, Netherlands: 208, US: 197), but East African publishers have grown substantially — Kenya went from 1 publisher in 2013 to 65 in 2024.
| Year | New | Cumulative |
|---|---|---|
| 2011 | +15 | 15 |
| 2012 | +64 | 79 |
| 2013 | +90 | 169 |
| 2014 | +89 | 258 |
| 2015 | +78 | 336 |
| 2016 | +111 | 447 |
| 2017 | +124 | 571 |
| 2018 | +274 | 845 |
| 2019 | +206 | 1,051 |
| 2020 | +157 | 1,208 |
| 2021 | +149 | 1,357 |
| 2022 | +123 | 1,480 |
| 2023 | +128 | 1,608 |
| 2024 | +91 | 1,699 |
If reporting expansion drives apparent fragmentation growth, the two curves should move in sync. They do. Both publisher growth and per-pair fragmentation growth accelerated during 2016–2019 and plateaued after 2020.
Publisher growth outpaced fragmentation growth by a factor of 14 in absolute terms. This makes structural sense: most new publishers are narrowly focused organizations that report in one or two countries. A Kenyan NGO joining IATI adds one publisher globally but only appears in Kenya’s data.
| Year | Publishers (indexed) | Avg orgs/pair (indexed) |
|---|
The strongest test is direct: when a new organization appears in a country-sector pair, was it a new IATI publisher (reporting expansion) or an existing publisher expanding coverage?
We cross-referenced organizations entering three governance pairs (Uganda, Kenya, Ethiopia) after 2015 with their IATI first-publish dates. The result is unambiguous: roughly half of entrants in each pair were organizations that had never published to IATI before.
| Pair | Total entrants | New publishers | Existing publishers | New % |
|---|---|---|---|---|
| Uganda Governance | 46 | 22 | 24 | 48% |
| Kenya Governance | 32 | 17 | 15 | 53% |
| Ethiopia Governance | 34 | 16 | 17 | 47% |
The other half — existing publishers appearing in new country-sectors — could represent genuine operational expansion, improved reporting completeness, or both. Without external verification, we cannot distinguish these.
Combined with the earlier findings in this series, the decomposition of apparent fragmentation becomes clearer:
At any point in time, roughly 21% of organizations counted in a typical country-sector pair are measurement artifacts — duplicated names, inactive organizations, or miscoded activities (Part 1).
Over time, about half of new organizations entering a pair are new IATI publishers who hadn’t reported before. Their appearance looks like fragmentation growth but is really reporting expansion (this analysis).
Among the activities that remain, structured coordination fields are inconsistently filled. Multi-funder activities literally don’t exist in the data. Coordination mechanisms are detectable only through keyword search (Part 3).
None of this means fragmentation isn’t a real coordination challenge. Even after full decomposition, each East African country-sector pair has 20–40 active reporting organizations, and financial concentration is extreme: the top 3 control 85–99% of disbursements. But temporal claims — “fragmentation is getting worse” — cannot be supported by IATI data without controlling for reporting expansion.
IATI was designed as a transparency tool: let organizations voluntarily report their activities so others can see who does what, where. It succeeds at this. But using transparency data for trend analysis requires accounting for the growth of the transparency system itself.
The publisher growth curve is not a flaw — it is a success. More organizations are reporting. But every analytical use of IATI data needs to grapple with the fact that more reporting looks exactly like more fragmentation when you count organizations per country-sector.
This is the final piece in a four-part methodological series. Together, the four analyses argue that IATI data, while valuable for transparency, requires careful decomposition before it can support claims about fragmentation levels, trends, or coordination effectiveness.