COP31 and Data: Making the Case From the Field

*The most important climate finance argument at Antalya will not be made by a negotiator. It was already made, years ago, by communities whose losses no one wrote down. This is what I learned trying to write them down.*

In Cox's Bazar, in the footprint of the world's largest refugee settlement, I helped run a cash-for-work programme that paid people to plant trees and stabilise slopes. The Rohingya influx had stripped the surrounding hills for fuel and shelter, and the host communities, Bangladeshi villages that had absorbed nearly a million people almost overnight, were living with the consequences. Eroding slopes, failing land, monsoon and cyclone risk pressing in from the Bay of Bengal. We partnered with local disaster management committees, set up the beneficiary targeting and the payment cycles, and wired the field monitoring back to the seasonal contingency planning for the camps.

The programme worked. But the conversation that stayed with me was about something we were *not* funding. The host communities had their own climate losses. Coastal erosion, salinity, storm damage accumulating year after year. Those losses were, in the language of the global system, undocumented. No threshold crossed, no international appeal triggered, no line in any database that a climate fund would recognise. The damage was real. The people were real. The evidence, by the standards of the architecture that moves money, did not exist. And so neither did the funding.

I have carried that asymmetry through every role since, and it is the reason I want to make a specific argument as the world moves toward COP31 in Antalya. The decisive climate-finance fight of this cycle is not really about the size of the pledges. It is about who can prove their losses. Data is not the back office of climate finance. Data is the gatekeeper. And I have spent ten years watching it decide who gets through.

When people hear "humanitarian data," they picture dashboards. The reality, in every country I have worked in, is closer to diplomacy.

In Afghanistan after 2021, before I could build anything, I had to find the data. It existed, held by the national disaster authority, by UN clusters, by line ministries, by meteorological services, in more than thirty separate places, in incompatible formats, governed by mutual suspicion. The technical work of integrating it into a single Humanitarian Spatial Data Centre, built to support risk-informed humanitarian response planning, was the easy part. The hard part was negotiating the data-sharing agreements that made the integration legal and trusted, and then producing the outputs in Dari and Pashto so the national authority could actually use them for its own loss reporting. We ended up running a platform that drew on those sources to serve more than two hundred partners. But the foundation was not code. It was trust, painstakingly assembled, that the data would be handled as a shared public good rather than extracted and taken away.

In Ethiopia, the lesson was about quality. I inherited a cash dataset where roughly forty per cent of records had no usable location. You cannot target a climate-shock-responsive programme on data that cannot tell you where people are. So I rebuilt the pipeline from the point of entry, validation rules, controlled fields, quality checks, not as a technical indulgence but because every downstream decision about who received support depended on it. When the Cash Working Group later positioned ahead of drought, releasing one CERF operation that reached around 185,000 people with most reporting improved livelihoods, it was that boring, invisible groundwork that made acting early defensible.

These are not war stories for their own sake. They are the texture of what "evidence-based climate finance" actually requires on the ground, and it is nothing like the clean abstraction the term implies in a negotiating text.

Step back from the field and the pattern becomes a structural injustice, and it is the heart of my case for COP31.

Almost every climate fund disburses against risk and loss evidence. The Loss and Damage Fund, the Green Climate Fund, the Adaptation Fund, the anticipatory-finance mechanisms, all of them, reasonably, want proof. But the proof flows from datasets that were never designed to see everyone. The most-cited global disaster database, EM-DAT, only records events that cross thresholds (ten dead, a hundred affected, an international appeal, a state of emergency) and even acknowledges itself that economic losses in poorer contexts are badly under-reported. The result is a single, devastating statistic I keep returning to: in the 2025 global disaster figures, Africa accounted for nearly nineteen per cent of disaster deaths but well under one per cent of recorded economic losses. African disasters are not cheap. African losses are simply not counted.

This is the visibility trap, and it closes in two stages. First, the small, recurrent, slow-onset hazards that wear down the poorest communities never enter the record. Then, because they are not in the record, the communities that suffer them cannot substantiate a claim on the funds designed to help them. Invisibility in the data becomes invisibility in the finance. The places with the greatest need end up with the weakest evidence, and the architecture, running exactly as designed, routes money toward the well-documented and away from the overlooked.

My host communities in Cox's Bazar were caught in the first stage of that trap. Multiply them by every under-instrumented district on Earth and you have the quiet, structural reason climate finance keeps missing the people it is most meant for.

If I had the floor at Antalya, I would make the case the way I have made it in field offices for a decade. Not as a plea for sympathy, but as an argument about plumbing.

Treat national disaster-data systems as climate-finance infrastructure, and fund them as such. Every dollar invested in a country's ability to record and substantiate its own losses unlocks many dollars of finance that can actually find their target, and protects against the far larger waste of money flowing to where the documentation is rather than where the need is. Build the DELTA migrations and, before them, the honest maturity assessments that decide whether a system will survive its own launch. Wire the funds' application processes to recognise disaggregated, extensive-risk evidence, so the grinding, recurrent losses count and not only the photogenic catastrophes. And make the technical-assistance channels, the Santiago Network among them, deliver data-system readiness as a named outcome, so the countries furthest behind are helped to compete rather than quietly disqualified.

None of this is glamorous. None of it will lead a communiqué. It is exactly the kind of unspectacular, foundational work that determines whether everything above it functions, which is precisely why it gets under-funded, and precisely why I keep insisting on it.

I think often about the host families in Cox's Bazar, planting trees on slopes they were trying to hold together, carrying climate losses that no system had bothered to write down. The tragedy was never that the world lacked compassion for them. It was that the world had no record of them, and in a finance architecture that runs on evidence, to be unrecorded is to be unfundable, no matter how real your suffering.

That is the case for data at COP31. Not data for its own sake. Data as the precondition for justice. The most important argument in Antalya will be about who gets to be visible enough to be helped, and we already know, from a decade of field experience, that visibility is something we build, deliberately, system by system, or fail to build and call it fate.

What gets counted gets funded. What gets missed stays vulnerable. Making the case for data is making the case for the people the data has always left out.