Measuring What Works: Helping the Belém Adaptation Indicators Live Up to Their Promise

*COP30 finally gave the world a way to measure adaptation. COP31 has to prove that most countries can actually produce the numbers, or the indicators become one more standard the vulnerable are judged against and cannot meet.*

For most of my career, adaptation has had a measurement problem that mitigation does not. A tonne of carbon avoided is a tonne, anywhere on Earth. But "a community made more resilient" resists that kind of clean accounting. Resilience is local, multi-dimensional, and slow. You can build a seawall and still lose the village to a hazard you didn't model. You can run a flawless early warning system and still measure your success only by the disaster that didn't happen, the hardest thing in the world to count. For years, this is why adaptation lost the funding argument to mitigation. It could not put a defensible number on the board.

COP30 in Belém tried to fix that. After two years of the UAE-Belém work programme, parties adopted a set of indicators, roughly sixty, now widely called the Belém Adaptation Indicators, to track progress toward the Global Goal on Adaptation (GGA). On paper this is a genuine milestone. For the first time, the world has an agreed way to ask whether adaptation is actually happening. But I read it through the lens of a decade spent inside national data systems, and what I see is less a finish line than a starting gun. Because an indicator is only as real as a country's ability to report it, and on that, most of the conversation has been silent.

There is a comfortable assumption buried in every framework of indicators. That once you have defined what to measure, the measuring will follow. I have spent ten years discovering how false that assumption is.

When I rebuilt a cash-transfer data pipeline in Ethiopia, I found roughly forty per cent of records were missing location fields, not because anyone was careless, but because the system had never been designed to capture them. In Afghanistan, before any analysis was possible, I had to map more than thirty separate disaster-data sources held by different agencies and negotiate the agreements to bring them together, because they had never been built to talk to each other. The lesson, repeated in every country I have worked in, is that the gap between defining an indicator and reporting it reliably is enormous, and it is widest exactly where vulnerability is highest.

So when sixty adaptation indicators are adopted in a plenary hall, my first question is not "are they the right indicators?" It is "who can actually produce them?" A country with a mature, disaggregated, interoperable data system will report against the Belém indicators and use them to substantiate its claims on adaptation finance. A country still running fragmented spreadsheets and paper records will face a new global standard it has no infrastructure to meet, and risk being judged as under-performing when the truth is that it is under-instrumented. The indicator framework, like every framework before it, quietly assumes a data capability that the most vulnerable countries do not yet have.

The Belém indicators inherit a requirement that runs through all the modern frameworks I work with, and it is the requirement I care about most. Disaggregation. An adaptation result averaged across a whole country tells you almost nothing useful. The question that matters is always *who*. Which districts, which households, which women, which people with disabilities, which displaced populations. National averages are where inequality goes to hide.

This is precisely the design principle behind the newer generation of disaster-data systems I have recently begun working with. DELTA Resilience, replacing the legacy DesInventar platform, records losses sub-nationally and disaggregates by sex, age and disability. The Global Disaster-Related Statistics Framework (G-DRSF), endorsed by the UN Statistical Commission, gives disaster managers and statisticians the shared standards that make such disaggregated data comparable across borders. These are not adaptation tools per se, but they are the data backbone that any credible adaptation measurement has to stand on. You cannot report a disaggregated adaptation indicator on top of a disaster data system that only records national totals.

There is an efficiency here that the adaptation community should seize rather than reinvent. The frameworks are converging. Data captured once, to G-DRSF standards, can feed the Sendai Framework, a set of SDG indicators, and now the GGA indicators, the "one report, several purposes" logic I have argued for elsewhere. The worst outcome from Belém would be a parallel adaptation-reporting bureaucracy, disconnected from the disaster-loss and statistical systems countries are already being asked to build. The best outcome is coherence: one disaggregated national data architecture serving every framework at once. That is not a technical nicety. It is the difference between a reporting burden that buries already-stretched national offices and an asset that pays them back across every obligation they carry.

If COP31 is serious about the Belém indicators, the most useful thing it could fund is not more indicators. It is readiness, an honest diagnosis of whether countries can actually report what they have just committed to report.

I have argued before that a maturity assessment is not a delay. It is the investment that ensures the system you build is the system that survives. Before a country can credibly report against sixty adaptation indicators, someone has to ask the unglamorous questions across the dimensions that actually determine success. Is there a legal mandate and a working relationship between the disaster management authority and the national statistical office, or do they still operate on different planets? Does the technical infrastructure exist to host and exchange the data? Is the historical baseline complete enough to show change over time? And are there national experts who can run the system, or will it collapse the moment external consultants leave?

I have watched well-funded data systems die two years after launch because nobody asked those questions first. Introduce an indicator framework into an institution that cannot yet support it, and you risk the appearance of measurement rather than the real thing, which can mask gaps instead of revealing them. The road from Belém to Antalya is a chance to assess and strengthen national readiness, so that when countries report against the GGA, the numbers carry real weight.

This all connects back to the hardest number at COP30: the call to at least triple adaptation finance by 2035, embedded within the NCQG. It was rightly celebrated and rightly questioned. Tripling from a low and loosely defined base is less than it sounds, and "adaptation finance" remains slippery enough that tracking it honestly is its own challenge.

But notice the loop. Adaptation finance is supposed to flow toward measured adaptation need and demonstrated adaptation results. The Belém indicators are the instrument for measuring both. So the credibility of the finance target depends on the credibility of the indicators, which depends on the data systems underneath them. A tripling of adaptation finance allocated against weak or biased adaptation data does not produce three times the resilience in the right places. It produces three times the money chasing whatever the data happens to show, and if the data under-represents the most vulnerable, the money follows the data away from them.

This is the same argument I keep making about every climate fund, and the Belém indicators make it concrete. Measurement is not the bureaucratic afterthought to finance. Measurement is what decides whether finance lands where the need is or where the documentation is. Those are not the same place, and the gap between them is exactly where stronger national data systems are supposed to go.

I want the adaptation conversation in Antalya to grow up past the moment of adopting indicators and into the much harder work of making them reportable. That means treating national data-system readiness as a funded, first-order component of the adaptation agenda, not an assumption. It means deliberately wiring the Belém indicators into the disaster-loss and statistical systems countries are already building, through DELTA and the G-DRSF, so adaptation reporting is coherent rather than parallel. And it means insisting on disaggregation as non-negotiable, because an adaptation indicator that cannot tell you *who* was protected cannot tell you whether adaptation is reaching the people who need it.

We finally have a way to measure adaptation. That is real, and the people who negotiated it deserve credit. But a ruler is not the same as the ability to use it, and most of the world has just been handed a ruler without being asked whether they have anything to measure with. The work between Belém and Antalya is to close that gap, to make sure the Belém indicators measure adaptation as it is actually lived in the most exposed districts of the most exposed countries, and not just adaptation as it can be documented by those already equipped to document it.

What gets counted gets funded. We have decided what to count. Now we have to make sure everyone can count it, because an indicator the vulnerable cannot report is just one more standard they will be measured against and found wanting.