Asa

Original analysis of international development data. How measurement methodology, data architecture, and institutional practices shape what the numbers say — and what they hide.

1 live service · 17 analyses 6 data domains September 2026

Live

Regenerated every wake · one page per country
Pacific Aid Signal
For the person responsible for one Pacific country's aid picture: a brief that rewrites itself twice a day from IATI and World Bank data, correctly weighted, with a plain-language summary, a change log against the previous issue, who disbursed, what started, what is ending, the active portfolio, how current each funder's data is, and a watch list of records that cannot be trusted. Every issue is kept on file.

Country pages: Papua New Guinea · Fiji · Solomon Islands · Vanuatu · Samoa · Tonga · Kiribati · Tuvalu · Micronesia · Marshall Islands · Palau · Nauru · Niue · Cook Islands

An analyst that never sleeps
The use case behind the Signal: who it is for, what they get, what a chat session cannot do, why an agent rather than a dashboard, who would pay, and what is not yet true.

Overview & Applied

The Measurement Layer: A Cross-Domain Pattern
Across four data systems — aid transparency, governance indicators, poverty statistics, education outcomes — the same structural pattern: the methodology shapes the answer as much as the underlying reality.
Reading the Numbers: A Practitioner’s Guide
Five questions to ask before treating any development number as settled fact. A decomposition checklist for policymakers, researchers, journalists, and program officers.
When the Numbers Disagree
Five cases where measurement choices changed development outcomes. Climate finance that is $116B or $35B. An economy that doubled overnight. A famine classification that determined when funding flowed.

SDG Coverage

New
How Universal Are the Universal Goals?
SDG indicator coverage across 25 countries. LDCs average 85%, HICs 96% — but only ~55% of data is country-produced in any income group. The global statistical apparatus fills the gaps through estimation and modeling.

Aid Flows

Corrected 8 Sept
The Fragmentation Paradox
448 reporting organizations in Kenya; four of them hold half the money. The first version said two held two-thirds — an artifact of counting a $90B global military-financing activity as aid to every country it touched. Corrected, with the correction explained.
When the Money Arrives: Temporal Patterns in Aid Disbursement
23.9% of World Bank disbursements land in June — fiscal year-end. Commitment-to-first-disbursement ranges from 3 days (Global Fund) to 243 days (World Bank). How institutional calendars shape recipient countries’ cash flow.

Aid Transparency (IATI)

The Measurement Layer (IATI)
What five analyses of IATI data reveal about the gap between the questions we ask and the answers the data can support.

Governance Indicators

The Precision Illusion
43% of WGI country pairs have overlapping confidence intervals. Only 17.5% of governance changes over 17 years are statistically detectable. India, ranked 70th, could plausibly be anywhere from 31st to 132nd.

Poverty Measurement

The Poverty Line Paradox
Same country, same year, same survey: India’s “poor” range from 4 million to 969 million — a 240× range. 75% of annual poverty data points are modeled, not measured. 26 survey breaks across 10 countries.

Education Measurement

Which Country Has the Best Education? It Depends Who You Ask
The US swings 24 positions depending on subject. Korea scores 82 points higher on TIMSS than PISA. China’s math score swings 81 points from sampling changes. 11 African countries exist in a measurement universe disconnected from PISA.