13.6.3 — Poverty Analytics — maturity: live

Subnational Poverty Indices

Constructing district- and village-level poverty indices by fusing satellite-derived proxies — nightlight intensity, built-environment density, crop yield, land cover — with ground survey data.

Auto-drafted reference page — content reviewed for shape, individual references not yet link-checked.

National poverty statistics published at country or province level mask the brutal variation that exists beneath them. A district 200 km from the capital can have child malnutrition rates three times the national average and appear invisible in aggregated figures, starving the local government of the evidence needed to justify budget allocations or trigger emergency transfers. Survey-based methods like the LSMS or DHS produce statistically reliable estimates only at regional scale; extending them to district or village level requires sample sizes that no developing-country statistics office can afford to field annually.

Satellite-derived proxies close that gap. Nighttime light radiance from VIIRS correlates strongly with household consumption; multi-spectral imagery captures rooftop materials, road density and agricultural productivity; SAR penetrates cloud cover to track seasonal crop stress. Stacked and calibrated against a sparse but representative ground survey, these covariates train small-area estimation models that produce poverty headcount ratios and multidimensional poverty index scores at a 1–5 km grid — updated annually at marginal cost once the constellation is in orbit.

The operational outcome is a living subnational poverty atlas that planners, finance ministries and social protection agencies can query in near-real time. Districts hit by flood or drought show poverty score deterioration within weeks, triggering pre-positioned response budgets before a new survey cycle is even commissioned. A sovereign constellation means the cadence, the methodology and the ground truth are owned by the state — not licensed from a commercial provider that can change pricing, coverage or data-access terms without notice.

What matters

Survey-based small-area estimation breaks down below provincial level without satellite covariates; grid-level indices require annual or more frequent refresh that no ground survey can deliver at scale.
VIIRS nighttime light radiance at 500m resolution and Sentinel-2 multispectral data at 10m are the two highest-signal inputs for consumption-proxy modelling — both require sovereign pre-processing pipelines to avoid dependence on third-party APIs.
Subnational poverty indices directly govern the formula-based fiscal transfers that move billions of dollars between central and local governments; methodological control is therefore a sovereign fiscal matter, not just a statistics question.
Commercial providers (Planet, Maxar) can suspend national-government data access under US EAR or ITAR export controls, breaking the index update cycle at precisely the moments — conflict, sanctions — when poverty tracking matters most.

Sovereignty score: 8/10 — A state that outsources its subnational poverty index to a foreign commercial data provider has surrendered control over which communities receive transfers, which regions attract investment, and how its fiscal constitution operates in practice.

Formula-based fiscal transfers in federal systems — worth billions annually — are directly keyed to subnational poverty scores; a methodology or data disruption imposed by a foreign vendor constitutes interference in sovereign budget allocation.
US EAR and ITAR controls on high-resolution commercial imagery (Maxar, Planet) can restrict or terminate data supply to governments under sanctions or geopolitical pressure, breaking index continuity at the worst possible moment.
Commercial API pricing for planet-scale annual multispectral coverage is prohibitive for low-income governments; a sovereign microsatellite constellation reduces marginal refresh cost by an order of magnitude after the initial capital investment.
National statistics offices are legally required in most jurisdictions to certify the methodology and provenance of official poverty statistics; delegating the core satellite input to a foreign commercial black-box pipeline makes that certification legally untenable.

Reference architecture

Payload: Multispectral imager, 8 bands (440–2200 nm), 5m GSD in panchromatic mode, 10m GSD multispectral; secondary VIIRS-class TIR channel for nighttime radiance at 500m GSD; optional SAR cross-calibration via hosted payload agreement with ICEYE or ISRO
Bus class: 16U cubesat or 40 kg-class microsat, 80W payload power; constellation of 8–12 units provides the revisit cadence needed for cloud-free annual composites over tropical and monsoon-affected regions
Orbit: Sun-synchronous LEO at 520–550 km, 10:30 AM descending node LTAN, 8-satellite walker constellation achieving 3–5 day cloud-free revisit at equatorial latitudes; monthly composites sufficient for annual index cycle
Ground segment: 2-station sovereign network (S-band TT&C + X-band downlink) co-located with national statistics office data centre; SatNOGS UHF/VHF backup for housekeeping telemetry; calibration vicarious site on stable desert target (e.g. national salt flat or tarmac pad) for radiometric normalisation
Software & data
Latency
Cost band
Lead time

References

World Bank – Poverty and Inequality Platform: Small Area Estimation Methods — The World Bank documents its small-area estimation methodology for subnational poverty mapping, noting that satellite-derived covariates are now a standard input when survey coverage is insufficient for direct estimation.
NOAA – VIIRS Day/Night Band Nighttime Lights — NOAA distributes VIIRS nighttime light composites used globally as proxy indicators of economic activity and electrification, with annual and monthly products at 500m resolution.
ESA – Sentinel-2 Mission Overview — ESA's Sentinel-2 twin satellites deliver 10m multispectral imagery on a 5-day revisit, providing the built-environment and land-cover data layers central to consumption-proxy poverty modelling.
FAO – Agricultural Stress Index System (ASIS) — FAO's ASIS uses satellite vegetation indices to detect agricultural drought stress at subnational scale, a key input for rural poverty index updates in agrarian economies.
UNDP – Multidimensional Poverty Index Methodological Notes — The UNDP and Oxford Poverty and Human Development Initiative describe how satellite-derived indicators of housing quality, electricity access and sanitation infrastructure can substitute for missing survey dimensions in national MPI calculations.