About Prism
A catalogue of African official statistics, with the receipts attached
National statistics offices publish good data in bad containers — PDFs, scanned tables, files that move. Prism reads those publications, catalogues what is in them, and keeps a permanent, citable page for every series with a link back to the page it came from.
- Pages displayed
- 100
- Slugs served — CSV, JSON, API
- 463
- Source datasets
- 413
- Country catalogued
- 1
What Prism does — and does not do
The second column matters more than the first. A statistics portal earns trust by being explicit about where it stops.
Does
- Catalogue statistics already published by official sources
- Preserve the source publication, table and page reference
- Show what is missing, and why, using the publisher's own reason codes
- Keep superseded vintages citable after a rebase or revision
Does not
- Produce, model or forecast statistics of its own
- Silently correct, smooth or interpolate a published figure
- Rank countries or states into league tables
- Charge for access to the underlying data
Built for three kinds of reader
The same series page serves all three. Pick a reader to see what the design owes them.
What Analysts & institutions get
- Full history CSV, JSON and a versioned API, per series or dataset
- Revision status on every observation
- Sticky-column matrices for wide state tables
- Machine-readable provenance on every series
Read the method before you cite us
How a figure travels from a source publication to a page on this site, and the standing rules that govern what Prism will and will not change along the way.