Growth Pole Map

Fifty-four country cards, the corridor seed, and the eight growth-pole chapters of Part VI on one map. Colour the countries by any card field; hatch the ones whose numbers you should not lean on; click a country, a corridor or a pole for what the card and the chapter say. Nine industry layers sit on top: mines, oil and gas, power, auto plants, financial centres, cable landings and data centres, tourism hubs, ports and border posts, and the data centres and AI power of Chapter 35's compute-inputs branch. Each is drawn from its sector chapter's own site table, each site carrying the source it came from.

Data confidence
Africa: growth poles, corridors and the country cards A map of the fifty-four UN-member African states coloured by a country-card field, with the book's corridor seed, its anchor cities and ports, and the eight growth-pole chapters drawn as envelopes. The same numbers are in the country-card table linked below this map.

Drag to pan, ctrl- or ⌘-scroll to zoom, or use the buttons on the map. Click a country, a corridor, a city or a pole label for what the card and the chapter say; Esc closes. The reader who would rather have a table than a map should use the country cards, which carry the same fields.

Layers
The book's own layers
Industry layers
Project status, corridors and power

What is drawn, and what is not

The country layer is the fifty-four UN-member states. Every popup is copied field for field from the same fifty-four country cards the card pages render, so the map and the card cannot disagree. A country with no value for the field on display is hatched and its popup says not yet fetched; it is never given the bottom colour of the ramp.

Morocco–Western Sahara is hatched, as everywhere else in the book. Western Sahara has no country card. The AU counts fifty-five members including the Sahrawi Arab Democratic Republic, and Chapter 7 says so; the cards and the aggregates here are the fifty-four UN members. Somaliland has no card either, so it has no border here: it is dissolved into Somalia rather than drawn and left unexplained.

The corridors are schematic. The seed's own note says it: six alignments carried over from The New Regional Economics and nine drawn through named intermediate cities and border posts, to be replaced with AfDB/PIDA published alignments before release. The status field takes four values: reformed, under reform, unreformed, planned. Each is a claim with a citation, and the citation is in every popup.

The project-status filter reads two vocabularies and translates one of them. Its four states are planned, under construction, operating and stalled. Chapter 14's power table answers exactly that question, and all four of its words translate: operating to operating, construction to under construction, pre-construction to planned, shelved to stalled. The corridor seed's status field answers a different one. Its words are reformed, under reform, unreformed and planned, and the sources behind them are transit days, roadblocks, one-stop border posts and concession agreements: they describe how a corridor that already carries traffic is run, not what stage a project has reached. Only planned is a stage of building, so only planned is translated; the other three leave the corridor without a project status, which the filter hides rather than guesses at, with the count of what it hid beside it. Each corridor keeps its reform word in its popup, because it is a real finding about a real corridor and simply an answer to another question. Where a power site's own row cites the Global Energy Monitor tracker, the status it carries is said to be that tracker's label; the layer panel reports how many rows do.

The pole envelopes are schematic too. Each is the union of that chapter's anchor cities, ports and corridors, buffered. It marks where a chapter's argument sits; it is not a boundary anybody administers, and it is not a functional-region estimate.

The industry layers are the sector chapters' own site tables, one per chapter, read straight off _figures/chNN/chNN_sites.csv. Every site carries the source string and the claim-ledger identifier of the row it came from, and both are in its popup. A site whose coordinates fall outside the map, whose ledger identifier does not resolve, or which is missing a property the layer requires is not drawn: it is listed, with the reason, in the layer file the build writes, so a gap is a recorded decision and not an oversight. The count under each layer's checkbox is what survived that check, out of the rows the chapter's table holds. Each popup links to the passage in its chapter where those sites are argued about, not to the top of the chapter.

A mine also carries its commodity's USGS position: the continent's share of world production, the world total, and, where the mine sits in the release's largest African producer of that commodity, that producer's own output and share. The release names no rank below the first and publishes no reserves table the book reads, so nothing further is claimed; the platinum-group metals, iron ore and uranium have no African production table in it at all and those sites carry no USGS line. A cable landing carries the systems its record names as landing there and the party that record names as running the station. Where a record names neither, the line is absent rather than filled with words: the silence is the record's own.

The ninth layer is Chapter 35's compute-inputs branch, and its rule is narrow on purpose: a hyperscale or AI campus, a hyperscaler cloud region, or a colocation campus whose operator publishes a critical IT load in megawatts. Chapter 17's layer holds the colocation halls as connectivity; this one holds the megawatts, which is the observable Chapter 35 says would tell you whether the compute-inputs case is happening or being announced. A capacity no operator publishes is empty, never a zero.

These are the sites the chapters name, not a cadastre. A layer holds the mines, plants, ports and landings its chapter argues about; it is not the population of everything of that kind in Africa, and the counts here should never be read as one. Nothing on this map is drawn from memory.

Sources

All fifty-four country cards as a table: the same numbers, sortable, one page per country. Sovereign Debt Tracker.