LP led in city centres. APC led in towns and villages.
This briefing classifies each polling unit in the 2023 presidential IReV record as a city centre, a town or city edge, or a village, then adds up votes for APC, LP, PDP and NNPP. LP’s lead is limited to the densest city centres.
37% of those votes
38% of those votes
43% of those votes
in villages
National totals
The same contest, counted three ways
Vote shares use APC, LP, PDP and NNPP only. Turnout is accredited voters divided by registered voters.
| Place | Polling units | Leader | APC | LP | PDP | NNPP | Turnout |
|---|---|---|---|---|---|---|---|
| City centre | 76,196 | LP, 37% | 2.96m | 3.19m | 1.90m | 0.64m | 24% |
| Town / city edge | 45,805 | APC, 38% | 2.23m | 1.34m | 1.78m | 0.55m | 28% |
| Village | 45,901 | APC, 43% | 2.31m | 0.82m | 1.97m | 0.27m | 28% |
8,507 units could not be classified because the coordinates were missing or invalid. In villages APC led and PDP was close; LP was not.
By state
The national split is not the story in every state
In Lagos, LP led the city and APC led outside it. In Enugu, LP led in city centres, towns and villages. In Kano, NNPP led in all three.
| State | City centre | Town / city edge | Village |
|---|---|---|---|
| Lagos | LP | APC | APC |
| FCT | LP | LP | APC |
| Enugu | LP | LP | LP |
| Kano | NNPP | NNPP | NNPP |
APC’s national lead in towns is driven mainly by large towns in the North, not by the edges of every state capital.
Method
How each polling unit was classified
The classification is added on top of the 2023 presidential IReV polling-unit file. Vote totals on those units are not changed.
Start from the published unit
The dataset has one row for each of 176,409 polling units, with votes for APC, LP, PDP and NNPP and, where available, a latitude and longitude.
Read the satellite settlement square under that point
The European Commission’s GHSL Settlement Model (GHS-SMOD, 2025) divides Nigeria into one-kilometre squares and labels each square by how built-up it is. The square that contains the polling-unit coordinates is the primary class.
Check the eight neighbouring squares
A 3×3 window around that square records how much of the immediate area is urban. Mixed edges are marked as transition zones. They keep the GHSL label of the centre square; they are not reassigned by hand.
Group the seven GHSL classes into three reporting labels
The seven satellite classes are rolled up so the national table can be read as city centre, town / city edge, or village. The mapping is fixed in advance and applied to every unit the same way.
Use GRID3 only as a check
GRID3 Nigeria Settlement Extents v4.1 shows whether a point falls inside a mapped settlement outline. It is stored as context. It does not override the GHSL class.
Leave broken coordinates unclassified
Missing coordinates, the dummy point 0,0, and points outside Nigeria are not given a settlement label. Those 8,507 units are excluded from the three groups above rather than guessed into a city, town or village.
| GHSL class | What it describes | Label used here |
|---|---|---|
| 30 Urban Centre | The densest continuous city area | City centre |
| 23 Dense Urban Cluster | A dense town or urban cluster outside the main city core | Town / city edge |
| 22 Semi-Dense Urban Cluster | An urban built-up area at lower density | Town / city edge |
| 21 Suburban / Peri-Urban | The built-up fringe around a city or town | Town / city edge |
| 13 Rural Cluster | A village or small rural settlement cluster | Village |
| 12 Low-Density Rural | Sparse rural settlement | Village |
| 11 Very Low-Density Rural | Very sparse rural settlement | Village |
Party totals
For each label, party votes are the sum of APC, LP, PDP and NNPP on the units in that group. The leader is the party with the largest share of that four-party sum.
Accredited / registered
Turnout is the sum of accredited voters divided by the sum of registered voters in the same group. Units without those fields are omitted from the rate, not filled in.
How Mundx reads IReV result forms is documented on Method. Companion briefings: how concentrated each candidate’s vote was, schools, halls and gates, one-party polling units, and 2023 governorship versus presidential votes. Only INEC declares official results.
Limits
This is a geographic reading of the published file, not a judgement on the official result.
The 2023 presidential dataset is a partial polling-unit record of four parties. A unit on a city edge can sit in a town square on the satellite map while serving a nearby village. The classification is only as precise as the coordinates and the one-kilometre GHSL square.
Open the 2023 map
The same polling-unit file drives the public presidential map. Each unit still links back to its published source record.