From analysis to grade

We divided Norway into hexagonal cells (H3) across all 357 municipalities. For each cell we calculate how many everyday destinations can be reached within a reasonable travel time, once by public transport and once by car. Together, those two values produce the final grade.

Everyday destinations fall into five categories: workplaces, groceries, schools, health and social. Workplaces are the sub-units in the Central Coordinating Register for Legal Entities, kept by the Brønnøysund Register Centre. We place them on the map using the addresses in Kartverket's Matrikkelen register. The rest come from OpenStreetMap. Destinations are scored on a saturating curve, where destinations close to the origin count more than those further away.

Each cell is weighted by how many people live there. Population comes from Statistics Norway's 250 m population grid, spread over the cells and reconciled to Statistics Norway's population by municipality, so each municipality counts in proportion to its population. Timetables come from Entur's national GTFS dataset, which covers public transport across the country.

Here is a high-level description of the process. If you want to know more, get in touch.

Two measured reaches

Using Entur timetables and the OpenStreetMap road network, we calculate how much people can reach by public transport and by car. Short trips count more than long ones, and trips over 90 minutes do not count.

Destinations

We count five kinds of everyday destination: workplaces, groceries, schools, health and social. The first destinations within reach count most towards the score.

Public transport quality

Public transport quality is a weighted measure of how many everyday destinations you can reach by public transport. The nearest count most, and the first destinations in each category count more than the ones after them. For each of the five categories we take the travel-time-weighted reach by public transport, score it on a saturating curve against an absolute constant, and add the five together with fixed weights.

0 = no everyday destinations within travel time · 100 = all five groups well covered

Transport choice

Transport choice measures how close public transport comes to the car. We calculate the same reach again by car, divide the difference between the two by their sum and invert it, so 100 means car and public transport reach the same and 0 means the car reaches everything and public transport almost nothing. Where combined reach is very low, the score is scaled down, so two equally poor options do not look like freedom.

Mobility freedom

Mobility freedom is the average of the two. Both have to hold at once: public transport quality asks whether the service is good in itself, transport choice whether it can compete with the car.

mobility freedom = 0.5 × public transport quality + 0.5 × transport choice

Example · Oslo

Public transport quality	81.8
Transport choice	87.7
Overall, the average of the two	84.8 · A

The A–G scale

Mobility freedom is the score behind the grade. High mobility freedom means public transport is good enough that people have a real choice about owning a car. Low mobility freedom means public transport is weak, the car has a large advantage, or destinations are difficult to reach by either mode.

The scale is absolute. The calculation never looks at where the other municipalities sit, so a municipality can only move if its own mobility changes.

70–100 · 10 municipalities · 29 % of the population

Walking and public transport cover practically as much of everyday life as the car.

55–70 · 33 municipalities · 23 % of the population

Walking and public transport cover most of everyday life, close to what the car does.

45–55 · 41 municipalities · 12 % of the population

Walking and public transport cover a large part of everyday life, helped by short distances.

30–45 · 114 municipalities · 22 % of the population

Walking and public transport cover part of everyday life, but far from what the car reaches.

22–30 · 66 municipalities · 8 % of the population

Walking and public transport cover only a smaller part of everyday life; the car reaches clearly further.

15–22 · 66 municipalities · 4 % of the population

Walking and public transport cover few of everyday journeys; the car is in practice necessary.

0–15 · 27 municipalities · 1 % of the population

Walking and public transport cover almost none of everyday life; the car is the only real option.

What improves the grade

There are several ways to improve the grade: shorter travel times to everyday destinations, fewer or faster interchanges, a new public transport service somewhere that had none, new stops closer to housing, and urban development that puts housing closer to destinations that are already there. New workplaces, grocery shops, schools, health centres and cultural venues also count, because they shorten the distance for all areas around them.

What the atlas does not capture (yet)

Congestion. Car times are computed on the road network without queues. The car's advantage is therefore overstated exactly where congestion is worst, and the densest municipalities score slightly lower than they should.

Demand-responsive transport. The calculation is built on planned departures, so services without fixed timetables are not captured.

The bicycle. Walking is included as the way to and from public transport, but cycling does not count as a mode in its own right.

The quality of the journey. Whether there is a seat, whether the interchange feels safe at night, whether the platform works with a walking frame or a pram. We measure travel time and reach, not how the journey feels.

The timetable, not reality. Travel times are computed on the planned timetables in Entur's data. Delays, cancellations and missed connections are not included, so public transport reach is what the timetable promises, not what you get.

Finally, the atlas measures supply, not behaviour. It says what can be reached, not what people actually do. Two municipalities with the same grade can have very different travel habits.

Data sources

Municipalities, counties and statistics follow the divisions used by Statistics Norway and Kartverket, so they can be checked against the sources' own documentation. These are the sources we use:

Entur: national GTFS dataset (all of Norway). Public transport supply: routes, stops and departures, used to calculate public transport quality.
OpenStreetMap: road network, points of interest and settlement areas. Walking and road networks for travel-time calculations, plus points of interest for groceries, schools, health and social destinations. Buildings, land use, hotels, sports facilities and other mapped sites also define settlement display footprints. The finer boundaries add no measurements or grades. Taken as the Norway extract from Geofabrik.
Statistics Norway (SSB) 07459: population. Population by municipality. The population weights are reconciled to these totals.
Statistics Norway (SSB): population on a 250 m grid. Where people live. The grid is spread over hexagonal H3 cells and reconciled to the population in 07459. It also supplies coarse display coverage where OSM is missing.
Statistics Norway (SSB) 14746: population projections. Projected population by municipality, the 2026 projections on the current 357 municipalities.
Brønnøysund Register Centre: sub-units in the Central Coordinating Register for Legal Entities. Workplaces in the destination basket: each sub-unit at its registered address, from a fixed extract taken on 15 September 2026.
Kartverket: addresses from the Matrikkelen register. Turns the registered address into a coordinate, so each workplace lands in the right place and the right municipality.
Kartverket: municipal and county boundaries. Official boundaries for 357 municipalities and 15 counties, used as analysis and map layers.
Kartverket: N500 map data. Coastline used to remove sea from the settlement footprints within statistical areas.
OFV: "Kostnader ved bilhold", Kalkyle 2024. Annual cost per car, 127,000 kr, weighted by the car fleet in Statistics Norway table 07849. Used in the car dependence indicator, not in the grade.