Calculate public transport quality classes

Covered is not the same as well served

Picture a village whose nearest bus stop is 300 metres away. The bus stops there twice a day. That village often counts as covered by public transport. For someone without a car who needs to get to work or to a doctor, it is not.

The reason is the measure. Most local transport plans in Germany draw a fixed radius around each stop. A radius says nothing about barriers on the way to the stop. It also says nothing about how often a bus or train actually calls there. The result is a map where almost everyone is covered and many people are still poorly connected.

The ÖV-Atlas by Agora Verkehrswende shows how big that gap is. Around 21 million people in Germany, about a quarter of the population, have poor public transport access or none at all. In cities above 100,000 residents, 77 per cent of people live in the two best classes. In municipalities below 1,000 residents, it is 1 per cent.

So transport authorities, transport associations and operators, municipalities, districts and the planning offices that work for them need a measure that rates distance and service together. With it, you can find underserved areas, for example for a local transport plan. You can show what a proposed service change does to access. And you can steer new development to places where the service already exists. In Switzerland, municipalities even use the classes to set how many parking spaces a building may have: better public transport means fewer spaces. Austria links its parking rules to public transport access in a similar way.

How public transport quality classes work

Public transport quality classes (ÖV-Güteklassen) come from Swiss planning practice. The Swiss Federal Office for Spatial Development (ARE) developed the method, the VSS standard SN 640 290 sets it out, and Austrian spatial planning uses it widely too. For a short definition, see our glossary. This page is about the calculation behind it.

It has two steps. First, every stop gets a category. The category depends on which mode serves the stop and how frequent the service is. Rail counts for more than tram, and tram for more than bus. The basis is the timetable of a typical weekday between 6 a.m. and 8 p.m.

Then the area around each stop gets a quality class. The closer a place is to a high-category stop, the better its class. A place that cannot reach a service at least every two hours within a reasonable distance gets no class at all.

Countries apply the framework in different ways. Switzerland uses four classes and measures distance as a buffer around the stop. Austria uses seven classes and measures distance along the real street network. For the ÖV-Atlas 2026, we calculated classes A to F.

It also matters what a class does not tell you. It describes the service where people live, not the journey they make. It says nothing about punctuality, crowding or travel time to a destination.

The full method, what the standard leaves open, and the results for Germany are in our whitepaper on public transport quality. You can download it further down this page.

Calculate public transport quality classes in GOAT

In GOAT, ÖV-Güteklassen is a tool of its own under Accessibility Indicators. You choose the day and the time window, for example a weekday from 6 a.m. to 8 p.m., or the evening hours. Each stop receives a category from I to VII, based on its mode and frequency. You measure distance either as a buffer or as an isochrone that follows the real street network, so detours and barriers that a radius ignores are taken into account. The GOAT documentation explains each setting.

The input is timetable data in GTFS format. Put the result on top of population data and you can see how many people live in each class, for every municipality and district.

We have done this at national scale. For Agora Verkehrswende, we calculated public transport quality classes for all of Germany, on the 100-metre grid of the 2022 census and with nationwide timetable data from DELFI. We used the most common daily service per line across 16 weeks, so that special services or holiday timetables do not distort the result. The outcome is one comparable figure for around 11,000 municipalities and 400 districts, which can be repeated every year for monitoring. With TU Wien and KCW, we calculated the classes for all of Austria. For NAH.SH in Schleswig-Holstein, we included on-demand services as well.

You can run the calculation yourself in GOAT, or ask us to do the analysis for your area. The projects are listed directly below.