Fewer buses. More alternatives. Increasing productivity in Leeds.

Tom Forth, 12 Oct 2022.

Leeds is a big city. A lot of people don't agree. They're wrong and I've built a tool to prove it.

Leeds is a big city

Select a radius. Click on the map. And it'll tell you the population around a point anywhere in the world, thanks to open data from the European Union's Global Human Settlement project.

2 million people live within 20km (12 miles) of Dewsbury station in West Yorkshire.

Try it yourself, play around with radiuses and locations. 1.9 million people live within 20km of Leeds, 1.4m in Lille, and 1.3m in Dublin.

Using Excel, I calculated the pattern of how the population accumulates for the three cities as you go out from their city centres. We see that the pattern of density is similar for all three. Then Leeds ends up biggest. Like I said.

Leeds, Lille, and Dublin have similar urban structures. Their populations increased similarly at all distances from their city centre in 2015. (Leeds' central density was underestimated at this time and has since significantly densified just in case you're thinking that its lower central density is significant.

Too many buses.

Another fantastic piece of open data that we now have for much of the world is public transport timetables. You may have enjoyed it in Google maps. And thanks to Google there is a widely adopted global standard for public transport timetables called GTFS which allows analysis and comparison.

Ireland's public transport timetables are available from Transport for Ireland's data pages.

France's timetables are published by the regions and cities that regulate and/or operate services and are well collated by the French government at transport.data.gouv.fr. For Lille the data is excellent.

Great Britain's data (Northern Ireland's transport is devolved, in part to facilitate whole-island interoperation) is now published in GTFS format. You can get all the bus data from the Department for Transport. Rail data is harder to work with, but you can get the GB Rail timetable in ATOC CIF format from The Rail Delivery Group and convert it to GTFS using the free converter I wrote. Or you can download my interpretation of the GB Rail September 2022 timetable right now.

Over the past few months I've been writing software to read these timetables for Dublin, Leeds, and Lille and count the number of public transport vehicles active throughout the day. Specifically I count the number of buses (all three cities run buses), trains (all three), trams (Lille and Dublin), and metros (Lille only) that are running at any time within 20km of the city centres.

While the population structure of the three cities are remarkably similar, except for Leeds being bigger (have I mentioned that yet?), their public transport by this measure is very different.

Let's look at buses first.

More buses run in Leeds, for more time, than in Dublin and vastly more than in Lille.

Two features of Leeds buses stand out. First, Leeds has loads more buses than Lille. Second, Leeds runs just as many buses throughout the day as it does at morning and evening peaks. This is in part because free bus passes which can be used from 09:30am now represent over a quarter of all journeys on buses in West Yorkshire.

The graph for combined rail services (train plus trams plus metro) is very different. Without a metro or a tram Leeds has very few services by this measure. Lille's autonomous metro, tram, and regional rail deliver a very large number of services with frequencies closely matched to demand. Dublin's trams, rail, and DART services are in between.

Lille's autonomous metro has many more active vehicles than trams in Dublin or trains in Leeds.

Low productivity Leeds.

It's time for some maths.

The area under the graphs above tells us how many vehicle hours are used to run the public transport systems in each city. Buses, trains, and trams each have one driver. Lille's driverless metros have no driver. Knowing this, we can convert vehicle hours to driver hours, which are the largest cost in a public transport system.

Separately, from public reports of transport agencies and national statistics, we know the number of passenger journeys by each mode in the three cities. Combining this two data sources we can calculate a very rough guess of the number of journeys provided by an average driver in an hour in each city.

The public transport systems in Lille and Dublin are used for about 40% more journeys than the public transport system in Leeds despite fewer vehicles being active for less time.

The public transport systems in Lille and Dublin are used for about 40% more journeys than the public transport system in Leeds. Inefficient uses of buses, and the inherent inefficiency of buses compared to trams, trains, and metros, means that the average driver in Dublin provides a third more journeys per hour at work than the average driver in Leeds. The average driver in Lille provides nearly three times as many journeys as the average drive in Leeds.

I've written often about how the poor public transport in Leeds makes us less productive. That is still true. I could write loads more about how Dublin and Lille have simple regulated fares and an integrated smartcard payment system, while Leeds doesn't. But I won't. For now let's focus on this angle. It's a directly measurable example of low productivity in the UK economy. We see that low capital investment in Leeds, specifically decades of decisions not to build a tram, a metro, or better electric railways like Dublin and Lille have done means that we have to use more driver time to provide public transport that fewer people use. It is the very definition of a low productivity economic system.

So what?

There are a lot of policy recommendations that come from this data. I've got my opinions, but maybe I'll write mine up later, once I've listened to yours.