Do Cities Like Minneapolis Have High Transit Ridership? (Part 1)

There are many cities in the world that are roughly the same size as Minneapolis. There are many cities in the world that have high transit ridership. But do these two sets of cities overlap? This article series is an exercise in two things: determining which cities are most like Minneapolis and determining what is the transit share of those cities. Here we tackle the first part, identifying cities that are like Minneapolis, from the perspective of population distribution.

Transit share in the United States is pretty universally low, except in New York City. Since Minneapolis is not very much like New York, it can be assumed that the sort of high transit ridership cities that some of us aspire for Minneapolis to be must exist outside the United States.

One way that I propose to determine similarity between Minneapolis and various non-US cities is by using a similarity score for the size of these cities at various set land areas. Using the zip codes around Minneapolis, we can start at Downtown and successively add the most densely populated zip codes to define a ‘city’ of a certain area.

For example, Starting with Downtown and East Bank, then adding zip codes 55408, 55407, 55405, and 55409, you get an area of 50.4 square kilometers. This is  roughly all of Minneapolis between Hennepin Ave and Highway 55 in the North, the lakes in the west, 46th Street in the south, and 26th Street in the east, and it has 197,462 people, according to the most recent 2016 American Community Survey estimate.

Going out to 100 square kilometers extends this region east to Downtown St. Paul and 331290 people; 200 square kilometers picks up most of the rest of Minneapolis and Richfield with 554323 people, etc.

With these estimates in hand, we can use similar city districts or census statistical areas in other countries to estimate their city sizes at these same intervals.

For another research project I performed this exercise for all cities with a metropolitan area over 2 million in the European Union, Canada, and Australia, as well as many other smaller cities. Many of these cities offer websites in English, allowing me to get data straight from the source. In other cases, Italian or Spanish language Wikipedia help to fill in the gaps.

Calgary Mass Transit

Cities in poorer countries are not, in general, good comparisons for Minneapolis; almost everyone in Minnesota can afford a car, something that cannot be said about Ukraine, much less Bangladesh.

Unfortunately, I was able to make little headway with wealthy East Asian cities of Korea, Japan, and Taiwan city websites and native language Wikipedia in non-Latin characters are particularly difficult to navigate.

In this Google Docs sheet, I have the results of this analysis. In addition to the smaller city areas, I include the population within 90 km of the central city, as a stand-in for metropolitan area population, since the definition of these areas is quite divergent between the US and other countries.

Some notes on the Google Docs: city names in italics are part of some other city’s 90 km metro area; cities that are highlighted in yellow have very questionable or non-official data sources so I had to do the best I could.

To calculate the similarity score between cities, I used the first order Minkowski distance, also known as Manhattan distance. The lower the score, the greater the similarity between cities. When ranked by similarity, we can select the ten cities that are most similar to Minneapolis from Europe, Canada, and Australia.

  • Calgary, Canada
  • Brisbane, Australia
  • Ottawa, Canada
  • Zurich, Switzerland
  • Nottingham, UK
  • Sheffield, UK
  • Mannheim, Germany
  • Stuttgart, Germany
  • Antwerpen, Belguim
  • Nuremburg, Germany

These cities are interesting bunch, some with qualitative characteristics that make them much like Minneapolis, some very much otherwise. Calgary, Brisbane, and Ottawa are like Minneapolis in that they are far from any other large cities; but while their densest 500 square kilometers is similar to Minneapolis, these cities have significantly smaller metropolitan populations.

Stuttgart and Zurich, on the other hand, are at the center of large Metropolitan agglomerations; Stuttgart in particular probably has more in common with Houston or Atlanta than Minneapolis.

Zurich Mass Transit

Nottingham, Sheffield, and Antwerp are connected to nearby and much larger cities; Nottingham is 45 miles from Birmingham, Sheffield is 30 miles from Manchester and Antwerp is 25 miles from Brussels. In some ways, Nuremburg is qualitatively the most similar to Minneapolis; the metropolitan region is about 3.5 million people and it is 100 miles from any larger cities.

Regarding the density of these cities, there is quite a distribution. Calgary and Nottingham have almost exactly the same population as Minneapolis at the 500 square kilometer range, Stuttgart is slightly more populous, while the other cities are all less populated.

Calgary, Brisbane, and Ottawa are all slightly less dense than Minneapolis in the city center, while the rest of the (European) cities are more dense, with Antwerp in particular almost twice as dense as Minneapolis’ central city.

Finally, as for climate, the cities range from as cold and snowy as Minnesota (Ottawa and Calgary) to tropical (Brisbane), with European cities both snowy and continental (Zurich and Nuremburg) as well as foggy and oceanic (the English cities and Antwerpen).

Brisbane Mass Transit

Overall, these cities exhibit a wide range of characteristics and come from widely distributed geographic area, which is a good thing for comparison. In the next article, I will break down the transit systems of these 10 cities to see if there is anything Minneapolis can learn from its foreign counterparts.

8 thoughts on “Do Cities Like Minneapolis Have High Transit Ridership? (Part 1)

  1. Aaron Berger

    I recently gave a talk on between-city comparisons for bicycling and walking to work. I understand that you’re specifically looking at international comparison cities so this is more of a side note than a response. For US city comparisons I found the Chicago Fed’s Peer City Identification Tool to be useful. They use identify most-similar cities in equity (racial and socioeconomic composition), resilience (economic change and labor market conditions), outlook (signs of economic and demographic future), and housing (affordability, tenure, and age of housing stock).

    https://www.chicagofed.org/region/community-development/data/pcit

  2. Scott E.

    Great article. I love city comparisons, and have always wanted to know this specific info. I’m a bit surprised about the cities in Europe that compare to MSP, as I had imagined some Scandinavian cities being on the list. Looking forward to the next article!

    1. Daniel HartigDaniel Hartig Post author

      Helsinki is right out of the top 10 (you can see from the Google Sheets link, if you care). Copenhagen and Stockholm are both _much_ denser than Minneapolis in their city core’s: both fit about a million people in 2/5 the size that Minneapolis does.

  3. Lou Miranda

    Really interesting, though it’s probably easier to make a like-for-like comparison to other cities in North America than any old world city, where cities are much closer together. Even the Northeast (US) has strings of dense cities close to each other that is completely different from the Twin Cities, which are an island unto themselves for hundreds of miles around. We’re pretty unique.

    Even so, we should definitely have far more emphasis on transit than we do. Vienna, Austria, in particular, has a concern for transportation & affordable housing that we should emulate.

  4. David MarkleDavid Markle

    Did you include Zip 55454 in downtown Mpls?

    Nuremberg has smaller cities of note not far away, namely Fuerth and Erlangen. (I seem to recall that Henry Kissinger came from the latter.)

    North American cities should provide the best comparisons for Minneapolis, being of roughly comparable age and having grown during the age of the automobile.

    1. Daniel HartigDaniel Hartig Post author

      Downtown is 55401, 55402, 55415, and 55454. East Bank is 55414 and 55455.

      Somewhat surprisingly, Calgary, Brisbane, and Ottawa are the 1, 3, and 4 comparables overall for Minnesota, with Denver as #2. Otherwise, eight other US cities intrude into this top 10.

      I’ll be doing a separate article on US comparables, eventually. In the case of the US I have a lot more Census data to use to make comparisons; in Europe it’s pretty much just population.

  5. helsinki

    After reading that Stuttgart and Atlanta have something in common (they do not) I think your methodology requires some modification. Specifically, it doesn’t seem like your “density” metric takes land use into consideration. For instance, Stuttgart is extremely hilly; much of the developed core is punctuated by undeveloped hills. Atlanta, by contrast, is as flat as a pancake, with very few gaps in the sprawl. Also, in Germany generally it is common to have open agricultural land very close to the city center, with “suburbs” not really contiguous to the city, but isolated as pockets of dense development surrounded by open farmland. The implications for transit use are obvious and important to articulate.

    1. Daniel HartigDaniel Hartig Post author

      Stuttgart and Atlanta have very little in common according to my methodology. I was simply pointing out that Stuttgart is at the center of a highly populated region of ~6 million, as is Atlanta.

      My metric accounts for land use at whatever geographic areas are available with information. In the US, that is zipcodes. In Germany, this is generally city districts (stadtbezirk, most commonly, as seen here: https://de.wikipedia.org/wiki/Liste_der_Stadtbezirke_und_Stadtteile_von_Stuttgart). If the zip code or stadtbezirke lump hills and forests into populated zones, there isn’t much I can do about that.

      Importantly, I use population weighted density measures to run similarity scores. That is, the density if each zip code or stadtbezirke is weighted by its population. A low density park area will not affect overall density much because of its low population.

      To see this in effect, the population of Stuttgart’s densest 500 km^2 is 1.09 million with weighed density of 2812 people / km^2. Minnesota has almost the same population, 1.04 million, but weighted density of 2431 people / km^2. So, you see that Stuttgart’s hills and forests are accounted for. Marseilles 1.06 million people are at 6598 people / km^2.

      To show that they are not similar, Atlanta’s numbers are 0.71 million and 1600 people / km^2. Stuttgart’s most similar city in the US is Seattle (1.14 million, 2927 people / km^2).

Comments are closed.