City Ranking Tool

Me in London

https://kentbutt.shinyapps.io/City_Ranking_Tool_Global

Every city wants to be number one in any metric it can, even if it’s a silly one. Even my hometown of Edmonton had a claim to fame with the world’s largest car park, holding a record of 20,000 vehicles with an additional 10,000 in overflow (Guinness World Records, 1981). While we’ve stopped chasing that record, we’re now in the running for the “most affordable major city in Canada” — as long as you define major city just right.

Personally, I care much more about affordability than parking lot size, and I expect you do too. Yet, other preferences may differ wildly, and there’s an endless number of factors that might push you towards or away from a city. If you’re trying to figure out where to move, or weighing job offers between different cities, city ranking tools can be extremely helpful — insofar as those tools align with your preferences. Given that I will soon be making one of these moves, I wanted a tool perfectly aligned with my preferences. I figured if I added some scrollable sliders (or better yet, a survey), it could align with yours too.

So that’s how this project came about — a tool where you fill out a survey or move sliders to rank cities according to your preferences. Unfortunately, international datasets are hard to find, and so the sliders aren’t very granular. The best dataset I found was the Oxford Economics Global Cities Index 2024, which ranks the largest 1,000 cities in the world according to five factors: Economics, Human Capital, Quality of Life, Environment, and Governance. Here is a breakdown of their methodology:

The Oxford Economics Global Cities Index is a holistic ranking of the 1,000 cities included in our Global Cities Forecasting Service. While our best-in-class economic forecasts underpin these rankings, cities are scored across five different categories to achieve a well-rounded comparison of locations. The five categories are Economics, Human Capital, Quality of Life, Environment, and Governance
Pg. 69. Oxford Economics Global Cities Index 2024
Oxford Economics Dataset Methodology

The Oxford Economics team collected data on all 1,000 cities, assigning the highest-scoring city a value of 100 and the lowest a 0, then normalized the values across that range to create an indicator for each subcategory. These indicators were then normalized again, totaled, and re-weighted. Finally, cities were ranked based on these new weighted scores, both for each factor and overall. (Pg. 71). The subcategories are as follows:

  • Economics
    • GDP Size: City’s gross domestic product size.
    • GDP Growth: 5-year GDP forecast for the city.
    • GDP per Person: GDP divided by the city’s population.
    • Employment Growth: 5-year employment forecast.
    • Economic Stability: Consistency of GDP growth over the past decade.
    • Economic Diversity: Relative size of each economic sector.
  • Human Capital
    • Population Growth: 5-year population forecast.
    • Age Profile: Ratio of residents aged 65+ to those aged 15-64.
    • Universities: Number of universities weighted by their rankings.
    • Corporate Headquarters: Number of the world’s largest 2000 corporations headquartered in the city.
    • Educational Attainment: Average of mean years of schooling for adults 25+ and expected years for children aged 6.
    • Foreign-born Population: Share of foreign-born residents.
  • Quality of Life
    • Income Equality: Gini coefficient for household income equality.
    • Income per Person: Household disposable income per person adjusted for purchasing power parity (PPP).
    • Housing Expenditure: Share of disposable income spent on housing and utilities.
    • Life Expectancy: Life expectancy at birth.
    • Internet Speed: Average broadband download speed.
    • Recreation & Cultural Sites: Number of recreation/cultural sites per capita.
  • Environment
    • Air Quality: Mean PM2.5 concentration.
    • Emissions Intensity: CO2 emissions divided by GDP.
    • Natural Disasters: Number of natural disasters since 2000.
    • Temperature Anomalies: Difference between daily max temperatures and long-run averages.
    • Rainfall Anomalies: Difference between total monthly rainfall and long-run averages.
  • Governance
    The Governance category contains indicators that measure the political stability of a city and the degree to which residents’ rights are protected.
    In recognition of the fact that national governments—not just those at the city level—have a significant influence on these outcomes, this category is measured at the national level, rather than at the city level. As a result, every city in a given country receives the same score.
    ” (Pg. 69)
    • Institutions: Aggregate score of institutional context and rule of law.
    • Political Stability: Aggregate score of likelihood of political instability and violence.
    • Business Environment: Aggregate score of ease of doing business and corruption control.
    • Civil Liberties: Aggregate score of political rights and civil liberties.

The dataset also accounts for differing definitions of cities and urban regions by using “Functional Urban Areas” methodology that the OECD typically uses. (Pg. 71)

Oxford Economic’s raw data that constitute each category, and the category scores themselves, are proprietary and I certainly don’t have the budget to buy them. Instead, I have only ordinal data — the publicly available rankings for each category. This normalizes the gap between each city ranking to a uniform distance, reducing the accuracy of this ranking scheme, but it’s the best I’ve got for now.

The original rankings were in a very hard-to-extract PDF format, so I used ChatGPT to read the whole thing and put it into a CSV for me. I then built a basic R Shiny web app to play with, which contains both a survey to rank the cities and sliders to adjust each category independently. The default rankings are based on what Oxford Economics considers important, however without access to the underling scores Oxford’s rankings are not replicable, and different cities move to the top, which is fun to see as when you play around with the tool.

After playing around with the tool, I’ve got tons of improvements in mind. If this gets a lot of interest (and depending on how much spare time I have with my thesis), my next step will be to include an optimizer to adjust the dials, so you can figure out what ranking scheme makes your city shine the brightest! I also want to take a more granular look at Canadian and US cities since that’s where I think I’m most likely to live.

I’d be excited to hear if you used this tool, or found it useful. Send me a message if you have any thoughts (and especially if you have any favourite city-level datasets I may not have seen), and please comment below with what you most value in a city!

Check it out at https://kentbutt.shinyapps.io/City_Ranking_Tool_Global

Code available on my GitHub.

Oxford Economics Global Cities Index 2024 available at https://www.oxfordeconomics.com/global-cities-index/

Leave a Reply

Your email address will not be published. Required fields are marked *