What City is The Best for You?: A comparison of global cities through data exploration
As a student at an internationally-minded university, I have often wondered why most graduates find themselves in the same few American cities after graduation. While New York or San Francisco might be great places to start off your career, college students tend to not explore other global cities, some of which might even be a better fit for them. In this post, I’ll be analyzing over 40 cities from around the world to demystify their pros and cons, broadening the opportunities for where you could consider living in the future. My analysis is based on data gathered from Numbeo, a platform that collects data on the cost of living in almost all cities every year, TomTom International, an organization that collects traffic data from over 416 cities annually, The World Cities Culture Forum, and Teleport, a company that builds software to help people explore cities based on their preferences. All the data analysis was done in R.
To start off, why would you want to live in a city in the first place? Urban areas tend to be overcrowded, more expensive and more polluted. However, cities offer a plethora of opportunities and amenities you can’t get anywhere else. The word cloud above shows a collection of the most frequently used words in the Wikipedia post about cities. Words like “economic”, “development”, “cultural”, “networks”, and “influence” depict the importance of cities. They are spaces of financial growth and cultural hubs, and they facilitate the creation of a multitude of connections with others.
Now that I have established why you would want to live in an urban area, I’ll be exploring the disadvantages and advantages that set specific cities apart using data. Unfortunately, while I would love to do an extensive comparison of all global urban areas, my data is limited to the 44 cities whose cultural amenities information are in the World Cities Cultures Forum datasets. The distribution of this data is fairly biased with most cities being located in Europe. As a full disclaimer, below is a pie chart showing the continental distribution of the cities I will be analyzing for the rest of this post.
As you can see, about 40% of the cities will be in Europe and almost a quarter of them will be in Asia. However, the data spans all 6 livable continents, giving us a variety of places to explore.
I was interested in knowing how the cost of living differed between cities and had changed over time using the Numbeo Cost of Living Index. The chart below shows a time series of the cost of living in 10 cities randomly chosen from the data set. Interestingly, it appears that almost all cities’ costs of living follow the trend of increasing between 2017 and 2018 but decreasing in 2019. The only city whose cost of living appeared to stably reduce throughout those three years is our very own Philadelphia.
While Philadelphia’s cost of living might have been stably reducing, it still had a high cost of living relative to other cities. Cape Town, South Africa and Lisbon, Portugal had the lowest costs of living by far. In fact, Cape Town’s cost of living is about half of the most expensive city to live in within the 10-city sample, Tokyo. If you are solely picking a city based on costs, it might be time to move to South Africa.
While a city might have a higher cost of living, if its average salary supports that lifestyle, it could still be a suitable option. After scraping the average salary of each city from Teleport, I combined this with the Numbeo data to create the scatterplot below.
The scatterplot shows a very clear correlation between a city’s cost of living and its average salary. This is good news because if you decide to live in a more expensive city, you are likely to receive a salary that matches the cost of living. This scatterplot also reveals some “great deal” cities. For example, Shanghai, Dubai, and Austin have relatively high average salaries relative to their cost of living. You should consider these cities if you want some extra disposable income. On the other hand, most European cities tend to have lower average salaries compared to their cost of living. For example, the average salary in Austin is more than double that in Rome and yet the cost of living is comparable. There’s a similar situation when comparing Brussels and Los Angeles and New York and Oslo. It should be noted that the average salaries are likely driven up by the especially high salaries of individuals with the highest income such as in cities like San Francisco, New York, Hong Kong, and Austin.
Given a city’s cost of living is largely driven by the amenities it offers, it is fascinating to check out which urban areas provide the largest “bang for your buck”. Scraping the Teleport site, I collected the number of sports venues and concert venues as proxies for the number of amenities. Below is a bubble plot with the bubble size based on the cost of living index.
Your idea of an ideal balance of a city’s amenities depends on your preferences. If you are not a big sports fan but love concerts, you might want to consider moving to Amsterdam or Brussels. If you want to be located where you will have the most variety in sports, London is your best bet. However, it is important to note that sports centers in different cities will be based on the most significant sports in that region. You’re unlikely to find many baseball centers in London! Three cities that stand out are Istanbul, Sao Paulo and Moscow. They have relatively high numbers of sports centers and concert venues, yet their cost of living indexes are fairly low. If you are looking for a city with a great “bang for your buck” then Istanbul, Sao Paulo and Turkey should be high on your list. On the other hand, it’s quite clear that Zurich has very few entertainment amenities relative to its very high cost of living which is probably not ideal for young workers.
Now that we have looked at entertainment amenities, we can explore the offerings of public amenities in the cities which gives us insight into how much the local government provides. To find this information, I collected data on the number of museums from teleport and the number of public libraries and community centers from the World Cities Culture Forum. WCCF only provides data on the number of community centers for 20 cities which are shown in the chart below.
In terms of the variety in overall public amenities, Paris and Seoul are on top. However, most of their amenities are specifically public libraries. For readers who are book lovers, Seoul offers the largest collection of public libraries, having 1077 of them. If you are mostly concerned about community engagement, then you should consider Melbourne or Los Angeles, each having over 450 community centers. I was personally surprised by the overwhelmingly large number of museums that San Francisco offered because I have never really considered it a “historic” city. Nevertheless, most of San Francisco’s public amenities are museums. Few cities, namely Helsinki and Edinburgh have an equal share of the different amenities. Finally, while we already established that Cape Town has a one of the lowest cost of living, its variety of public amenities is on par with most cities, again proving that this is a great city to get a “good deal” on the optionality of amenities.
Although it is great for a city to have many amenities, these amenities would be useless if you are not able to reach them due to traffic. For this reason, I collected data on cities’ congestion levels by scraping the TomTom International website. Because pollution is also an important aspect to consider especially with regards to health, I combined the pollution index data from Numbeo with the congestion data to create the scatterplot below.
The scatterplot shows that there is a positive correlation between congestion levels and pollution although this relationship is not as strong as that between salary and cost of living. One of the reasons for this might be the fact that the pollution index does not only include air pollution but also water pollution and overall cleanliness of the city. Because the data from TomTom was listed as the number of hours lost driving in rush hour, I am not sure if it is biased against cities that might have poor driving infrastructure but very good public transportation. Nonetheless, a clear winner in the case of congestion and pollution are Scandinavian countries: Helsinki, Stockholm, and Oslo. These countries have the lowest levels of congestion and pollution. Previously, we found that Moscow had some of the highest number of sports and concert venues. Unfortunately, it appears to also have a very high level of congestion which could make reaching these areas difficult. Some cities with fairly bad pollution despite not being very congested are Dubai, Philadelphia, Barcelona, Shanghai and Beijing. These cities’ high pollution might be coming from a source other than the transportation sector. On the other hand, some cities with high driving congestion but pretty low pollution are Zurich, Edinburgh, Dublin and Tel Aviv. If you are willing to forgo driving and care more about clean air, you might want to consider these locations.
Finally, although the amount of amenities might attract you to certain cities, it is essential that you are comfortable there, especially if you identify as part of a minority group. Scraping data from Teleport, I collected the LGBT equality index as well as the tolerance for minorities scores which I scaled by 10. The results of 10 cities randomly chosen from the dataset are shown in the grouped bar chart below.
Some cities have enormous discrepancies between the two indexes. For example, while we have observed Cape Town as a city with a great “bang for your buck” in terms of amenities, specific minority groups might not feel comfortable there. Across the Indian Ocean, New Delhi has pretty low scores for both LGBTQ equality as well as tolerance of minorities. While Taipei has the highest scores for both groups in this 10-city sample set, its score for tolerance of minorities is much larger than that for LGBTQ-identifying individuals which is a similar case for Hong Kong. Contrastingly, Shanghai has pretty low tolerance compared to Hong Kong and Taipei. On the other hand, while Barcelona has a very high LGBTQ equality score, it has a fairly low tolerance for minorities.
Overall, my analysis shows that there are several global cities with a plethora of advantages to consider. Cities like Shanghai, Dubai, Austin, and Buenos Aires offer higher average incomes relative to their cost of living, possibly leaving you with more disposable income. Istanbul, Moscow, and Sao Paulo also have a lot of entertainment amenities considering their relatively low cost of living. If you are a book lover, you will probably want to consider London and Seoul which have enormous levels of public libraries. Cape Town specifically has a relatively good level of public amenities given how low its cost of living is compared to other global cities. However, Cape Town has a really low score for minority tolerance. Of the 10 cities chosen from the data set, Barcelona had the highest LGBTQ equality index and Taipei had the highest minority tolerance level. Unfortunately, many cities had big differences between their tolerance for LGBTQ individuals and minorities. It is important to ensure that given your various identities you would feel comfortable and accepted in whatever city you choose to live in. Altogether, the analysis provides opportunities for recent graduates to explore new cities to move to in the future that might be a better fit for them than they might expect!
Ayina Anyachebelu is a junior from Lagos, Nigeria in the Huntsman Program at the University of Pennsylvania. She is studying Business Analytics and International Studies.
This project was done for Dr. Prasanna Tambe’s OIDD 245 course: Analytics and The Digital Economy at the Wharton School.