There are countless rankings of cities out there. And most of them probably don't mean very much. But the concept of a "global city", as coined by Saskia Sassen in the early 90s, is still immensely fascinating to me. And that's because there is, in fact, an order. There are cities that are less and more important to the global economy.
To this end, Resonance Consulting has just released their annual ranking of the world's best cities. And this year, they've introduced something new to their methodology: perception data. For this, they partnered with Ipsos and asked more than 22,000 respondents across 30 countries the following three broad and open-ended questions:
What are the top 3 towns or cities you would most like to live in someday?
What are the top 3 towns or cities you would most like to visit in the next 12 to 24 months?
What 3 towns or cities do you believe currently offer the best job opportunities?
The intent with these questions was to not anchor people to a specific list of places, and to not necessarily anchor people to big global cities. Maybe the best job opportunities are believed to be in small towns that most people aren't thinking about. The result is that thousands of different towns and cities were mentioned during the survey period.
While this didn't necessarily impact the cities and usual suspects that you would expect to see in a ranking like this -- cities like London, New York, and Paris -- it did change certain things and offer some interesting insights. For example, the strong global perception of Sydney helped to move it into the top 10 for the first in the ten-year history of this report.
On the other end of the spectrum, negative sentiment (outside of China) toward Hong Kong caused the city's ranking to drop precipitously. It is now ranked 97th, behind cities like Naples (Italy), Birmingham (UK), and Rochester (US). Singapore, in case you're wondering, is ranked 5th:
Boy, population densities can be so misleading. The typical approach is to just take the number of people and divide it by a given area. This then gives you something like X number of "people per square kilometer." The problem with this approach is that there are countless factors that can skew your result.
