fivemack: (Default)
Tom Womack ([personal profile] fivemack) wrote2006-05-18 11:43 pm
Entry tags:

Crowdedness metric

I've just been introduced, on [livejournal.com profile] james_nicoll's journal, to the Gridded Population of the World dataset: some kind demographers have divided the world into six-square-kilometre chunks and counted the people in them.

This lets me calculate the crowdedness metric, which is the average over each person in the population of the population density of the region that person lives in. This means that Canada Russia a country containing three incredibly crowded cities amidst endless desolate uninhabited wasteland gets a figure corresponding to the cities rather than the wasteland; it's about the metric that I imagine people actually using when complaining that their country is too crowded.

High-ranking places: well, of course the city-states of Hong Kong, Singapore and Macao are at the top. The next few are


  1. Egypt (mostly desert and everyone lives in Cairo)

  2. South Korea

  3. Jordan (a small country, but 95% desert)

  4. Yemen (ditto)

  5. Greece

  6. The Philippines

  7. Bangladesh

  8. Peru

  9. Japan

  10. Vietnam

  11. Indonesia

  12. Taiwan

  13. Lebanon



A biased sample of countries whose position in the ranking by density and perceived-density is similar:

Thailand, Nigeria, Hong Kong, Singapore, Denmark, Iraq, China, Taiwan, India, Israel, Hungary, UK, Austria

Countries with low density but high perceived density (that is, very concentrated population, or, if you prefer, vast tracts of desolation):

Russia, Gabon, Argentina, Canada, Peru, Australia, Brazil, CentralAfricanRepublic, Surinam, Paraguay, South Africa, Chile

Reasonable-sized (population >4 million) countries with high density but low perceived density (that is, very uniformly spread population):

Serbia, Bosnia, Burundi, Syria, Rwanda, Nepal, Germany, Slovakia, Czech Republic, Puerto Rico, Netherlands, Turkey

Jared Diamond writes of Burundi and Rwanda as being pretty much at carrying capacity at tech level, and I could imagine the same being said of Nepal (and Bhutan, which would be on the list were it larger); that's very strange company in which to find Germany. Albania would also be high on the list were it larger; is it ridiculous to imagine Tito and Hoxha as having pursued deliberate projects of decentralisation?

[identity profile] fivemack.livejournal.com 2006-05-19 08:02 am (UTC)(link)
It took 34 minutes and 88 lines of perl.

http://tom.womack.net/demog

[identity profile] james-nicoll.livejournal.com 2006-05-19 02:08 pm (UTC)(link)
Thank you. May I nick some of this for my LJ?

[identity profile] fivemack.livejournal.com 2006-05-19 03:10 pm (UTC)(link)
Naturally, that was roughly why I did the work.

Do with it as you will.

[identity profile] james-nicoll.livejournal.com 2006-05-19 04:34 pm (UTC)(link)
I meant to ask: what are the units?

[identity profile] mindstalk.livejournal.com 2013-01-09 04:04 pm (UTC)(link)
I would say "obviously square kilometers" but the numbers turn out to be odd. 2.87 million for Argentina, vs. 2.78 on Wikipedia. Multiplying Wikipedia Russia area by land percentage gives 14.79, vs. 13.7 in the table. Wikipedia gives 316 km2 for Malta's land area, the table has 586. France: 674,800 vs. 494,000, or 540,000 for Metropolitan France. 9.8 million vs. 8.3 million for the USA.