The Populated Density metric: a distribution-based view of country populations
Using high-resolution datasets to create a new view on population patterns around the world
Introduction
Australia is apparently one of the most sparsely populated countries on the planet. Ranking tables of countries by population density as presented in Wikipedia, the World Bank and the CIA Factbook often place Australia near the bottom with a density of just 3 people/km². The United States by comparison has an overall population density of 35 people/km². This suggests the USA is over 10x more densely populated than the land down-under.
However, many residents of Australia live in apartment blocks, and the skylines of major cities such as Sydney, Melbourne and Brisbane feature many tall skyscrapers similar to large US cities:
If Australia is such a sparsely populated country, then why do so many residents live and work in such high-density buildings?
Much of Australia is covered by desert and the vast majority of the population are concentrated in coastal cities. The vast central deserts are very sparsely populated, giving the impression of a low overall population density:
Comparing to the population density map of USA below we can see that in areas where both countries are populated there is a similar mixture of reds and yellows — certainly not different by a factor of 10x:
Thus, typical measures of population density such as (total population) / (total land area) may not reflect the experience of a typical resident of the country. Is there another metric we could develop that would capture the experience of a typical person living in a particular country?
We could look at the distribution of population density across populated areas only — a “Populated Density” metric. This article describes using a high-resolution population density dataset to create these distributions, and the sometimes surprising results.
Methodology
The Kontur Population Dataset provides a high-resolution count of population down to areas of 0.7km² — approximately the size of a small neighbourhood. This large-scale dataset can be combined with country boundaries from Natural Earth using cloud-based analytics computing platforms such as Snowflake to create an high-resolution density dataset of approximately 32 million cells. Thus we can now describe the population as a distribution across an entire country instead of a single number.
The next challenge is how do we compare countries? One advantage of using (total population)/(total land area) is that we have a single number to construct a ranking. The inherent assumption with this metric is that the population is uniformly distributed across the entire land area of a country.
Now that we have a variable density across the country we could use a weighted average to construct the Populated Density metric across all n cells within a country:
As we will see in the results the distributions of population density across countries takes on various shapes (see Technical Notes for more information). Different combinations of densities for particular countries could result in the same Populated Density value but combination of densities across the country may be quite different. Thus it can also be helpful to describe distributions using a set of categories to compare and contrast countries.
A collection of categories was constructed to summarise the population density distributions into a number of recognisable characteristics based on typical density. The cutoffs for each category were calibrated manually and represent characteristic types of housing at each density level:
Results
Top 30 Countries
The Populated Density for countries of the world are presented below. For reference the population density rankings from Wikipedia (based on United Nations data), the World Bank and the CIA Factbook are also presented for comparison.
Macao, Maldives, Hong Kong and Singapore are consistently ranked as high-density countries across all ranking systems. Interestingly however, Monaco is listed as the second most-dense country on Wikipedia, but is 23rd on the Populated Density scale. The density category breakdowns for the top countries highlights substantial differences in how the populations are distributed within each country. The top 5 ranked countries each have over 40% of the population living in ultra-high density surroundings:
The categorical distribution shows that whilst a large proportion of land area of Monaco is populated it has fewer ultra-dense high-rise buildings as seen in locations such as Singapore and Hong Kong. Images of the skylines of these locations correlates with these results:
Full results can be found in the accompanying results article, or also explore the full table on Tableau Public.
Australia vs. United States
Australia and the United States are relatively similar in terms of population density profile. Approximately 2.5% of the population in both countries live in high-density apartments, and close to 50% live in suburban housing:
Australia has 1.75x more population living in townhouses and low-rise apartments, whereas the USA has 1.7x more population living in low-density neighbourhoods and rural areas. A substantial proportion of the Australia actually lives in slightly more dense conditions than their American counterparts. Hence, Australia is not a paticularly sparsely populated country in terms of actual experience of residents.
Greece vs. South Korea
Interestingly South Korea and Greece both have approximately 25% of their populations living in high-density neighbourhoods. This would initially seem surprising that an EU country would have similar density to a populous country in South-East Asia, especially given the population density of South Korea reported on Wikipedia is 518 persons/km² compared to just 79 persons/km² for Greece:
A substantial proportion of the population of Greece is concentrated in Athens. Furthermore, Athens itself has a high proportion of the city population living in relatively high-density conditions:
Images of typical Athens neighbourhoods bear resemblance to areas in densely populated Asian cities such as Tokyo:
North America & South America
Mexico had the highest proportion of population living in high-density areas out of North American countries. The proportion of population in South American countries living in high-density conditions is often higher than countries in North America, with Colombia and Peru having the highest percentage in this category:
Asia & Middle East
Asian countries had a high number of countries where a significant proportion of the population live in ultra high-density neighbourhoods. Macao, Singapore and Hong Kong all have a significant majority of their populations living in high-density scenarios. Yemen and Afghanistan also have very concentrated population density, higher than countries such as Japan and South Korea:
India and China had relatively similar population density profiles with 11%-17% of the population in high-density conditions and the remainder of the population living in lower-density settings.
Africa
Many African countries have a surprisingly high proportion of their population living in very high-density neighbourhoods. Eritrea, Sudan and Somalia all have a majority of their population living in the high-density category, despite being placed 191st to 203rd in the Wikipedia density rankings:
In the cases of Sudan and Somalia desert geography has concentrated the populations to cities such as Khartoum and Mogadishu which have substantial areas of high population density. Interestingly, the areas of high density in cities such as Khartoum and Mogadishu are typically characterised by vast expanses of medium-height apartment buildings with relatively little open spaces. This stands in contrast with high-density areas in some other cities which have tall residential skyscrapers punctuated by public areas and green spaces:
Europe
European countries showed a wide range of density situations, with Monaco, Estonia and Greece having the highest proportion living in high-density neighbourhoods. Despite its vast land area, Russia had a similar density profile to countries such as Spain, Italy and Portugal:
Oceania
Oceania represents a small fraction of the world’s population with island nations representing most countries. Thus suburban homes and low-density neighbourhoods are typical population settings for the majority of countries in this region:
Limitations
The population density categories presented in this article represent an alternative approach to characterising an element of life in various countries. There are some known limitations to this approach which can impact interpretation of the results and can be improved upon in future iterations of the metric:
- The source data is not from official government sources. Thus there may be data quality issues with both the population counts and country boundaries used in this analysis.
- The overall population density metric depends on the resolution of the dataset. The H3 Level 8 resolution from the Kontur Population dataset has been used in this article, with each cell representing approximately 0.737 square kilometres. Results will differ with lower-resolution data, particularly the ability to identify ultra-high population situations.
- Categories such as “ultra-high density” could represent several different styles of high-density living. For example high-rise towers with substantial space between buildings would show a similar density result to large numbers of medium-height buildings in close proximity with little public areas.
More details on methodology and limitations can be found in the accompanying Technical Notes article.
Conclusions
The Populated Density metric uses high-resolution population distributions to generate a view on the lived experience of people in countries around the world. Broad density averages reported by institutions such as Wikipedia and World Bank do not capture this aspect of population density and assume an entire population is uniformly distributed across the land area of a country — an assumption that is challenged in many cases.
Countries such as Australia and the United States have relatively similar density profiles. Many countries around the world with large land areas have populations concentrated in a small number of densely-populated cities — an effect not reflected in broad density averages. Expressing population density as a distribution enables grouping together countries based on similar density profiles to provide a richer method of comparison.
