Environment & Urbanization

World leading environmental and urban studies journal

Will Africa have the world's largest cities in 2100?

David Satterthwaite
07 Oct 2016

A new report suggests that most of the world's largest cities in 2100 will be in Africa – including many with over 40 million inhabitants. This blog suggests growth in numbers will hinge more on the extent of economic development.

Lagos in 2010. Under different scenarios, Lagos could have 61 million inhabitants in 2100, or over 100 million (Image: Ricardo Gomes, Creative Commons, via Flickr)

The world's urban population is growing rapidly – and current projections suggest it will increase by 2.9 billion between 2015 and 2050, and it could grow another 3 billion by 2100. But where will this vast growth in the world's urban population live?

A new paper by Daniel Hoornweg and Kevin Pope projects the world's largest cities and their populations in 2100 under three different scenarios:

The world's five largest cities in 2100, according to different socio-economic scenarios

  1. The world makes relatively good progress towards sustainability, with sustained efforts to achieve development goals, while reducing resource intensity and fossil fuel dependency (rapid development of low-income countries; reduced inequality; rapid technology development; open, globalised economy). Global population of 7 billion, 90 per cent urbanised.

    Under this scenario, projected populations (millions) in 2100 would be: Lagos – 61.0; Mumbai – 52.3; Kinshasa – 48.8; Karachi  – 44.4; and Delhi – 44.3.

  2. This assumes extension of current trends in urbanisation, along with middle of the road assumptions about population growth, technological change and economic growth. Global population of 9.5 billion, 80 per cent urbanised.

    Under this scenario, projected populations (millions) in 2100 would be: Lagos – 79.8; Dar es Salaam – 62.3; Kinshasa – 60.3; Mumbai – 57.7; and Karachi – 49.9.

  3. Fragmentation: Urbanisation follows the slow pathway. In high-income countries, low population growth (especially ageing), slow economic growth and technological changes reduce the incentives for urban expansion. Elsewhere, the population grows rapidly, particularly in rural areas, causing significant land use change and environmental degradation. Global population of 13.5 billion, pockets of extreme poverty and moderate wealth, many countries with rapidly growing populations.

    Under this scenario, projected populations (millions) in 2100 would be: Lagos – 100.2; Dar es Salaam – 77.5; Luanda – 69.2; Karachi – 52.9; Mumbai – 52.6.

The five largest cities in 2100 do not vary much but their populations do – under scenario 1, Lagos has 61.0 million inhabitants in 2100 but over 100 million under scenario 3.

The Hoornweg and Pope paper also gives estimates for the population of the world's 101 largest cities in 2100 based on extrapolations of past growth rates. These present some surprises: many of the world's largest cities are in Africa. Indeed, 13 of the world's largest cities in 2100 may be in Africa (see table 2).

The world's 20 largest cities in 2100 (based on projected populations)

City Population 2010  Extrapolated population in 2100
 Lagos  10.8  88.3 
 Kinshasa  9.4  83.5
 Dar es Salaam  3.9  73.7
 Mumbai  19.4  67.2
 Delhi  21.9  57.3
 Khartoum  4.5  56.6
 Niamey  0.94  56.1 
 Dhaka  14.7  54.3
 Kolkata  14.3  52.4 
 Kabul  3.7  50.3
 Karachi  14.1  49.1
 Nairobi  3.2  46.7
 Lilongwe  0.73  41.4
 Blantyre  0.70  40.9
 Cairo  16.9  40.5
 Kampala  1.6  40.1
 Manila  11.9  40.0
 Lusaka  1.7  37.7
 Mogadishu  31.4  36.4
 Addis Ababa  2.9  35.8

(Italics: cities in Africa) 

Words of caution

Large cities only exist because they have large economies. Mega-cities need mega economies, even if, for some, much of their economy is informal. There are some exceptions – for instance cities that become 'large' because of influxes of refugees or internally displaced persons.

So an alternative approach to predicting the world's largest cities in 2100 would be to predict the world's largest economies and then predict the spatial distribution of their economies as the basis for identifying the largest cities. Then estimates can be made for the population of these cities, based on projections for natural increase and for net in-migration up to 2100.

But of course, predicting the world's largest economies in 2100 and their size and nature is even more fraught with uncertainty than population predictions.

Rapid economic growth, booming populations

So what could change in the 84 years up to 2100?  If we had viewed China in 1977, who would have predicted it to have among the world's most rapid economic growth, and its largest cities – many now among the world's most populous – see such rapid rise in numbers?

Who would have predicted the population of Shenzhen going from under 30,000 inhabitants in 1970 to over 10 million inhabitants today (and many million more if Shenzhen's migrant workers were counted as city residents)?

Who in the aftermath of the Vietnam war would have predicted the economic success of the country's largest cities? Syria's largest cities would have figured in a long list of the world's largest before the terrible civil war that still continues there.

From modest to massive

Instead of looking 84 years ahead to 2100, why not go back 84 years to 1932: who would have predicted that the modest city of Sao Paulo with less than a million inhabitants in 1932 would come to be the largest industrial city in South America by 1980? 

Or in the United States, how a whole new generation of towns in the South would produce many of the world's fastest growing cities during the 20th century. Or that the proportion of European cities among the world's largest would see a sharp decline with more decentralised urban development and falling population growth.

Or that the long-established dominance of India's four largest cities (Kolkata, Mumbai, Delhi and Chennai) would be challenged by a new generation: Surat, Pune, Hyderabad, Bangalore and Ahmedabad  ̶  if not already, on their way to becoming very large cities. Or to go back to Brazil, a new generation of smaller (but becoming larger) cities such as Porto Alegre and Curitiba would challenge Sao Paolo and Rio de Janeiro for new investment.

Niamey, Lilongwe and Blantyre populations are projected to increase more than fiftyfold between 2010 and 2100; is this likely? Looking back, from 1930 to 2010, many cities had populations that grew at this rate including Dhaka, Chittagong and Khulna in Bangladesh and the largest city in most sub-Saharan African nations. Some US cities too, including Phoenix and Miami. But all these were relatively small cities in 1930; many had less than 20,000 inhabitants.

Caveat: economic development

So can we imagine the economies needed to support very large cities in countries that at present have small economies?  

To go back to the predictions for 2100, what must happen in Malawi for its two largest cities to multiply their population more than fiftyfold in the next 84 years? What would have to change to make Mogadishu – torn apart by civil war – a city with more than 30 million inhabitants in 2100? Dar es Salaam and Kampala are successful cities within their nation but it is difficult to imagine an economy developing that can support a city of 74 million for Dar and 40 million for Kampala.

But for these African cities, their projected population assumes very high population growth rates for their nation will continue. Niamey having 47 million inhabitants in 2100 does not seem to make any sense – but if Niger reaches its projected national population of 215 million in 2100, this projection for its capital does not seem so far off.

Sustained growth?

So, for the urban future. The list of the world's largest cities and their populations depend so much on where private capital (local, national, global) investment chooses to locate (and where competent city governments develop to attract this).  

Where these concentrate in large cities we may well get many cities with 30 million or more inhabitants. But competent local governments in smaller cities can draw some of this away from the super-large cities. Coastal cities at high risk from climate change may lose new investment to inland cities.

Total urban populations for all nations will be heavily influenced by the scale of natural increase in populations but will the population growth rates used in estimating city populations really be sustained?

We can anticipate (and even see already) more decentralised patterns of urban development in many regions. What would a world committed to climate change adaptation and mitigation, as set out in the Paris Agreement, bring to the future of our cities?  

And looking to the Sustainable Development Goals, with their ambition for cities that are inclusive, safe and resilient, how would the urban future change if governments really took these goals seriously?

David Satterthwaite (david.satterthwaite@iied.org) is a senior fellow in IIED's Human Settlements research group and visiting professor at the Development Planning Unit, University College, London. This blog follows two blogs by Daniel Hoornweg; the first was on how rapid urbanisation, if managed sustainably, could ease the pressure of exponential growth set for urban areas, and the second was on the shifting power of cities.

This blog draws on Population predictions for the world's largest cities in the 21st century, by Daniel Hoornweg and Kevin Pope, to be published in the April 2017 issue of Environment & Urbanization. The paper is available under open access until 10 October 2016.

This blog post also appears on www.iied.org. Please follow this link to view and post comments: