Africa Ascendant

The stagnation of the eighteenth and nineteenth centuries coincides with the peak of the Atlantic slave trade, which removed millions of people, overwhelmingly young adults, from West and Central Africa, and with the violence and disruption that accompanied it. Historians continue to debate how much of the demographic standstill the trade caused directly and how much reflects other factors, but it is clear that Africa missed the population growth that much of the rest of the world experienced in the same period. Colonial conquest and its epidemics extended the low point into the twentieth century.

Growth began in earnest after 1920 and accelerated after 1950, as mortality fell — especially among infants and children — while fertility remained high. Africa's share recovered to 9 per cent by 1950 and 10 per cent by 1975, and passed its seventeenth-century level of about 13 per cent around 2000. In other words, Africa's share of humanity at the start of this century was roughly where it had been four hundred years earlier; what is new is what comes next.

By 2024 the share had reached 18.6 per cent, and the UN projects 25.5 per cent by 2050 and 37.5 per cent by 2100. In the default scenario settings, the share passes half of humanity in the late 2350s. The arc, then, is a U-curve followed by a long climb: a return to Africa's historical weight in the first decades of this century, and a rise far beyond it over the centuries that follow.

Africa's rise is the mirror image of decline elsewhere. The UN projects the world's population to peak in the 2080s at about 10.3 billion and to be slightly lower, 10.2 billion, by 2100, as Asia, Europe and Latin America shrink. Africa is the only region still growing strongly at the end of the century, adding about 1.35 billion people between 2050 and 2100 alone. By 2100 more than one in three people alive will be African, and among children under 15 the share will be almost one in two.

The growth is concentrated. Nigeria is projected to reach 477 million by 2100, roughly double its 2024 population of 233 million; the Democratic Republic of the Congo grows from 109 to 431 million, Ethiopia from 132 to 367 million and Tanzania from 69 to 263 million. Angola, Sudan, Uganda, Mozambique, Kenya and Côte d'Ivoire all pass 100 million. Several of these will be among the most populous countries on Earth, and the cities growing within them will be among its largest.

Age tells the same story from another angle. Africa's median age was 17 in 2000 and 19 in 2024, against 43 in Europe; the UN projects 24 by 2050 and 35 by 2100, when Africa will still be the youngest region. Africa's share of the world's working-age population rises from 16 per cent in 2024 to 26 per cent by 2050 and 41 per cent by 2100. Whether that becomes a demographic dividend depends on whether the jobs, schooling and infrastructure are built for it — the subject of section 7.

The sensitivity is the finding. Holding the rest of the world at the Africa Ascendant setting, Africa's 2500 population is 0.7 billion if its long-run fertility settles at 1.8, 2.5 billion at 2.0, 4.6 billion at 2.1, 5.6 billion at 2.14 and 8.0 billion at 2.2. A difference of two-tenths of a child per woman, invisible in any single year, decides whether Africa in 2500 is smaller than today's China or larger than today's world.

The "UN 2100 rates" preset shows what happens if every region simply stays at the fertility the UN projects for the end of this century — 2.02 for Africa and lower elsewhere. Both populations then decline after 2100, the rest of the world much faster than Africa, so Africa's share keeps rising even as its numbers fall. That is a robust result: Africa's share of humanity rises in almost every combination of settings, because its fertility is projected to remain above the rest of the world's for longer.

None of these paths is a forecast. Fertility four centuries from now will depend on economies, technologies, cultures and choices that cannot be foreseen, and the simple growth model here omits age structure, mortality shocks and migration between regions after 2100. What the engine offers is discipline: any claim about Africa's population in 2300 or 2500 is, implicitly, a claim about long-run fertility, and it should be stated as one.

The future of the migrant layer is easier to bound than to predict. If the share of Africans living outside the continent stays close to today's level, the number abroad in 2100 is about 56 million; at 2 per cent it is 76 million, and at 3 per cent, a doubling of today's rate, it is 114 million. These are large numbers in absolute terms, but small against the 3.8 billion projected to live in Africa. Migration will matter enormously to the countries that send and receive migrants; it will not, on its own, change where most Africans live.

The historical layer shows why counting migrants understates the diaspora. The roughly 10 million people who survived the Middle Passage have, over a dozen generations, become ancestors of a population twenty times larger, and of many more who carry some of their ancestry without identifying with it. The same process is beginning for today's migrants. Their children and grandchildren are not migrants and do not appear in migrant statistics, but they carry the ancestral link forward.

That third layer — descendants and, especially, the descendants of mixed unions — grows by a different logic from migration. It depends on how often families form across ancestral lines, and over generations it spreads African heritage far more widely than migration does. Sections 5 and 6 model it, because no census measures it: most countries do not record ancestry at all, and those that do record self-identification, which is not the same thing.

At the default setting — a group making up 10 per cent of a population, with one child in five born to randomly paired parents — prevalence rises from 10 per cent to 22 per cent after five generations, 43 per cent after ten and 67 per cent after fifteen, roughly 450 years. Throughout, the average share of the group's ancestry stays at exactly 10 per cent, and by the fifteenth generation only about one person in seven has a quarter or more. Most of those who carry the ancestry carry a little of it. That is what "prevalence, not proportion" means.

Genetic studies show the pattern in today's populations. In a 2015 study of more than 160,000 Americans, Bryc and colleagues found that about 3.5 per cent of people who identified as European American carried at least 1 per cent African ancestry; in South Carolina and Louisiana, the share was about 12 to 13 per cent. In Brazil and much of Latin America, African ancestry is widespread among people of every self-identified colour. Self-identification and ancestry measure different things, and ancestry is the wider of the two.

Over longer horizons the logic becomes striking. Genealogical models by Rohde, Olson and Chang suggest that all people alive today share common ancestors who lived only a few thousand years ago, because the number of a person's ancestors doubles every generation. On the scale of centuries, the same arithmetic means that any ancestry entering a well-connected population tends, eventually, to be carried by most of it. The question is not whether African ancestry spreads but how fast — and that depends on how often families form across lines.

With the population paths of section 3 and the central base, the share of humanity with any recent African ancestry in 2100 ranges from about 44 per cent at slow mixing (0.1) to about 54 per cent at fast mixing (0.6); at the default moderate setting it is 45 per cent, and the Africa Ascendant figure of 48 per cent corresponds to a mixing intensity of about 0.36. Because Africa itself will hold 37.5 per cent of humanity, most of the 2100 figure is the continent's own population; mixing adds between six and sixteen percentage points.

By 2500 the range widens and the role of mixing grows. At slow mixing the worldwide share reaches about 75 per cent — close to the Africa Ascendant figure of 77 per cent — and at moderate mixing about 89 per cent. At the mixing intensity that reproduces the dashboard's 2100 figure, prevalence outside Africa approaches saturation by 2500, and nearly everyone alive would have some recent African ancestry. The dashboard's two published figures therefore imply that mixing is faster this century than later — a plausible pattern, but an assumption.

The robust conclusion is qualitative. Under any positive rate of intermarriage, the population with some African ancestry grows faster than the population of Africa, and over centuries it tends toward a majority of humanity — while the average African share of any one person's ancestry outside Africa stays small. The exact year in which it passes half of humanity depends on assumptions no one can verify; the direction does not.

The most immediate consequence is economic. By 2100 Africa will hold about 41 per cent of the world's working-age people and nearly half of its children. Whether that becomes a demographic dividend — rising output per person as the share of workers grows — depends on investment made in the next few decades: in schooling, health, electricity, transport and the cities that will absorb most of the growth. The experience of East Asia shows how large the dividend can be when the investment comes; the experience of countries where it did not shows the cost of missing it.

The second consequence is geopolitical and cultural. A world in which a third of humanity lives in Africa, and in which several of the largest countries and cities are African, is a world in which African languages, music, literature, religion and politics carry more weight in global life. That shift is already visible in culture and markets; demography will deepen it. Migration and the diaspora will amplify it, as they have for every population whose members and descendants have spread across the world.

The third is a question of identity. If, over centuries, most people come to carry some recent African ancestry, the categories through which societies have thought about race and heritage will fit the world less and less. Ancestry will be shared far more widely than identity; heritage will be held in common more than in exclusive parts. That is a prospect, not a prediction, but it follows from the arithmetic of prevalence — and it is one of the more hopeful implications of a long-run demographic story too often told only as a story of pressure.

Population, fertility, age structure and median age for 1950–2100 are from the UN World Population Prospects 2024, medium variant, via Our World in Data; Africa is the UN region. Historical populations from 1500 to 1940 are from the HYDE reconstruction as published by Our World in Data. Migrant stocks are from UN DESA's International Migrant Stock 2024, for people born in Africa by region of residence. Census, survey and genetic figures are cited in the text and listed below.

Beyond 2100, each region's annual growth converges, with a time constant of 50 years, from its UN 2099–2100 rate to ln(TFR / 2.1) / 29, the growth rate of a stable population with the chosen long-run fertility, a replacement level of 2.1 and a generation length of 29 years. The defaults (Africa 2.136, rest of the world 2.098) are calibrated to the Africa Ascendant scenario's 2500 anchors of 5.6 and 5.135 billion. The ancestry model starts from an assumed population outside Africa with any African ancestry in 2025, which grows with that population, spreads at m / 30 × a × (1 − a / N) a year, and gains new African-born arrivals at the chosen mobility ratio times Africa's population over an average stay of 30 years. Everyone in Africa is counted as having African ancestry.

The prevalence illustration tracks the full distribution of ancestry fractions in bins of 1/256 across generations. "African ancestry" throughout means recent ancestry, from people of the continent in historical times: all humans share deeper African ancestry from the dispersal out of Africa tens of thousands of years ago. All post-2100 and all ancestry figures are scenarios, not forecasts or counts. Related in this series: the last third, on development across the Global South.