Why Birth Rates Crash: The Demographic Transition

Early in 1929, in a small office at the Scripps Foundation for Research in Population Problems in Oxford, Ohio, an American demographer named Warren Thompson sat sorting a stack of national vital-statistics tables. They were dull documents, columns of births and deaths gathered from governments around the world, but Thompson noticed something in them that almost no one had named clearly before. Countries did not simply have high birth rates or low ones. They appeared to be moving, all of them, along the same path, just at different points on it. The pattern he was about to publish in the American Journal of Sociology would organize how the next hundred years read the population question.

It is a strange thing to realize that the explosive growth of the human population, from roughly one billion people in 1800 to more than eight billion today, and the sudden plunge in birth rates now emptying maternity wards in Seoul and Tokyo, are two faces of the same process. That is the claim the demographic transition makes. This article walks through what the model says, who built it, what actually drives each step, and why its most modern chapter remains one of the liveliest arguments in the field.

The Single Idea Behind a Century of Population Change

The Demographic Transition Model is, at its core, a descriptive account of how birth rates and death rates fall together as a society industrializes. The crucial word is together, but not at the same time. Death rates fall first. Better food supplies, cleaner water, basic public health, and eventually modern medicine mean that fewer infants die and more adults live to old age. Birth rates, however, are slow to follow. People go on having large families out of habit, religious conviction, economic logic, and simple inertia, long after the old rationale (that many children will die young) has stopped being true.

The result is a gap. For a stretch of decades, deaths drop while births stay high, and into that gap pours an enormous surge of population growth. This is not because people suddenly start having more children. It is because, for the first time, most of the children they have survive. Eventually birth rates do catch up, falling as the social and economic costs of children change, until births and deaths roughly balance again at a new, low level. The model traces, in other words, the journey from one kind of stability (high births, high deaths, slow growth) to another (low births, low deaths, slow growth), with a population explosion sandwiched in the middle.

Drawn on a single page, it is two curves over time. One line is the crude birth rate, the other the crude death rate, both measured per thousand people per year. The two lines start high and close together on the left. The death-rate line plunges first; the birth-rate line plunges later. The space between them swells in the middle stages and closes again at the end. Almost the entire history of modern population growth lives inside that widening and narrowing gap.

From Thompson's Tables to Notestein's Stages

Thompson published the first version of the idea in his 1929 paper, titled simply Population. He sorted the world's countries into three groups based on their growth patterns, an early and somewhat rough sketch rather than the polished diagram students learn today. What he had identified was the shape of the thing, the sequence of falling death rates followed by falling birth rates, even if he had not yet filled in the machinery.

The fuller version came from Frank Notestein, who worked at the Office of Population Research at Princeton. In his 1945 essay Population: The Long View, Notestein elaborated the four classical stages and, just as importantly, supplied the social mechanisms behind them. He argued that high fertility in pre-industrial societies was held in place by deep cultural supports, religious teaching, moral codes, family structures, all of which made sense when death rates were savage and large families were the only insurance against extinction. Industrialization, urbanization, and rising education slowly dissolved those supports, and fertility fell once children became an economic cost rather than an economic asset. Notestein gave the model its explanatory spine, and the framework that bears no single inventor's name is really Thompson's pattern fleshed out with Notestein's reasoning.

Walking the Stages One at a Time

The classical model has four stages, each defined by a characteristic relationship between birth and death rates. In Stage 1, the pre-industrial condition that covered almost all of human history, both rates are high and roughly balanced. Births are high because families need them; deaths are high because famine, disease, and war keep killing. The population is large in its fluctuations but grows little on average.

Stage 2 is where the transition begins and the population explodes. Death rates fall sharply, driven by improvements in food supply, sanitation, and public health, while birth rates stay stubbornly high. This is the widening-gap stage, the one responsible for the great surge in human numbers. In Stage 3, birth rates finally begin their own decline. As cities grow, as children go to school instead of to work, and as women gain education and access to contraception, the economic and social calculus of having many children inverts. The gap between births and deaths narrows, and population growth slows even though numbers are still rising. Stage 4 is the new equilibrium: both birth and death rates are low, the two lines run close together again, and the population stabilizes at a high level.

To this four-stage scheme demographers have added a contested fifth stage to describe something the original authors did not anticipate, a world in which birth rates fall not just to the replacement level but well below it, so that populations begin to shrink. This is the situation now unfolding in Japan, South Korea, and a growing list of high-income countries, and it is where the model's tidy story starts to fray.

What Actually Pushes a Society Through

It is tempting to read the diagram as a smooth automatic escalator, as if every country simply rides the curve from Stage 1 to Stage 4 by some law of nature. The truth is messier and more interesting, because each transition is driven by a different and not-guaranteed mix of changes. The fall in death rates that opens Stage 2 is largely a public-health and agricultural story: better nutrition from improved farming, clean drinking water, sewage systems, vaccination, and the eventual arrival of antibiotics. None of these requires people to change their beliefs or behavior; they simply stop dying.

The fall in birth rates that defines Stage 3 is a far deeper social transformation, and that is why it lags. It depends on urbanization, on the spread of mass education, and above all on the changing status of women, who in industrial economies marry later, work outside the home, and gain control over whether and when they bear children. It depends on the shifting economics of childhood, since a child on a farm is a pair of working hands while a child in a city is years of expensive schooling. Contraception makes the choice possible, but the desire for smaller families comes first. Naming these drivers separately matters, because it shows that the transition is not automatic at all. A country can lower its death rate quickly through imported medicine while its birth rate stays high for generations, which is exactly the demographic predicament much of the developing world faced in the twentieth century.

A Map of the World, Sorted by Stage

The real power of the model is that it turns an abstract diagram into a working map of the population world as it stands in 2024. Different countries are simply at different points along the same curve. Niger sits firmly in Stage 2, with death rates already falling but birth rates still very high and its population growing fast. Bangladesh and India occupy Stage 3, where fertility has dropped substantially from its peak but populations are still expanding. The United States and Brazil are in Stage 4, with low birth and death rates and roughly stable populations. Japan and South Korea have moved into the proposed Stage 5, where births have fallen below deaths and the population is contracting.

The numbers underneath this map are striking. The total fertility rate, the average number of children a woman would bear over her lifetime at current rates, now spans more than a tenfold range across the world. Niger stood near 6.6 children per woman in 2023, while South Korea recorded 0.72 the same year, the lowest national figure ever measured. To put that in perspective, the replacement level, the rate at which a population exactly replaces itself, is about 2.1 children per woman. Niger is at three times replacement and South Korea is at roughly a third of it, and both are described by the same model, separated only by where they sit along its single path.

How Fast, and Will Stage 5 Last?

The speed of the transition varies enormously, and this is one of the model's most important lessons. Britain, the first industrial nation, took roughly 150 years to move through its transition, an unhurried slide spread across generations. Japan and South Korea, latecomers who industrialized in compressed time after the Second World War, ran through the same sequence in only forty to fifty years. Some sub-Saharan African countries, meanwhile, show fertility plateaus that defy the textbook expectation, with birth rates that have declined and then stalled at a high level rather than continuing smoothly downward. The transition is a useful description of what tends to happen, not a schedule that every nation is obliged to keep.

Whether Stage 5 is a genuine new stage or a temporary phase is the active debate in contemporary demography, and it is worth being honest that the question is unresolved. Japan has been losing population since around 2011, and South Korea's fertility rate is the lowest ever recorded anywhere. China joined the club after its population peaked in 2022 and began to fall, a turning point of immense consequence for the world's most populous society of recent memory. Reading the model against these specific countries shows what the contested fifth stage looks like on the ground: shrinking workforces, aging populations, and economies bracing for a future with fewer young people than old. The doubt is whether sub-replacement fertility is a permanent floor that high-income societies have fallen through, or a deep but recoverable trough from which birth rates might one day partly rebound. Notestein's original model assumed populations would settle at balance, not slide below it, so the very existence of Stage 5 is a sign that the framework is still being rewritten by events its authors never saw coming.

Key Takeaways

The Demographic Transition Model, sketched by Warren Thompson in his 1929 paper Population and given its social mechanisms and four classical stages by Frank Notestein in 1945, describes how birth and death rates fall together as a society industrializes, with death rates dropping first (through better food, sanitation, and medicine) and birth rates lagging behind until the changing economics of childhood, mass education, and the shifting status of women bring them down too. The widening gap between falling deaths and still-high births in Stage 2 is what produced the modern population explosion, while Stage 4 represents a new low-level equilibrium, and a contested Stage 5 has been added to describe the sub-replacement fertility and outright population decline now seen in Japan, South Korea, and China. Today the same single curve maps a world stretched across a tenfold fertility range, from Niger near 6.6 children per woman to South Korea at 0.72, with transition speeds ranging from Britain's leisurely 150 years to East Asia's compressed forty to fifty, all of which reminds us that the transition is a powerful description of what societies tend to do, not an automatic escalator they are guaranteed to ride.