How many of us, and how old.

The solid line is the UN's reconstruction of the past, from censuses (the last in 2011), the Sample Registration System and household surveys. The dashed line is a projection: what happens if fertility, death rates and migration follow the UN's central assumptions. The shaded bands show how uncertain that is. The wider band is the range the UN's model gives a 95% chance of containing the true figure.

A projection is not a forecast of policy or a target. It mostly follows from people already born: the women who will be mothers in 2045 are alive today.

Each bar is one five-year age group. A country with many births and early deaths draws a triangle: wide at the bottom, narrow at the top. As births slow and people live longer, the bottom stops growing and the middle and top fill out. Where the marigold tick sits beyond a bar, that group is projected to be larger in 2050; where it sits inside, smaller.

The median age splits the population in two: half are younger, half older. It tells you whether a country's typical person is a schoolchild, a young worker or a parent of grown children.

When child deaths fall before births do, a country first gets younger: more babies survive. That happened in India until the 1970s. Once families have fewer children, the median age climbs steadily.

Economists call the years when the working-age share rises a "demographic dividend": more earners for every mouth to feed. Fifteen to 64 is a convention, not a fact about Indian lives. Many 15-year-olds are in school, many 64-year-olds still work a field, and the share counts people of working age, not people with work.

The dividend is a window, not a payment. It pays only if those of working age find jobs; the jobs guide shows how many do.

A dependency ratio counts people too young or too old to be expected to earn, for every 100 who are of working age. Children cost school fees and food; older people cost medicine and care, and in India most of that care is given by their own families, not by pensions.

The fertility rate is a snapshot of this year's behaviour, not a count of any real woman's children. Replacement is a little above two because some children die young and slightly more boys than girls are born. Below it, each generation is smaller than the one before, though the population keeps growing for decades while the large generations born earlier have their own children.

The national figure hides wide differences between states. The NFHS guide has India's own survey measure; the gender guide has the Sample Registration System series.

The gap between the two lines is the population's natural growth. Deaths rise even as health improves, simply because there are more old people. When the lines cross, the population shrinks unless more people arrive than leave.

Fewer births is already visible in classrooms: fewer children enter school each year, which frees room for spending per child, or closes village schools.

Life expectancy at birth is the average lifespan a baby would have if this year's death rates held for its whole life. It is pulled down hard by deaths in infancy, which is why most of its rise since 1950 came from children surviving, not from old people living longer.

"Urban" follows the Census of India: statutory towns with a municipality, and census towns with at least 5,000 people, a density of 400 a square kilometre and three-quarters of working men outside farming. Many large villages that look and work like towns are counted as rural, so this share understates how urban India's way of life has become.

The UN's urban figures were last revised in 2018 and, without a census since 2011, everything after rests on projection.

Cities grow three ways: their own births, people moving in from villages, and villages being reclassified as towns when they cross the census thresholds. The figures do not separate the three. A jump at a census year is partly reclassification.

Net migration is a balance, and it hides how many move each way. It counts people crossing India's borders to live, not the far larger movement within India, from village to city and state to state, which the census and surveys measure separately. Money sent home by Indians abroad is in the external account guide.

The variants change one assumption at a time. Low and high move fertility half a child below or above the medium path; constant fertility holds today's rate forever; zero migration closes the borders. None is a prediction. Together they show which assumption matters most, and for India that is the number of children families choose to have.