Chapter Three: Two Kinds of More

Imagine two towns. Call them Springhill and Sunset.

On paper, in every official record, the two towns are identical in the one respect we usually care about most. Each has exactly ten thousand residents. They return the same figure to the census. They appear the same size on the map, claim the same line in the county's population table, qualify for funding under the same formulas. If all you knew about Springhill and Sunset was the number ten thousand, you would treat them as the same place, and so would almost every economist, planner, and politician who looked at them.

Now walk down the main street of each.

In Springhill, the average resident is thirty-one years old. The school is overcrowded; there is a waiting list for kindergarten and a bond measure to build a second elementary. The pediatrician's office is booked solid. There are strollers outside the coffee shop and a youth soccer league with too many kids and not enough fields. The hardware store is busy because people are renovating, expanding, adding a bedroom for the baby coming in spring.

In Sunset, the average resident is seventy-one years old. It is quieter, and in its way lovelier — well-kept gardens, no traffic, a certain unhurried peace. But the school closed nine years ago for lack of children and the building is now a community center where the bridge club meets. There are three pharmacies and a thriving home-health-aide business. The busiest enterprise in town is the medical clinic, and the second busiest is the funeral home. There are no strollers. There is no kindergarten waiting list, because there is no kindergarten.

Same number. Ten thousand and ten thousand. And yet you already know, without my telling you, that these are not the same place, that they do not have the same economy, and above all that they do not have the same future. Springhill, barring catastrophe, will be larger in thirty years than it is today, because it is full of people who are making more people. Sunset will be smaller in thirty years, and smaller still in fifty, and the arithmetic by which this happens is not in dispute and cannot be argued with. The residents of Sunset are not going to have children. Most of them are past the age where that is possible. The ten thousand is not a population so much as a balance — a quantity that will be drawn down and not replenished.

The number lied. Or rather, the number told the truth about the present and said nothing whatsoever about the future, and we mistook the one for the other. It is the single most important mistake in the entire field of how we think about human numbers, and once you see it clearly you cannot unsee it.

Principal and Interest

Here is the distinction. There are two completely different ways a population can get bigger, and we use the same word — "growth" — for both, which is the root of nearly all our confusion.

The first way is that new people are born. Call this real growth. A child is born, and the population goes up by one. But that is the least of what has happened, because the child is not merely a unit added to a total. The child is, in time, a parent — a source of further people, who will themselves be sources of further people. Real growth compounds. One becomes two becomes four. Every new person is not just an addition to the population; they are a multiplier of all future populations. Real growth is principal in a savings account: it earns interest, and the interest earns interest, and the whole thing snowballs forward through the generations under its own power.

The second way a population can get bigger is that the people already alive stop dying — they live longer, so that at any given moment more of them are still around. Call this ambient growth. This, too, makes the number go up, and it can make it go up dramatically. If everyone who used to die at forty now dies at eighty, then at any given moment you are counting two generations of living adults where you used to count one. The population swells. But notice what has and has not happened. You have added years to existing lives. You have not added new sources of life. A man who lives to ninety instead of forty is, demographically, counted for longer — but he produces no more children at eighty-five than he would have, which is to say none. Ambient growth is a one-time deposit into the account. It raises the balance, once. It does not compound. It earns no interest, because longer life does not, by itself, produce a single additional human being who would not otherwise have existed.

Now you can see why the distinction matters so much. The two engines of the twentieth-century explosion did two different things. The conquest of early death — the part that saved children before they grew up — was real growth, because a rescued child becomes a parent. But the back half of the same revolution, the part that extended the lives of adults who had already had their children, or were never going to have more, was ambient. It was longevity. It made the headline number climb just as steeply as births did. On the graph, the two are indistinguishable — a person is a person, and the census counts them the same whether they are a newborn or a centenarian. But under the graph, in the part that determines what happens next, they could not be more different. One is principal. The other is a lump sum that sits there and is gradually spent.

The Mask

Now the thing all of this has been building toward, and it is genuinely alarming.

For most of the twentieth century, the two kinds of growth ran together, and the ambient kind — the longevity, the lump sum — was large enough and steady enough to mask what was happening underneath it. Because here is what was happening underneath it: the real birth rate, the principal, the only number that actually compounds and actually determines the long-run future, was falling. It had been falling, in the developed world, for decades. Family sizes were shrinking. The number of children the average woman would have over her lifetime was sliding downward, year after year, through the whole period in which the total population was visibly booming.

How can the population boom while the birth rate falls? Because the ambient engine was covering the gap. As long as people kept living longer — as long as the lump sum kept being deposited, as the lifespan revolution rolled through the population adding decades to everyone's life — the total number kept rising even as the engine of real, compounding growth was quietly losing power. The longevity gains were so large that they hid the fertility collapse the way a rising tide hides the fact that the river feeding the bay has slowed to a trickle. The bay is full. The bay has never been fuller. And the river is dying.

Demographers have a precise way of seeing through the mask, and it is worth knowing. They don't look only at the total population; they look at its shape — the distribution of ages, drawn as what they call a population pyramid. A healthy, genuinely growing population looks like a pyramid in the literal sense: a broad base of children, narrowing as you go up through the older ages, because in any reproducing society there are more young than old. That is what real growth looks like, drawn as a picture. But run your eye over the age pyramids of the developed world today and you will not find pyramids. You will find shapes that bulge in the middle and at the top and pinch in at the bottom — narrow bases, wide crowns. Japan's looks less like a pyramid than like an inverted one, balanced precariously on a thin point of children. These are not the silhouettes of growing populations. They are the silhouettes of Sunset: a large quantity of the already-old, resting on a small and shrinking quantity of the young. The headline number can stay flat, even rise, for a while. The shape tells you it is a balance being drawn down, not a principal compounding forward.

You can feel the mask slipping in our institutions before you can see it in the totals. Consider Social Security, and every pension system like it anywhere in the world. These systems were designed on an unspoken assumption that the base of the pyramid would always be broad — that there would always be many young workers paying in for each old retiree drawing out. That assumption was real growth made into public policy. And the reason these systems are everywhere in slow-motion crisis is that the assumption has quietly failed: the ambient engine kept the retirees alive and drawing benefits for longer than anyone planned, while the real engine failed to produce the workers to pay for them. The funding gap in your national pension scheme is not an accounting error. It is the ambient-versus-real distinction, presenting its bill.

When the Lump Sum Runs Out

So put the pieces together, because they assemble into something with a hard and specific conclusion.

The lifespan revolution, the great longevity boom, the ambient engine — it is a one-time transition that has very nearly run its course. The lump sum has very nearly finished being deposited. In the longest-lived countries the gains have slowed to months per decade; the door from forty to eighty is closed behind us and the door from eighty to ninety barely opens. Which means the masking effect is ending. The tide that has been hiding the dying river is going back out.

And when it does, what gets revealed is the real number — the principal, the compounding engine, the actual rate at which new human beings are being created. That number, stripped of the ambient camouflage that has flattered it for a century, is below the level required to replace the existing population almost everywhere in the developed world, and in a growing share of the developing world besides. In some countries it has been far below replacement for a generation already, hidden the entire time beneath the longevity bulge. This is the floor that the population has been resting on without knowing it, and as the ambient support is withdrawn, the total settles down onto that floor — and the floor is lower than almost anyone has been willing to say.

And this is why the distinction has to be drawn before a single forecast is attempted. Nearly every projection you have ever read about the human future treats population growth as a single thing — a trend, a slope, a number that has been going up and will keep going up until it gently levels off around nine or ten billion and plateaus there forever, comfortable and stable. That picture is built on the mask. It takes the masked number at face value. It does not separate the principal from the lump sum, the river from the tide, Springhill from Sunset. And the moment you do separate them — the moment you insist on asking not "how many people are there?" but "how many of them are making more?" — the comfortable plateau dissolves, and what stands behind it is a descent.

There is one more thing standing in the way of seeing this clearly, and it is the most audacious obstacle of all. To strip out the ambient growth and find the real number, you have to trust the underlying population data — you have to believe the census figures that the separation is performed upon. And there is very good reason to believe that the largest population figure on Earth, the one belonging to the country we are all assured is the most populous nation in human history, is not a measurement at all.

It is a fiction.