China's Population in the 21st Century

Griffith Feeney

Presentation at the University of Rostock, Rostock, Germany, Tuesday, July 8, 1997.


Good morning! It's a pleasure to be here from half way around the world and I thank you for the opportunity to address you this morning.

My talk is about demographic change in China and is divided into three parts. First, what has happened over the past 50 years. Second, what will happen over the next 50 years. Third, what these demographic changes mean for China and the rest of the world.

The Past 50 Years

Fifty years ago China had the demographic characteristics typical of a pre-modern society: high mortality, high fertility and young age distribution.

Mortality fell rapidly after 1949, but fertility remained high for two decades, with women having an average of about six children per woman at the end of reproduction. In consequence, the country experienced rapid population growth and an even younger age distribution.

Then in 1970 the government initiated an anti-natalist policy, the "later-longer-fewer" policy. In the following decade, fertility in China declined more rapidly than it ever has in a major country.

Here's a picture of this fertility decline.

Figure 1: China Age-Specific Birth Rate TFRs

shows a time series of total fertility rates, or TFRs. This is the standard measure of fertility used by demographers throughout the world. It projects current rates of childbearing into the future and tells us the average number of children women will have if current behavior continues into the future.

The plot gives a convenient indication of magnitudes and timing. In 1970 the TFR was about six children per woman. By 1977 it had declined to about 2.6 children per woman. This is an extraordinarily rapid decline. Fertility declines in other countries have taken several decades to half a century or more to do what China did in less than a decade.

There is however one problem with this picture, and a very serious problem. In 1979, flushed with the success of their earlier anti-natalist policy, China introduced a new policy, the "one-child family policy," that encouraged every family in China to have only one child. Enormous resources of all kinds were allocated to implementing this policy. In the picture you are looking at, however, fertility decline stops just before the policy was introduced and hardly declines at all subsequently. To judge from this picture--and it is, please remember, the standard way to represent fertility decline--the one-child family policy was a total failure.

Did the policy fail, or is there something wrong with this picture? The answer, it turns out, is that there is something wrong with the picture. The familiar total fertility rate of demography textbooks does a very bad job of representing demographic and social reality in this case.

Here's another picture showing a different rate that does a better job.

Figure 2: Two Pictures of China's Fertility decline

It is a statistic I invented specifically to study the impact of China's one-child family policy. The general idea and metric, average children per woman, are the same, but this measure controls for "parity" (how many children a woman has had at any given time) and birth order, which the TFR does not. The TFR series shown before is shadowed here for comparison.

The picture is dramatically different. The rate of decline is slightly lower, the decline continues all the way through to 1984, and bottoms out at just over two children per woman.

The fertility measure you see here is based on statistics called "parity progression ratios," which are proportions of women having a given number of children who advance to another child. The parity progression ratios provide very concrete evidence of the demographic impact of the one child family policy.

Here they are.

Figure 3: China Period Parity Progression Ratios

The line at the top shows the proportion of married women who have at least one child. This is high and constant throughout the period. There was no decline in motherhood during China's fertility decline.

The four lower series in the picture show the parity progression ratios for higher order transitions, 2nd to 3rd birth, 3rd to 4th birth, and so on. The declines are large and cover the whole period shown.

By far the most interesting and important series is for progression from 1st to 2nd birth. Before the one child policy in 1979, nearly all women had a second child. Within five years, however, the policy had brought the percentage of women going on to a second child down to just over 60 percent.

There is an important moral to this story of recent demographic trends in China: the necessity of presenting a reasonably full statistical picture when addressing complex social questions. The first TFR series we looked at suggested that the one-child family policy failed. The second TFR series suggested that the policy had a substantial demographic impact. The parity progression ratios indicated the impact very specifically, but they also showed that the policy did not eliminate all second births, and indeed that significant proportions of women continued to have three or more children. If we were allowed to pick and choose among these statistics, without presenting the whole picture, we could argue equally well that the policy failed or that it was a success.

The Next 50 Years

Now let's turn to the next 50 years. "To predict is difficult," Einstein is supposed to have said, "especially the future." We cannot predict the demographic future in China with absolute certainty. We can however say a surprising large number of useful things with a surprising degree of assurance.

The place to begin is the age distribution. Here is the age distribution of the whole country from the 1990 census.

Figure 4: China 1990 Age Distribution Compared with Stationary

Each point shows the number of persons (vertical axis) at a given single year of age (horizontal axis).

This age distribution tells a lot bout the future if we know how to read it. Here's how to read it. Imagine how this picture will change in one year's time. Each point will move to the right one year, because we all get one year older every year. Norman Ryder once called this "the most important fact in demography." The point for age zero reflects the number of births to the population during the year.

Some of us do die, however, and because of this each point moves not only to the right, but also down. The downward movement is very slight at younger ages, when mortality risks are low, and higher at older ages, when risks are high. Given current mortality levels in China, there isn't much downward movement until around retirement age.

This takes a bit of hard thinking if its the first time you've seen it, but the effort is worthwhile, for we immediately get an important conclusion about China's future. The typical number persons under age 35 is 22-23 million. Over the next 35 years, all of these persons will get 35 years older, i.e., will be aged 35-69, and few of them will die. Now look at the numbers of persons aged 35-69 in 1990. The typical numbers here are much smaller. Conclusion? There will be a dramatic increase in the number of persons aged 35+ in coming decades.

Doing the demographic arithmetic, we calculate a growth of 400 million persons aged 35 and over over the next 40 years, or about 100 million per decades. This is a lot of people. In purely people-counting terms, it's more than another Europe.

Moreover, this growth of population is about as close to certain as we ever get in such matters because the people in question have already been born. The only thing that could prevent this population increase would be catastrophically high mortality; hundreds of millions of excess deaths in coming decades.

Conclusion 1 400 million increase in population 35+ over next 40 years

What about total population growth in coming decades? This is less certain because it depends on future fertility levels, but a plausible scenario is that numbers of births entering the population in future decades will be similar to numbers in past decades. If this is the case, the number of persons under age 35 won't change much, and total population growth will be the same as growth of the population over age 35, about 400 million over the next 40 years.

Conclusion 2 400 million total increase over the next 40 years.

We can say more about these 400 million new Chinese. Because the one-child family policy was quite fully implemented in China's cities, the urban TFR is about one child per woman, and has been at this level since the mid-1980s. The growth pattern in the cities is thus very different from that in the country as a whole.

We can see the impact by looking at the age distribution of China's cities. Here it is in Figure 5.

Figure 5: Age Distribution of China Cities, 1990

The number of persons under age 20 is about two thirds the number aged 20-39. We can see from the shape of the age distribution that there will be a large increase of older persons in the cities, as in the county as a whole. In the cities, however, this increase will be approximately canceled out by declines in the number of younger persons. With respect to natural increase (migration is another matter), cities won't grow much at all over the next 40 years. From which we conclude that the 400 million new persons will be born into what are, now, at least, rural areas.

Conclusion 3 The 400 million increase will be concentrated in rural areas

We can say more about where these 400 million new Chinese will be by looking at a map of population densities, shown here.

Figure 6: China Population Density
Figure 5 PDF

This isn't a cartographically rigorous display, but it is a useful picture and will serve our purposes here. The essential conclusion is immediate. There is an enormous concentration of Chinese population to the right in the picture. This is the area in which traditional Chinese agriculture can be practiced. The (almost) empty spaces to the north and west cannot support this agriculture and are sparsely settled.

Most of the future population increase will be where the population is concentrated already. The sparsely settled areas to the north and west may experience very large increases relative to their own (very small) population, but they simply can't absorb the very large numbers of persons that will come on the scene in future decades.

Conclusion 4 The increase will be concentrated in agrarian China

The final conclusion is important, but a bit more complex. To explain it, we 'zoom in' on agrarian China and look closely at the density patterns. Here's the picture.

Figure 7: China Population Density
Figure 7

From the work of G. William Skinner we know that agrarian China can be divided into nine "macroregions." These correspond fairly closely to large scale drainage basins. Each macroregion displays a "core-periphery" structure, with a dense inner core and surrounding, roughly concentric rings, with density declining as we move from inner core to the outermost periphery.

Most of the macroregions can be seen fairly clearly (one does have to look closely) on the density map. The three macroregions of the Yangze River, for example, are here (Upper Yangze), here (Middle Yangze) and here (lower Yangze). The densest (darkest) areas are the core, with lightening indicating movement toward the periphery.

The inner cores of macroregions are already so dense that they will be strained by additional population. The far peripheral regions are topographically unsuited to dense populations and are unlikely to to show much growth overall. By elimination, then, we expect most future population increase to occur in the intermediate ranges of the core-periphery structure of each macroregion.

So. We actually know quite about about the shape of China's demographic future. We can't predict with certainty and precision, but neither do we face a blank wall of uncertainty. We can know more than enough about the demographic future to worry about its implications. Which is what we will do in the third and last part of the presentation.

China in World Perspective

Now let's turn to Part 3, China in world perspective. A world density plot is a useful starting point. Here it is .

Figure 8: World Population Density, 1995
Figure 8
Bigger GIF 37k | PDF | Source: The Global Demography Project

What is instantly apparent is how China and India-Pakistan-Bangladesh dominate the world population map. This is true today, and since the current demographic situation in these countries pretty much guarantees that they will add roughly a billion persons to world population over the next half century or so, it will be even more true in the future.

What does this mean for them? What does it mean for us? How these countries will absorb the additional population they are nearly certain to get and still prosper? Or will things go terribly wrong?

Let's bring the focus back to China. Most of the additional population will be born into (what are now) rural areas in which most of the labor force is engaged in agriculture. If there is anything we can be reasonably sure of, it is that these areas are already overpopulated. We can't be precise, but in this perspective we can manage without precision. China's agricultural productivity would probably improve by the removal of a few hundred million persons from the agricultural labor force. And this is under current conditions. In every developed country agricultural workers as a percentage of the labor force drops with development from way over half to under ten percent. We can expect a similar decline in demand for China.

So the population growth is not needed in, and could not easily (or at all) be accommodated in agriculture. By elimination, they need to be absorbed into a modern economy. What China needs over the next half century is the development of a modern economy on the scale of the US or Europe to accommodate future population growth.

Some elements of the "Asian Economic Miracle" are in place for China. One key aspect, however, is not: export-led growth. Exports can't play the same role in China that they have in Taiwan, Korea, or even Japan, because China is so big.

How will China cope? One possibility is that "export-led growth" will develop internally; that the poorer parts of the country will develop by "exporting" to the more developed parts of the country. This could have important implications for trade with China. Some opportunities will lie in the expanding consumer market, but this will remain much smaller than the whole country. Other opportunities may lie in facilitating development of the less developed parts of the country by helping them "export" to the more developed parts of the country.


We've covered a lot of ground in a short time. China's fertility decline and the role of it's anti-natalist policies; some key aspects of the demographic future in China; and the significance of these demographic changes for China and the rest of the world. This is the end of my presentation. Obviously much has been left unsaid, but I believe we have some time for questions. Thank you for your attention.