6 billion - and counting

Posted: 7 September 2000

Author: Carl Haub

Author Info: Carl Haub is Senior Demographer at the Washington-based Population Reference Bureau.

Can the world sustain a population of 8-12 billion in the current century? Maybe, says Carl Haub, but there is no certainty that population growth can be kept within these bounds.

Without question, one of the 21st century's vital issues is that of population growth. It will see the playing out of humanity's great expansion which in the last century has witnessed the most far-reaching demographic change in the history of human species.

In 1900, a relatively scant 1.5 billion people inhabited the Earth, only one billion of those in the developing countries of Asia, Africa and Latin America. By the end of the century, we had passed six billion and the population of the developing countries itself was nearing the five billion threshold.

Just as the population explosion of the 20th century was a result of the mortality revolution, the final resolution of that growth must come from a fertility revolution. The first took place with unprecedented speed and so must the second. Why is this so?

Fertility revolutionSimply put, life expectancy, which took centuries to rise in the industrialised countries, did so much more swiftly in the developing countries. So swiftly, that their social preference for larger families did not have time to change. Mortality decline in the developed countries took place during a progressive shift from predominately rural societies to urban ones. But, in many of today's developing countries, rapid declines in death rates were not at first accompanied by declines in birth rates. As a result, population growth took off at a pace never before seen.

Crowded street in Dhaka, Bangladesh,a country where fertility has fallen fast.© Shehzad Noorani/Still Pictures

The story may be familiar, but it is worth emphasising - what happened in the developing countries in this century did not simply repeat what had happened in Europe and North America. It happened much more quickly. That, of course, is why we had six billion at the 20th century's end, up from less than two billion in 1900.

Rising life expectancy and falling infant mortality produced growth rates that could climb to three per cent or higher when combined with historically high birth rates. This is the situation many developing countries found themselves in during the early post-war years. If the situation persisted, countries could easily find their numbers doubling every 20 years. Nearly vertical growth curves were quite possible.

There is, of course, only one desirable way to slow such growth: through a decline in the birth rate. But many developing countries faced a truly difficult situation - not only would birth rates have to be reduced when the socio-economic situation was unfavourable, they would have to do so much more quickly than the wealthy countries ever did.

Thirty years ago, it was difficult to foresee what might take place. Not only was rapid population growth a new phenomenon, effective means of contraception were only just becoming available.

But over the course of the past 30 years, birth rates have come down. Few, however, have declined to the precise level needed to actually bring growth to zero, about two children per woman, or "replacement level" fertility.

Fertility decline is, in fact, the foundation of all projections that we so often cite. That is something most users know, but the importance of the nature of that decline is not as obvious, or as well understood as it should be.

Is a sustainable population possible? There have, of course, been many estimates and definitions of a sustainable global population. But for argument's sake let's take the median of the low and high estimates cited by Joel Cohen in his survey of the subject, How Many People Can the Earth Support?1. These range from 7.7 to 12 billion.

On this basis one can only conclude that sustainability remains in jeopardy.

Global picture In order to understand how that can be, we must take a close look at global population projections and how they should be interpreted. Projections are often used to make a writer's point, without full understanding of what they actually mean. Unquestionably, the United Nations comprehensive series of world population projections are the most widely referenced. For that reason, we will take a closer look at these.

The United Nations Population Division publishes its projections every two years2. The very fact that it publishes a broad spectrum of possibilities should tell the user something. If future world population size could truly be known, only one would be required. The five main series of projections vary considerably. In the long term (to 2150), they project an astounding range of possibility - from under 4 billion population to 27 billion. Why such a great variation?

Most users have a strong tendency to approach projections backwards. The question typically asked is: what size population has the UN projected for 2050? for 2150? Is it more or less than their previous efforts? But those are not the questions we should be asking. When using projections, our first inquiry should be: what did the UN assume in making each projection? Why do they differ?

The answer is simple. The projections diverge for just one reason: fertility. At the global level, this is the only source of variation. Since the range of the results falls well outside the bounds of the population size we have proposed as "sustainable" for the purposes of this article, we are duty bound to pay strict attention to the assumptions the UN makes regarding fertility.

Since the UN publishes five projections, an odd number, there is necessarily a "middle" series. Often, that makes the task of citing the results far too easy for the user. Just use the middle series. The other four series are simply regarded as outliers, or even wild improbabilities, and are rarely considered.

The UN middle series projection assumes that all world regions will converge on the same fertility level, a total fertility rate (TFR) of 2.1 children per woman by 2050-55. This very fact alone should signal that projections often contain highly stylised assumptions, used more for purposes of illustration than for prediction. In the case of the middle series, the illustrative future fertility level is, of course, exact replacement level fertility. Each couple simply replaces itself, two for two.

This assumption has the added advantage of resulting in an eventual stationary population size, one that is neither increasing nor decreasing at some point in the distant future. Should that assumption be exactly borne out by events, world population will rise to 9.4 billion by 2050 and ultimately stabilise at about 11 billion.

Recognising the rather pat assumptions of the middle series, we are led to investigate the UN's other variants, ranging from the high to the low. The UN itself puts it this way: "The range of potential demographic outcomes underscores the difficulty in focusing on any particular scenario and also highlights the critical importance of current policies and actions for the long-range future of world population". In other words, seemingly small differences in what one assumes in a projection will make very large differences in the results.

Falling fertility What, for example, would be the result if fertility did fall in all world regions along the general lines of the medium series, but did not come exactly to 2.1, but, say, to 10 per cent above or below that value? Then, we would have a population of 18.3 billion if the global TFR were 2.3 or only 6.4 billion with a TFR of 1.9. These two projections alone show how very sensitive future world population size is to small adjustments in assumed fertility. Here, the projections are telling us that, if we postulated a future world TFR of 2.1 and were only "wrong" by 10 per cent, we would have a range in world population of about 12 billion from the high to low.

What if the ultimate TFR varies even more from 2.1 than that? The answer to this question is contained in the UN's High and Low projections. These two assume a difference of half a child from 2.1, so that the High projection assumes a future world TFR of 2.6 and the Low, 1.6. Then, future world population size in 2150 varies from about 27 billion (and still growing) down to 3.6 billion (and declining).

Such a wide variety of possibilities often presents too confusing an array of potential futures. But that is the central point of this article. The future course of the birth rate will play a final, unknown role in determining an ultimate size of world population or its tendency to grow continuously or to enter a period of decline.

The good news is that birth rates are expected to decline in developing countries and they already have in many, to a greater or lesser degree. The rationale for this expectation is somewhat vague, but generally is based upon the fact that birth rates have historically declined as economies develop and as family planning becomes more widely available. But it should be obvious that the expectation of decline itself is not the point. The real questions are: when? how fast? to what level? Population giant India, which recently became the world's second "population billionaire" offers an eloquent example of just how important future fertility trends are.

Children of ChinaGangfeng Wang/Panos PicturesIndian choices The UN medium projection assumes that India will achieve replacement level fertility in 2010. This results in a population size for India of 1.5 billion in 2050 and, ultimately 1.7 billion. In short, India will add a number equal to its entire population in 1980. In comparison, the US Census Bureau's single projection assumes replacement fertility would be reached in 2035. That results in a population of 1.7 billion in 2050 (as far as the Bureau's projections go) and, a likely size closer to two billion later on. But, both of those projections cited assume that replacement fertility will, in fact, be reached. What if Indian couples prefer something more than 2.1 children, as has already happened elsewhere? What if they prefer one-half child more? Will that make a difference worth bothering about? The UN high projection provides the answer.

If Indian couples decide on 2.6 children on average, then India's population will rise and pass 4.7 billion, the population of the entire Earth in 1983. Here we can see why one should never accept the results of any single projection as a given. The simple mathematics of population dynamics is quite unforgiving when it comes to small differences in projection assumptions.

Do past trends offer any clue as to what we might expect? Here, we see that different countries provide, in fact, different clues. Among the developing countries, we have seen uninterrupted decline to TFRs of two children per woman or less (Cuba, South Korea, Thailand, Taiwan). We have seen TFR declines that greatly slow or stall above two children per woman (Argentina, Costa Rica, Jamaica, Colombia). And, finally, we have seen countries with little or no decline (Madagascar, Nigeria, Iraq, Pakistan).

In Europe, fertility declines have fallen to historically low levels to the point where women now average but 1.5 children each and population decline has become a reality. Can we expect developing countries to emulate the European experience? The United States has not, holding firm at two children per woman. Can we seriously expect Niger to emulate Germany's very low fertility given the vast social, cultural and economic differences between those two countries? Perhaps that may happen in the distant future, but medium fertility projections assume that replacement level fertility will be achieved in Africa only 40 years from now.

The fact that these questions cannot now be answered does not obviate their importance. It is on these questions that the final outcome of humankind's sudden numerical outburst depends. For now, we can only observe the progress that has been made, monitor current trends and carefully assess what all variants of population projections are telling us so that we can at least prepare. Sustainable population? Demography tells us that it is in our grasp, but can so easily slip away.

See The Numbers Game

In its last biennial projections in October 1998, the UN showed a medium projection of world population in 2050 about 5 per cent smaller than the 1996 long-range projections cited here. Sadly, most of this difference was due to increased levels of AIDS in Africa - not fertility.

1 Cohen, Joel. How Many People Can the Earth Support?, New York, W. W. Norton and Company, 1995.
2 United Nations Population Division.
World Population Prospects: The 1996 Revision, United Nations, New York, 1998. The UN also publishes long-range projections on a periodic basis. Those cited here are from the same source, World Population Projections to 2150.