Good morning. I am indeed very happy and honored to be with you here today. As was said, I arrived late last night. Before that I was in Ethiopia working on African issues of population, AIDS, and sustainable development. When you travel from one part of the world to another, there are indeed very significant differences that you all are aware of, but once you have gone through different stages within one week, you become aware of them again and again.

We have heard already about the historical dimension of population growth, so I will speak briefly on this historical dimension. Next, I will give a description of where we stand today—what are the reasons for some of the trends that we see. Then I will move on to the future: what can we assume about the future trends of the components of population change, birth rates, death rates, and migration. Finally, I will discuss one of the dimensions of population change that I increasingly believe has possibly the largest impact on the world of tomorrow: namely, the educational composition of the population. Here we do not just look at the numbers of people but the skills, the training that people have—so to say, the human capital.

For millennia, world population has been growing at a very slow rate. It was not until 1800 that the population reached its first billion (one billion is roughly the current population of Europe and North America together). It took more than a century, until 1930, to add the second billion, but it took only 30 years, until 1960, to reach the third billion. That is the figure I still remember from school. I think everybody remembers a certain number about world population from the time he or she went to school, and since then, we have had a further increase in the speed of growth. The fourth billion was added between 1960 and 1975, taking only 15 years. The five billion mark was reached in 1987, and the six billion mark late last year—a 12-year period for adding one billion.

There are, of course, different estimates. Some people claim that six billion was reached in the summer of 2000; others claim it was reached in October. I think you can appreciate that counting people is more difficult than counting votes. This is because people move around all the time. Just imagine trying to catch all of the butterfly ballots flying around in a big room—you have no idea whether you have already counted one of them. For instance, if you think of street children in Calcutta, how are you going to make sure you counted all of them once and none twice?

There is a certain margin of uncertainty in population figures. Fortunately, it is not too big—a few percentages up and down. There is also a difference among countries. I believe that in the more developed countries with accurate registration systems, the count is more reliable. But take, for instance, South Africa, where there was a recent census. Some people estimated 37 million people; others came up with 44 million. That is quite a difference in the estimates about the current population.

On a global level, despite the apparent acceleration in the time needed to add an additional billion, both the annual growth rate of world population and the number of persons added each year have passed their peaks and are expected to continue to fall. The growth rate peaked at 2.1 percent per year in the late 1960s and has since fallen to about 1.5 percent. The annual absolute increment of population peaked at about 87 million per year in the late 1980s. It is now around 80 million people per year.

This does not mean, however, that little additional population growth is to be expected. We will talk about this seemingly contradictory trend next. On the one hand, we are experiencing the most rapid population growth in history; yet on the other, we expect this growth to slow down and eventually stabilize or even turn negative. It may well be that we will experience a global population decline toward the end of this century.

When we compare maps of the world that plot population densities of countries in 1960 (a world of three billion) and 2000 (a world of six billion), we see stunning differences. In 2000, one can still see the low density of Canada and Russia, but Latin America has a significantly higher population density, between 20 and 50 people per square kilometer. Note the very high-density areas of China, India, and Southeast Asia. The density has also increased in Europe and Africa. Only the Sahara and other desert areas continue to show very low population densities. Comparing such maps shows that the doubling of the population has largely been concentrated in what we call the developing countries, or the south, although there was some growth in the north.

What is behind this trend? Is there something we can simply extrapolate? That is what many people would do when there is such a strong increase and the belief that this condition will continue. Indeed, we have good reasons to assume it will not continue.

There has been much talk about carrying capacity—that it cannot continue indefinitely because there simply is not enough food, enough living space for everyone to have a decent life for an ever-increasing number. But there are also internal reasons, social reasons, why we believe that world population is going to stabilize.

There is something we call the theory or the paradigm of demographic transition. Figure 1 shows these two curves for my home country, Austria, for a long time series. We start in 1820. The thick line shows the birth rate—the number of births to 1,000 of the population. The ups and downs are typical for what we call pre-modern societies before World War I, as is the case in this example.


Figure 1. Crude birth and crude death rates, Austria and Kenya (1819 to 1989).

Below that thick line is the death rate—deaths per 1,000 of the population. It shows some peaks, some epidemics and wars. On average, the death rate was a little below the birth rate, which means that the population grew, but it was a very moderate growth. Around 1870 there is a remarkable beginning of a downward trend in the death rate in all of Europe as well as in the United States. This was not due to modern medicine—1870 was well before antibiotics or efficient vaccinations were discovered—but rather to improved sanitary conditions. In many of the big European cities, sewage systems were built. People started to use soap, which is probably the single most important factor in bringing down the death rates, infectious diseases, and the relative changes in lifestyle together with a better nutritional status of the population, which also contributed to a better health status.

As seen in Figure 1, the birth rate even increased a little. This can be attributed to the improving health of women who now could have more children if they wanted, and indeed, they still had a high desire for large families. Fertility did not follow the downward trend until early in the twentieth century. Figure 1 shows a widening gap between the death rate and the birth rate, resulting in population growth.

But this growth was never more than, say, 1 percent or 10 per 1,000; that is the difference between the birth and death rates. As we know from European and American history, many of these surplus births moved to the New World. They could not find living space in Europe. This was a time of very strong immigration from the Old World to the New.

Continuing with Figure 1, World War I shows a peak in mortality and, of course, a very low level of fertility, followed by some recovery. We see what the German-speaking countries call “the Nazi baby boom.” During the Nazi time, some pronatalist policies caused the birth rate to jump. Next comes World War II and the postwar baby boom. This is not different in most countries of the industrialized world.

Figure 1 also shows an example of a rather typical developing country, Kenya. When the data starts in 1950, we see a very high birth rate of more than 50 per 1,000 people, which is much higher than what we have ever seen in Europe due to universal and early marriage. The death rate had already declined, owing to the introduction of modern medicine, antibiotics, malaria eradication, and so on to the developing countries after World War II.

For Kenya, Figure 1 shows a 3 to 4 percent increase of the total population per year. This means a doubling of the population in something like 20 years, which is a level that we have never seen in European history. As mentioned above, Europe had at most a 1 percent increase along with the possibility for out-migration that Kenya does not have.

One developing country, for which we have good data and where I have done some fieldwork, is the island of Mauritius. Here we see essentially the same phenomenon. Figure 2 shows the death rate from 1875 to be even more erratic than in Europe. There were some epidemics, malaria, Spanish Flu, and some ups and downs. As seen in Figure 2, the birth rate in Mauritius remained at roughly the same level. There was no population growth in the early part of the twentieth century. But then, within a few years after World War II, the death rate almost halved due to malaria eradication and antibiotics. At the same time, the birth rate jumped due to better health of women. During the time when Mauritius had an annual increase of 3 to 4 percent, Richard Titmuss and J. E. Meade, the famous British economist, went to the island to study what they thought was a textbook example of a country trapped in a vicious circle of poverty and rapid population growth. At the time, they were very pessimistic that Mauritius would not have a good future. But indeed, during the 1960s, Mauritius had a very steep decline in fertility, almost halving the number of children per woman within seven or eight years, to a level that today is as low as that of North America.


Figure 2. Birth and death rates in Mauritius, 1871–1991.

Before we move to the reasons for this remarkable trend that we can see in similar form in all countries of the world, let me quickly mention the global trends since 1950. Table 1 shows life expectancy at birth for the different continents. All parts of the world show an increase, with the strongest increase in Asia. At the beginning, Asia was close to Africa, but now Asia is closer to the developed countries.

Table 1. Regional population sizes, mean number of children (TFR) and life expectancies at birth (both sexes), 1950–2000 (Source of data: World Population Prospects. The 1998 Revision. Volume I: Comprehensive Tables. New York: United Nations, 1999).


Population Size


Life Expectancy at Birth










Northern Africa









Sub-Saharan Africa









Eastern Asia









South Central Asia









South Eastern Asia









Western Asia









Latin America & Caribbean









Northern America










Table 1 also shows a decline in fertility rates (the number of children per woman) for all parts of the world. Even in Africa, where fertility had been at a very high level—between six and seven on average—we see a recent significant decline. In Latin America and Asia, this remarkable decrease began in the 1970s.

What are the reasons behind this process called the demographic transition? It would take too long to describe all of the theories and empirical data behind this remarkable fertility decline. Instead, I would like to use the words of one of the most famous American demographers, Ansley Coale, who was the head of Princeton University’s Office of Population Studies. After a very extensive review of fertility declines, mostly in historical Europe but also in developing countries, he came to the conclusion that there are three preconditions for a sustainable fertility decline.

The first is a mental precondition. Coale says that fertility must be within the realm of conscious choice. You must be able to think rationally with intention about the number of children that you have, not just take them as God-given, as has been the case in many of the traditional societies. If you interpret this in terms of policies, it implies emphasis on education, toward more rational behavior. But one must be cautious. You cannot say that previous behavior was irrational; it was embedded in a social rationality. But here we talk about individual rationality; you actively think about the number of children that you want to have and do not take it as something naturally given to you or as suggested by traditional societal norms.

The second precondition that Ansley Coale mentioned is that smaller family size must be advantageous to you; there must be a reason why you want to have fewer children. On the policy side, this brings in the issue of costs and benefits. What benefits do children bring in terms of helping on the farm, etc., versus the cost? As we all know, in a modern society in an urbanized environment, children are much more of an economic cost than an economic benefit. There are, of course, noneconomic benefits, such as emotional benefits, which are the main reasons people still have children at all. It is unlikely that anyone would have children for purely economic reasons, so there needs to be some differentiation other than purely economic reasons. But these noneconomic, emotional benefits can usually be met by just one or two children. When someone has eight, nine, or ten children and begins to think about having fewer, these economic benefits play a role. If you want good schooling for your children, for example, you can afford to have only a smaller number. The size of the family has to be advantageous to you.

The third important precondition is that there must be an acceptable means for limiting your family size—reproductive health considerations, family planning services. What is considered an acceptable means depends on the culture.

It is important to acknowledge that these three preconditions go hand in hand. This is why in some cases—for example, in Kenya in the 1960s—there were strong family planning efforts that showed no effect whatsoever, simply because the other two preconditions were not met. Another example is Mauritius where, up to 1965, there was a rather high female literacy rate, so the first precondition was met. The second was also there: it was advantageous to have fewer children, but the people in Mauritius in the early 1960s did not yet have acceptable means to limit family size. A family planning program was implemented by the government, and even the Catholic Church in Mauritius was heavily involved in spreading the idea of voluntary family planning by so-called natural methods. Because the two other preconditions had been met, there was a very strong response to these family planning efforts.

This was a very brief attempt to give a concise picture of the preconditions for a lasting fertility decline. Despite tremendous variations in different parts of the world, the bottom line remains that these three preconditions must be met.

I would like to say a word on Africa because it has recently been in the headlines because of the AIDS epidemic—perhaps pandemic is the better word. There are indications that AIDS is not only killing many people but that it may also have a significant fertility-reducing effect. At this point, we cannot anticipate the longer term impacts of AIDS on the entire infrastructure and society of the countries affected. There are devastating consequences at all levels, but one of the consequences may be to enhance the decline of fertility in Africa.

Let’s move on. In 1974, there was a world population conference in Bucharest where the developing countries came up against the industrialized countries, who were pushing family planning. The tenor of the developing countries was, “We don’t need your family planning. We need development.” “Development is the best contraceptive,” was their rallying cry.

Figure 3 shows that economic development alone does not mean lower fertility. Here are time series for 21 countries, with the number of children (TFR) on the one axis and per capita income on the other. This is a nonrelationship. There are all kinds of patterns. Some countries, like Mauritius, have a strong fertility decline at a very low income. Others have very high increases in income without any change in fertility. The simplistic statement that an increase in a country’s income will automatically bring down fertility does not seem to hold.

Figure 3. Relationship between per capita income and fertility (TFR) in a sample of developing countries (1970–1990) and in Mauritius (1950–1990). (Source: Population-Development-Environment. Understanding Their Interactions in Mauritius, W. Lutz (ed.). Berlin, Springer-Verlag, 1994, p. 369).

Is there a better predictor of fertility? I would say yes; my favorite candidate is the female literacy rate. Look at Figure 4, at the same time series of the same 21 countries. It is amazing. One can see that up to a female literacy of about 50 percent, nothing changes. Fertility stays at a high level. After that, most of the countries seem to show a fertility decline. I should add here that a female literacy rate of 50 percent essentially means that most of the younger girls are educated while elderly women are still illiterate, because education tends to happen at a younger age. Once the more educated women come into reproductive age, fertility rates start to decline. At the end of this presentation, we will come back to the issue of education.


Figure 4. Relationship between female literacy rates and total fertility rates in a sample of developing countries with a total fertility rate above 6.0 in 1950 (1950–1990). (Source: Population-Development-Environment. Understanding Their Interactions in Mauritius, W. Lutz (ed.). Berlin, Springer-Verlag, 1994, p. 370).

There are also significant impacts of fertility and mortality on the age structure of the population. Figure 5a gives us the example of sub-Saharan Africa—a steep population pyramid with exponentially increasing young age groups. Figure 5b shows the opposite in the example of Western Europe, which does not resemble the form of a pyramid at all.

Figure 5. Age pyramids of (a) sub-Saharan Africa (top) and (b) Western Europe (bottom) in 2000.

To briefly explain the pyramid, age is on one axis, with women and men on both sides of the pyramid. These figures show how the world is divided today. There are very young populations in which more and more young people will enter the school system and later will enter the labor force looking for jobs. This is one of the reasons for the huge unemployment problems in developing countries, whereas in the north and in Western Europe, the sizes of the younger age groups are shrinking, causing unemployment to improve. There are, of course, many other consequences on the economy and on society resulting from the process of population aging, and we will hear more about this later.

Let’s quickly move on to the future. Demographers have an easier task than economists or meteorologists to project the future because we have to worry about only three factors that determine the future size and structure of the population: fertility or the birth rate, mortality or the death rate, and migration. Fertility, mortality, and migration are influenced by the physical, economic, social, cultural, and political contexts, and each of these is hard to forecast. Through the process of population dynamics, inputs in terms of fertility, mortality, and migration are then translated into certain population characteristics at a subsequent point in time: population size, population density, growth rate, age distribution, sex ratio, and regional distribution. All of these characteristics feed back to the social, economic, and natural environment. In population projections, we have to make assumptions about these three main determinants of population change.

How do we make assumptions? The best way to start is by making alternative assumptions—see what would happen to world population if we had a low path of fertility as compared to a high path. But this sort of sensitivity analysis is only of limited usefulness. It does not tell us what is likely to happen and what is unlikely. For assessing the likelihood of certain trends we need substantive arguments and their evaluation by experts. At the International Institute for Applied Systems Analysis (IIASA) in Austria in 1996, we produced a 500-page documentation of alternative views about the future paths in fertility, mortality, and migration for different parts of the world. We tried to ascertain what can be assumed today based on empirical evidence, knowledge about fertility intentions, likely improvements in life expectancy, and other possible future trends.

Figure 6 shows the population path from 1950 until about 1996, when we made the projections. For the future, it combines the most likely fertility and mortality paths. Fortunately, on a global level we don’t have to consider migration. As long as we don’t have any immigration from outer space, we can leave it out. However, this is not entirely true, because if a lot of people move from a high fertility continent to a low fertility continent and adopt the new low fertility level, it affects the world population size as well.

Figure 6 goes here

Figure 6. Unavoidable and possibly avoidable world population growth to 2050, based on 1996 IIASA projections. (Source: The Future Population of the World. What Can We Assume Today? W. Lutz (ed.). London, Earthscan, 1996, p. 432).

At the bottom of Figure 6 we find something that we might call the momentum of population growth plus the inertia of fertility—fertility cannot change too rapidly. Alternatively, we might call this the unavoidable population growth because it would be unrealistic to assume that tomorrow in every country of the world, the number of children per woman would drop, let’s say, to 2.1, which is approximately the level of fertility it takes to replace one generation. Hence, there will be some unavoidable population growth over the coming decades.

You will see that if one varies only the fertility rates, thus accounting for uncertainty in future trends as is done, for instance, in the United Nations’ forecasts or in some of the projections of the other agencies, it does not capture the whole picture, because mortality uncertainty is quite significant. For instance, if you compare the two bottom lines in Figure 6, they combine identical low fertility paths, in one case combined with low mortality, i.e., improvement in life expectancy (which means fewer people dying), and in the other case with high mortality. When more people survive, the population is larger. If you have a higher mortality rate, which can be due to AIDS or other reasons, then population size peaks and declines thereafter.

The possibilities are many, and one cannot really say which path the world population will take. For this reason, we developed a model that we call probabilistic population projections, where we try to attach probabilities to alternative trends. Our findings—that a doubling of world population is unlikely—were published in Nature magazine in 1997.[1]

Using an improved method of probabilistic forecasting in a new projection recently published in Nature[2], we showed that there is around an 85 percent chance that the world’s population will stop growing before the end of the century. There is a 60 percent probability that the world’s population will not exceed 10 billion people before 2100, and around a 15 percent probability that the world’s population at the end of the century will be lower than it is today. For different regions, the date and size of the peak population will vary considerably.

The inner area in Figure 7 gives the 95 percent uncertainty interval in 1996. We assume that about 95 out of 100 cases fall into this range. About 60 percent of all future trends fall in the lighter shaded area and 20 percent in the inner dark area.

Figure 7. Forecasted distributions of world population sizes (fractiles). For comparison, the United Nations medium scenario (white line) and 95 percent interval as given by the NRC[3] on the basis of an ex post error analysis (vertical line in 2050) are also given. Reprinted by permission from Nature 412 (2 August 2001): 544, copyright 2001 Macmillan Publishers Ltd.

Figure 7 shows the distribution of simulated world population sizes over time. The median value of our projections reaches a peak around 2070 at 9.0 billion people and then slowly decreases. In 2100, the median value of our projections is 8.4 billion people with the 80 percent prediction interval bounded by 5.6 and 12.1 billion. The medium scenario of the most recent UN long-range projection[4] is inserted in Figure 7 as a white line. It is almost identical to our median until the middle of the century but is higher thereafter due to the UN assumption of universal replace-level fertility, i.e., two surviving children per woman.

A stabilized or shrinking population will be a much older population. At the global level, the proportion above age 60 is likely to increase from its current level of 10 percent to around 22 percent in 2050. This is higher than it is in Western Europe today. By the end of the century, it will increase to around 35 percent, and extensive population aging will be experienced by all world regions. The most extreme levels will be reached in the Pacific OECD (mostly Japan), where half of the population is likely to be age 60 and above by the end of the century, with the 80 percent uncertainty interval reaching from 35 to 61 percent. Even sub-Saharan Africa in 100 years is likely to be more aged than Europe today. Compared to the medium scenario of the UN long-range projections of the proportion 60 and above, the trend of our median is almost identical up to 2050 but shows significantly stronger aging thereafter. This confirms recent criticism that conventional projections tend to underestimate aging.[5], [6] The extent and regional differences in the speed of population aging—the inevitable consequence of population stabilization and decline—will pose major social and economic challenges.

It needs to be recognized that population numbers are only one aspect of human impact and that in some of the world’s most vulnerable regions, significant population growth is still to be expected. Nevertheless, the prospect of an end to world population growth is welcome news for efforts toward sustainable development.

In conclusion, I would like to say a few words about the educational composition of the population. I believe that when we consider the impacts of the population on the environment and the controversy associated therein, education really is what some people call a win-win strategy—something that is good for the future population as well as good for the environment. It really may be the best solution out of some of the vicious circles that we see in the world today.

At IIASA we have developed a demographic method that we call multistate population projection with which we not only project the population of one country and see how it will develop in the future, but we break it down into different subcategories, which are the educational categories in Figure 8. We see South Asia, essentially India, where the illiterate population without any education is shown in white [yellow in color]; the population with at least one year of primary schooling is black [red in color]; those with some secondary education are lighter [blue in color]; and those with some tertiary education are gray [green in color]. We also see that India has a large gender gap; females have a very high proportion—let’s say, 25 to 29 percent. Half of the females in their 30s have never had a single year of schooling, whereas schooling for males is a little better. This is a very poorly educated society, even today.

Figure 8. Age and education pyramids for South Asia in 2000 and in 2030 according to “American” scenario.

Now we calculate several scenarios. In one that we call the “American scenario,” we assume that South Asia will slowly move toward North American school enrollment rates, which means very high secondary and even tertiary education (see Figure 8).

At the younger ages, this makes a great difference. In a constant scenario, the gender gap remains large, and the higher proportion of people with very low education is perpetuated in India to the year 2030. But even if one makes a tremendous effort to increase the Indian school system to American enrollment ratios, it affects only the younger generation. In this case the gender gap narrows, and there is much higher secondary and tertiary education at the younger age groups.

Because you are educating only children and possibly young adults, and not the older people above age 30 or 40, it makes no difference for the skills of the working population in the short run. Education of the labor force is something that is very inert. If you invest in education today, it takes 20 or 30 years to translate into a better education of the labor force that will have an impact on productivity and all the other beneficial economic consequences, but the cost of education needs to be spent now. This is why one needs to have a long-time horizon for societal investments.

An educated society is likely to be more productive and better off. It can also more easily cope with and adapt to climate change conditions and all kinds of environmental challenges that will come up in the future.

Thank you.

[1] Lutz, Wolfgang, Warren Sanderson, and Sergei Scherbov. 1997. Doubling of world population unlikely. Nature 387: 803–805.

[2] Lutz, Wolfgang, Warren Sanderson, and Sergei Scherbov. 2001. The end of world population growth. Nature 412: 543546.

[3] National Research Council. 2000. Beyond Six Billion: Forecasting the World’s Population. Panel on Population Projections. John Bongaarts and Rodolfo A. Bulatao (eds.). Committee on Population, Commission on Behavioral and Social Sciences and Education. Washington, D.C.: National Academy Press.

[4] United Nations. 1999. Long-Range World Population Projections: Based on the 1998 Revision. New York: United Nations, ESA/P/WP.153.

[5] Tuljapurkar, S., N. Li, and C. Boe. 2000. A universal pattern of mortality decline in the G7 countries. Nature 405: 789–792.

[6] Vaupel, J.W. and H. Lundström. 1996. The future of mortality at older ages in developed countries. Pages 278–295 in W. Lutz (ed.), The Future Population of the World. What Can We Assume Today? Revised Edition. London: Earthscan.