E6-147-22
E6-147-22
E6-147-22
Hania Zlotnik
Population Division, Department of Economic and Social Affairs, United Nations
Contents
1. Introduction
2. The United Nations population projections: History and relevance
S
TE S
3. Methods and assumptions to project national populations
R
AP LS
3.1. The Projection of Fertility
3.2. The Projection of Mortality
3.3. The Projection of International Migration
4. Past and future population trends
C EO
4.1. Trends and Prospects for Fertility Change
4.2. Trends and Prospects for Mortality Change
4.3. Population Growth and Distribution
4.4. Population Ageing
E –
5. Conclusion
H
Glossary
PL O
Bibliography
Biographical Sketch
M SC
Summary
The Population Division of the Department of Economic and Social Affairs of the
SA NE
United Nations Secretariat has been producing population projections for the world and
the countries that constitute it since the 1950s. This chapter focuses on the work of the
Population Division in producing estimates and projections of the world population and
on the implications of such work for our understanding of the relevance of population
U
On the basis of the results of the 2006 Revision, this chapter discusses population trends
and prospects until 2050 in terms of the expected changes in fertility and mortality; the
resulting changes in population growth and size, and trends in population ageing. The
focus is on trends and prospects at the world level, the major development groups and
the major areas of the world. (Note: The temporally relative statements in this chapter
with such terms as ‘the past/last’, ‘the next’, have to be understood with reference to the
date of this writing, that is, the year 2008)
1. Introduction
S
TE S
Thus, in order to project a population, it is necessary to know the initial distribution of
R
AP LS
the population by age and sex and then have a way of finding the number of births, the
number of deaths by age and sex, and the net number of migrants by age and sex that
the population will experience over a period following the initial observation of the
C EO
population. Population specialists have developed models and techniques that allow the
projection of populations, often over lengthy periods. In the United Nations, the
Population Division of the Department of Economic and Social Affairs has been
producing population projections for the world and the countries that constitute it since
the 1950s. This chapter focuses on the work of the Population Division in producing
E –
estimates and projections of the world population and on the implications of such work
H
for our understanding of the relevance of population dynamics for development.
PL O
At present, in addition to the United Nations Population Division, the main institution
M SC
producing, on a regular basis, population projections for each country in the world is the
US Census Bureau. In 1985, the US Census Bureau released its first comprehensive set
of estimates and projections, covering more than 200 countries and areas. Since then,
SA NE
the US Census Bureau has routinely updated the estimates and projections for particular
countries as new data become available. The population estimates and projections
prepared by the US Census Bureau are based on methodology similar to that used by the
United Nations Population Division. A full description of the US Census Bureau
U
The World Bank disseminates population projection data and produces country
projections as needed by its country programs. The site where the data can be accessed
(http://go.worldbank.org/H4UN4D5KI0) provides minimal documentation on data
sources and virtually no explanation of the methodology or assumptions underlying the
projections.
This paper focuses on the population projections produced by the United Nations
Population Division because, as the paper argues, they have been influential in shaping
government views on population dynamics and in guiding action to modify those
dynamics. For a full assessment of the United Nations population projections in
comparison to those produced by the other institutions see Bongaarts and Bulatao
The most recent set of population estimates and projections for the world, its major
areas and regions, and all countries or areas is the 2006 Revision of World Population
Prospects (United Nations, 2007). The 2006 Revision is the twentieth set of population
estimates and projections prepared by the Population Division since it was established
in 1946.
The first set of United Nations projections was published in 1951 and it provided a
“long-term view” of future population growth by projecting the population for eleven
groups of countries from 1950 to 1980. The second set appeared in the proceedings of
S
TE S
the first conference on population held under the auspices of the United Nations, the
1954 World Population Conference held in Rome, Italy. In preparing both sets of
R
AP LS
population projections, analysts at the Population Division were guided by the view
that: “Under the conditions which exist at present in regions like the Near East,
mortality can be reduced greatly without any radical social changes; certain relatively
C EO
simple measures of economic development and sanitary reform can lead to a stable
death rate much below the present level.” (United Nations, 1951, p. 4). Consequently,
both the 1951 and 1954 Revisions projected an increase in the population growth rates
of Asia and Latin America and the Caribbean, consistent with a reduction of death rates
in those regions, though they were not equally optimistic about Africa.
E –
H
The next revision, completed in 1957 (United Nations, 1958), was the first to present
PL O
population projections based on the component method, that is, prepared by making
independent assumptions about the future path of fertility (which determines the number
M SC
of births) and mortality (which determines the number of deaths). Overall, the results of
the 1957 Revision turned out to be closer to actual trends than those of previous
Revisions. Thus, its medium variant projected a world population of 4.2 billion in 1980,
SA NE
fairly close to the 4.5 billion estimated today. In addition, the results of the 1957
Revision made plain that the projected reductions in mortality would produce an
acceleration of population growth in developing regions: it projected a population
increase of 1.5 billion in developing countries as a whole between 1950 and 1980,
U
In the 1960s, the 1963 and the 1968 Revisions extended the projection period to 2000
and projected reductions in fertility based on the view that “a […] calamity would
eventually ensue, at least in the remote future, if fertility decline were to be postponed
indefinitely while mortality decline continued” (United Nations, 1966, p. 4). The report
of the 1963 Revision noted that the high growth rates caused by declining mortality
could not be sustained forever and, consequently, fertility would have to decline.
Therefore, both Revisions projected declining fertility in most developing regions even
if the onset of fertility reductions were projected to occur relatively late in some regions
(mainly in sub-Saharan Africa) and the projected pace of decline was, in retrospect,
modest for some countries. Nevertheless, the medium variant of the 1963 Revision
projected with considerable accuracy that the 2000 world population would be 6.1
billion, while that of the 1968 Revision overshot the mark by projecting a population of
6.5 billion in 2000. Both Revisions noted that, were fertility to remain constant at 1960s
levels, the 2000 population would surpass 7 billion persons.
The results of these early Revisions were influential in raising awareness about the
implications of the demographic transition, the process whereby a population passes
from a regime of high mortality and high fertility to one of low mortality and low
fertility, with mortality declining first and fertility decline following. The acceleration
of population growth that the population projections documented set the stage for major
developments in the policy arena. In December 1967, at the United Nations, the
Secretary-General received the World Leaders Declaration on Population, a document
that signaled the political acceptability of governmental policies to influence population
growth. The Declaration, signed by 30 heads of State or Government, noted that the
world population had taken all of recorded time to reach a billion at about 1850 but that
S
TE S
the second billion had been added in just 100 years and it had taken only thirty years to
add the third. It stated that “At today’s rate of increase, there will be four billion people
R
AP LS
by 1975 and nearly seven billion by the year 2000.” Convinced that rapid population
growth seriously hampered efforts to raise living standards, to further education, to
improve health and sanitation, and to provide better housing and transportation, the
C EO
signatories expressed the belief that the “population problem must be recognized as a
principal element of long-range national planning if governments are to achieve their
economic goals and fulfill the aspirations of their people”.
This Declaration and the report on “World Population: A Challenge to the United
E –
Nations and its System of Agencies” (Berelson, 1978), were important elements leading
H
to the decision of the Secretary-General to transfer in 1969 the recently established trust
PL O
fund on population to the United Nations Development Programme where it grew into
the United Nations Fund for Population Activities (UNFPA), a key institution
M SC
supporting the family planning programs that would underpin the reduction of fertility
in many developing countries.
SA NE
The United Nations population projections also served to set the stage for the first inter-
governmental conference on population held at the global level, namely, the United
Nations World Population Conference, held in Bucharest, Romania, in 1974 (United
Nations, 1975). The World Population Plan of Action adopted by the Conference used
U
the results of the 1973 Revision to present recent trends and prospects in regard to
population goals and policies, providing guidance on how best to achieve moderate or
low population growth and setting goals for the reduction of mortality. During the
general debate at the Conference, participants referred to the accelerating population
growth that developing countries were experiencing. Expecting that mortality would
continue to decline, many speakers believed that fertility was also bound to decrease but
considered that the timing and tempo of this reduction would be contingent on advances
in social and economic development. The representatives of some developing countries
spoke about initiatives to develop policies and programs having a direct bearing on
fertility and many stated that the age distributions of their populations were not
beneficial for development, displaying high dependency ratios that would not decline
significantly in the medium-term. Representatives of developed countries noted that
their populations were beginning to age, that the number of old people was increasing
and that their populations were highly concentrated in urban areas with deleterious
effects on the environment (United Nations, 1975).
After the World Population Conference, the need to monitor the implementation of the
Plan of Action that emerge from it added relevance to the task of preparing
comprehensive and comparable estimates and projections of population. As the Group
of Experts on Population Projections convened by the Population Division in 1977
noted, the United Nations projections were useful because: (1) they were made for every
country and territory of the world; (2) they were internationally comparable; (3) they
filled the gap left by the fact that over two thirds of countries, having about three
quarters of the world population, had never prepared official population projections of
their own, and (4) the reports accompanying the projections presented the methods,
assumptions and the resulting demographic parameters in a comparable fashion (United
Nations, 1979).
S
TE S
In response to increased demand for timely population estimates and projections,
R
AP LS
revisions of World Population Prospects began to be issued biennially starting with the
1978 Revision. Furthermore, the projection horizon was expanded to 2025 starting with
the 1980 Revision and ending with the 1992 Revision, after which it was extended to
C EO
2050. In extending the projection horizon to 2025, fertility in developing countries was
assumed to decline until it eventually reached replacement level and stabilized at that
level if it was reached before 2020. High-fertility countries, however, did not reach
replacement level by 2025. The start of the decline in fertility was established from
existing evidence or projected on the basis of socio-economic and policy information
E –
for the countries concerned. Starting with the 1984 Revision, the fertility of countries
H
with below-replacement levels, most of which were in the developed world, was
PL O
maintained below replacement level for a short period and was then assumed to increase
slowly so as to reach replacement level by 2025. Since fertility in developed low-
M SC
fertility countries actually continued to drop, the medium variants of the revisions
prepared in the 1980s tended to overestimate fertility in the more developed regions.
SA NE
The 1992 Revision marked a turning point in the projection of mortality because it made
for the first time explicit allowance for the effect of the HIV/AIDS epidemic in the most
affected countries (15 at the time). Furthermore, its results set the stage for the
consideration of population issues at the International Conference on Population and
U
Development (ICPD) held in Cairo, Egypt, in 1994. Thus, the Programme of Action
resulting from the Conference cited the results of the 1992 Revision and used it as a
guideline to set goals regarding the reduction of mortality.
Americas (including Brazil and the United States), five in Asia (including China and
India), four in Europe (including the Russian Federation), and one in Oceania. In these
countries the epidemic was expected at best to dampen potential reductions in mortality
or, at worst, produce outright increases in mortality. Overall, the results of the revisions
modeling the impact of the HIV/AIDS epidemic have contributed to raise awareness of
the devastating impact of the disease. After the establishment of the Joint United
Nations Programme on HIV/AIDS (UNAIDS) in 1996, close collaboration between
UNAIDS and the Population Division has ensured that the methodology used to derive
estimates of previous trends in the epidemic is consistent with that used to project its
future path. Furthermore, as efforts to control the epidemic and provide effective
treatment for those infected bear fruit, the models used make allowance for the effect of
anti-retroviral treatment in prolonging life.
Given that longevity continues to increase in most countries of the world, it has been
S
TE S
necessary to expand existing mortality models so that they can be used in projecting all
populations over several decades. To that end, existing families of model life tables
R
AP LS
have been extended to cover ever wider ranges of life expectancies. For the revisions
issued after 1980 and before 1998, the highest life expectancy at birth in the model life
tables used for projection purposes was 75 years for males and 80 years for females. In
C EO
the 1998 Revision, the highest life expectancy was raised to 82.5 years for males and
87.5 years for females, and in the 2002 Revision, it was raised again to 92.5 years for
each sex. Although the highest life expectancy at birth is not reached by any country
during the projection period, it acts as the mathematical limit toward which the life
expectancy of each country tends. However, the highest life expectancy used for
E –
projection purposes is not meant to represent an actual limit for human longevity.
H
PL O
With respect to the projection of fertility, the 2002 Revision marked an important
turning point because it was the first to drop the assumption that fertility would stabilize
M SC
at replacement level (that is, at 2.1 children per woman) over the medium-term future.
Instead, it was assumed that fertility would converge to a level of 1.85 children per
woman, a value close to the average fertility of countries that had below replacement
SA NE
fertility during 1990-2000. When a country’s recent fertility was above 1.85 children
per woman, fertility was projected to decline until it reached that level, either within the
projection period, if the starting fertility level was moderate, or beyond the projection
period if the starting fertility level was high. When a country’s recent fertility was
U
below 1.85 children per woman, it was projected to increase slowly until it reached 1.85
children per women and it then remained constant until the end of the projection period.
This change in the limit to which fertility tended was prompted by the fact that, for
many populations, the transition to low fertility had led to prolonged periods of below
replacement fertility. However, there is considerable uncertainty about how long that
period will last, especially in countries where fertility has not yet reached very low
levels. The revised fertility assumptions underpinning recent Revisions are contributing
to raise awareness about the consequences of sustained low fertility in terms of
population ageing and reductions in population size.
The United Nations population projections are prepared by making assumptions about
future trends in fertility, mortality and international migration. In all cases, a method is
used to establish future levels of the relevant variable and another is used to obtain a
distribution of its effects by age. The previous section has described the major changes
made over time in key characteristics of the models used and has set those changes in a
historical perspective. This section will focus on the methods underlying the most recent
set of projections (the 2006 Revision). Population trends and prospects will be discussed
in the next section on the basis of the results of this most recent Revision.
Assumptions about future fertility are set differently for two major categories of
countries: high-fertility and medium-fertility countries, on the one hand, and low-
fertility countries on the other. All countries in the first group had a total fertility in
2000-2005 above replacement level (that is, higher than 2.1 children per woman). Low-
S
TE S
fertility countries are those with total fertility at or below replacement level in 2000-
2005.In the medium variant, fertility is assumed to converge eventually to 1.85 children
R
AP LS
per woman, although some high-fertility countries do not reach that level before 2050.
Fertility in high-fertility and medium-fertility countries is assumed to follow a path
derived from models of fertility decline established on the basis of the past experience
C EO
of all countries whose fertility decreased during 1950-2000. The models relate total
fertility during a period to its expected decline during the next period. Once total
fertility reaches 1.85 children per woman it remains at that level until 2050.
replacement level during the projection period and moves towards 1.85 children per
H
woman by 2045-2050. In countries with total fertility below 1.85 children per woman in
PL O
2000-2005, fertility is assumed to increase linearly at about 0.1 children per woman per
decade after a transition period of 5 or 10 years and it may not reach 1.85 children per
M SC
woman by 2050.Once the future path of total fertility is established for the medium
variant, that of the low variant is produced by letting total fertility be approximately half
a child below that of the medium variant and that of the high variant is produced by
SA NE
making its total fertility be half a child above the fertility of the medium variant.
To project the age pattern of fertility, interpolation is used to move from the most
recently observed or estimated proportionate age pattern of fertility to a model
U
proportionate age pattern of fertility at a future period. The model pattern is attained at
the end of the projection period or when the lowest level of total fertility is first reached
and, in this latter case, is kept constant thereafter. Model patterns were derived from the
experience of low-fertility countries and reflect the variety of patterns observed today in
those populations.
-
-
-
Bibliography
Berelson, B. (1978). In Memoriam: John D. Rockefeller 3rd, 1906-1978, Studies in Family Planning, vol.
9, No. 7 (July), pp. 178-179. [Note on the role that John D. Rockefeller 3rd played in engaging world
leaders and the United Nations to make acceptable governmental policies to influence population
growth].
Bongaarts, John and Rodolfo A. Bulatao, eds. (2000). Beyond Six Billion: Forecasting the World’s
Population, National Academy Press, Washington, D.C. [Compares and assesses the different sets of
comprehensive population projections available at the time].
Bongaarts, John (2008). Fertility transitions in developing countries: Progress or stagnation? Working
Paper No. 7. Population Council. [An examination of the pace of fertility change in developing countries
with multiple Demographic and Health Surveys].
O'Neill, Brian C., Deborah Balk, Melanie Brickman and Markos Ezra (2001). A guide to global
population projections, Demographic Research, vol. 4, No. 8 (June 13), pp. 203-288. [Drawing on new
thinking on uncertainty, the authors discuss who makes projections and how, and the key assumptions
S
TE S
upon which they are based; they compare methodology and recent results, providing a guide to sources of
demographic information]
R
AP LS
Shapiro, David and Tesfayi Gebreselassie (2008). Falling and stalling fertility in Sub-Saharan Africa.
Paper prepared for the IUSSP Seminar on Human Fertility in Africa: Trends in the Last Decade and
Prospects for Change to be held in Accra, Ghana from 16 to 18 September 2008. [The paper assesses the
current status of the fertility transition in Africa based on data from the Demographic and Health
C EO
Surveys].
UNAIDS Reference Group on Estimates, Modelling and Projections (2002). Improved methods and
assumptions for estimation of the HIV/AIDS epidemic and its impact: Recommendations of the UNAIDS
Reference Group on Estimates, Modelling and Projections. AIDS, vol. 16, pp. W1-W14. URL:
http://www.epidem.org. [Presents the model used to simulate the demographic impact of the HIV/AIDS
E –
epidemic].
H
United Nations (1951). The past and future growth of world population: A long-range view, Population
PL O
Bulletin, vol. 1 (December), pp.1-12. [Presents the results of the first set of global population projections
produced by the United Nations Population Division].
M SC
United Nations Population Division (1955). The past and future population of the world and its
continents. In Proceedings of the World Population Conference, 1954, vol. III, E/CONF.13/415, United
Nations publication, Sales No. 1955.XIII.8 (vol. III), pp. 265-281. [Presents the results of the second set
of global population projections produced by the United Nations Population Division].
SA NE
United Nations (1958). The Future Growth of World Population. United Nations publication, Sales No.
58.XIII.2. [Discusses the results of the third set of global population projections prepared by the United
Nations Population Division].
U
United Nations (1966). World Population Prospects as Assessed in 1963. United Nations publication,
Sales No. 66.XIII.2. [Presents the results of the 1963 revision of global population projections prepared
by the United Nations Population Division].
United Nations (1973). World Population Prospects as Assessed in 1968. United Nations publication,
Sales No. 72.XIII.4. [Presents the results of the 1968 revision of global population projections prepared
by the United Nations Population Division].
United Nations (1975). Report of the United Nations World Population Conference, 1974. Bucharest 19-
30 August 1974. E/CONF.60/19. [Presents the outcome documents of the first major United Nations
intergovernmental conference on population].
United Nations (1977). World Population Prospects as Assessed in 1973. United Nations publication,
E.76.XIII.4. [Presents the results of the 1973 revision of global population projections prepared by the
United Nations Population Division].
United Nations (1979). Prospects of Population: Methodology and Assumptions. Papers of the Ad Hoc
Group of Experts on Demographic Projections. United Nations publication, Sales No. E.79.XII.3.
[Discusses the methodology and assumptions that the United Nations Population Division used in the
1970s to prepare the global estimates and projections of population].
United Nations (1980). Selected Demographic Indicators by Country, 1950-2000: Demographic
Estimates and Projections as Assessed in 1978. ST/ESA/Ser.R/38. [Presents the results of the 1978
revision of global population projections prepared by the United Nations Population Division].
United Nations (1981). World Population Prospects as Assessed in 1980. (United Nations publication,
Sales No. E.81.XIII.8). [Presents the results of the 1980 revision of global population projections
prepared by the United Nations Population Division].
United Nations (1984). Report of the International Conference on Population, 1984. Mexico City, 6-14
August 1984. E/CONF.76/19. [Presents the outcome documents of the second major United Nations
intergovernmental conference on population].
United Nations (1985). World Population Prospects: Estimates and Projections as Assessed in 1982.
(United Nations publication, Sales No. E.83.XIII.5). [Presents the results of the 1982 revision of global
population projections prepared by the United Nations Population Division].
S
TE S
United Nations (1989). World Population Prospects 1988. (United Nations publication, Sales No.
E.88.XIII.7). [Presents the results of the 1988 revision of global population projections prepared by the
R
AP LS
United Nations Population Division].
United Nations (1991). World Population Prospects 1990. (United Nations publication, Sales No.
E.91.XIII.4). [Presents the results of the 1990 revision of global population projections prepared by the
United Nations Population Division].
C EO
United Nations (1993). World Population Prospects. The 1992 Revision. (United Nations publication,
Sales No. E.93.XIII.7). [Presents the results of the 1992 Revision of global population projections
prepared by the United Nations Population Division].
United Nations (1998). World Population Prospects. The 1996 Revision. (United Nations publication,
E –
Sales No. E.98.XIII.5). [Presents the results of the 1996 Revision of global population projections
H
prepared by the United Nations Population Division].
PL O
United Nations (2000). World Population Prospects. The 1998 Revision, vol. III: Analytical Report.
(United Nations publication, Sales No. E.99.XIII.10). [Presents an analysis of the results of the 1998
Revision of global population projections prepared by the United Nations Population Division and a
M SC
Biographical Sketch
Hania Zlotnik is Director of the Population Division of the United Nations. Having joined the Population
Division in 1982, she directed its Mortality and Migration Section from 1993 to 1999 and its Population
Estimates and Projections Section from 1999 to 2003 when she became Assistant Director. She holds a
Ph.D. from Princeton University and is a graduate of the Universidad Nacional Autónoma de México
(UNAM). Her work has spanned the field of demography, covering the analysis of fertility, mortality,
international migration and urbanization with especial emphasis on their quantitative aspects. She has
prepared manuals on demographic estimation techniques and on the collection of international migration
statistics. She drafted the United Nations Recommendations on International Migration Statistics. She has
edited or written numerous reports published by the United Nations, including studies of female
migration, population distribution and migration, population estimates and projections, and levels and
trends of urbanization. She has published over 35 articles in books or refereed journals and has
collaborated in editing books on international migration and international migration statistics. She has
been an active member of several professional associations, having served as Board Member for the
Population Association of America and as Vice-President of the International Union for the Study of
Population (IUSSP) from 2001 to 2005.
S
TE S
R
AP LS
C EO
E –
H
PL O
M SC
SA NE
U