Figure 3-2: Most recent world population projections from the four main demographic organizations, including high and low variants. Two of the projections (USCB and World Bank) will not be published. Variations in future world populations are largely determined by different assumptions concerning the demographic transition in developing countries as well as long-range fertility rates worldwide.
Figure 3-3: Central projections extracted from Figure 3-2 show excellent agreement among the central projections over the next 100 years (with the possible exception of the IIASA projection).
Figure 3-4: Industrial (IND) and developing (DEV) countries' population projections from the medium (central) variants. Note that regional definitions differ in IIASA compared to UN projections.
Figure 3-5: Projected world average TFRs from the UN, IIASA, and World Bank. The older 1994 World Bank projections are shown, because the latest (1996) projection methodology is not published. USCB fertility rates are not readily available and so are omitted. UN fertility assumptions are shown only for the 1996 Revision as full detail of the 1998 data were unavailable at the time of drafting this report.
World population projections are currently generated by the following institutions:
The main defining characteristics of the individual projections and a detailed description of each projection is given in Gaffin (1998).
Figure 3-2 displays these latest world population numbers from all four demographic organizations. Two of the four projections, from the World Bank and USCB, contain only one central estimate and are unpublished. These are generated currently for internal organizational purposes and are not further considered here.
The only population projections that currently incorporate variation of the long-range fertility rates are those produced by UN and IIASA (Gaffin, 1998). For the UN, these variants result in four additional projections to the UN medium projection and are referred to as "low,""medium low," "medium high" and "high." The designations reflect the average world fertility rate relative to the medium projection. For IIASA, the primary variants also refer to world average fertility (and mortality) rate relative to the medium projection and are referred to in Figure 3-2 as "high" and "low."
The central projections in Figure 3-3 show excellent agreement
over the next 100 years, with the exception of the "overshoot" in the IIASA
projection, discussed below. Although significant, such an agreement does not
imply certainty or accuracy. Rather, it reflects the use of similar methodologies
and databases, the dominance of population momentum associated with a young
population age structure, and the assumption of replacement-level fertility
in the long term.
Figure 3-4 shows the breakdown of population growth projections in industrial countries (IND) and developing countries (DEV) as defined by the UN. The key conclusion to draw from Figure 3-4 is the dominance of the DEV population on future world population growth.
The most important variable to determine future population levels is fertility. To a lesser extent, future population also depends on mortality and migration rates. Figure 3-5 shows the world-average total fertility rate (TFR), the average number of births per woman, assumed for the various projections. Overall, a fairly broad range of assumptions about future world average fertility is encompassed by the projections. Future population size is very sensitive (see Figure 3-2) to comparatively small changes in the long-range fertility rate. For example, within the UN projections, a decrease of the asymptotic fertility rate of less than half-a-birth per woman (from 2.1 to 1.7) decreases population in 2100 by 46% (UN, 1998). Only the replacement-level fertility of about 2.1 results in a stable population in the long run.
Such sensitivity to small asymptotic fertility-rate changes indicates the high and low projections from UN and IIASA are all feasible scenarios of the future population (Gaffin and O'Neill, 1998).
Up to 2050, the IIASA central TFR is high compared with the UN and World Bank central projection. Later the IIASA central TFR declines to below replacement-level fertility rate, as depicted in Figure 3-5. As a result, the IIASA central forecast lies considerably above all the others throughout the 21st century. After that, IIASA's below-replacement, long-term fertility assumption causes the projection to decline and to "hit" the World Bank and UN curves in 2100. This "overshoot" occurs because the IIASA central scenario assumptions rely on an expert poll conducted in 1993 (Lutz, 1994), whereas other projections incorporate more recent information into their future assumptions (more rapid recent fertility declines in many developing countries; see Courbage, 1998). However, among demographers disagreement persists on the timing and rates of demographic transition in the developing countries, in particular between the demographers at IIASA and those at the UN and World Bank (see the discussion in Lutz, 1994) on the differences among alternative population projections). This disagreement reflects important uncertainties in the projection of future demographic developments. The below-replacement, long-term fertility level in the IIASA central estimate is a hallmark of the institute's work. Demographers at IIASA strongly adhere to the view that there is little reason to expect developed nations to return to replacement-level fertility, while much evidence suggests that low long-term rates will persist (Lutz et al., 1996). One issue here, however, is that the TFR measure is currently depressed in Europe, in part because women are delaying their childbearing until later ages (Bongaarts, 1998). Thus, the TFR may rise if these women do have children, and such a correction has to be incorporated into the long-term TFR assumptions (e.g., see Courbage, 1998). Important uncertainties are by how fast and how much the TFR will rise.
Other reports in this collection