An important aspect of the usefulness of population projections and forecasts is their accuracy.
Although other aspects, such as the information content (e.g. has only total population been projected,
or also age groups? which regional level?) and the usefulness for policy purposes (e.g. does the
projected trend imply immediate policy measures?) are relevant as well, the degree to which the
forecast may be expected to reflect real developments in the future is a key factor in assessing its
quality.
For industrialized countries we know some common characteristics concerning the accuracy of
historical forecasts and projections, see Keilman (forthcoming) for an overview. Ex-post comparisons
between projected and observed trends in population variables have revealed that the forecast
accuracy of fertility is better than that of mortality - behaviourally determined variables are difficult to
forecast. Yet large errors have been found for both the young and the old after a forecast period of 15
years (errors up to +30 per cent for the age group 0-4, and -15 per cent or lower for women aged 85+
are not uncommon). This suggests that those old forecasts supplied useful information perhaps up to
10-15 years ahead, but certainly not longer. Finally, detailed studies for a few countries have found
only a weak association between improvements in forecast accuracy and the introduction of more
sophisticated forecast methods.
These findings relate to the accuracy of forecasts produced for industrialized countries. Much less is
known about the reliability of population forecasts for developing countries. Inoue & Yu (1979)
investigated the errors in total population size of six rounds of United Nations projections, with base
years from 1950 to 1970 and observed data for the period 1950-1975. They found a consistent
overestimation of the projected growth rate in developing countries after 1960, which was explained
to a large extent by the rapid slowdown of population growth in China. They also concluded that
errors in the base population and in the growth rate of population immediately preceding the starting
year were important determinants for errors in the projected population size of developing countries.
Keyfitz (1981) and Stoto (1983) analysed, for various countries in the world, errors in projected
population growth rates in projections made by the United Nations during the 1950s and 1960s.
Important findings were that errors varied strongly by region and by base year: regions in which
population growth was high had large errors, as did forecasts made in the early 1950s. Moreover,
Keyfitz concluded that the error in the growth rate was more or less independent of forecast duration.
These conclusions were confirmed by Pflaumer (1988), who analysed the predicted growth rates in
101 countries with at least a million inhabitants (excluding China). Forecasts were those made by the
UN between 1963 and 1978, and actual growth rates applied to the period 1960-1980. Pflaumer found
also some evidence for an improvement over time in the accuracy of the projected growth rates.
Furthermore, errors were relatively small in countries with large population sizes.
The purpose of this report is to extend the analyses of the United Nation's projections mentioned
above, which were focused on growth rates and total population sizes. I investigate the accuracy of
the UN projections of the age structure and birth and death rates in seven major regions of the world:
Africa, Asia, Europe, the USSR, Latin America, Northern America, and Oceania. I also include
findings for a few large countries which may dominate their region: China and India (Asia), and the
USA (Northern America). UN-forecasts made between 1951 and 1988 have been evaluated. Projected
numbers on total population size, crude birth rate, crude death rate, age structure in five-year age
groups and dependency ratios for the period 1950-1990 were compared with corresponding ex-post
observed numbers. I try to answer two broad questions in this report. First, does accuracy differ
strongly among regions? In other words, are population trends in some regions easier to project than
those in other regions? And second, did the UN-projections improve over time? The results indicate
that the latter was indeed the case, not only because base line population estimates were improved,
but also because unforeseen declines in birth rates became less important for projection errors.
Furthermore, to prepare projections is more difficult for some regions than for others. Age structure
projections for the former USSR and for Asia show larger errors than on average. For Asia this is
explained by errors in base populations. When errors in the base population are removed and hence
one considers errors caused by wrong assumptions regarding fertility and mortality only, the age
structure of the former USSR is still very inaccurate, but also that of Oceania, Northern America and
Europe.
In Section 2, I first present a number of simple error measures that have been used for the evaluations.
Next I list the UN-projections that have been selected for the evaluation and give a brief historical
account of main aspects of these projections. I discuss the problem of which data should be used as a
yardstick against which the projected numbers can be compared. I opt for the most recent data, but the
consequence is that I am confronted with a problem which is caused by the continuous revision of
"observed" population numbers. The last issue taken up in Section 2 is a solution to this problem.
Section 3 contains the main findings. The accuracy of total population size is briefly presented. Most
attention is given to errors in crude birth and death rates and in the age structure of the subsequent
forecasts in the various regions. In addition to errors in forecast results by five-year age group,
duration, base year, and region I present observed and forecasted values for the young and the old age
dependency ratios. A number of implications of the findings for population forecasting, as well as
recommendations, are given in Section 4.
The main text focuses often on the accuracy results for the world as a whole for various indicators
(total population, birth and death rates, five-year age groups, and dependency ratios). Errors in these
indicators have also been computed for each of the seven major regions, and sometimes for India,
China and the USA as well. The region-specific results are presented in this report in the form of
tables and figures in the Appendix, but only briefly referred to in the text.
