An appraisal of Hong Kong population projections.

Both the accumulation of detailed demographic data covering a long period of time and the development of research techniques for population analysis have improved the reliability of assumptions regarding fertility, mortality, and migration in Hong Kong, in turn facilitating improved population projections. In the past 2 decades, the Hong Kong Government has prepared 5 sets of projections on population size and structure. This article examines the accuracy of each of these projections and the formulation of their underlying assumptions. For example, recent projections have recognized the fact that women in different age groups exhibit different fertility patterns. The projection of the birth order-specific fertility rates also has been adopted, with the net reproduction rate used to check accuracy. In formulating mortality trends, standardized sex-cause-specific death rates for broad age groups were first projected into the future and then summed over various causes to derive the projected age-sex-specific death rates. Migration remains the most difficult component to forecast, and is the main source of error. Projections prepared in 1961 and 1971 (for 1971 and 1976) turned out to be far above actual population levels, whereas the 1976-based projections (for 1984) were overly conservative. The most recent set of projections, covering the period 1981-2001, was based on reasonably good assumptions. The population size is projected to grow from 5.13 million in 1981 to almost 7.11 million in 2001. The sex ratio will be 1074 males:1000 females. The rates of population growth is expected to decrease from 2.1%/year in the early 1980s to just over 1% in 2001. In terms of the age structure, the proportion aged 65 years and over is projected to increase from 7% in 1981 to 10% in 2001; the dependency ratio, however, is expected to decrease from 456/1000 to 444/1000.