Demographic Trends and Population Policy 3 sidered more reliable than the earlier information about China's population that was based on the traditional household registers. A Committee on Population and Demography, appointed by the National Academy of Sciences of the United States and chaired byAnsley Coale, has carefully sifteS the available data from China. The resulting adjusted estimates are preferable to the official data that suffer from varying degrees of underestimation of birth and death rates. Of course, a large country such as China has sizeable inter-provincial variations in the quality of data and also in the manner in which various policies announced at the central or national level are implemented. This rich variety of differences cannot, however, be covered in the present paper which relies for most purposes on the adjusted estimates. KEY DEMOGRAPHIC DATA Table 1 summarises the key statistics about China's population according to the censuses conducted in 1953, 1964 and 1982. The average growth rates for the inter-censal period can be compared with the rates of natural increases shown in the second segment of the table. In the absence of significant international migration, the rate of natural increase equals the growth rate. Also, the estimated death rates are supplemented by the information on the life expectancy at birth or the average age at death. The data on the size and inter-censal growth of China's population suggest that the post-liberation period witnessed a significant rise in the growth rate, which was much higher than the estimated average rate of 0.5 per cent during 1900-50 (and zero during 1850-1990)2. Besides, the average rate of growth during the 18 years between 1964 and 1982 was higher than during 1953-64; but the vital rates for the intervening years need to be examined to understand some of the momentous changes. Within the first inter-censal decade, the birth rate was between 40 and 44 per 1,000 population up to 1957 and the death rate was on the decline from a peak of about 29 in 1954.^ During 1958-61, the years of the Great Leap Forward and the ensuing crisis, the birth rate declined and the death rate rose sharply so that during 1959-61, the population actually declined. The lost births are reflected in the age pyramid of the population enumerated by the 1982 census in the form of sharp declines in the age group 21-24 (see Figure I). The peak death rate during 1960 was almost 39 and the birth rate during the same year was 25. The number of excess deaths was 27 million for the three-year period. But for these excess deaths and birth losses, the rate of inter-censal growth during 1953-64 would have been higher. The birth rate reached a low of 22 during 1961 but recovered to 41 during 1962,47 during 1963 and 41 again during 1964. The death rate declined sharply during 1964-66 and has continued to show a downward trend since then. The result was an acceleration of the rate of population growth to an unprecedented 2.7 per cent during 1964-66. It stimulated a resumption of