The use of census data for estimating international migration.

Technique for using census data to estimate international migration were described and applied to census data from Kuwait for illustrative purposes. These estimation technics can be used to assess migration flows in countries where migration data is not available. Census data which can be used to analysis migration patterns include age and sex distributions, foreign residence of relatives, birthplace, nationality, residence at some specific time in the past, and lenght of residence. The relative value and reliability of these different types of census data examining migration were discussed. Census infromation an age and sex distributions of the population is very useful because selective migration will be revealed by distortions in the distribution. An examination of the 1970 age and sex distribution for Kuwait revealed that the sex ratios, number of males to females, were higher than expected for ages 10-59. Subtraction of the number of females, aged 10-59, in the population from the number of males, aged 10-59, indicated that there was an excess of 97,359 males. This figure represented the estimated number of surviving male immigrants in the 1970 Kuwait population. This estimate was further refined by comparing the 1970 age and sex distribution with a stable population distribution. The comparison revealed that the 1st estimate of male immigration was probably too low, and that female immigration was also a factor. Another technic can be used if data from 2 censuses are available. The data from the 1st census can be used to project the age and sex distribution of the 2nd census. Then the projected distribution can be compared to the actual distribution observed in the 2nd census. Differences between the observed and projected censuses will serve to identify distortions in the distribution of the 2nd census. Errors in age reporting and differences in census data collecting procedures may bias the results of the analysis. This technic was applied to Kuwait census data on males only for 1965 and 1970. The estimated number of surviving male immigrants in the population in 1970, derived by applying this technic, was 70,598. Estimates of immigration were considerably different when age and sex distribution and place of birth were used jointly to estimate the size of the immigrant group. The estimated number of female immigrants was 106,000, and the estimated number of male immigrants was 200,000. When information on duration of residence was added to the other information, it was possible to estimate immigration by time periods. Several other technics were discussed and suggestions for altering were provided.