Introduction

Demographic data tend to be incomplete and inaccurate in some degree even at the best of times. Quite often the statistical material available for study is deficient, and attempts to remedy the trouble by means of new enquiries may be frustrated by the smallness of practical samples or the refusal of respondents to co-operate. For remote subjects, for developing countries or for the past the information may indeed by very patchy and very difficult to interpret. In spite of these handicaps, demographers persevere and often produce meaningful results. The first need is clearly to test the data for completeness and accuracy, and this is an important aspect of population analysis. A number of methods for doing so have been evolved, and the earlier part of the present chapter is concerned with these.

If completeness or inaccuracy has been established, unless the degree of shortfall or error is small the ideal procedure is to discontinue all demographic analysis and devote all available resources to the collection of more representative material. Often, however, this is impracticable; or it would take so long to achieve the necessary improvement in the available information that demographic studies urgently required – perhaps for policy formulation – ought to proceed by other means.

The second part of this chapter is devoted to these alternative means. Many demographers have described what they have done in particular circumstances, but it is clear that the best solution in any given instance depends very much on the nature of the subject under investigation and on the seriousness of the deficiency in the data.