ABSTRACT
The problem of how to construct an index number is as much one of economic theory as of statistical technique. Frisch (1936), p. 1
It is true that some applications of index numbers are not strictly economic but occur in (more or less) distantly related subjects ranging from demography to technology. Examples are easily found: standardised birth, sickness or death-rates; crop yields. It remains true, however, that the main uses of index numbers are in economics and hence that the theory is best developed in an economic conte-xt. There is, then, little difficulty in extending the application of the index-number technique to other fields. It is important to avoid the trap of divorcing the economic and the statistical aspects of index numbers, of attempting to consider an index number in the abstract. Index numbers are essentially practical constructs. The two approaches to them, the economic and the statistical, must be used together and from the outset.
