Construction of Economic Index Numbers With an Incomplete Set of Data
Keith C. Brown
K. Rao Kadiyala
Review of Economics and Statistics 65, 1983, pp. 520-524
In this paper, the Joint Equation Estimation Method (JEEM) of predicting missing data is introduced and applied to the problem of constructing economic index numbers. The main implication of this investigation is that any properly designed regression method, particularly the JEEM, can be a valuable tool in reconstructing a partial series of economic data. It is documented that beyond the superior theoretical properties of the JEEM relative to single equation approaches, there are also substantial empirical gains in terms of both a reduction in predictor variance and measured predictive ability. When this is coupled with the fact that there is virtually no additional computational difficulty, the advantage of the grouped equation technique is that much more pronounced.
Download this paper (PDF format)