Pakistan Journal of Applied Economics

71-101

Authors: Ashok PARIKH*

The estimation and forecasting of trade shares is attempted in a world framework for major trading countries on the basis of data on bilateral flows for the period 1965 – 1980 for manufactures and chemicals [Standard International Trade Classification (SITC), (5) to (8)J . The assumptions of separability and homotheticity are made to obtain an operational Hickman-Lau model from Armington's approach in which products are distinguished not only by their kind but by their place of production. The problem of heteroscedasticity is encountered and estimates of elasticities of substitution are obtained using the most general form of the mode1. The results are compared with the Gana, Hickman, Lau, and Jacobson (1979) study. The estimates of substitution elasticities based on 1965“ 1980 data are lower than the ones obtained using 1960- 1970 data and it is argued that the flexible exchange rate (and the agents' responses) could be responsible for low substitution elasticities when the period after 1973 is included.

125-129

Authors: A. Hafeez SHAIKH* & Asad ZAMAN**

The availability of computer packages has put a variety of econometric techniques within the reach of nearly everyone. The sophisticated mathematics and modern technology required lends a magical quality to these tools. One has the feeling that if only one could pronounce the right magic word (or, use the right regression technique) all econometric problems would disappear. The truth, as one learns from experience, is far from this fantasy . Because strong assumptions are made by econometric theory, it is not at all uncommon for the most sophisticated of techniques to generate the crudest of errors. Therefore it is extremely important to understand : a) The overall limitations of econometric techniques: what can and cannot be done; and b) factors not included in the model (such as politics, strikes, outliers) which may influence results.