An econometric estimation of the aggregate import demand function for Bangladesh: some further results

Abstract

This study empirically estimates some critical parameters of the aggregate import demand function for Bangladesh for Bangladesh using quarterly time series data and by applying the Johansen–Juselius multivariate cointegration technique. Cointegration results indicate that the import demand function is dominated by income and relative prices. The income elasticity is significantly positive and exceeds unity indicating that aggregate imports are to be considered as ‘luxury’ goods. The effect of the relative price variable is significantly negative but its elasticity coefficient is less than unity.

Acknowledgements

The authors would also thank Dr Robert Jarrett of the English Department at the University of Houston-Downtown for editing the English grammar and style in the paper. However, the usual disclaimer applies.

Notes

 For an excellent discussion on the import demand functions in developed countries, see Murray and Ginman (1976), Stern et al. (1976), Khan and Ross (1977), Boylan et al. (1980), and Goldstein and Khan (1985), among others.

 For relevant literature on developing countries, see Hemphill (1974), Khan (1974), Sarmad (1988; 1989), Moran (1989), Shilpi (1990), Emran and Shilpi (1996), among others.

 For studies on Bangladesh, see Shilpi (1990), Emran and Shilpi (1996), and Dutta and Ahmed (1999).

 For details about the derivation of this equation, see Emran and Shilpi (1996).

 The quarterly data on GDP is not available in Bangladesh. However, quarterly data on this variable was needed in order to calculate the ratio of foreign exchange to GDP. However, a quarterly series was obtained from Mr A. Kashem of the Research Department of Bangladesh Bank, the central bank of this country, who compiled this quarterly data series using a linear interpolation technique.

 Johansen and Juselius (1992) used stationary inflation rate variables in their cointegration model along with non-stationary variables for explaining the Purchasing Power Parity theory for the UK.

 The details on the empirical test results for the lag length tests and for the deterministic components in the data were not reported here because of space limitation; however, they can be provided to anyone upon request.