Abstract
In their comment on Muscatelli, Srinivasan and Vines (MSV) (I992), Athukorala and Riedel challenge our conclusions regarding the estimated longrun elasticities of export demand for Hong Kong, and our suggestion that estimating a system of export demand and supply could provide a solution to the 'normalisation problem'. Unfortunately, by focusing only on a subset of the issues raised by MSV, Athukorala and Riedel tend to ignore some of the key econometric evidence against their preferred 'small country assumption', and their chosen normalisation (price-dependent export demand). There are three key pieces of evidence which tend to point against the small country assumption.' First, let us turn to the normalisation issue. As Athukorala and Riedel note, least-squares methods such as the Phillips-Hansen procedure will not yield estimates of cointegrating vectors which are independent of normalisation, and that is why MSV point out that the Phillips-Hansen procedure may only alleviate not 'solve' the 'normalisation' problem. Ideally, when estimating a simultaneous cointegrated model (SCM) one should attempt to test and impose identifying restrictions on the H matrix in a Johansen-type multi-variate VAR framework:
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
