Trade liberalisation in developing countries over the last 20 years has oft
en been implemented with the expectation of growth being stimulated; yet th
e evidence on its growth enhancing effects is mixed. This paper argues that
problems with mis-specification and the diversity of liberalisation indice
s used are in part responsible for the inconclusiveness. Using a dynamic pa
nel framework and three different indicators of liberalisation, the paper f
inds that liberalisation does appear to impact upon growth, albeit with a l
ag. The evidence points to a J curve type response and this finding is robu
st to changes in specification, sample size and data period, (C) 2002 Elsev
ier Science B.V. All rights reserved.