Abstract
This paper criticizes Cho and Graham's argument that poor countries converge from above their steady-state income levels, based on their misspecification of formulating the steady-state income by omitting the variation in the base period technology across countries when estimating steady-state income. This paper also questions the cross-country regression methodology, which generally ignores the changes in variables over time. A time-series approach is employed to analyse the long-run behaviour of actual and steady-state income levels for a group of seven developing countries, which are observed to be above their steady-states in Cho and Graham (1996). An error-correction-based-test is used to examine the existence of cointegration. The results suggest that these countries' actual and steady-state income per capita tend to move together over time, which is consistent with the Solow model's prediction.