Pooling time series and cross-sectional data: An Application to Turkish export demand analysis

Abstract

In this study. Pooling of time series and cross sectional data is used for constructing a demand model for the Turkish Exports. Two regression models are employed and compared by their fitness to the proposed pooling arrangements and demand relations. 25 Year time series (1963-1985) and cross sectional data covering top 10 exporters from Turkey are used for this purpose. Multiple regression analysis is conducted over different pooling arrangements and properness of pooling and fitness of model is tested by means of a series of F tests.