Abstract
We construct an aggregate data panel to estimate price and income elasticities of the Arab countries imports from and exports to Euro zone. We study the nonstationarity of our series and verify the cointegration hypothesis among the variables using Pedroni's (Citation2004) heterogeneous panel cointegration tests. The panel data circumvent the problem of short span sample and increase the power of the nonstationarity tests. Then, we estimate the idiosyncratic and panel cointegrating vectors using dynamic ordinary least squares (DOLS) (Kao and Chiang, Citation2000), fully modified ordinary least squares (FMOLS) (Phillips and Hansen, 1990) and group mean DOLS and FMOLS developed by Pedroni (Citation2000, Citation2001). Our variables are shown to be cointegrated. Arab imports from Euro zone countries are income inelastic, but price elastic. Results of export function are not conclusive and depend on the used estimator.
Acknowledgments
The author would like to thank Peter Pedroni, an anonymous referee and participants at Unit Root and Cointegration Testing Conference in Faro, Portugal (September 29 to October 1, 2005) for helpful comments.
Notes
1 When we consider Arab exports to Euro zone countries (i.e. Euro zone imports from Arab countries), we consider only seven nonoil exporters countries which are: Egypt, Jordan, Lebanon, Morocco, Sudan, Syria and Tunisia. When imports from Euro zone countries, we consider eight more countries which are Algeria, Kuwait, Libya, Oman, Qatar, Saudi Arabia and UAE.
2 MENA stands for Middle East and North African countries which are in alphabetical order: Algeria, Bahrain, Comoros, Djibouti, Egypt, Iran, Iraq, Kuwait, Lebanon, Mauritania, Morocco, Oman, Qatar, Saudi Arabia, Somalia, Sudan, Syria, Tunisia, Turkey, United Arab Emirates and Yemen.
3 Islam prohibits the use of interest rate in financial transactions.
4 The euro zone includes 12 countries. However, since in the Directions of Trade Statistics, imports and exports of Luxembourg are aggregated with Belgium's, we consider both countries as one member in our panel.
5 Since CPI series are not available for those countries in the period considered here, we use GDP deflator instead.
6 IPS (2003) stands for Im et al. (Citation2003) as mentioned in the introduction.