Dear Joseph

For the 4 indepenent variables you cite that has different variables; which versions do you have (for example, for inflation is it consumer price index, producer price index, etc.). Then other researchers can focus on which are the better measures. Some related literature is:

http://www.ccsenet.org/journal/index.php/ijbm/article/viewFile/7472/6488

Article The foreign direct investment-economic growth nexus

http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9361.2007.00426.x/full

Article The effects of Internet usage, financial development and tra...

http://onlinelibrary.wiley.com/doi/10.1111/rode.12130/full

What do you mean by different values of the variables? does it mean different measurement scale?

if the measurement scale of the variables is different, it does not prevent you from running the regression analysis. Because, all the variables used in the regression need not be of same measurement scale.

further discussion in this regard is most welcome.

Thanks a lot DRs for your response. The independent variables I have are Inflation GDP deflator (annual %), Official exchange rate (LCU per US$ period average), FDI net (BoP , Current US$), Trade (% of GDP) and GDP growth (annual %) as the dependent variable. Is ok to run the regression with these figures. or which measurements is best to use for these variables?

i have a paper in the same title " Determinants of Growth GDP per ca pita through Panel Data Analysis in Selected Arab Countries" you can read this paper .

Thanks , doing that now.

Your 4 independent variables look good. I might tweak the FDI to be (Net FDI/ Total GDP) and the F-X rate to be (% change in average F-X rate from year-to-year) to have them on a relative basis.

Mostly appreciated. DR Raymond.

It seems that you have not yet formulated a theoretical model about the question you want to analyse. But that needs to be done before econometric regression. For example, how can a FDI of a year, which is only a financial, not a real flow, influence the growth rate of the same year. Even in the longer run, the GDPeffect of FDIs is indeterminate. Some of them could bring additional real investment and technical progress, some could bring disinvestment (closing of a plant), and others could be neutral. Almost all your explaining variables are themselves dependent on GDP or its growth, as GDP ratios are, in principle, negatively correlated with growth. And there are also correaltions between the explaining variables: it is unlikely that the exchange rate would not influence GDP deflator and trade. I assume that trade is the sum of imports and exports, with - by GDP definition - imports with negative and exports with positive impact on GDP. This would lead to an inconclusive coefficient. For the trade balance (X-M), the coefficient will be positive, as in regressions with absolute data, it should be near 1.

Thanks a lot Anton Rainer. your reply is really helpful. so one wonders why most other research on same subject uses the same variables as I am using. can you please suggest variables that is best for measuring the impact of trade liberalization on economic growth, dependant and independent

Dear Joseph,

It was not clear to me that the motive of your regression should be to find out the impact of trade liberalisation. I think to do that, one has at first to find or develop a measure for this liberalisation. That is a difficult task, because liberalisation does not only mean lower taxes on trade, but also non-measurable measures like ad,omostratove easing for imports and exports and standardisation of products. Maybe that you can use the receipts of customs duties of your country (or the countries which you cover in your analysis) and dummies for the non-tariff part of treaties. I would not regress these data on GDP, but on imports and exports, which have a direct effect on GDP by definition. The problem is that the impact of liberalisation, in general, takes time, and that it is different for different industries. In the long run, it is likely to be advantageous for economic development (higher productivity by international division of labour), but in the short run it could be negative, because it could cause a disappearance of sectors of production (replaced by imports). Therefore, I doubt that you will find clear results by regressing aggregate variables, even if you found good "measures" for liberalisation. For the reduction of duties, one has also to take account of their budgetary effect, because the loss of reciepts has to be compensated somehow.

Best wishes

Most appreciated Anton Rainer. your reply is very helpful for my research and I intend to dig a bit deeper, but I have noticed most research on this use openness as a measure of liberalization which they proxy by the ratio of net trade to GDP. I totally agree with the difficulties associated with the choices of variables as many researcher's use different variables and therefore different results. Many thanks for the insight.