i'm applying a panel data analysis. I have four variables; real GDP, trade openness, energy price and energy consumption. I took ln (log)of trade openness,real GDP and energy price. but i cannot decide whether ln(log) of energy consumption should be taken or not. please inform me on this. Thank you

Note: Kg of oil equivalent per capita is used to measure energy consumption.

EN= energy consumption

Y=Economic growth

T= trade openness

EP=energy price

W= country fixed effect EN=f(Y,T,EP,W)

enter image description here

enter image description here

Thank you.

  • $\begingroup$ Why are you considering the logarithms in the first place? What is the reason? Why you have no hesitation in taking the log of trade openness for example, (and for all the other variables), but you have doubts especially about energy consumption? You should include such information in your question, not as a comment-reply, because as it stands your question cannot be answered and should be closed. $\endgroup$ Nov 26, 2015 at 13:12
  • $\begingroup$ Except from energy consumption, other variables are price. But, energy consumption is not price. if unit of variables is price, I have to take log. But, I dont have applied before such a variable, unit of whose is different from price. Thus, I'm confused. Please inform me on this topic. thanks @AlecosPapadopoulos $\endgroup$
    – 1190
    Nov 26, 2015 at 14:49

1 Answer 1


Prior to any trasnformation:

  1. Energy consumption is a per capita variable, specifically "Kgr per capita"
  2. Real GDP is a per capita variable
  3. Trade Openness is a per capita variable
  4. Energy price is a unit price of "crude oil"

I see two problems: Is "Energy Price" the "price per Kgr" -because usually oil prices are not quoted per Kilogram. But it should expressed here "per Kgr" so as to stand for the price of the dependent variable.

If it does, it is only natural to consider the logarithms of the variables, dependent variable included.

But one more thing that needs clarification is what $Y$ stands for. The authors write "$Y$(economic growth)". If this is the case then for this variable you don't just take the logarithm of real GDP but the first-difference of the logarithm (which approximates the rate of growth).

  • 1
    $\begingroup$ It appears to be the price of crude oil per barrel in USD. Also, the paper does not look like refereed properly. It is available here: sciencedirect.com/science/article/pii/S0301421514000986 $\endgroup$
    – london
    May 17, 2016 at 16:52
  • $\begingroup$ @london I will have a look at the paper, thanks for the link. Why are you doubting the peer-review process? You doubt the journal or something specific to the article? $\endgroup$ May 17, 2016 at 17:35
  • $\begingroup$ The quantitative methods applied without considering or motivating the research question within an economic context, the text gives limited information about the variables and units of measurement are not clear and the overall structure. Energy demand is measured in heat content, but not the price, and the price is actually oil price, not energy price. Anyway, it appears to be poorly edited paper. $\endgroup$
    – london
    May 17, 2016 at 18:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.