within tranformation fixed effects
Dear members,
I'm trying to regress the per capita GDP at the hand of the service ecport sophistication index EXPY and other control variables.
My data includes 22 countries from the Euro Area from 2005 - 2013. The paper I (firstly ened to replicate) a 3 year dynamic panel
The regression formule goes as follows:
y(i,t-)- y(i,t-1=) αy(i,t-1)+β^' SE(i,t-1)+φ^' GD(i,t)+μt+ϑt+ε_i,t)
Where y(i,t-) is the natural logarithm of GDP per capita, y(i,t-1) is the level of GDP per capita at the beginning of each three year period, SE(i,t-1) is the dynamic service export sophistication (EXPY) measure, GD(i,t) is the set of relevant control variables for growth determinants, μt represents time dummies, ϑt stands for country fixed effects and ε(i,t) is the error term.
I'm know quite a bit about fixed effects but I fail to implement this within transformation over 3 years periods (in a larger span of 9 years) with stata.
I tried this:
gen lagGDP = l3.GDP if B==2005 | B==2008 | B==2011
gen lagEXPY = l3.EXPY if B==2005 | B==2008 | B==2011
gen logGDP = log(GDP), , gen logEXPY = log(EXPY)
gen loglagGDP =log(lagGDP)
gen difGDP = logGDP - loglagGDP
The difference and lagging of initial EXPY and initial GDP are to eliminate the endogenity thath comes with growth regressions. control variables are average years of schooling (barro & lee), m2 real money + quasi money for financial development, Rule of Law for institutional quality and Trade(%) for trade openness.
Fixed effects with dummies:
reg difGDP loglagGDP loglagEXPY logHC logM2 LAW c1 c2 c3 c4 c5 c6 c7 c8 c9 c10 c11 c12 c13 c14 c15 c16 c17 c18 c19 c20 c21 c22 y1 y2 y3 y4 y5 y6 y7 y8 y9
Is this the right way to run a within transformation?
thank you in advance,
Joost Persyn
I'm trying to regress the per capita GDP at the hand of the service ecport sophistication index EXPY and other control variables.
My data includes 22 countries from the Euro Area from 2005 - 2013. The paper I (firstly ened to replicate) a 3 year dynamic panel
The regression formule goes as follows:
y(i,t-)- y(i,t-1=) αy(i,t-1)+β^' SE(i,t-1)+φ^' GD(i,t)+μt+ϑt+ε_i,t)
Where y(i,t-) is the natural logarithm of GDP per capita, y(i,t-1) is the level of GDP per capita at the beginning of each three year period, SE(i,t-1) is the dynamic service export sophistication (EXPY) measure, GD(i,t) is the set of relevant control variables for growth determinants, μt represents time dummies, ϑt stands for country fixed effects and ε(i,t) is the error term.
I'm know quite a bit about fixed effects but I fail to implement this within transformation over 3 years periods (in a larger span of 9 years) with stata.
I tried this:
gen lagGDP = l3.GDP if B==2005 | B==2008 | B==2011
gen lagEXPY = l3.EXPY if B==2005 | B==2008 | B==2011
gen logGDP = log(GDP), , gen logEXPY = log(EXPY)
gen loglagGDP =log(lagGDP)
gen difGDP = logGDP - loglagGDP
The difference and lagging of initial EXPY and initial GDP are to eliminate the endogenity thath comes with growth regressions. control variables are average years of schooling (barro & lee), m2 real money + quasi money for financial development, Rule of Law for institutional quality and Trade(%) for trade openness.
Fixed effects with dummies:
reg difGDP loglagGDP loglagEXPY logHC logM2 LAW c1 c2 c3 c4 c5 c6 c7 c8 c9 c10 c11 c12 c13 c14 c15 c16 c17 c18 c19 c20 c21 c22 y1 y2 y3 y4 y5 y6 y7 y8 y9
Is this the right way to run a within transformation?
thank you in advance,
Joost Persyn
Comment