Hello!
I'm reading economics in São Paulo (Brazil) and studying the per capita income convergence between Brazilian states. For that, I estimated the "Barro Regressions" as it follows: "(1/T)*log(yend/ybeg)i=β0+β1*log(ybeg)i+e" ["(1/t)*log(yend/ybeg)i" is the annual growth rate of per capita income for i state; "ybeg", the per capita income of the beginning year]. Benigno Valdés, for instance, at "Economic Growth: theory, empirics and policy" book, writes the estimated "β1" as "((e^(λT)-1)/T)" and λ as the speed of convergence. I've estimated β1 but don't know if there is a way to have the λ attached to the β1 at Stata.
I have this (what I called "VRPC" is the "(1/T)*log(yend/ybeg)i" and "LRPC90" is "log(ybeg)i"):
. reg VRPC LRPC90
Source | SS df MS Number of obs = 26
-------------+---------------------------------- F(1, 24) = 5.94
Model | .000892961 1 .000892961 Prob > F = 0.0226
Residual | .003606052 24 .000150252 R-squared = 0.1985
-------------+---------------------------------- Adj R-squared = 0.1651
Total | .004499012 25 .00017996 Root MSE = .01226
------------------------------------------------------------------------------
VRPC | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
LRPC90 | -.0104969 .0043058 -2.44 0.023 -.0193837 -.0016101
_cons | .0415848 .0099832 4.17 0.000 .0209805 .062189
------------------------------------------------------------------------------
So the estimated β1=-0.0104969.
I'm reading economics in São Paulo (Brazil) and studying the per capita income convergence between Brazilian states. For that, I estimated the "Barro Regressions" as it follows: "(1/T)*log(yend/ybeg)i=β0+β1*log(ybeg)i+e" ["(1/t)*log(yend/ybeg)i" is the annual growth rate of per capita income for i state; "ybeg", the per capita income of the beginning year]. Benigno Valdés, for instance, at "Economic Growth: theory, empirics and policy" book, writes the estimated "β1" as "((e^(λT)-1)/T)" and λ as the speed of convergence. I've estimated β1 but don't know if there is a way to have the λ attached to the β1 at Stata.
I have this (what I called "VRPC" is the "(1/T)*log(yend/ybeg)i" and "LRPC90" is "log(ybeg)i"):
. reg VRPC LRPC90
Source | SS df MS Number of obs = 26
-------------+---------------------------------- F(1, 24) = 5.94
Model | .000892961 1 .000892961 Prob > F = 0.0226
Residual | .003606052 24 .000150252 R-squared = 0.1985
-------------+---------------------------------- Adj R-squared = 0.1651
Total | .004499012 25 .00017996 Root MSE = .01226
------------------------------------------------------------------------------
VRPC | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
LRPC90 | -.0104969 .0043058 -2.44 0.023 -.0193837 -.0016101
_cons | .0415848 .0099832 4.17 0.000 .0209805 .062189
------------------------------------------------------------------------------
So the estimated β1=-0.0104969.
