Forum: help

RE: Production Elasticity [ Reply ] By: Arne Henningsen on 2012-05-22 08:34	[forum:5704]
Hi Tomas Your formula for calculating the elasticities from Translog models seems to be correct. Can you please provide a reproducible example which demonstrates that your manually calculated elasticities differ from the elasticities calculated by translogEla()? Your approach for obtaining elasticities at the sample mean seems to be correct. Best, Arne

Production Elasticity [ Reply ]
By: Tomas B on 2012-05-22 07:15

[forum:5702]

Hello,

Does the function translogEla() rely on the following computation of the elasticity:
E_{ji} = b_{j} + b_{j1} ln x_{1i} + ... + b_{jn} ln x_{ni},
where i denotes the i-th firm and j stands for the j-th input?

As I tried to check the results manually they seemed to be obtained in a rather different manner.

Secondly, is it appropriate to create a data frame consisting of the mean values, say DFmean, and then obtain t-values for these elasticities by the virtue of :
estEla<- translogEla( c( "logX1", "logX2", "time" ),
data = DFmean, coef = coef( model ),
coefCov = vcov( model ),dataLogged=TRUE )
estEla / attributes( estEla )$stdDev .

Regards,
Tomas B.