Hello all,

I am having difficulties with the interpretation of the regression results from estimating a growth regression in a dynamic panel data set-up, estimated using Stata's xtabond and xtabond2 commands. (i'm using difference GMM and system GMM estimators by Arellano-Bond and Arellano-Bover/Blundell-Bond respectively).

For my thesis I am interested in the effect of a set of regressors X on growth. My specification is a follows:

Ln(Y_i,t) = a*Ln(Y_i,t-1) + b*X_i,t + e_i,t

- Is it correct that b represents the effect on the growth rate(as the lagged income per capita is also incorporated in the model)?

- Also if the outocome is for example 0.5 how do i calculate the actual effect?

- How do i calculate the rate of convergence in this model if stata reports 0.990 as coefficient for the lagged income variable?

Thanks in advance, your help is much appreciated!

Maarten

I am having difficulties with the interpretation of the regression results from estimating a growth regression in a dynamic panel data set-up, estimated using Stata's xtabond and xtabond2 commands. (i'm using difference GMM and system GMM estimators by Arellano-Bond and Arellano-Bover/Blundell-Bond respectively).

For my thesis I am interested in the effect of a set of regressors X on growth. My specification is a follows:

Ln(Y_i,t) = a*Ln(Y_i,t-1) + b*X_i,t + e_i,t

- Y measures income per capita
- X is the independent variable of interest, which represents a percentage where 1% is written as 0.01 for example.

- Is it correct that b represents the effect on the growth rate(as the lagged income per capita is also incorporated in the model)?

- Also if the outocome is for example 0.5 how do i calculate the actual effect?

- How do i calculate the rate of convergence in this model if stata reports 0.990 as coefficient for the lagged income variable?

Thanks in advance, your help is much appreciated!

Maarten

## Comment