How to interpret a simple regression coefficient?

[Jahangir Alam]

I have the following regression equation:

log(y_t/y₀) = B₀+B₁Trade_share + e_t

where t = 1 ... 5
Y is productivity
Trade_share (>0) is exports plus imports as a share of output.

B₁=0.07

Should I interpret as a log-level model?

[Carlo Lazzaro]

Jahangir:
you seemingly have a log-linear regression model.
Hence: one-unit increase in -trade- increases the dependent variable by:

(exp(0.07)-1)*100%)=7.25%

[Jahangir Alam]

@Carlo Lazzaro, since the variable Trade measures as a share, can I use one-unit change?

I am thinking like this:

How much dependent variable changes with ten percent increase in trade share? What do you think?

Thanks Carlo,

Jahangir

[Carlo Lazzaro]

Jahangir:
thanks for clarifications.
The interpretation that you ask for goes like this:
when adjusted for the other predictors, a 10% point increase in -Trade- increases the dependent variable by roughly 72.50%.

[River Huang]

I have the following regression equation:

log(y_t/y₀) = B₀+B₁Trade_share + e_t

where t = 1 ... 5
Y is productivity
Trade_share (>0) is exports plus imports as a share of output.

B₁=0.07

Should I interpret as a log-level model?

It seems to me that the dependent variable is nothing but the growth rate over the 0 to t period. So, please treat/interpret this variable as growth in a linear regression (not a log-linear model).
In addition, the interpretation depends on how the trade share data measured. Say, trade_share = 56 (the unit is in %) or trade_share = 0.56?

[Jahangir Alam]

My trade share is 0.56!