How can I test in a multiple regression model whether a drop of 1% in $x_1$ will cause a larger effect than a 1% drop in $x_2$, given that I used the growth rates of my dependent and independent variables?
Let me give you a hint since this question seems to be a kind of homework question. Suppose the following model:
$\ln y = \beta_1 \ln x_1 + \beta_2 \ln x_2 + \epsilon$.
In this model, $\beta_1$ and $\beta_2$ are elasticities and one can be tested against the other with a $F$ statistic.