Good morning,
I allow myself to ask for help once again with some methodology.
I am doing research the Kuznets curve hypothesis and obtained the following results:

Code:
xtreg lnWaste lnGDP lnGDP2 lnHhsize lnUrban lnEducated, re

------------------------------------------------------------------------------
     lnWaste |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       lnGDP |   1.503827   .9565295     1.57   0.116    -.3709363     3.37859
      lnGDP2 |  -.0589236   .0487258    -1.21   0.227    -.1544245    .0365773
    lnHhsize |    .859486   .2017066     4.26   0.000     .4641484    1.254824
     lnUrban |   .4833079   .2190035     2.21   0.027     .0540689     .912547
  lnEducated |   -.111818   .0470411    -2.38   0.017    -.2040168   -.0196192
       _cons |   -3.71604   4.732251    -0.79   0.432    -12.99108    5.559002
-------------+----------------------------------------------------------------
I computed where the turning point lies and would like to see whether the relationship between Income (GDP) and Waste increases at low values and decreases at high values of the interval.
According to Lind (2010), "as the distribution of the estimated slope at any point is readily available, a test for the sign of the slope at any given point is straightforward." However, I am unsure how to test for the sign of this slope if I already obtained my regression. I am unsure whether this is a simple mathematical operation or whether I should use Stata for this.

Could anyone give me a hand with this?
Wishing you all a good week-end.