Page 1 of 1

Technological Unemployment

Posted: Sun Mar 22, 2015 5:51 am
by IanKennedy1
Hi,

I am currently trying to observe the effects of patents (innovation) on the unemployment rate, accounting for many control variables such as GDP, Inflation, Interest Rates, etc. Unfortunately, I do not know which path to take. Would it be better to regress all variables as a multiple regression, treating them all as independent endogenous variables, or undertake a 2SLS regression? In all examples of 2SLS regression I have seen, the exogenous variables are influencing a independent variable on the RHS, however I want to control these variables against unemployment which I am unsure how to do.

Any assistance would be greatly appreciated.
Kind Regards.

Re: Technological Unemployment

Posted: Mon Mar 23, 2015 5:41 am
by IanKennedy1
bump

please

Re: Technological Unemployment

Posted: Mon Mar 23, 2015 5:52 am
by trubador
Yes, you'd better use a method (like 2SLS) that takes into account the endogeneity among the dependent variables.