Testing instrument validity - ordered probit

[Dinushka Herath]

Hi,

I am looking at the relationship between financial inclusion and poverty. my dependent variable is income quintile (inc_q) which is an ordinal variable and i have some control variables along with the binary financial inclusion variable. I used extended ordered Probit regression to address the problem of endogeneity in the financial inclusion (fin_inc) variable. I ran the following code

eoprobit inc_q female age educ emp_in time , endogenous(fin_inc = fin14, probit)

where fin14 is my instrument and it gives data on if an individual has made mobile phone transaction or not. I want to test if this a good instrument. how do I test for for instrument exogeneity?


Any advice on this is appreciated


D. H

[Zairihan Halim]

Hi,
I have the same question as posted by Dinushka. Is there a syntax to test for the validity/exogeneity of instrument(s) after the eoprobit estimation? I read the materials on eoprobit, but did not find any on such tests. Please advise.