Using lagged independent variable as an instrument (in panel data)

[Felix Zhu]

Hi everyone,

I wonder if it is possible to use a lagged independent variable as an instrument variable to resolve endogeneity issue. Endogeneity comes from the correlation between the explanatory variables and the error term (at time t for individual i). So my rationale is that, a lagged term would hence have nothing to do with the error term at time t.

I also see that this method might be potentially hampered by serial correlation. I wonder if anyone could provide any insight on this issue.

Thank you!

Felix

[George Ford]

People do it all the time.

But,

https://marcfbellemare.com/wordpress/wp-content/uploads/2019/05/WangBellemareLaggedIVsMay2019.pdf