Respected Scholars, I am running a panel analysis and there is a shock in my FDI data. However, the reviewers asked me to check for the existence of structural break and try to remove it if exist. But as far I know no structural break for panel analysis. Any suggestion please?
Dear Musibau Hammed
In my opinion structural breaks should not be a matter of concern if you are using a micro panel with short time series (e.g.,5 years). Normally, Chow test is used for structural breaks in time series, but Gretl software allows chow test for panel data also. However, Bai (2006, 2010) have outlined a method for common breaks in panel data. Consider the attached articles and the following Stata forums for further guidance.
https://www.statalist.org/forums/forum/general-stata-discussion/general/1460428-structural-break-in-panel-data
https://www.statalist.org/forums/forum/general-stata-discussion/general/1292309-structural-break-analysis-with-panel-data
Regards
Still on the above discussion (Detection and removal of structural break in panel model), I found that zandrews may be applied to a single time series within a panel. Then I followed this technique to detect structural break in my series. We have following outcomes.
zandrews FDI if country ==12, break(trend) graph
Zivot-Andrews unit root test for FDI
Allowing for break in trend
Lag selection via TTest: lags of D.FDI included = 0
Minimum t-statistic -2.476 at 2011 (obs 450)
Critical values: 1%: -4.93 5%: -4.42 10%: -4.11
We found a structural break in 2011 base on the above result and its graph. Please, what is the next thing to do? I mean how to remove the structural break in this model? Thanks.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running