How to do the unitroot test for the panel data with gaps?

[Jacky Wei]

I want to use the "xtunitroot" to do the unitroot test for the panel data, but I find that the command can't handle the panel data with gaps.Do you have any other options? Thanks for your reply!

[Justin Niakamal]

Take a look at page 8 in the pdf entry for xtunitroot. There are a couple of tests that allow unbalanced panels.

https://www.stata.com/manuals/xtxtunitroot.pdf

[Jacky Wei]

Take a look at page 8 in the pdf entry for xtunitroot. There are a couple of tests that allow unbalanced panels.

https://www.stata.com/manuals/xtxtunitroot.pdf

Thanks a lot , but "unbalanced panels" and "panels with gaps" do not have the same meaning. I've read the pdf you're talking about before and that's why I know xtunitroot can't handle the panel data with gaps.

[Justin Niakamal]

xtunitroot fisher does not require strongly balanced data, and the individual series can have gaps.

webuse pennxrate, clear
drop in 17/19
drop in 84/86
xtset id year, yearly


xtunitroot fisher lnrxrate, dfuller lags(3) drift


 xtset id year, yearly
       panel variable:  id (unbalanced)
        time variable:  year, 1970 to 2003, but with gaps
                delta:  1 year

.
.
.  xtunitroot fisher lnrxrate, dfuller lags(3) drift

Fisher-type unit-root test for lnrxrate
Based on augmented Dickey-Fuller tests
---------------------------------------
Ho: All panels contain unit roots           Number of panels       =    151
Ha: At least one panel is stationary        Avg. number of periods =  33.96

AR parameter: Panel-specific                Asymptotics: T -> Infinity
Panel means:  Included
Time trend:   Not included
Drift term:   Included                      ADF regressions: 3 lags
------------------------------------------------------------------------------
                                  Statistic      p-value
------------------------------------------------------------------------------
 Inverse chi-squared(302)  P       917.0841       0.0000
 Inverse normal            Z       -18.8585       0.0000
 Inverse logit t(759)      L*      -19.5772       0.0000
 Modified inv. chi-squared Pm       25.0274       0.0000
------------------------------------------------------------------------------
 P statistic requires number of panels to be finite.
 Other statistics are suitable for finite or infinite number of panels.

[Scott Merryman]

The Fisher test can accommodate gaps.

[Jacky Wei]

xtunitroot fisher does not require strongly balanced data, and the individual series can have gaps.

webuse pennxrate, clear
drop in 17/19
drop in 84/86
xtset id year, yearly


xtunitroot fisher lnrxrate, dfuller lags(3) drift


xtset id year, yearly
panel variable: id (unbalanced)
time variable: year, 1970 to 2003, but with gaps
delta: 1 year

.
.
. xtunitroot fisher lnrxrate, dfuller lags(3) drift

Fisher-type unit-root test for lnrxrate
Based on augmented Dickey-Fuller tests
---------------------------------------
Ho: All panels contain unit roots Number of panels = 151
Ha: At least one panel is stationary Avg. number of periods = 33.96

AR parameter: Panel-specific Asymptotics: T -> Infinity
Panel means: Included
Time trend: Not included
Drift term: Included ADF regressions: 3 lags
------------------------------------------------------------------------------
Statistic p-value
------------------------------------------------------------------------------
Inverse chi-squared(302) P 917.0841 0.0000
Inverse normal Z -18.8585 0.0000
Inverse logit t(759) L* -19.5772 0.0000
Modified inv. chi-squared Pm 25.0274 0.0000
------------------------------------------------------------------------------
P statistic requires number of panels to be finite.
Other statistics are suitable for finite or infinite number of panels.

I guess I didn't read it carefully enough and focus on IPS.
Thank you very much!

[Jacky Wei]

The Fisher test can accommodate gaps.

Thank you!