Analyzing the Behavior of GDP Across Countries Around Crisis Periods

[hari venkatesh]

How can I create a graph to display the average behavior of GDP across countries around crisis periods?

I have a panel dataset with cross-country data, including a crisis variable (0 and 1) and GDP growth.

I want to show how the average GDP behaves during pre-crisis, crisis, and post-crisis periods (-3, -2, -1, 0, 1, 2, 3). H

Help me to create the STATA code for this graph?

The example dataset attached.

Thank you.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int year str18 country byte crisis float loggdp
2010 "Argentina"          1 1.4875462
2011 "Argentina"          0 1.4915873
2012 "Argentina"          1 1.4933988
2013 "Argentina"          1 1.4953402
2014 "Argentina"          1 1.4971235
2015 "Argentina"          0 1.4933127
2016 "Argentina"          0  1.494806
2017 "Argentina"          0  1.485393
2018 "Argentina"          1 1.4798278
2010 "Austria"            0  1.499721
2011 "Austria"            0 1.4997633
2012 "Austria"            0 1.4997697
2013 "Austria"            1  1.499822
2014 "Austria"            0 1.4999533
2015 "Austria"            0 1.4999458
2016 "Austria"            0 1.4999447
2017 "Austria"            0 1.4999433
2018 "Austria"            0 1.4999447
2010 "Brazil"             0 1.4243143
2011 "Brazil"             0 1.4236982
2012 "Brazil"             0 1.4281896
2013 "Brazil"             1 1.4218334
2014 "Brazil"             0 1.4249254
2015 "Brazil"             1 1.4411718
2016 "Brazil"             0  1.463491
2017 "Brazil"             0 1.4621078
2018 "Brazil"             0  1.459924
2010 "Dominican Republic" 1 1.4371136
2011 "Dominican Republic" 0 1.4373026
2012 "Dominican Republic" 0 1.4593494
2013 "Dominican Republic" 0 1.4698063
2014 "Dominican Republic" 0  1.472264
2015 "Dominican Republic" 0  1.477959
2016 "Dominican Republic" 1 1.4796697
2017 "Dominican Republic" 0  1.479243
2018 "Dominican Republic" 0 1.4438057
2010 "Ethiopia"           0      1.48
2011 "Ethiopia"           0      .785
2012 "Ethiopia"           0      1.08
2013 "Ethiopia"           0      .715
2014 "Ethiopia"           1      1.08
2015 "Ethiopia"           0      1.28
2016 "Ethiopia"           0       .78
2017 "Ethiopia"           0      1.35
2018 "Ethiopia"           0      35.5
2010 "Gambia"             1 1.4736544
2011 "Gambia"             0 1.4903766
2012 "Gambia"             1 1.4713308
2013 "Gambia"             0 1.4311203
2014 "Gambia"             0  1.421347
2015 "Gambia"             1 1.3998796
2016 "Gambia"             0 1.1941377
2017 "Gambia"             1  1.219466
2018 "Gambia"             0  1.208385
2010 "Kenya"              0  1.412585
2011 "Kenya"              0  1.408473
2012 "Kenya"              0  1.443087
2013 "Kenya"              0  1.447311
2014 "Kenya"              1 1.4678593
2015 "Kenya"              1 1.4575927
2016 "Kenya"              0  1.414596
2017 "Kenya"              0  1.425446
2018 "Kenya"              0  1.437257
2010 "Kyrgyz Republic"    0  1.477767
2011 "Kyrgyz Republic"    0 1.4195933
2012 "Kyrgyz Republic"    0  1.411935
2013 "Kyrgyz Republic"    1 1.4326514
2014 "Kyrgyz Republic"    0 1.3464007
2015 "Kyrgyz Republic"    0 1.3087792
2016 "Kyrgyz Republic"    0  1.444839
2017 "Kyrgyz Republic"    0 1.4339454
2018 "Kyrgyz Republic"    1  1.437184
2010 "Lebanon"            0 1.4883842
2011 "Lebanon"            0 1.4666675
2012 "Lebanon"            0  1.470216
2013 "Lebanon"            0 1.4735326
2014 "Lebanon"            1 1.4736426
2015 "Lebanon"            1 1.4731468
2016 "Lebanon"            0 1.4751496
2017 "Lebanon"            0 1.4806542
2018 "Lebanon"            0 1.4959004
2010 "Liberia"            1  1.492647
2011 "Liberia"            0  1.492153
2012 "Liberia"            0 1.4928873
2013 "Liberia"            0  1.492431
2014 "Liberia"            1 1.4945573
2015 "Liberia"            0  1.492199
2016 "Liberia"            0  1.496289
2017 "Liberia"            1   1.49454
2018 "Liberia"            0  1.494272
end

[Andrew Musau]

The crises years do not coincide across countries. So aggregating the data is therefore not straightforward without having to make some probably unreasonable assumptions. You can have a graph matrix depicting each country separately.

* Example generated by -dataex-. For more info, type help dataex
clear
input int year str18 country byte crisis float loggdp
2010 "Argentina"          1 1.4875462
2011 "Argentina"          0 1.4915873
2012 "Argentina"          1 1.4933988
2013 "Argentina"          1 1.4953402
2014 "Argentina"          1 1.4971235
2015 "Argentina"          0 1.4933127
2016 "Argentina"          0  1.494806
2017 "Argentina"          0  1.485393
2018 "Argentina"          1 1.4798278
2010 "Austria"            0  1.499721
2011 "Austria"            0 1.4997633
2012 "Austria"            0 1.4997697
2013 "Austria"            1  1.499822
2014 "Austria"            0 1.4999533
2015 "Austria"            0 1.4999458
2016 "Austria"            0 1.4999447
2017 "Austria"            0 1.4999433
2018 "Austria"            0 1.4999447
2010 "Brazil"             0 1.4243143
2011 "Brazil"             0 1.4236982
2012 "Brazil"             0 1.4281896
2013 "Brazil"             1 1.4218334
2014 "Brazil"             0 1.4249254
2015 "Brazil"             1 1.4411718
2016 "Brazil"             0  1.463491
2017 "Brazil"             0 1.4621078
2018 "Brazil"             0  1.459924
2010 "Dominican Republic" 1 1.4371136
2011 "Dominican Republic" 0 1.4373026
2012 "Dominican Republic" 0 1.4593494
2013 "Dominican Republic" 0 1.4698063
2014 "Dominican Republic" 0  1.472264
2015 "Dominican Republic" 0  1.477959
2016 "Dominican Republic" 1 1.4796697
2017 "Dominican Republic" 0  1.479243
2018 "Dominican Republic" 0 1.4438057
2010 "Ethiopia"           0      1.48
2011 "Ethiopia"           0      .785
2012 "Ethiopia"           0      1.08
2013 "Ethiopia"           0      .715
2014 "Ethiopia"           1      1.08
2015 "Ethiopia"           0      1.28
2016 "Ethiopia"           0       .78
2017 "Ethiopia"           0      1.35
2018 "Ethiopia"           0      35.5
2010 "Gambia"             1 1.4736544
2011 "Gambia"             0 1.4903766
2012 "Gambia"             1 1.4713308
2013 "Gambia"             0 1.4311203
2014 "Gambia"             0  1.421347
2015 "Gambia"             1 1.3998796
2016 "Gambia"             0 1.1941377
2017 "Gambia"             1  1.219466
2018 "Gambia"             0  1.208385
2010 "Kenya"              0  1.412585
2011 "Kenya"              0  1.408473
2012 "Kenya"              0  1.443087
2013 "Kenya"              0  1.447311
2014 "Kenya"              1 1.4678593
2015 "Kenya"              1 1.4575927
2016 "Kenya"              0  1.414596
2017 "Kenya"              0  1.425446
2018 "Kenya"              0  1.437257
2010 "Kyrgyz Republic"    0  1.477767
2011 "Kyrgyz Republic"    0 1.4195933
2012 "Kyrgyz Republic"    0  1.411935
2013 "Kyrgyz Republic"    1 1.4326514
2014 "Kyrgyz Republic"    0 1.3464007
2015 "Kyrgyz Republic"    0 1.3087792
2016 "Kyrgyz Republic"    0  1.444839
2017 "Kyrgyz Republic"    0 1.4339454
2018 "Kyrgyz Republic"    1  1.437184
2010 "Lebanon"            0 1.4883842
2011 "Lebanon"            0 1.4666675
2012 "Lebanon"            0  1.470216
2013 "Lebanon"            0 1.4735326
2014 "Lebanon"            1 1.4736426
2015 "Lebanon"            1 1.4731468
2016 "Lebanon"            0 1.4751496
2017 "Lebanon"            0 1.4806542
2018 "Lebanon"            0 1.4959004
2010 "Liberia"            1  1.492647
2011 "Liberia"            0  1.492153
2012 "Liberia"            0 1.4928873
2013 "Liberia"            0  1.492431
2014 "Liberia"            1 1.4945573
2015 "Liberia"            0  1.492199
2016 "Liberia"            0  1.496289
2017 "Liberia"            1   1.49454
2018 "Liberia"            0  1.494272
end

encode country, gen(country2)
gen year2= year-2000
bys country2: egen max= max(loggdp)
xtset country2 year2
xtline loggdp, byopts(note("") yrescale ixaxes) addplot(dropline max year2 if crisis, msy(none)) ///
plotregion(margin(zero)) xtitle("") leg(order(2 "Crisis"))

[hari venkatesh]

Dear Andrew,

Thank you for your response and advice.

Instead, I would like to illustrate how GDP behaves around crisis periods such pre-crisis, crisis, and post-crisis periods (-4, -3, -2, -1, 0, 1, 2, 3, 4)

Please find the attached example for your reference.

I would greatly appreciate your assistance in creating the code for this graph.

Thank You.

[Andrew Musau]

I already gave my reasons in #2 why I think that this won't be possible with the data that you have. Others having an interest in the question and a different opinion will continue the thread.