Dear all,
I'm trying to estimate certain exporter-time and importer-time specific characteristics in the gravity model. I closely follow Heid ei al. (2021), "Estimating the effects of nondiscriminatory trade policies within structural gravity models", which identifies unilateral and nondiscriminatory trade policies in the presence of importer and exporter fixed effects by using intra-national trade.
here's my command
and it's what Heid et al. (2021) do
where "timetoexport" is exporter-time specific and "lsavg_x_inter" is importer-time specific.
when running my data and code, i always get warning:
note: import_ci omitted because of collinearity over lhs>0 (creates possible existence issue)
but in Heid et al. (2021)'s data and code, everything is fine. i check the data, the only difference is some non-zero values.
what's the problem? it almost drives me crazy.
Thank you so much.
I'm trying to estimate certain exporter-time and importer-time specific characteristics in the gravity model. I closely follow Heid ei al. (2021), "Estimating the effects of nondiscriminatory trade policies within structural gravity models", which identifies unilateral and nondiscriminatory trade policies in the presence of importer and exporter fixed effects by using intra-national trade.
here's my command
Code:
ppml_panel_sg trade_value export_ci import_ci, exporter(exporter_code_2digit) importer(importer_code_2digit) year(year) cluster(exporter_code_2digit importer_code_2digit)
Code:
ppml_panel_sg trade_rescaled lsavg_x_inter timetoexport, exporter(exporteriso316613) importer(importeriso316613) year(year) cluster(importeriso316613 exporteriso316613)
when running my data and code, i always get warning:
note: import_ci omitted because of collinearity over lhs>0 (creates possible existence issue)
but in Heid et al. (2021)'s data and code, everything is fine. i check the data, the only difference is some non-zero values.
what's the problem? it almost drives me crazy.
Thank you so much.
Comment