singelton problem with panel data

[Nancy Barakat]

HI all
I am studying the effect of firm-level adoption MM ( a binary variable ) on its energy intensity using panel datasets from WBES for three countries (unbalanced panel data). I have run fixed effect model and now I am running fixed-effect 2SLS model. I tried to use xtivreg2 but it run on only 32 observations out of 680 due to dropped singelton observations. so, should I shift to ivreghdfe or ivreg2, and if so should I cluster the standard error by firm_id or should I resort to two-way clustering by including industry type or region of establishment. Actually, I am confused about the singelton issue, am I wrong for defining my data as panel or can I resort to ivreghdfe and consider the absorb option of multi fixed effects.

Here is the code that I have tried so far:

 ivreghdfe EE20152  finance_measure R_D    human_capital labprod2015 exporter  foreign_ownership ln_age  firm_sizeWBES industry_type   i.year  (MM = new_var_2 ) [pweight=average_wstrict  ], robust cluster (unique_id_01  ) endog (MM)

ivreg2 EE20152 finance_measure R_D  human_capital labprod2015 exporter  foreign_ownership ln_age  firm_sizeWBES industry_type   i.year (MM = new_var_2 ) [pweight=average_wstrict  ], endog(MM) robust

Thanks in advance

[FernandoRios]

HI Nancy
This is related to your other question about RIF. You do not have panel data. Your variable unique_id_01 is unique to almost every observation in your data. To say your data is panel, you need to have at least 2 observations per Panel id
F

[Nancy Barakat]

Thanks Fernando for your reply
Actually I have doubted that but when I tab panel year, there are considerable number of firms which are tracked over same periods and fixed effect model, using xtreg, fe, worked perfect on whole sample.
here is the tab results by country:
For uganda

tab panel year

Panel: Firm |
interviewed in these | Year of survey
years | 2006 2013 | Total
----------------------+----------------------+----------
2013 only | 0 553 | 553
2006 only | 354 0 | 354
2006 and 2013 | 209 209 | 418
----------------------+----------------------+----------
Total | 563 762 | 1,325

For Kenya


Panel: Firm |
interviewed in | Year of survey
these years | 2007 2013 2018 | Total
--------------------+---------------------------------+----------
2007 only | 506 0 0 | 506
2013 only | 0 397 0 | 397
2018 only | 0 0 708 | 708
2007 and 2013 only | 91 91 0 | 182
2013 and 2018 only | 0 233 233 | 466
2007, 2013 and 2018 | 60 60 60 | 180
--------------------+---------------------------------+----------
Total | 657 781 1,001 | 2,439


For Tanzania


Panel: Firm |
interviewed in these | Year of survey
years | 2006 2013 | Total
----------------------+----------------------+----------
2013 only | 0 698 | 698
2006 only | 304 0 | 304
2006 and 2013 | 115 115 | 230
----------------------+----------------------+----------
Total | 419 813 | 1,232

Thanks a lot in advance