Testing joint significance of fixed effects with ppmlhdfe: testparm after ppmlhdfe, absorb

[Noemi Seng]

Dear community,

after doing my ppmlhdfe regression with year fixed effects I would like to test whether the year FE are jointly significantly different from zero. I did the regression using the following command:

 local gravity_sectorlevel lngdp_o lngdp_d lndistw lnsumgdp comcol col45 comlang_off lnsmp_dest
ppmlhdfe TotalassetsthUSD_green `gravity_sectorlevel', absorb(year country_origin_sector_encode country_dest_sector_encode) d cluster(country_pair_encode)

The command testparm i.year (which I know from xtreg) doesn't work here, as the absorb option works without factor notation. Is there any alternative command available?

Thank you in advance!

Best,
Noemi

[Andrew Musau]

ppmlhdfe is from SSC, as you are asked to explain (FAQ Advice #12). Use -i.year- in place of -absorb(year)-.

ppmlhdfe ... i.year, absorb(country_origin_sector_encode country_dest_sector_encode) ...
testparm i.year