Hello, I am doing a 3sls regression on int remittance and poverty among developing countries. So, following previous literature, I created regional dummies to control for the fixed effects. However, after regressing, it shows that one of the dummies has been omitted due to collinearity.
1)What should I do in such situation?
2)If I have to change the specification of the model, which one should I use? or Is it okay if I don't use the regional dummies at all?
I have attached the regression table below.

------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
1 |
lremit | -.1175261 .0166862 -7.04 0.000 -.1502305 -.0848217
lgdp | -1.391645 .0430366 -32.34 0.000 -1.475995 -1.307295
lgini | 3.730252 .2090265 17.85 0.000 3.320568 4.139937
dum1 | .1253087 .1321924 0.95 0.343 -.1337836 .384401
dum2 | -.1683417 .1319286 -1.28 0.202 -.426917 .0902335
dum3 | -.3516937 .1528049 -2.30 0.021 -.6511858 -.0522016
dum4 | -.4744162 .1442135 -3.29 0.001 -.7570694 -.1917629
dum5 | 0 (omitted)
_cons | .328668 .8032379 0.41 0.682 -1.245649 1.902985
-------------+----------------------------------------------------------------
2 |
lpov1 | .0330895 .0105028 3.15 0.002 .0125043 .0536747
lremit_L1 | .9526584 .0067366 141.42 0.000 .939455 .9658618
ltrade | .1192078 .0241448 4.94 0.000 .0718848 .1665309
lFD | -.0146556 .021529 -0.68 0.496 -.0568517 .0275405
_cons | -.4492467 .1362539 -3.30 0.001 -.7162994 -.182194
------------------------------------------------------------------------------
Endogenous variables: lpov1 lremit
Exogenous variables: lgdp lgini dum1 dum2 dum3 dum4 dum5 lremit_L1 ltrade
lFD
------------------------------------------------------------------------------