Can a lack of data for a variable lead to 'omitted because of collinearity'?

Hi All,

Thank you in advance for reading.

I am currently running a fixed effects model for my dissertation, studying GDP per capita growth and a number of explanatory variables. My sample is 25 transition countries over a period of 20 years, 1990-2010. As with a study of this kind. a number of countries / variables have data gaps but I was told, for the most part, this is not an issue.

I am focusing on how each decade is different in terms of its economic growth determinants and so am running two separate OLS regressions for each decade. This means included all of the country and time dummies in my regression. The issue in question comes in the results where 2 of my countries are 'omitted because of collinearity'.

So for example I type (excuse the capital notation):

reg GDP FDI GOVT CAPITAL CAPITALLAG1 EDU INFLATION INTERNET LFPR EBRDSSP EBRDPL EBRDCP Armenia Azerbaijan Belarus Bulgaria Croatia Czech Estonia Georgia Hungary Kazakh Kyrgyz Latvia Lithuania Macedonia Moldova Poland Romania Russian Serbia Slovak Slovenia Tajikstan Ukraine Uzbekistan DUM1991 DUM1992 DUM1993 DUM1994 DUM1995 DUM1996 DUM1997 DUM1998 DUM1999 DUM2000 DUM2001 DUM2002 DUM2003 DUM2004 DUM2005 DUM2006 DUM2007 DUM2008 DUM2009 DUM2010 if YEAR>1999

My reference categories for both dummy categories are Albania and Dum1990 respectively.

I have looked into what could be the reasoning and I'm finding it quite hard to apply explanations to my example. If I looked in my original data set, two variables for both the omitted countries completely lack data and so what I'm asking is, could this be the reason?

Thanks again,

Jamie