[EB] Problems Matching respondents of separate EB-files with cumulative file

[Otto Hermkens]

Hello Fellow Eurobarometer researchers,

I am a student of sociology presently employed at the Social and Cultural
Planning Office in Rijswijk, Holland. I am working on my thesis, and using
Eurobarometer datasets (EB's). 
My research focuses on trends in environmental concern. Therefore I am
interested in using several EB's. I am trying to match environmental 
concern variables from separate EB-files (EB 5,6,10,20,28, 31a,35,36,40)
to the cumulative EB-file (1970-1992), so as not to have to harmonize 
all demographic variables myself, but profit from earlier labor by the
constructors of the cumulative file. 
The problem is that I am meeting serious matching problems:

(1) Duplicate respondent numbers in various files: eb 5,6 and 20. In EB 20
this problem can be solved by matching on several other variables as well,
so it looks like different respondents with the same respondent number
for eb20, but for eb 5 and 6 the respondents with the same respondent
numbers seem to be complete duplicates (at least on all variables in the
cumulative file), suggesting old-fashioned weighting by card duplication
(a suggestion by Nero Matthias of the Steinmetz-archive).  

(2) Different respondent numbers between the separate EB-files and the
cumulative file: this seems to go for EB's 20,28,31a, and for eb 36 for
greece only. For all Eurobaromters, separate files have the same number of
respondents as the cumulative file.

My question to you is the following:
Does any of you know of the above mentioned problems, and of a solution?
I would be very pleased to hear from you.

Thank you very much for your attention. 

Best Regards,

Otto Hermkens