2003 PMSA and Re: merging

>
Dav,
Thanks for your response. I was running into about 20+ variables, that according to SPSS which references exact name, were not common between the 1998 and 2002 Metro AHS datasets. I can try the merge again and write down all of the "unpaired" variables, but off the top of my head it seemed to be several of householder and interviewee demographic variables.
 
On another issue...for the 2003 National set, I'm looking at PMSA as recorded through the SMSA variable (label = 1980 design PMSA code). Over 36,000 of the total 71,000 are listed as "9999". Is this because these areas were not considered part of any PMSA in 1980 - or is this a disclosure issue? Is there another variable that would provide a more complete locational variable, other than Census region?
 
Thanks again,
Kristen Crossney
PhD Candidate, Urban Planning and Policy Development
Rutgers University
Crossney@eden.rutgers.edu