I am trying to aggregate data at the msa level using the IPUMS 2000 5% sample and census msa figures. However, there is a lot of variation in the codes used. There are also some MSAs in the IPUMS 2000 5% that do not appear in the 2000 Census delineations, and vice-versa. One example is Corvallis, OR MSA (fips code 1890), which is in the Census 2000 delineations but not in the IPUMS 5% 2000. I am wondering if there is anything available to help with this translation.
The METAREA codes in IPUMS-USA are actually FIPS codes, however, they represent the coding and naming structure used in the 1990 OMB definitions with minor adjustments to reflect hierarchical relationships among metro areas across time. For a full discussion of METAREA codes and labels, so the METAREA Variable Description and Comparability Statement. MSA’s that are not represented in the data are likely not identifiable due to population restrictions, such as in your example of Corvallis, OR which did not have a sufficiently large population to be identified in the Public Use Data. To see how the original MSA codes for the 2000 samples nest within the METAREA codes you can see the METAREA Translation Table (I recommend reading the Introduction to Translation Tables to better understand how these files are constructed).
I hope this helps.
Not an answer but a follow-up question. I understand that some metarea codes are on the 1990 OMB list (see attached) but not in the IPUMS sample due to confidentiality reasons. However, how are there metarea codes in IPUMS that are *not* on the 1990 OMB list? For example, Akron OH (metarea = 80) is in the IPUMS sample but not in any OMB documentation that I can find. Any idea what’s going on here? Thanks!
In the text file you linked to, Akron, OH can be found listed under PMSA FIPS code 0080.
I hope this helps.
Thanks! What about METAREAs that are neither in the 1990 OMB documentation now the METAREA translation tables? Any idea where these are coming from? A few are listed below.
NAME (METAREA)
Bend, OH (900)
Bowling Green, KY (1130)
Coeur d’Alene, ID (1700)
El Centro, CA (2300)
Farmington, NM (2540)
Harrisonburg, VA (3260)
Kingston, NY (3830)
Las Vegas-Paradise, NM (4130)
Madera, CA (4700)
Michigan City-La Porte (5020)
Monroe, MI (5220)
Ocean City, NJ (5840)
Prescott, AZ (6470)
Salisbury, MD (7130)
Valdosta, GA (8700)
Vero Beach, FL (8740)
The METAREA values that you list are available in the IPUMS-CPS data, but they are not available in the IPUMS-USA data. Since the Translation Tables linked to above pertain to the IPUMS-USA version of the METAREA variable, those metro areas cannot be found in the Translation Table. As well, those metro areas were not part of the 1990 OMB delineation, which is why they are not included in the OMB file. You can, however, find these metro areas in more recent versions of the OMB delineation file, e.g. Dec 2009.
Hope this helps.