I am hoping to aggregate IPUMS data to link metropolitan characteristics. My understanding is that the data is able to be aggregated and representative at the state level through the application of weights. I’m unclear about lower levels of geography though.
Reference:
Parks, Virginia. 2012. “The Uneven Geography of Racial and Ethnic Wage Inequality: Specifying Local Labor Market Effects.” Annals of the Association of American Geographers 102(3): 700-725.
Link to this article:
Yes, you are still able to aggregate representative statistics to lower level geographies (such as MSAs) with IPUMS USA data. Of course, you’ll need to use the correct sample weight variables - typically PERWT for person-level analysis and HHWT for household-level analysis. Also note, that we don’t generally expect estimate from the PUMS to exactly match published figures. They will be close, and likely within some margin of error, but rarely exact.
Important note: not all MSAs are identified by IPUMS geographic variables _and_ even where MSAs are identified, there are some mismatches between IPUMS MSA codes and the actual official MSA delineations. The description for the MET2013 variable explains the situation in detail. I recommend reading it before aggregating microdata by MSA.