The 2012 5-year ACS dataset corresponds to two different sets of PUMA Codes and definitions (one from 2008-2011 and another from 2012). If I wanted to compare variables from the 2000 Census to the 2012 5-year ACS by PUMA, can I do anything within the 2012 5-year ACS dataset to make valid comparisons?
Since PUMAs are the smallest geographical unit available in both the 2000 Census PUMS and the 2012 5-year ACS PUMS the only way to make valid comparisons by 2000 Census PUMAs would be to only consider PUMAs that did not change their boundaries (note that they still may have changed their codes even if they did not change their boundaries). The MABLE/Geocorr12 website allows you to compare two geographies including PUMA(2000) and PUMA(“2012”). You can select multiple states using Shift-Click or Ctrl-Click. The resulting table gives the allocation factor of one geographic unit within another (a 1 indicating that the source geographic unit is completely contained within the target geographic unit). There is also an option to generate a 2nd allocation factor (AFACT2) representing the portion of the target geocodes in the source geocodes. If both allocation factors are 1 than the PUMA did not change. If the PUMAs did change, to create a constant measure of a geography, you would need to combine PUMAs from both 2000 and 2012 until the allocation factors are both 1.
Since the only year in the 2012 5-year ACS file that does not use 2000 Census based PUMAs is 2012 I would recommend using the 2011 5-year ACS file if this is acceptable for your research. The 2011 5-year ACS contains the exact same information for 2008-2011, with the only difference being a lack of 2012 and the addition of 2007.
I hope this helps.
This is very helpful. Thank you for the rapid response, and for pointing out the MABLE resource as well!