A colleague pulled data in January 2013 from the 2009-2011 multi-year ACS sample that included HHINCOME. I pulled that same data this week, and found that, although the sample is the same in both data sets (same number of records and all serial numbers match), the value in the HHINCOME field is different for the 2010 and 2011 portions of the sample. It is higher in the newer file, and each record is higher by about the same factor. Did the inflation factor change, or was there some other change?
As noted in the Errata and Revisions note for the April 8th data release, under Corrections, the adjustment factor was corrected for the 2011 and 2010 cases in the 2011 multi-year files. This would effect HHINCOME. To get the corrected data for your extract you can either resubmit your extract (from your Download or Revise Extracts page), or merge the new values onto your existing data using the variables YEAR, DATANUM, SERIAL, and PERNUM to uniquely identify each record.
I hope this helps.