Why don't MIGPUMA1 codes in the crosswalk file agree with frequency distributions?

I’m using the 3-year ACS 2009-2011 to do analysis of migration. I’ve selected New York and done a frequency of MIGPUMA1. The crosswalk file shows values for MIGPUMA1 from ‘001’ to ‘043’ which makes sense and agrees with other PUMA references. The problem is that my frequency distribution shows additional MIGPUMA1 codes of ‘044’ to ‘053’ and then additional values (not consecutive) up to ‘082’. Why are additional codes in the file and how do I reference where they are?

The variable MIGPUMA1 represents the PUMA from which respondents moved. Therefore, the MIGPUMA1 codes 001-043 correspond to MIGPLAC1==36 for New York. The MIGPUMA1 codes beyond this range represent PUMAs of Migration from other states. In fact, because MIGPUMA1 codes are shared between states, you have to use this variable with MIGPLAC1 in order to accurately represent where, within the U.S. a person moved from. If you are interested in people who moved to New York, you would keep everyone from STATEFIP==36, New York (using PUMA to determine their current location to the smallest geographic boundary) and then use MIGPLAC1 and MIGPUMA1 together to determine their PUMA of origin.

I hope this helps.