in the industry classifications tables provided on the website, 00000000 codes are indicated to be N/A - less than 16 years old, unemployed, etc. But in the 2002 to 2004 ACS samples (IPUMS-USA), some of these observations have IND codes corresponding to various industries. does it mean that IND codes have a larger coverage?
more generally, I’d be interested in finding correspondence tables between the IND classification used in the ACS samples and commonly used classifications, such as ISIC or SITC codes. Any help in this would be much appreciated.
I created an extract for the 2002-2004 ACS samples including both IND and INDNAICS. As far as I could tell, all cases coded as “00000000” for INDNAICS were coded as 0000 for IND. If you’re still seeing otherwise, could you please send us a few serial numbers of problematic cases?
I’m not sure which precise mapping you are hoping to create, but there are several resources for corespondence tables. The two URLs below will provide a good start:
http://www.census.gov/people/io/metho…
http://www.census.gov/eos/www/naics/c…
Other tables can often be found by performing a Google search. Sometimes an exact crosswalk is not available and will have to be constructed from two or more other available crosswalks.
You are right. I don’t know what happened, I must have made some coding mistake yesterday. Sorry about this.
I am trying to convert these codes into the GTAP classification of industries, and I have a mapping of ISIC codes into GTAP codes. It seems like I will have to use the NAICS codes in the ACS data, convert these into the ISIC classification, then GTAp. The links you provided to crosswalks seem to confirm that this would be the only way, as there is apparently no direct crosswlak from the census IND sectors to any classication other than the NAICS.
Thank you for your help and apologies again for that mistake.