I would like confirmation of my notion that the most recent reliable source of comparable data that would contain variables relating to ethnic origin, visible minority status and labour force status is the 2006 census.
I know that the data from the NHS might be usable at the national or provincial levels (which is what the user wants), but that it wouldn't be comparable to earlier data from the census. I don't see that the LFS contains ethnicity or vismin variables, even in the master files. Have I missed anything?
I consulted a few subject matter divisions, and it doesn't appear that you have missed anything.
However, I had also considered SLID, and while the PUMF does not have the variable information you are looking for, the author division confirmed that the master file does.
The SLID microdata file does contain these variables as well as various labour-related variables. The client can consult the SLID Electronic Data Dictionary: <http://www.statcan.gc.ca/pub/75f0026x/2013000/t03200-eng.htm#etho1c15> for a complete list. The client would have to access the data through the Research Data Center or request a custom table on a cost-recovery basis.