I’m creating a dataset that combines the ACS, Census, and CPS samples for certain years. Can I simply append one on top of the other and assume that the industry and occupation codes are the same within years (ignoring year-to-year coding changes–asuming they’re applied identically to all datasets)? In other words, can I assume that an occ code of, say, 6 in 1987 is identical between all datasets?
Yes, within years the occupation and industry codes follow the same coding scheme. It is important to note, as you do above, that this only holds within years, as codes change quite a bit over the years.