Do you have any suggestions for how to use YRIMMIG variable to determine exact year of immigration?
Are there any imputation techniques recommended by IPUMS? The codebook notes comparability issues across survey years and suggests using last year in interval as conservative estimate.
There is no way to determine the exact year of immigration in the public use CPS data. As you note, the comparability documentation suggests using the provided code as it is the latest possible year for each range as a conservative estimate. You may wish to compare your results using the latest possible year with a midpoint year for each range and see how it changes your results, or consult the literature on this topic for other ideas.
Thank you for letting me know! I am already planning to use provided code as conservative estimate and midpoint year for each range too. I will continue to look for alternative methods in the literature too in the meantime.