Dramatic jump in some IND1990 in 2016


Hi, I downloaded IPUMS-CPS for 1980’s, 1990’s and 2000’s and limited the cases to five industries (IND1990 = 421, 432, 762, 770, 901, which are air transport, transport services, hotels, lodging, gov) and noticed a significant jump in 2016 of the number of observation in each of these sub-industries. For example, air transport has 322 for 2015 and similar number prior, but 3055 for 2016 and after. Could you please explain?



From what you have described, it sounds like your analysis may inadvertently include additional sample months after 2015. Just using ASEC samples, you should find the following unweighted counts for IND1990=421

2015: 322
2016: 314
2017: 337
2018: 336

I would recommend that you revise the extract you are working with and select “change” next to the samples you have selected. From there, if you only wish to work with the ASEC samples, click on the Basic Monthly and Supplement Topics tabs to ensure you do not have additional samples selected.