To specifically demonstrate the problem I looked at males with college degree working 35+ hours per week and 50+ weeks per year in the 2015 data. At age 23, I find 71 observations with Census data and their Table Creator Tool. With IPUMS I find 64. Moreover, there appear to be 2 observations in the IPUMS data for which there is no analogue in the Census data. I am concerned that the microdata is not consistent. Any ideas on what could be driving this?
Comparing 2015 ASEC CPS IPUMS with Census data and Table Creator, I note differing record counts for FTYR earners
It is most likely that this discrepancy is driven by slightly different filters and/or universe statements defining the observations in the two tables you are comparing. The issue is unlikely to be due to differences in the underlying data, since both tables use the same data source. If you could send additional details, to email@example.com, about how you are specifying your tables (screenshots are very helpful), then I can take a closer look into the causes of this discrepancy.