ACS Data Users Group

Is it appropriate to compare data from the 1990 long-form (STF3) to data from the 2015 5-year ACS? For instance, one can find data on numbers of persons below the poverty level from both the long-form and the ACS. While these are both survey-based sources, is it OK to compare totals from these two?

There are many ACS data tables that should be comparable with the 1990 Census long form tables. The Census Bureau has a webpage on "Comparing 2015 American Community Survey Data" that provides guidance on comparing ACS data with Census 2000 data. You could use this as a starting point for comparing the latest ACS 5-year estimates with 1990 data.