Longitudinal Comparisons Start Year 5-year ACS

Hi: My research team is seeking input on comparing 5-year ACS data for aggregates of census tracts. Our question is whether or not we use a baseline survey year to compare to all subsequent non-overlapping 5-year survey OR compare contiguous surveys? Here are the choices we are mulling over:

Example A: 2006-2010 as baseline year, such as 2006-2010 comparing to 2011-2015, but also to the other subsequent surveys that do not overlap with 2006-2010 such as the most recent 2012-2016 survey.


Example B: comparing contiguous years, such as 2007-2011 to 2012-2016 to 2017-2021

Does anyone know a good reason why one of these options would be preferred over the other?

Thank you!