I am attempting to determine voter registration rates by race within specific precincts in my county. The local Supervisor of Elections' Office provides data online of the voter registration rates by race within each precinct for the 2012 General Election. I'm then comparing this data to the overall population over 18 years of age (qualified voters) within each precinct using 2010 US Census Data.
My question is: how do I account for the discrepancies created by the two-year gap between the data sets? Within that two-year period, hundreds of people could have turned 18 and be qualified to vote, left the precinct, passed away, etc. I'm trying to collect this data for a statistical report I'm compiling, and am unsure how to handle this issue? Can I model projected population rates using existing data, or should I just note the significant margin of error?