Indiana: According to technical documentation compiled by the source of our numerator data (the Catalist voter file), the number of voters in this aggregated voter file for Indiana is smaller than the total number of votes cast as documented by the office of the Indiana Secretary of State. The variance is -3.72% (i.e., Catalist file contains almost 86,000 fewer votes cast than the official count). In order to adjust for this variance, we increased our voter counts by 3.7%, and our estimated turnout reflects this adjusted votes cast.
Massachusetts: According to technical documentation compiled by the source of our numerator data (the Catalist voter file), the number of voters in this aggregated voter file for Massachusetts is smaller than the total number of votes cast as documented by the office of the Massachusetts Secretary of State, due to incomplete data provided by some townships. The variance is -6.68% (i.e., Catalist file contains almost 184,000 fewer votes cast than the official count). In order to adjust for this variance, we increased our voter counts by 6.68%, and our estimated turnout reflects this adjusted votes cast.
Utah: According to technical documentation compiled by the source of our numerator data (the Catalist voter file), the number of voters in this aggregated voter file for Utah is smaller than the total number of votes cast according to the office of the Utah Secretary of State due to a change in state law which provides voters the option to make their information unavailable to those accessing voting information from a state. The variance is -10.01% (i.e., the Catalist file contains almost 108,000 fewer votes cast than the official count). In order to adjust for this variance, we increased our voter counts by 10.01%, and our estimated turnout reflects this adjusted votes cast. However, because the option to keep information from voter registration private was on the voter registration form starting with the 2018 election, it is reasonable to believe that new voters, many of whom would be young voters, are overrepresented among those who chose to keep their information private and therefore underrepresented by the voter file.
