Some weekend projects start with a simple question, this past weekend was no different.

“How much churn is there in Florida’s voter file?”

We wanted to get a sense of how much churn was in the voter file leading up to the 2020 election. We took the voterfile from Jan 2019 and compared it to the voterfile in December of 2019. (We went county by county using the R library CRAN compareDF) For the period of 2019, we wanted to explore how many records changed, how many were added, and how many were removed.

In the aggregate, it would appear there isn’t much churn or turnover in the voterfile. The number of registered voters grew about 2% for the year. However, that masks the story.

On average, 7% of the voters had changes in their data (this can be anything from switching parties, address changes, having the precinct number change, etc), 7% were additions, and 5% were removed during that time.

In 2020, Florida added 979,146 voters and removed 741,312 voters.

But my take away is this – remember there are differences in macro- behavior and micro-behavior – and macro level changes maybe masking much larger micro-shifts.

Below are the county breakdowns, some of the counties that exhibit high percentages of change were shifting / changing precinct numbers.

[table id=13 /]