PA, FL, WI & MI Vote by Mail Modeled Projections

Today, Citizen Data is releasing the results of new vote by mail models in Pennsylvania, Florida, Wisconsin, and Michigan based on vote history, current ballot requests, and turnout likelihood.

Across the board, our team found projected levels of vote by mail at rates higher than seen in 2016.

Key findings include:

  • Voting by mail will be much more prevalent in all 4 states relative to their recent history, with half of voters in Florida voting by mail and a majority doing so in Michigan.
  • While both parties will vote by mail at increased rates, Democrats will vote by mail at much higher rates. Conversely, Republicans will vote in person at much higher rates.
  • Urban areas will see higher volumes of turnout, as well as much higher vote by mail rates. 
  • Levels of Covid-19 concern and trust in the USPS correlate with vote method.

Here are some state-specific highlights:


  • Pennsylvania is expected to have 33.6% voting by mail, a dramatic increase from the less than 4% who did so in 2016, with the Philadelphia metro area likely to see the highest rates of mail voting. 
  • Democrats in Pennsylvania are likely to vote by mail at more than 2x the rate of Republicans.
    • Mail: 68% Democrat, 27% Republican, 5% Independent
    • In-person: 39% Democrat, 51% Republican, and 10% Independent


  • Florida voters are expected to cast their ballots 50% vote-by-mail, either by mailing back or dropping off, and 50% in person. This represents a dramatic shift from the 27% who voted by mail in 2016.
  • A plurality of mail-in voters will be Democrats and in-person voters Republicans.
    • Mail: 49% Democrat, 32% Republican, 19% Independent
    • In-person: 30.6% Democrat, 46.6% Republican, and 22.7% Independent


  • Wisconsin is expected to have 43% by mail voters and 57% in person.
  • The state does not collect party ID; however, per survey data, we project voters will breakdown roughly:
    • Mail: 63% Democrat, 29% Republican, 6% Independent
    • In-person: 24% Democrat, 63% Republican, and 10% Independent


  • Michigan is expected to have 62% voting by mail and drop-off. This represents a dramatic shift from the 25% who voted by mail in 2016.
  • Democrats are likely to vote by mail at more than 2x the rate of Republicans.
    • Mail: 68.7% Democrat, 24.6% Republican, 6.7% Independent
    • In-person: 27.5% Democrat, 52.4% Republican, and 20% Independent

In preparation to produce these models, Citizen Data surveyed 30,000 likely voters nationally between September 3-8 via IVR/P2P on their vote method intentions and to assess contributing factors to their likelihood to vote by mail for the 2020 general election. You can view results of the national survey here.

The survey data was then matched to Citizen’s voter file of all registered voters, weighted, and stratified. Then, leveraging the survey results, Citizen also modeled the likelihood that each registered voter would vote by mail or in person using 2016 and 2018 turnout as separate variables and then combining scores. Using our live feed of absentee ballot requests, we trained an additional set of machine learning algorithms to predict the likelihood that those individuals who have not yet requested a ballot would do so. For the remainder of registered active voters who had not yet requested a ballot, we assumed that individuals would request a ballot if their ballot request likelihood as predicted by the model was above 50%.

The insights from these models are part of an ongoing Citizen Data effort to support election administrators and non-partisan non-profits with the credible, unbiased data they need for strategic resource allocation in the months and weeks leading up to the election. On an ongoing basis, Citizen will release updates to these state models as well as new projections in additional states and other key insights and analyses.

For this project, Citizen is partnering with a number of organizations — including the National Vote at Home Institute and the Stanford-MIT Project on a Healthy Election.