This week, Wisconsin held its primary election as planned despite burgeoning Covid-19 infections and a state-wide stay-at-home order, because of Supreme Court rulings denying postponing in-person and absentee voting.
While we’ll have to wait for official data to see the exact impact of Wisconsin’s decision on turnout and public health, we can assume that the impact will be sizable.
But assumptions aren’t helpful when making tough decisions. We at Data 2 the People prefer to use data. And when the perfect data is not available, we go to the next best data that is available. This is why we were so excited to see the results of recent surveys from 1Q. The surveys aimed to assess public behavior and sentiment in the context of Covid-19, and included questions on Americans’ preferred election format in this unprecedented time.
We dug into this data and share our findings in the interest of preserving the health and safety of both our democracy and its constituents. Our hope is that public officials across the country can use this data to help inform equitable decisions that optimize for both public health and election integrity.
Some notes before we share the results:
This data reflects snapshots in time. Surveys reflect the opinions, knowledge, and trends at the time they are conducted. The Covid-19 situation is evolving rapidly, and sentiment expressed in these survey may likewise change with time. The survey we primarily reference was conducted on April 8th, and we use data from a survey on March 22nd to look at trends.
Respondents are only a sample of the US population. The April 8th survey does a great job of census-balancing the demographics of the respondents. Still, the sample size is small so there is uncertainty in how the results generalize.
What people say is not necessarily what they will do. This is always the case, but is especially important to keep in mind during a period of time that is especially stressful and emotional for many.
Here is what the data shows:
1. Respondents expressed low likelihood of voting in person.
When asked their likelihood of voting in person at a voting location if their state were holding elections next week, only 27% of respondents said they would be “likely” to vote. The majority of respondents - 55% - indicated they would not be comfortable voting in person.
2. Respondents were generally happy with mail and online voting options.
57% of respondents expressed likelihood to vote by mail and 72% expressed likelihood to vote through a secure online link.
Notably, vote by mail would make voting more accessible for about a third of respondents. 34% of respondents are “likely” to vote by mail where they were “unlikely” to vote in person.
Some states already have 100% vote by mail - Colorado, Hawaii, Oregon, Washington and Utah. We are unaware, however, of any state using or considering online voting methods in this primary season for the general population (some states have considered online methods for absentee or military voting), so the online option is purely hypothetical. (But interesting for the future!)
3. The interest in voting by mail appears consistent across demographics.
We did not see meaningful differences in likelihood to vote through different channels when looking across gender, race, education level, and income. This suggests that encouraging vote by mail options could be an equitable option during this time when in-person voting presents a health risk to many.
4. Interest in voting by mail does differ by political party, but this may be due to recent politicization of the topic.
Results from the most recent survey conducted on April 8th indicate that Republicans are less interested in voting by mail than Democrats, though the groups have similar rates of aversion to voting in person. This is a change from the survey on March 22nd, however, when Democrats had a slightly stronger aversion to vote by mail.
On April 8th, 22% of Republicans reported they were unlikely to vote by mail in the next week, while only 13% of Democrats reported such. This difference was significant, with only a 4.3% likelihood of chance. This trend was also reflected in the “incremental” mail-in voter population - the group unlikely to vote in person, but likely to vote by mail. 39% of Democrats are “incremental” mail-in voters, vs. 35% of Republicans.
What is particularly interesting is that results differ from a similar survey conducted on March 22nd which showed a dramatically less polarized environment. At that time, Democrats actually had a stronger aversion to voting by mail than Republicans. It’s possible that the partisan dialog and news coverage surrounding this topic in the past few weeks swayed responses. We’ll note that the March 22nd poll surveyed a different population (and a smaller population), but we believe the trend is worth noting.
We'd like to emphasize the universal benefit mail-in voting provides. Roughly a third of Democrats and Republicans report that they are unlikely to vote in person, but would vote by mail. As a reference point, only 23% of Americans cast a mail-in vote in 2016. (Only 17.7% were absentee ballots.) By supporting mail in options during this pandemic, we will include large legions of voters on both sides of the aisle.
We hope this data will help states decide quickly what changes to make to voting, so they can begin the required operations and communication processes as soon as possible.
As always, our goal is to help current and future public leaders leverage the best data available to make the best possible decisions. For further details and analysis, please see the appendix file, here. If you have any questions or would like to know more about our analysis, please reach out to us at data2thepeople.org.
For further information on how you can use mobile surveys for rapid data collection, please reach out to 1Q.com. Many thanks to 1Q for sharing their survey data.