The problem of "likely" voters
As we mentioned we’d seen some recent research around how we model the likelihood to vote. TL/DR: it’s quite hard, and broadly speaking, the political polling establishment hasn’t quite figured it out. Partly it’s a reporting issue, but it’s also an issue of our own discourse of voter/non-voter.
here is the post: The Problem of “Likely” Voters
———————————————————-
Here are a few things we think you should know about likely voters based on the paper’s findings. If you want more details, check out our post here explaining these points further.
There is no standard definition (yet) of likely voters that all pollsters use. Based on the definition and approach, there’s at least a 7 point spread between how much a model under or over reports likely voters.
It’s probably better to think about voters as frequent and infrequent, and that takes the form of a “churn rate” of 1/4 (people who vote one election, but not the next and vice versa).
Over-reporting on voter turnout happens most by individuals who appear to experience strong social pressure to vote and are well enfranchised (education, income, longevity within a community, participation in church and public affairs). One key finding for us has been the performances of shame and guilt in our culture of voting, and this is an example of what it does.
Our Take:
Infrequent, unlikely, and uncertain voters are not that well understood, and current research is starting to identify the scope of the issue but not its shape. Methodologies need to be further developed to understand the nuance of likelihood to vote.
There’s a large question to be addressed in understanding intent, and how we engage with intent to vote in our popular discourse. Are current GOTV (get out the vote) approaches really tackling the right question?
Could “intent to vote” serve as an indicator for civic health itself? Rather than being a screener for a separate set of partisan questions, how could we use and understand intention to vote as a way to understand an individual’s civic engagement.
Here’s the Github of the research paper and its data and research to date
This is the question text for the ANES voter turnout analysis
Pew Research’s backgrounder on the Perry-Gallup scale for measuring voter intention
Potential hiatus next week! But we’ll be back - I’m just without internet for a week.
An early draft working paper explores how we might understand the impact of the Women’s March, and the political value of protests.
Research on civic engagement by non-citizens - some lit review and some initial findings.
Some results and analysis from our design probes around local community and engagement.
Thank you to all those who have already dropped a line and shared your thoughts. And thanks in advance to sharing this with anyone else you think might be interested.
If you think of feedback, suggestions, or people we should meet, please get in touch! You can email us at: 1stpersonprojects@thedifferenceengine.co