[OC] Non-participation rates consistent across safe and competitive states, red or blue (2020 election)
September 19, 2024
[OC] Non-participation rates consistent across safe and competitive states, red or blue (2020 election)
Posted by ptrdo
12 comments
[OC] According to U.S. Census surveys, the main reasons people do not vote include disinterest, dissatisfaction with the candidates, being too busy, illness, or disability. However, one might expect that eligible voters in a competitive state with a significant impact on the national outcome would be more motivated to participate. Despite this, non-participation rates in the 2020 election were generally consistent, regardless of whether states were safe or competitive, red or blue.
Data was aggregated in Excel, charted in R ggplot, and then finessed in Adobe Illustrator.
That said, any visualization that directly compares South Dakota, New York and California margins of victories as percentages is definitely skewing the presentation of the underlying data and creating a false equivalency.
While the effect isn’t large, there is a pretty clear 5% point difference between blue states and red states, especially at the extremes. It would be much better to present this on an X-Y scatter plot rather than bar graphs. Any correlation would be much easier to visualize there.
I’m not sure “margin of winner” is correct, it’s “vote share”. Unless in West Yorkshire trump got ~66% of the votes.
Non-participation is a form of voting… at least for people who are sick of both parties and have a shred of self respect.
Is it Thursday already? I wonder if we should allow politics all week long next month
Should’ve made it increasing then decreasing for it to look Gaussian
I see this sort of take all the time, but my mind always goes back to sample size. Someone correct me if they want, because I’m going to throw out some random numbers here:
Polls take sample sizes of like 1000-2000 people and then apply that to the entire 300M+ population of the US. They gives something like a 3-5% accuracy with 95% confidence. Well, if we are taking a sample size of 50-60% of the voting population, shouldn’t that give us like 1% accuracy with 99.9% confidence? Just seems like when people say “if just 1000 more democrats voted, Hillary would have won this state!” or “if just 1000 more republicans would have voted, Trump could have won that state!” But the reality is, the next 1000 votes would have probably been VERY similar to the previous 1 million votes, therefore it probably wouldn’t have closed the gap.
Either way, that’s what I always think when I see these things, but not sure if it’s the right mindset.
This is the only reason Texas is a red state. Over 40% of the eligible population doesn’t vote.
I mean… just looking they seem higher in red states
It’s a weird way to show that…
I’d say maybe do a % of voting eligible population that voted for each – and a darker area to show the margin.
The way it is now, the data is mixed – a margin up top and a percent below.
And what I’d be SUPER interested in seeing… The % of people who didn’t vote, what their likely preference would have been based, I suppose, purely on demographics.
Texas is easily in play if more eligible Democrats vote.
Meanwhile, I am not voting this year because North Carolina voter ID laws mean it would take far too much time to register.
12 comments
[OC] According to U.S. Census surveys, the main reasons people do not vote include disinterest, dissatisfaction with the candidates, being too busy, illness, or disability. However, one might expect that eligible voters in a competitive state with a significant impact on the national outcome would be more motivated to participate. Despite this, non-participation rates in the 2020 election were generally consistent, regardless of whether states were safe or competitive, red or blue.
Data was aggregated in Excel, charted in R ggplot, and then finessed in Adobe Illustrator.
Election Results
[https://www.fec.gov/introduction-campaign-finance/election-results-and-voting-information/federal-elections-2020/](https://www.fec.gov/introduction-campaign-finance/election-results-and-voting-information/federal-elections-2020/)
General Election Turnout Rates
[https://election.lab.ufl.edu/dataset/2020-general-election-turnout-rates/](https://election.lab.ufl.edu/dataset/2020-general-election-turnout-rates/)
Census Survey Data
[https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-585.html](https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-585.html)
Non-participation is a problem.
That said, any visualization that directly compares South Dakota, New York and California margins of victories as percentages is definitely skewing the presentation of the underlying data and creating a false equivalency.
While the effect isn’t large, there is a pretty clear 5% point difference between blue states and red states, especially at the extremes. It would be much better to present this on an X-Y scatter plot rather than bar graphs. Any correlation would be much easier to visualize there.
I’m not sure “margin of winner” is correct, it’s “vote share”. Unless in West Yorkshire trump got ~66% of the votes.
Non-participation is a form of voting… at least for people who are sick of both parties and have a shred of self respect.
Is it Thursday already? I wonder if we should allow politics all week long next month
Should’ve made it increasing then decreasing for it to look Gaussian
I see this sort of take all the time, but my mind always goes back to sample size. Someone correct me if they want, because I’m going to throw out some random numbers here:
Polls take sample sizes of like 1000-2000 people and then apply that to the entire 300M+ population of the US. They gives something like a 3-5% accuracy with 95% confidence. Well, if we are taking a sample size of 50-60% of the voting population, shouldn’t that give us like 1% accuracy with 99.9% confidence? Just seems like when people say “if just 1000 more democrats voted, Hillary would have won this state!” or “if just 1000 more republicans would have voted, Trump could have won that state!” But the reality is, the next 1000 votes would have probably been VERY similar to the previous 1 million votes, therefore it probably wouldn’t have closed the gap.
Either way, that’s what I always think when I see these things, but not sure if it’s the right mindset.
This is the only reason Texas is a red state. Over 40% of the eligible population doesn’t vote.
I mean… just looking they seem higher in red states
It’s a weird way to show that…
I’d say maybe do a % of voting eligible population that voted for each – and a darker area to show the margin.
The way it is now, the data is mixed – a margin up top and a percent below.
And what I’d be SUPER interested in seeing… The % of people who didn’t vote, what their likely preference would have been based, I suppose, purely on demographics.
Texas is easily in play if more eligible Democrats vote.
Meanwhile, I am not voting this year because North Carolina voter ID laws mean it would take far too much time to register.