Will Jordan
@will-jordan.bsky.social
Dem pollster with GSG. Bakersfield native. @williamjordann on Twitter
It’s fair that voters who shifted hardest right in '24 -- e.g. low-propensity Hispanic voters -- probably turned out at lower rates in NJ/VA, so the “persuasion” angle is obscured.
But this Echelon Insights chart makes a key point: data shows Trump right now is *weaker* with low-propensity voters
But this Echelon Insights chart makes a key point: data shows Trump right now is *weaker* with low-propensity voters
November 25, 2025 at 4:15 AM
It’s fair that voters who shifted hardest right in '24 -- e.g. low-propensity Hispanic voters -- probably turned out at lower rates in NJ/VA, so the “persuasion” angle is obscured.
But this Echelon Insights chart makes a key point: data shows Trump right now is *weaker* with low-propensity voters
But this Echelon Insights chart makes a key point: data shows Trump right now is *weaker* with low-propensity voters
Pre-election prediction markets gave Sherrill 5% odds of a margin exceeding 14 points. The odds of this didn't change Sep-Nov despite Trump's approval rating dropping 5 points during that period.
November 18, 2025 at 8:09 PM
Pre-election prediction markets gave Sherrill 5% odds of a margin exceeding 14 points. The odds of this didn't change Sep-Nov despite Trump's approval rating dropping 5 points during that period.
I don't think "passive news consumption" is a new or a social media phenomenon. But the growing role of recommendation engines means everyone is getting more of what they "seem to want" so passive/active may be come a more politically salient difference than it ever has been.
October 27, 2025 at 6:20 PM
I don't think "passive news consumption" is a new or a social media phenomenon. But the growing role of recommendation engines means everyone is getting more of what they "seem to want" so passive/active may be come a more politically salient difference than it ever has been.
Something somewhat counter-intuitive is "passive news consumer" is not a very distinctive type of social media user--tho there are notable exceptions (less X/Reddit, more TikTok). But they get less news online probably due to a mix of who they follow & what algos think they want.
October 27, 2025 at 6:20 PM
Something somewhat counter-intuitive is "passive news consumer" is not a very distinctive type of social media user--tho there are notable exceptions (less X/Reddit, more TikTok). But they get less news online probably due to a mix of who they follow & what algos think they want.
These news consumption habits are closely related to political engagement, but not identical. "Do you seek out the news" means something slightly different than "is politics important to you."
October 27, 2025 at 6:20 PM
These news consumption habits are closely related to political engagement, but not identical. "Do you seek out the news" means something slightly different than "is politics important to you."
I continue to love this simple survey question as a way to divide the public. It cuts across partisanship -- Rs may hate "the media" but they have engaged/attentive voters on favored platforms (Fox/X) just like Ds.
October 27, 2025 at 6:20 PM
I continue to love this simple survey question as a way to divide the public. It cuts across partisanship -- Rs may hate "the media" but they have engaged/attentive voters on favored platforms (Fox/X) just like Ds.
House battleground very different in 2026 vs 2018
'18
Rs defend 65 seats @ Trump+10/bluer
Ds defend 30 @ Clinton+10/redder
'26
Rs defend 30 seats @ Trump+10/bluer
Ds defend 62 @ Harris+10/redder
BUT
'18
Ds need 24 (37% of R defensive seats)
'26
Ds need 3 (10% of R defensive seats)
'18
Rs defend 65 seats @ Trump+10/bluer
Ds defend 30 @ Clinton+10/redder
'26
Rs defend 30 seats @ Trump+10/bluer
Ds defend 62 @ Harris+10/redder
BUT
'18
Ds need 24 (37% of R defensive seats)
'26
Ds need 3 (10% of R defensive seats)
August 7, 2025 at 2:34 AM
House battleground very different in 2026 vs 2018
'18
Rs defend 65 seats @ Trump+10/bluer
Ds defend 30 @ Clinton+10/redder
'26
Rs defend 30 seats @ Trump+10/bluer
Ds defend 62 @ Harris+10/redder
BUT
'18
Ds need 24 (37% of R defensive seats)
'26
Ds need 3 (10% of R defensive seats)
'18
Rs defend 65 seats @ Trump+10/bluer
Ds defend 30 @ Clinton+10/redder
'26
Rs defend 30 seats @ Trump+10/bluer
Ds defend 62 @ Harris+10/redder
BUT
'18
Ds need 24 (37% of R defensive seats)
'26
Ds need 3 (10% of R defensive seats)
This is a great question to ask. As someone left of center, if you think the answer is the second one, think hard about what needs to change or accept you believe there is basically no chance of Democrats controlling the Senate and passing real laws or nominating judges for the next decade or so.
July 29, 2025 at 2:25 PM
This is a great question to ask. As someone left of center, if you think the answer is the second one, think hard about what needs to change or accept you believe there is basically no chance of Democrats controlling the Senate and passing real laws or nominating judges for the next decade or so.
This new Navigator tracking survey has Democrats taking the lead over Trump in H2H on taxes, inflation, and the economy for the first time in I’m not sure how long (a while). Trump led on all 3 in late February.
May 3, 2025 at 1:53 PM
This new Navigator tracking survey has Democrats taking the lead over Trump in H2H on taxes, inflation, and the economy for the first time in I’m not sure how long (a while). Trump led on all 3 in late February.
This debate about opt-in vs opt-out/news engagement is not just about reaching difficult to reach audiences. The “news comes to me”/opt-out voters care most about different issues, may respond to diff message/branding. Less about concept of “democracy, foreign policy—more about taxes & health care.
April 30, 2025 at 2:22 PM
This debate about opt-in vs opt-out/news engagement is not just about reaching difficult to reach audiences. The “news comes to me”/opt-out voters care most about different issues, may respond to diff message/branding. Less about concept of “democracy, foreign policy—more about taxes & health care.
Dems have to contend w/the “engagement divide” too. We found the *only* major demo/political group that trusted “the media” was solid Dems. Disengaged voters are deeply distrustful. Related factors. But professional Ds (like me) must view world less like top bar in this chart, more like the bottom.
April 30, 2025 at 2:20 PM
Dems have to contend w/the “engagement divide” too. We found the *only* major demo/political group that trusted “the media” was solid Dems. Disengaged voters are deeply distrustful. Related factors. But professional Ds (like me) must view world less like top bar in this chart, more like the bottom.
But disengaged voters, which are mostly on the margins of Trump’s coalition, are the ones to watch now. See what we found below: DOGE/Musk/Econ policies earn much less support from new, less engaged Trump voters. These issues likely become a wedge for Trump coalition.
April 30, 2025 at 2:17 PM
But disengaged voters, which are mostly on the margins of Trump’s coalition, are the ones to watch now. See what we found below: DOGE/Musk/Econ policies earn much less support from new, less engaged Trump voters. These issues likely become a wedge for Trump coalition.
Trump won most of the disengaged voters in 2024. The sort of voter who thought Obama was cool but didn’t love “politics” (or politicians). This made Trump’s otherwise narrow victory unusually disruptive of politics as usual. Political engagement seems closely related to news engagement (no surprise)
April 30, 2025 at 2:14 PM
Trump won most of the disengaged voters in 2024. The sort of voter who thought Obama was cool but didn’t love “politics” (or politicians). This made Trump’s otherwise narrow victory unusually disruptive of politics as usual. Political engagement seems closely related to news engagement (no surprise)
Rather than follow the lines of politics, news engagement cuts across them. And importantly it does a lot to divide the voters we (and they) need to solidify and persuade from those who are bought in. This divide has probably always existed, but media trends have likely deepened it. From our study:
April 30, 2025 at 2:11 PM
Rather than follow the lines of politics, news engagement cuts across them. And importantly it does a lot to divide the voters we (and they) need to solidify and persuade from those who are bought in. This divide has probably always existed, but media trends have likely deepened it. From our study:
What’s different about this question from many Qs in politics, is the *lack* of polarization—by party or by race. This dimension, which we can call “news engagement,” cuts across other dimensions, and tells us something your party & demographics alone cannot. Makes a potentially powerful tool.
April 30, 2025 at 2:07 PM
What’s different about this question from many Qs in politics, is the *lack* of polarization—by party or by race. This dimension, which we can call “news engagement,” cuts across other dimensions, and tells us something your party & demographics alone cannot. Makes a potentially powerful tool.
We at GSG discussed this recently in a report reviewing media consumption trends, which I’ll link below. Here’s a question we’ve started asking more often: When it comes to news about current events and politics, do you … seek it out, or does news come to you?
What do you notice?
What do you notice?
April 30, 2025 at 2:04 PM
We at GSG discussed this recently in a report reviewing media consumption trends, which I’ll link below. Here’s a question we’ve started asking more often: When it comes to news about current events and politics, do you … seek it out, or does news come to you?
What do you notice?
What do you notice?
Enjoyed this piece from @robflaherty.bsky.social Recently we have been studying a similar concept to the “opt in” and “opt out” voters in our polling, and I think it’s incredibly useful for understanding how current trends in political behavior and media habits are related…
April 30, 2025 at 2:03 PM
Enjoyed this piece from @robflaherty.bsky.social Recently we have been studying a similar concept to the “opt in” and “opt out” voters in our polling, and I think it’s incredibly useful for understanding how current trends in political behavior and media habits are related…
A striking set of results from Navigator Research. A lot of the time simple language tweaks make little difference, but here, voters are much more comfortable agreeing with a position once it criticizes the status quo.
April 25, 2025 at 1:35 PM
A striking set of results from Navigator Research. A lot of the time simple language tweaks make little difference, but here, voters are much more comfortable agreeing with a position once it criticizes the status quo.
To hold the power to pass major legislation over next decade, Democrats need to run the sort of candidates and be the sort of party that can win/defend Senate seats in every swing state and win seats in Trump+10/+15 states. It’s about necessity as much as opportunity.
April 8, 2025 at 3:04 PM
To hold the power to pass major legislation over next decade, Democrats need to run the sort of candidates and be the sort of party that can win/defend Senate seats in every swing state and win seats in Trump+10/+15 states. It’s about necessity as much as opportunity.
Really seems British voters like to pretend to pollsters they live in a democracy with a proportional representation system until there’s an election and then, eventually, they voter for a party that can win in their constituency. Epistemologically difficult to read the mood.
April 8, 2025 at 10:53 AM
Really seems British voters like to pretend to pollsters they live in a democracy with a proportional representation system until there’s an election and then, eventually, they voter for a party that can win in their constituency. Epistemologically difficult to read the mood.
At the regional/county level, the 2023->2025 swing is very similar to the 2020->2024 swing.
(Notice Crawford and Harris both did relatively well in suburban WOW)
(Notice Crawford and Harris both did relatively well in suburban WOW)
April 2, 2025 at 3:02 PM
At the regional/county level, the 2023->2025 swing is very similar to the 2020->2024 swing.
(Notice Crawford and Harris both did relatively well in suburban WOW)
(Notice Crawford and Harris both did relatively well in suburban WOW)
Similarly, at the county level, there is no real correlation between Wisconsin counties with higher turnout and better or worse performance by the liberal candidate (Crawford)
April 2, 2025 at 1:56 PM
Similarly, at the county level, there is no real correlation between Wisconsin counties with higher turnout and better or worse performance by the liberal candidate (Crawford)
So far there is not a clear relationship between special election turnout and Democratic over performance.
If anything higher salience/turnout specials like FL-01/06 have been better for Democrats than the average (average swing is +11%D so far in 2025)
If anything higher salience/turnout specials like FL-01/06 have been better for Democrats than the average (average swing is +11%D so far in 2025)
April 2, 2025 at 1:33 PM
So far there is not a clear relationship between special election turnout and Democratic over performance.
If anything higher salience/turnout specials like FL-01/06 have been better for Democrats than the average (average swing is +11%D so far in 2025)
If anything higher salience/turnout specials like FL-01/06 have been better for Democrats than the average (average swing is +11%D so far in 2025)
While voter ID passed in Wisconsin the “yes” vote (62.7%) was narrowly under the lowest support ever recorded by Wisconsin’s premier pollster, suggesting some mix of:
-very lib electorate
-real support lower than polls
-thermostatic backlash
-voters do love 2 vote “no”
-very lib electorate
-real support lower than polls
-thermostatic backlash
-voters do love 2 vote “no”
April 2, 2025 at 12:48 PM
While voter ID passed in Wisconsin the “yes” vote (62.7%) was narrowly under the lowest support ever recorded by Wisconsin’s premier pollster, suggesting some mix of:
-very lib electorate
-real support lower than polls
-thermostatic backlash
-voters do love 2 vote “no”
-very lib electorate
-real support lower than polls
-thermostatic backlash
-voters do love 2 vote “no”
The Musk triple threat:
- universal name ID/salience
- infuriates liberal base, but not equally adored by conservative base
- backlash with marginal Republican voters
- universal name ID/salience
- infuriates liberal base, but not equally adored by conservative base
- backlash with marginal Republican voters
April 2, 2025 at 12:47 PM
The Musk triple threat:
- universal name ID/salience
- infuriates liberal base, but not equally adored by conservative base
- backlash with marginal Republican voters
- universal name ID/salience
- infuriates liberal base, but not equally adored by conservative base
- backlash with marginal Republican voters