What the polls show post-Biden withdrawal
Harris makes significant cuts into Trump's lead, but some warning signs emerge among key demographic groups
We now have several polls since Joe Biden announced his withdrawal from his re-election campaign, and Kamala Harris assumed the mantle of the presumptive Democratic nominee. A few data points in these surveys stood out …
1. Trump’s lead has narrowed.
We now have several publically released surveys since Biden’s announcement to step down that 1) show how Harris is now performing versus Trump and 2) have results from before Biden’s announcement that provide a point of comparison. The new surveys show …
Trump maintains a +2 or +1 point lead nationally in a two-way face-off against Trump (without other candidates).
Since Harris assumed the mantle of the Democratic nominee, Trump’s advantage over the Democratic candidate has declined by -3 or -4 points.
Trump has more or less stayed where he was when Biden was his opponent … or declined slightly by a point. Most of the movement has been from Harris gaining support.
2. Harris maintains Democratic support
The recent NY Times/Sienna and CNN polls show that Harris has maintained her support among self-identified Democrats or increased it slightly by a few points. The WSJ survey paints a different picture, with Harris rallying support +8 points among Democrats.
Bottom line: In either case, the picture is remarkably similar across these surveys: Harris has support from 92% to 94% of Democrats.
3. Non-whites have rallied behind Harris
NY/Times Sienna, CNN, and WSJ/Harris all show Harris gaining ground among non-white voters.
NY/Times Sinna shows a +13 increase in support among Hispanic voters compared to Biden’s previous polling. That’s a substantial increase.
CNN shows a +8 increase in “people of color,” and WSJ/Harris shows a +12 increase in support among non-white voters.
Bottom line: Harris is looking more like a traditional Democratic presidential candidate in per performance with non-white voters. However, exit polls in 2020 showed Biden capturing 65% of the Hispanic vote … which Harris appears to still be behind.
4. Harris has a surge in youth support … but conflicting data on older voters
The latest NY Times/Sienna poll shows dramatic changes for Harris among voters under 65, including…
Harris has rocketed from Biden’s +3 advantage among voters 18-29 to a +21 point advantage over Trump.
She’s reversed Democrat’s fortunes among 30 to 44-year-olds, going from -6 to +4.
She’s narrowed Trump’s advantage among 45 to 64-year-olds from -18 to -6.
It’s in the oldest age category, where the picture becomes more muddled. Here, NYT/Sienna shows Harris losing Biden’s +3 advantage to trailing Trump by -5 points.
The CNN survey uses different age categories. But it aligns with the NY Times/Sienna poll in that it shows …
Harris now performs better among the youngest voters.
Harris has pushed middle-aged voters into her column.
Harris is trailing among voters over 65+.
However, the CNN poll shows Harris losing ground on voters 50 to 64 compared to Biden, while NY Times/Sinna shows Harris narrowing the gap considerably. CNN also shows Harris improving slightly among voters 65 and older compared to Biden, while NY Times Sienna shows a slight decrease.
Bottom line: How Harris performs with older voters will be a deciding factor in the election. Voters 45-64 comprised 38% of the electorate in 2020, and voters over 65 comprised 22%. Whether Harris is gaining or losing momentum among these voters is a key data point … but we have conflicting data right now.
5. The battle over suburban voters rages on
We only have one recent data point by living density, which comes from the NY Times/Sienna survey. It shows Harris increasing her advantage among suburban voters to +5 over Trump, where Biden only had a +1 advantage before he stepped aside.
Bottom line: With suburban voters in crucial swing states set to be one of the deciding groups in the election, Trump’s slide among suburban voters should worry his team.
6. Whites without a college degree remain unchanged
White voters without a college degree show little to no shifts in opinion since Biden’s withdrawal. And while no one is expecting the Democrats to win this group … the margin by how much Harris loses will play a determining factor in her ability to carry key swing states. After all, these voters comprise over 50% of the electorate in key swing states such as Pennsylvania, Ohio, Wisconsin, and Michigan.
In the latest NY Times/Sienna Poll, 29% of whites without a college degree support Harris … virtually unchanged from the 28% who supported Biden previously.
In the CNN poll, 30% of whites without a college support Harris ... a slight increase of +3 from the 27% who previously supported Biden.
Bottom line: While it does not appear Harris has lost ground among whites without a college degree compared to where Biden was polling prior to his decision to step aside, she remains behind where exit polls placed Biden’s performance among white working-class voters in 2020 (around 33%). Harris’s ability to minimize her losses among this group in key midwestern states will be a deciding factor in the election.
Closing notes
These initial polls show that Harris has managed to narrow the election considerably and staunch the bleeding following Biden’s debate performance. She’s generating a huge amount of enthusiasm from traditionally Democratic-leaning groups. But her ability to make inroads in key swing demographic groups will be the deciding factor in this election. In this respect, she’s not where she needs to be.
A note of caution
Many of these surveys are of registered voters, and not all registered voters will vote. And the surveys conducted among likely voters are difficult even in this stage of the election: it’s hard to predict who will vote and who will not. And on top of it all, the assumed margin of error for many of these demographic groups is large and subject to swings based on sampling error that may not reflect any changes to underlying attitudes. That’s why I tend to focus on the direction and movement of the surveys.
An introductory note on this substack
This is my first post on Substack. I plan on posting as regularly as I can, offering insights on polling and similar data. I’ve been a polling commentator for a variety of publications over the years before starting my own polling company. If you have any suggestions or comments, I would love to hear them. Most of the posts will be available to everyone. Eventually, the goal is to use support from founding members and supports to conduct custom surveys. And if you have any polling needs, feel free to check us out here.