Marquette pollster Charles Franklin on Kamala Harris’ lead in Wisconsin, and more

Greetings and welcome to the Reconbagulation Area podcast.

My name is Dan Schaeffer, thank you for joining us.

The Marquette University Law School of Poll is the gold standard poll for elections

and public policy in the state of Wisconsin and the latest poll was released just yesterday.

Joining us to discuss the results as he does after every poll in this election year is Charles

Franklin, the executive director of the poll, Charles.

Thank you so much for joining us and welcome to the Reconbagulation Area.

Thanks for having me. It's always fun to be here.

All right, so obviously there's going to be a lot of attention on this type of poll at this time

of year. We are 53 days now away from election day here in the state of Wisconsin, one of really

the seven states that will decide the outcome of the presidential election as a swing state

in the Electoral College. Let's start at the top of the ticket, the highlight item of the poll,

Kamala Harris with a four point lead over Donald Trump with both registered and likely voters.

What stood out to you within that head-to-head matchup between the two candidates for president?

Yeah, there's sort of two things. One is that's a little bit higher than a month ago before the DNC,

but after Biden dropped out, we had Harris down by one 50 to 49 with registered voters,

but up by one 50 to 49 with likely voters. And when you included the other third party candidates

in the race, she was up by two points. So take your pick down one up one or up two.

Looks like a very slim Harris lead a month ago. Now at four points with both registered and likely,

that's still inside the polls margin of error. So one shouldn't make too much of that,

but it does look like maybe a two or possibly a three point gain compared to a month ago.

If you just kind of average those results from a month ago,

polling averages for the state with other people's polls included have been showing Wisconsin

at about a three point Harris advantage. So we're at one point above that, but we were about a point

a point or two below it a month ago. So some gains for her, you mentioned those polling averages.

You know, I watch those two and I think it's really interesting to see that Wisconsin has

probably been the strongest swing state for Democrats of those polls. And I think that's

interesting because Wisconsin has been to the right of Michigan and Pennsylvania in the last few

election cycles, but the polling right now shows Wisconsin, you know, a tick to the left of

those other states. What do you make of that? Yeah. And I find that surprising too. And we

could toss out Nevada in that list as well as a, but even more Democratic state, but one where

Trump has held the lead in Nevada almost all year. You know, I've been tracking those averages

at key points in the race. And Wisconsin sometimes has been the best for the Democrats,

only because it was tied in Wisconsin or only slightly behind in Wisconsin. Since Biden dropped

out, we do seem at least in the averages across all the polls to be a little bit more Democratic than

Michigan and Pennsylvania. And then the other toss up states are either very, very close or

leaning towards Trump. You know, I'd have to say that if you look at recent elections in Wisconsin,

there's reason to believe Democrats are doing better here. If you include the 2022 gubernatorial

election, Democrats lost, of course, the Senate race, but by only one point. And then the Supreme

Court race that the liberal candidate won by 11. So I do think there's reason to believe the

party is doing pretty well here in the state. But having said that, I especially remember the

polling was too optimistic for Democrats in 2016 and again in 2020. So I think there is good reason

to take the results, including mine, unfortunately, with some extra measure of caution because of that

underestimate of Trump's vote in the past. Has that has that changed significantly over the years

that the ability to, you know, properly estimate the Trump vote? Yeah, it's a mixed picture.

Our worst performance was 2016, where we were off by more than six points. We still missed,

but got the winner right in 2020 when we were off by about four points, which was better than the

average of the polls, but still was, was an error. But if you look at the two midterms 2018 and 2022,

our average error in those four races was below our usual average of 2.2 points is our average error

over all the races we've done. And we were less than that in the 2018 and the 2022 races. So it looks

more like it's related to Trump on the ballot, rather than being related to polling being broken

in general. Otherwise, we shouldn't have gotten 18 and 22 as well as we did. This year we'll find out

in November, but between now and then all we can do is do everything we can methodologically to

get it better. But hope. And we'll find out in November for sure. Right. Yeah. It's going to be

interesting. So as you know, I write a more in depth breakdown of of every new version of the

Marquette University Law School poll. And you can find that now on the Reconvibulation Area and

on the Civic Media Network. That's new from from our past breakdowns here. And one of the things

that really stood out to me in this breakdown is just how wide the gender gap is. You know,

obviously we have a woman candidate running for president and a man running for president. So

there some of that is a little bit to be expected. But do you think this is more pronounced now than

ever in the state of Wisconsin? This gap where men typically back Republicans and women typically

back Democrats. That gender gap has been with us for a long time. Women as a group are significantly

more democratic in their partisanship than men are. And so part of the gender gap is because women

identify as Democrats more than men do. And that's for good political reasons. It's not just a fluke

there. But it does mean when we talk about the gender gap, some of it is built into partisanship

and some of it is reaction to these specific candidates. Right now we're seeing in this

month's poll about a 14 point gap in support for Harris men at 44 women at 58 in this latest

poll with registered voters. And that's substantial. But the caution I'd raise is in previous

election cycles. We've seen those gap numbers bounce around from month to month. They do seem

malleable or if you will in the way of all cross tabs, they're subject to a greater margin of

error. So we should expect them to move around more. That said, this is pretty large at 14 points.

One of my to-do list things is to go back and look at how it's varied in every election since 2012.

But I haven't quite gotten to that. Suffice it to say it's a substantial gender gap this year.

Whether it's historically high or merely high, I think we have to wait until we run the data for

everything. But it still shows up as a very important feature of this year's election

that genders playing such a big role. And it tends to play a big role when you break it down into

other categories as well. Gender among young and old voters, for example. Yeah, how does that

break down? Well, the gender gap continues to show up among both the young and the old.

There's a little bit less of a gap among the older voters. I'm talking about basically over 40 or

45, just breaking age down into two categories. We have seen, and there've been some national

reports on this, that younger women are moving more solidly in a liberal direction and a more

democratic direction. Their male counterparts, not as much though the younger men also are somewhat

more liberal and a bit more democratic than their elders are. So the gender gap among the young

may be more to how much women have moved in the younger age groups rather than men and women pulling

apart in that younger age group. Yeah, that's interesting. So I went through a lot of the cross tabs

in this poll and compared it to the last one. Obviously, we haven't had a lot of polling between

Harris and Trump, just given the change in July. But one of the things that I thought swung pretty

wildly and I'm wondering if you can attribute this to anything is the independent voters. So you

had a big change in independent voters that were more backing Trump in the August poll and now in

this poll, more backing Harris. And I looked at the favorability numbers in that too. There was a huge

swing in favorability among independence for both Harris and for Trump. So you've talked about

this a number of times how independent voters can be pretty tough to pin down on where they stand on

things. But what does that say to you about how much seems to be changing with the new candidate

at the top of the ticket for Democrats? I think it does reflect a real shift due to Harris rather

than Biden. Each individual poll independence, pure independence who don't lean towards one

party or the other is the group we're talking about. That's only about 8 to 10% of the sample.

So it's a small sample in a small group. But in a state as close as we are, it's always a group

that we pay a lot of attention to. If you look back over the Biden versus Trump polls,

say the last two before he dropped out, independence were leaning towards Trump having earlier in

the cycle lean towards Biden. And that shift among independence was a contributing factor to

Trump pulling ahead of Biden by one or two points in our late polling for him. But with Harris

in the race, independence were more evenly balanced in our July poll. And then this poll are

substantially towards Harris over 60%. I wouldn't put much stress on the exact numbers there because

again, big margin of error with a small group. But it is the group that because they don't have

partisan ties, tend to be moved by the recent political wins. So I think if we tell this story

and kind of believe it's true, you would see that independence were leaning towards Trump

for against Biden for at least the last two polls this spring and early summer.

They moved some of the way towards Harris in July, just a week after she got in the race,

were actually the week that she got in the race. And now post DNC, but before the debate when

this latest poll was done, had moved substantially towards Harris. I have to say when I see that

group up in the 60% range, I think that's not very likely to actually happen on election day.

But I think if you take the broader movement, it looks like Harris's entry has brought independence

her direction and maybe a solid majority of them. But I'd still wait right up until election day

to see, does this group settle down and quit moving around so much? Because that's the other thing

to watch for. When you see big change from one poll to the next, it could be sampling error,

but it could be real movement. It's when we look at it as we were just talking about through two

Biden polls and two Harris polls. I think you can say with some confidence independence have

shifted in Harris's direction. But I don't think you can be very confident about what the magnitude

of that shift is just because of sample size. Yeah. Well, one thing that does clearly seem to be

shifting in Democrats favor within the last two polls since Harris has been it been included

is enthusiasm. The enthusiasm for Democrats was down around 40% very enthusiastic when in the

final poll with Joe Biden, that jumped up over 60% in August and now over 70%. What does this

mean ultimately? Obviously, we know the enthusiasm. We talked about the kind of post convention

bounce that Harris is getting. Why does enthusiasm matter? I think it's one of the most important

findings of the last four or five polls. And it's the one that I think is most convincing

in substantive terms that the things we're seeing in the polls lines up with sort of the way the

election is shaping up all year long until Biden dropped out. Overall, enthusiasm was running

15 to 20 points below where it had been four years ago at the same time in the 2020 election cycle.

But then Biden drops out and in that first poll after he drops out, we get the big rise that you

just talked about to not quite the 2020 levels, but close to it this month, we're even closer to

the 2020 levels just like one point below 2020. And so we fully recovered that overall enthusiasm that

had been missing all year. But the partisan shift is also terribly important because Republicans were

running solidly ahead of Democrats in enthusiasm under Biden. They almost caught up in July and

now they've surpassed Republicans in September. I do want to say that Republican enthusiasm rose

after the RNC as well. Their enthusiasm went up seven points after the Republican convention.

But that same month, Democrats went up 20 points to almost catch them and now to exceed them.

Republican enthusiasm's basically stable this month compared to the post RNC period.

So there's no drop off there. Those pesky independence come back in having been low in

enthusiasm all year that perked up a little bit after the RNC, but now it's dropped back down.

So how they're being engaged by the campaign, I think is still real iffy. But the partisan gap is now

an advantage to the Democrats having been a big advantage to Republicans previously.

You see it by looking at the vote by enthusiasm with Biden and the race, those who are very

enthusiastic were substantially preferring Trump. And those not so enthusiastic or not at all

enthusiastic had a substantial Biden lead. And I talked about that in June because this is a risk.

Turn out goes down with lack of enthusiasm is this a way for Democrats to underperform.

Now the shoe is on the other foot. It's been completely reversed. Harris is doing better,

though not extraordinarily better, but about six points better with the very enthusiastic and

the somewhat enthusiastic. It's only the not at all enthusiastic where Trump has a big lead.

So this is in one sense exactly what you'd expect. Democrats have become much more enthusiastic

than they used to be. They strongly support Harris. The result is this inversion of the

enthusiasm vote relationship. But I think that both enthusiasm levels are now up to levels

where I think we can plausibly assume a very high turnout, whether we match the records of 2020,

I think is still a little bit up in the air. But when Biden was the candidate, I was seriously

thinking that we would see a drop off in turnout in this fall's election. Now I don't think

that's as likely or if it does drop, it doesn't drop by very much. And that's why the enthusiasm

metrics really matter turnout. And you've talked about this in previous polls.

There's a lot of data that suggests that the more enthusiastic that your voters are,

the more likely they are to go vote. And the more likely they are to encourage others to vote,

or just get involved in the election more generally. And I think that is the part that we,

you know, it's one thing to just say, hey, Democrats are fired up for Harris. They had a great,

you know, the convention seems to have, seems to have had an impact. But it's another thing to

recognize the connection there is to turnout. And it is to driving those numbers up. And I think

we've seen that in Wisconsin, you know, 2016 was a low turnout year and Democrats lost. 2020 was

a high turnout year and Democrats won. You know, there's some shifting things underneath the surface

with, you know, kind of the likelihood of voters for each party, you know, we know Republicans

have under Trump have targeted more of the low propensity voters in their strategy as well.

And you've taken some steps in the Marquette poll to try to understand the likelihood

of voting as well. You talked about that in this week's presentation. I was wondering if you

could expand on that a little bit for our listeners here at the recon population area.

Sure. I'm happy to. And you're exactly right turnout is the behavioral consequence of enthusiasm.

And it's not automatic, right? They're enthusiastic people that don't vote. And there's some sort

of less enthusiastic people that go and do their civic duty. But shifting turnout in a state where

21,000 votes determines the winner is an absolutely critical aspect of our elections.

What we've done in the Marquette poll is in 2023, more than a year ago, we shifted our approach

instead of using random digit telephone calls to draw the sample. We started sampling from the

voter list, the list of registered voters in the state. And that has a variety of advantages.

But the one that's relevant to what we're talking about is the voter list. I want to be clear,

because someone asked this at the presentation, the voter list does not include who you voted for.

That is private and secret. But it does include whether you voted or not. That's valuable additional

information that we didn't have in previous years of the Marquette poll, where all we knew was

your phone number. Now we know for our final sample how they've or if they voted in 2016 and 2020

and 2022. And in the primary elections, those years. So that gives us a much more concrete

measure about who is most likely to actually vote this year. Now it's still an estimate.

But we run models using data from the voter file and other adjacent data to make a prediction of

how likely is it that a person on the voter file actually voted in 2020 using the 2016 voting

history or in 2022 using the 2016 and 2020 voting history. And now we're using that to predict

turnout in this year's election. What we find is Democrats have an advantage in how frequently

they've turned out to vote in recent elections. Republicans turned out frequently. I don't mean to

make it sound like there's a big discrepancy here. It's a point or two percentage point or two.

But Democrats have a little bigger advantage and have a little bigger advantage in voting in all four

of the last possible elections they could have voted in compared to Republicans. Now this is

partly Trump does a good job of mobilizing low propensity voters, low probability of voting

voters. And so we see those folks may be showing up only in the 2020 election, but not voting in

22 or not voting, having voted in 2016. This all assumes they were eligible in all those

elections, of course. So it looks like Democrats have a bit of an advantage. Back in the spring

when Democratic enthusiasm was so low, I was saying there's a bit of a

disconnect here between the partisan enthusiasm favoring Republicans, but the vote history favoring

Democrats by just a little bit as I say. So what we've done is use that model of past voting

behavior to provide an alternative likely voter model that's based on age and when you registered

and other things and and past actual turnout to create this model of likely voters. And that means

that now we're a very high turnout state and among registered voters upwards of 90% actually

vote in our presidential election. So people think of America as a low turnout place, but

if you narrow it to those who've registered to vote, we're quite high in turnout.

But this measure lets us also include people with an 80% chance of voting or a 70% chance of voting.

And we model that so that the 70% chance of voter doesn't get a full 100% of the vote of the weight,

but they get included in the model. Now the truth is that when we just ask people, are you

absolutely certain to vote? Most of those people have also voted in all these past elections and

have a high chance of voting. But there are some that haven't had as much of a history of turnout,

even though they say they'll sure to vote this time. But the reverse is also true. Some of the people

that have voted every time tell us, oh, I'll probably vote or maybe I won't vote this time.

So both of these measures contain some people that look a little different from the other measure.

That said, what we've done, we've been developing this all year and we just implemented it

in the survey this year or this month, rather, is to use that history of voting as our likely voter

model instead of the simple, are you absolutely sure you'll vote? Last month it made a difference

of tenths of a percentage point, which measure you used, this month it makes a difference of

one percentage point or less. So we're not getting different substantive results depending on which

method we use. I'm using the history of voting turnout measured the new one in part because I think

it's a little more stable over time. We shouldn't see as much fluctuation in whose included in the

likely voters. And I think that's a little bit of an advantage. If it made a big difference between

the two, I'd worry a little bit about which one to trust. But I think the history of voting is

a more stable measure that probably won't fluctuate as much from one wave of the survey to the next.

Very interesting stuff. A couple more things I wanted to get to here. So this is the first poll

that has included polling, extensive polling at least, on Minnesota Governor and Vice Presidential

candidate Tim Walsh. Now you often joke at these presentations that there are no popular candidates.

Everybody's has just about every candidate has a net negative favorability rating.

The only real exception of that for the past year has been Governor Tony Evers,

who's been in the black on that one. And now Tim Walsh joins him in that category. So the two

Democratic Midwestern governors have positive favorability ratings. And as I went through it,

you know, there's a lot that I think Democrats should be excited about when looking at Tim Walsh's

numbers and what he can add to the ticket here. Yeah. I think that's right. And it is interesting.

He's a net plus six on this. Harris is a net minus four. I think it is. It might be minus three.

So slightly negative. And Trump is at minus 10 or negative 13 something like that. So that sort of

gives you a sense of the range of these things. Baldwin, for example, is a plus one also just barely

net positive this month. But Walsh stands out as different from the rest. Now it is true that 19

percent still say they don't know enough about him to have an opinion. So that is probably helping

him a little bit on the from the don't know is 16 percent don't have an opinion of JD Vance.

Though he's a little more negative than Trump is actually. But we also saw things like in terms of

was he an excellent or good pick for vice president among Democrats 90 percent say he is an

excellent or good pick with Vance. It was 75 percent of Republican saying he was excellent or

good right after the RNC. So Walsh is being welcomed into his party a little bit more strongly

than Vance, though both parties were united pretty much in in their VP selections.

The other things we're seeing with was is that first of all he is in fact a little bit better known

in the western part of the states to get Minnesota TV. And it's not huge, but it is an advantage.

On the other hand, those are some of the most Republican areas of the state. And so his net

favorability is actually negative in the western Minnesota broadcast areas. Then it is in the east

where he's net positive, but he is better known there, which you would expect if you're getting

Minneapolis news, for example. So that holds up reasonably. The other thing we did for the first

time is ask people to rate the candidates on how liberal or conservative they are on a scale

from very conservative somewhat conservative moderate and so on to very liberal.

People put JD Vance very much in the very conservative category. But with Walsh, they put him

considerably less in the very liberal category. That is the largest group. So people definitely

see him as a liberal or a progressive, but they see him less so than the way they see Kamala Harris

and much less so than the way they see on the conservative side JD Vance or that matter Donald

Trump. So a lot of the debate about Walsh has been does his progressive record as governor in

Minnesota make him a far left politician as Republicans have tried to pay him. Or is he the

Minnesota dad who will fix your car tire or push you out of a ditch. And it's still too early to

see how this plays out over the whole campaign. But here in this early going after the DNC, I think

he has some real advantages in favorability and people seeing him as more moderate and less of

a very liberal candidate, even then they see Kamala Harris. Harris and Trump are about equal in

the very conservative or very liberal category, but Walsh is different. Yeah, as I wrote my breakdown,

I think Walsh is checking just about every box you might want to see for a candidate running in

Wisconsin in the position he's in. Before we don't have a ton of time left here, but I do want to

talk about the Senate race as well, because that one tightened a little bit from previous polling.

Their Eric Havde's name recognition seems to have gone up. And now we've typically seen Baldwin

polling better than the top of the ticket, Harris or Biden. And now we see that pretty much even.

So Baldwin 52 Havde 48 in this poll. So what are you seeing for why that race has tightened a little

bit? I think on that, it's a little more that the presidential race has moved up rather than the

Senate has necessarily moved down. They do match at four points. In our press polling to last three

or four polls, we've had Baldwin's lead range from four points to as much as seven points,

depending on how you ask the question and whether third party candidates are included.

Here, it's either four or five points. So it is a little tighter than that range we saw before.

Havde is much better known now. He's at 22% who say they don't know enough. He was at 80% didn't know him

in January. Also, for the first time, his favorables moved up a little more than his unfavorables did.

He'd been getting worse off on the favorable unfavorable balance. But this month,

he's narrowed the gap, though it's still a net negative nine versus Baldwin's plus one.

The other thing for the first time we asked about how do you perceive these candidates?

And the narrowest division is is someone who will work to solve national problems. Baldwin has a

very small advantage on that measure over Havde. Then we ask cares about people like you and Baldwin's

advantage is still small, but a little bit bigger. And then the last one will work for the concerns

of Wisconsin. That's the largest Baldwin lead of these three. So she holds a lead on all of them,

but is strongest on that connection to Wisconsin and is the slimmest lead on work to solve national

problems. Yeah, the connection to Wisconsin that does seem to make sense because we're always

hearing in the news, Tammy Baldwin talking about, you know, bringing home federal dollars for

certain infrastructure projects are proposing certain bills. And I think, you know, that has

certainly been a hallmark of her time in the Senate that constituent service and working for

Wisconsin. Absolutely. And these come in part from what the campaigns are talking about.

Those messages from her as well as the Havde lives in California kinds of attacks,

but on Havde side, he's talked about he's tired of the gridlock in Washington and wants to work

to solve our problems. So that's where those questions come from and cares about people like you,

it's just a standard measure of how well candidates relate to voters.

Yeah. So last question here and we just got a couple minutes, but I really appreciate that you

have continued to include questions on deportation policies. You get into the issues in a lot of

different ways in this poll, but the deportation one I think is really important that you continue

to include. And you ask the question a couple different ways about, you know, deporting immigrants

who are living in the United States legally. And then you also include pieces of that on whether

that immigrant might have a criminal record. You know, quickly, what are you seeing

any difference between those questions there? So one is worded more or less the way Trump presents it,

deport all illegal immigrants who are in the US illegally period. That's getting over 60% support

and has been all year is much higher than you would have expected four years ago or eight years

in our second version of the question, it's framed the same way at the beginning,

but then at the end we say, even or including those who've been in the US for a number of years

have jobs and no criminal record. So that's a softer version of it recognizing that some of these

immigrants have been here for a long time and your neighbors and your your your daughter, your

children's classmates, their support drops into the 40% range, but it ticked up this month,

it had been 45 favoring deporting even those who've been here a long time, 55 opposed,

ticked up to 49 51 this month. That's not much of a movement, but that's something to keep watching

and especially if this issue goes on past the election. Absolutely. Well, Charles Franklin,

thank you so much for joining us in the Reconbobulation Area. We always appreciate your insights.

Thank you.

Thank you for listening to the Reconbobulation Area podcast, now part of the Civic Media Network.

For more analysis of the latest edition of the Marquette University Law School of Full,

find my full breakdown on the Reconbulation Area and at Civic Media.

Thanks again for listening. We'll see you next time.

Thank you very much for joining us in the Reconbobulation Area.