Welcome to FiveThirtyEight’s weekly politics chat. The transcript below has been lightly edited.
sarah (Sarah Frostenson, politics editor): Former Vice President Joe Biden’s team is talking a big game about an expanded electoral map with Arizona, Georgia and Texas in play, even though those states haven’t voted for a Democratic presidential nominee in two decades.
So let’s talk about just how feasible this strategy is. How competitive are those three states at this point? And what’s more, how does this strategy complement — or counteract — Democratic efforts to pick up Midwestern battleground states like Michigan, Wisconsin and Pennsylvania, or perennial swing states like Florida?
First up, Arizona. What do we think? Does Biden have a shot there?
geoffrey.skelley (Geoffrey Skelley, elections analyst): Of the three states we’re looking at, I think it’s pretty clear that Arizona is the most in play — and that Biden may even have the lead there, based on the limited polling we have.
President Trump won Arizona by 3.5 points in 2016 while losing the national popular vote by 2 points. So it stands to reason that if Biden is up 6 points or so nationally, Arizona is a toss-up, and that’s before we consider other things that may have shifted between 2016 and now.
nrakich (Nathaniel Rakich, elections analyst): I agree, although I have been surprised at the degree to which Arizona seems to have moved to the left since 2016.
sarah: What other evidence do we have that Arizona has moved to the left since 2016?
geoffrey.skelley: Well, unlike in Georgia and Texas, Democrats actually won major statewide contests in Arizona in 2018 — including the state’s marquee Senate race — and election turnout was nearly as high as the 2016 presidential contest, meaning that performance may reflect a broader shift toward the Democrats rather than just a side effect of the midterms’ blue wave.
nrakich: G. Elliott Morris of The Economist had an interesting newsletter item recently that showed how much various states have moved left or right since 2016, based on the 2020 polls so far. Arizona had the starkest movement.
And Geoffrey’s right that, if Arizona were still 6 points redder than the nation and Biden led by 6 points nationally, we’d expect polls of Arizona to show a tied race. But Biden has consistently led in Arizona polls so far.
|May 10-14||Redfield & Wilton||45||41||D+4|
|May 9-11||OH Predictive Insights||50||43||D+8|
|April 7-8||OH Predictive Insights||52||43||D+9|
|March 6-11||Latino Decisions||50||42||D+8|
|March 3-4||OH Predictive Insights||49||43||D+6|
|March 2-3||Public Policy Polling||48||47||D+1|
On the other hand, I’m still somewhat skeptical of the idea Arizona has moved that much to the left. Some of the higher-quality polls, like from Marist and Monmouth, do have the race closer to a tie, whereas the polls suggesting Arizona has gotten significantly more Democratic (e.g., by showing Biden up by 8 points) are not coming from gold-standard pollsters.
sarah: One other thing about Arizona that makes me think it might be fertile ground for Democrats in 2020 is that Democratic Senate challenger Mark Kelly seems to have the upper hand against Sen. Martha McSally, and if that race ends up close — or flips blue — that bodes well for Democrats in the long run, as it’s more evidence that Arizona might be becoming more of a blue state.
nrakich: Yeah, Kelly has been a monster fundraiser. He’s taken in more than $31 million since the beginning of last year.
Although I don’t think a down-ballot race is likely to drive turnout for the presidential. If anything, Kelly might run ahead of Biden because of his money and great bio. 👨🚀
geoffrey.skelley: That’s fair, but it’s worth remembering that every Senate seat that was up in 2016 went for the party that carried the state at the presidential level, so the fact a Democrat is polling that well in the Senate contest is probably a decent sign for the party’s chances as a whole.
sarah: For sure. It’s less that a down-ballot race would affect the top of the ticket, but more that Arizona really might go blue in 2020.
It sounds like we agree with the Biden campaign’s assessment that Arizona is in play, so does it make sense for them to campaign there?
Or is there an argument to be made that they should keep an eye on it, but maybe not commit fully?
nrakich: I mean … both?
It’s a spectrum.
I definitely think Biden should spend more time and money in Arizona than in Georgia and Texas. But I still think Arizona is unlikely to be the tipping-point state, and Biden should spend even more time and money in must-win states like Wisconsin, Pennsylvania and Michigan.
geoffrey.skelley: Oh, they should definitely fully commit. Arizona gives them another possible path to 270 in the Electoral College. Arizona’s worth 11 electoral votes, so it could sub in for, say, Wisconsin (10 electoral votes) if Trump were to narrowly carry the Badger State.
nrakich: Now you have me questioning myself, Geoffrey! *whips out calculator*
Hmmm, Florida and Wisconsin were 3 points to the right of the nation in 2016. Arizona, as discussed, was 6. That’s not a big gap at all; maybe they do converge this year?
geoffrey.skelley: Another thing to keep in mind is that Democrats have been making inroads in the suburbs and dominating urban areas. Maricopa County (Phoenix and its environs) was the most populous county in the country to vote for Trump in 2016, but Trump only won it narrowly by about 3 points, and in 2018, Democratic Sen. Kyrsten Sinema carried it by 4 points. So Democrats may be hoping for a repeat in 2020. Win Maricopa, win Arizona.
sarah: OK, it sounds like focusing on Arizona is smart for the Biden campaign, but maybe we’re a bit more skeptical of Georgia and Texas, the other two states the campaign has included in its “expanded” electoral map?
nrakich: Yeah. Georgia was 7 points to the right of the nation in 2016, and Texas was 11 points to the right. Given long-term trends, they have both probably moved a little to the left, but they have further to go than Arizona.
That said, Biden may well win those states — take a look at the polling there:
|May 11-13||BK Strategies||46||48||R+2|
|May 4-7||Public Opinion Strategies||47||46||D+1|
|March 31-April 1||Battleground Connect||46||48||R+2|
|May 8-10||Emerson College||48%||52%||R+3|
|April 27-28||Public Policy Polling||47||46||D+1|
|April 18-27||University of Texas at Tyler||43||43||EVEN|
But if he does, he will probably already have clinched the Electoral College in the Midwest, Arizona or Florida.
geoffrey.skelley: Georgia is interesting. On the one hand, Biden could target the increasingly Democratic suburbs of Atlanta. On the other hand, it’s one of the most inelastic states in the country — meaning voters there are among the most likely to stick with their usual party regardless of which way the rest of the country swings — in part because its white voters remain predominantly Republican and its large black population is heavily Democratic, and there just isn’t a ton of movement there.
Additionally, if Democrats couldn’t carry Georgia in 2018 when the electoral environment was very pro-Democratic, that makes me skeptical they can win it in a presidential year, when partisan conditions could be more balanced. That said, if Biden is winning by 6 or 7 points nationally, that might be enough to put Georgia in his column, as Trump only carried it by 5 points in 2016. But as Nathaniel was saying earlier, that’s not a situation where Georgia is an integral part of Biden winning 270 electoral votes. It’s gravy at that point, though maybe it helps Democrats in the two Senate contests there.
nrakich: Yeah, Georgia is definitely inelastic. But on the other hand, Georgia has inched leftward (relative to the nation as a whole) in the last three presidential elections. And I think there is room for more suburban whites to move toward Democrats, not only in Georgia but also in Texas and Arizona.
sarah: That’s a good point, and I think a real question determining whether Georgia and Texas will be competitive is just how much the trends of 2018 — namely, suburban white voters moving to the Democratic Party — hold true.
This is an extreme hypothetical, but earlier this year, Nathaniel looked at what would happen if a state’s presidential vote was based strictly on how rural or urban the state is, and he found that Georgia would remain in the R column, but both Arizona and Texas would swing blue:
What do we make of this? Might Texas actually turn blue before Georgia?
nrakich: We have a tendency to think about elections through the lens of the decisive voters in the previous election, which for 2018 was suburbanites. But as I showed in that urbanization article, Georgia does have a lot of rural voters too, and there is still room for them to move even more toward Trump. So, actually, maybe those two trends will cancel each other out.
geoffrey.skelley: OK, but Georgia was still notably closer to going for Clinton than Texas — Trump won Georgia by 5 points and Texas by 9 points, which is a fairly sizable difference. And while Georgia may be more inelastic than Texas, Texas is not that elastic. Our 2018 elasticity score for Texas was 1.03 — not that far above the baseline of 1 — while Georgia’s was 0.90.
Texas is changing, but Barack Obama lost it by 12 points in 2008, which was a really good environment overall for Democrats.
nrakich: Yeah, there’s just too far for it to go.
geoffrey.skelley: As is often the case with questions about when Texas could go blue, it depends on how fast the political environment changes, but it still probably won’t happen until sometime after 2020, given what we know currently.
sarah: People seem to agree that the Biden campaign shouldn’t invest too much in Georgia and Texas if it comes at the expense of other battleground states in the Midwest or Florida. Is that fair?
nrakich: I think there’s a case for keeping your options open in Georgia. But the Biden campaign would be foolish to invest significantly in Texas. If Texas votes Democratic, Biden will already have won virtually every other swing state and, therefore, the election. It’s simply not a part of his path to 270 electoral votes — more like a part of his path to 400.
Also, Texas is an extremely expensive state in which to campaign, so it just wouldn’t be an efficient use of his money.
geoffrey.skelley: If Trump really is doing a lot worse among older voters than in 2016, it would be foolish for Biden to abandon Florida, which has one of the oldest populations in the country.
I could see reasons for Democrats to worry about Florida being a mirage after they failed to win the gubernatorial and Senate races there in 2018, but it’s just been too close in recent presidential elections to actually give up on it. Trump only won it by 1 point in 2016!
nrakich: Oh, I have strong feelings about Florida.
nrakich: Florida is definitely still a swing state; it’s not as inelastic as the 2018 results implied. The Democratic nominees for governor and senator, Andrew Gillum and Bill Nelson, still outperformed Hillary Clinton in most counties; they just underperformed Clinton in a few key areas, especially Miami-Dade County. (This article by Florida Democratic consultant Matthew Isbell does a great job showing that.)
The reason for this is probably that their Republican opponents, Ron DeSantis and Rick Scott, did a lot better among Hispanic voters than Trump did. According to exit polls, Trump got 35 percent of the Latino vote in Florida in 2016, while DeSantis got 44 percent and Scott got 45 percent. In 2020, I don’t think Trump will be able to match DeSantis’s and Scott’s numbers.
So if Biden can pair Clinton’s performance among Hispanic Floridians with Nelson’s and Gillum’s among other voters, he can absolutely win Florida.
geoffrey.skelley: We’ve talked a lot about how Biden might be able to expand his electoral map, but he can’t afford to give up on Michigan, Pennsylvania and Wisconsin. In 2016, they were collectively decided by 78,000 votes, and who wins them in 2020 will likely be consequential as well.
The bigger questions in the Midwest and Rust Belt are probably whether to invest in Iowa and Ohio, which Trump carried by about 9 and 8 points, respectively. Those two states might be harder for Democrats to win back considering how they swung hard toward the GOP in 2016 after backing Obama in 2012.
That said, Iowa does have some history of being pretty swingy. It’s also cheaper to advertise in Iowa than Ohio, and if we’re talking down-ballot races, there is more at stake there, too. Potentially four competitive House races and a Senate seat in Iowa, whereas Ohio has no Senate race and is likely to have only one or two close House races.
nrakich: Yeah, if Biden wants to be an effective president, he’ll need a Democratic Senate. IMO, that means he should give extra credit to Georgia and Iowa when deciding where to allocate his resources.
sarah: The balancing act that the Biden campaign will inevitably have to engage in isn’t entirely clear to me yet. How much will they actually invest in states like Arizona, Georgia and Texas versus doubling down on states like Michigan, Pennsylvania and Wisconsin?
Much of this will inevitably boil down to what the tipping-point state is in 2020, but one thing that’s hard to figure out is how much of the map already realigned in 2016. Put another way, does Biden have his eyes on states like Arizona because winning states like Wisconsin back will be difficult?
nrakich: But I think that’s the needle we need to thread: Arizona might be moving in one direction and Wisconsin in the other, but even in the “realigned” (really more “recalibrated”) 2016 map, Arizona was redder than Wisconsin.
geoffrey.skelley: It’s curious because some of this comes down to the national environment. Maybe Wisconsin is a point or two redder than it was in 2016, but if Biden wins by 4 or 5 points nationally, maybe that’s enough to carry it even if Wisconsin is continuing to move toward the GOP.
But how exactly that plays out in each state is hard to say.