The Case for Blue Arizona
Why Arizona might be one of Biden's better swing states, despite the polling.
A few weeks ago, I wrote my first piece on a specific swing state, Nevada, as part of a series I’ll be doing for each battleground state. Today, we turn to Nevada’s neighbor, Arizona. The two states actually have quite a bit in common, and in 2020, they converged politically, both voting for Biden after three straight presidential elections of supporting different candidates. Yet, as I’ll explain, I think Arizona will not only vote more Democratic than Nevada does this fall (which last happened in 1996), but that more generally, it could be one of Biden’s better swing states this year.
First, it’s crucial to grasp Arizona’s political geography, which is quite unique. Like Nevada, Arizona is a fairly expansive state in which the overwhelming majority of residents live in just two metropolitan areas: Maricopa County (Phoenix) and Pima County (Tucson). Below, I have a map of Arizona’s counties, shaded according to whether they voted for Biden (blue) or Trump (red) in 2020.
By far, the most important county to understand in Arizona is Maricopa County, roughly in the center of the state. In 2020, it cast 61% of the total vote in Arizona, more than any other county. The city of Phoenix, the urban core of Maricopa, cast a little under a third of the county’s total vote in 2020. Thus, while Phoenix is the single most important city in Maricopa (and aruably the state), it does not dominate the county’s politics as much as one might assume. In fact, Maricopa is home to another major city, Mesa, as well as large suburbs such as Chandler, Scottsdale, and Tempe.
Politically, the trends in Maricopa County have been unmistakenly positive for Democrats: in 2012, Obama lost Maricopa by close to 11 points. Then, in 2016, Hillary Clinton came within 3 points of winning the county. In 2020, Biden won Maricopa by 2 points, the major reason why the state flipped from red to blue. The table below, showing how Maricopa voted in each presidential election since 1948, helps make the long-term trend fairly clear.
It might be tempting to focus entirely on Maricopa, given its size and the fact that the county has become a bellwether for the state as a whole (which wasn’t the case historically). But, it’s important to also consider Arizona’s other metro area: Pima County, home to Tucson. Although much smaller than Maricopa, Pima County typically casts a still-significant 15% of Arizona’s total vote. It has long been Democratic, having last voted for a Republican presidential candidate in 1988, whereas Maricopa voted Democratic for the first time in 72 years in 2020. As with Maricopa, Pima has been trending Democratic in recent years, though it started from a somewhat bluer baseline.
Outside of Maricopa and Pima County, the remainder of the state is mostly rural, though it’s not quite as Republican as one might expect. Notably, it’s also not trending Republican in the way that rural areas elsewhere in the country have been. In 2016, Clinton lost the counties outside of Maricopa and Pima by about 17 points, while Biden lost them by 16.5 points. Part of that stability can be explained by Arizona’s Native American community, which makes up about 13% of the population of the non-metro portion of the state. Native Americans are most concentrated in northern and northeastern Arizona, and have been reliably Democratic for decades.
Basically, Arizona can be thought of as three units for the purposes of electoral analysis: blue Pima County, purple Maricopa County, and red non-metro areas. It should come as no surprise that in the aggregate, the state is as evenly divided as can be. In 2020, Biden’s margin of victory was just 0.3%, or roughly 10,000 votes out of about 3.4 million cast.
So, why does Arizona look good for Biden this fall, despite his incredibly thin margin four years ago? Part of it is simply that the Democratic trend is so evident in the largest (and growing) parts of the state, while the areas that are trending Republican include a few small counties along the southern border with Mexico.
More broadly, it’s worth emphasizing just how much Arizona is a poor fit for Trump, even though the state has a long Republican lineage. Consider this: in the 2008 election, when Obama won in a landslide nationally, he lost Arizona by 8 points. Given that his opponent was Senator John McCain, an Arizona institution, that was actually a decent showing for Obama. But, it made clear that the state was firmly in the Republican column, having voted about 15 points to the right of the country as a whole. Four years later, Romney won Arizona by a similar margin as McCain had, confirming that the state was still not particularly competitive.
It was arguably Trump’s nomination in 2016 that single-handedly made Arizona a purple state, which it has remained ever since. In fact, from 2012 to 2016, Arizona shifted away from Republicans more than nearly any other state did. Even though Clinton did not win its 11 electoral votes, she came closer to winning Arizona than any other Democrat had in 20 years.
Why did Arizona react so poorly to Trump? McCain’s disdain for Trump is legendary, but I think more fundamentally, it’s about education. Arizona is a fairly well-educated state, and as I’ve mentioned in previous posts, that single factor currently tells us more than nearly any other factor about how a state is trending. The state’s suburbs, particularly around Phoenix, contain a critical mass of voters who perhaps long thought of themselves as traditional conservatives, but who have become cross-pressured voters since Trump won the Republican nomination.
I don’t want to overstate my case here, but I almost can’t imagine Maricopa County, the state’s electoral powerhouse, not voting at least slightly more Democratic this fall than it did in 2020. In making this claim, I’m very much thinking back to how Maricopa voted in the 2022 midterms, particularly in Arizona’s Senate race, as voting patterns in senatorial and presidential elections have become increasingly correlated. Two years ago, Democrat Mark Kelly won Maricopa by about 6.5 points over his Republican challenger in what was a neutral-to-slightly-red midterm election nationally.
Even in Arizona’s gubernatorial race in 2022, which was nearly a tie statewide, the Democratic candidate (and now governor), Katie Hobbs, won Maricopa County by 2.5 points, a margin that Biden could exceed this fall. Maricopa is the type of county that is arguably bluer at the presidential level than in gubernatorial races, which has become a somewhat common pattern in well-educated counties across the country.
In fact, even the slightest improvement for Biden in Maricopa County would be electorally significant, and would likely come in tandem with a small improvement in Pima County, given the characteristics these two counties share. From 2016 to 2020, both Maricopa and Pima moved in lockstep, each voting 5 points more Democratic in 2020 than they had in 2016. For Republicans to have both Maricopa and Pima County, which cast a combined 76% of Arizona’s vote, trending away from them is simply unsustainable.
Essentially, for every point that Biden improves in these two counties, Trump would have to improve by about 3 points in the rest of the state to offset Biden, and as alluded to earlier, the areas outside of Maricopa and Tucson have been fairly stable in their recent voting patterns. It’s possible (and perhaps even likely) that Trump will perform slightly better in non-metro Arizona this fall, but it’s not clear that this alone would be enough to flip the state, given how little of the total vote comes from this area.
There is, however, a major caveat to all of this: Biden’s polling in Arizona has been rather poor. We can’t simply dismiss the polls as being skewed or somehow unfair to Biden, because as I noted last time, he’s doing surprisingly well in polls of red states, such as Alaska and Utah, among others. And even in other swing states, like Pennsylvania and especially Wisconsin, Biden is polling decently.
Below, I have a screenshot from 538, which keeps tracks of both national and state polls, showing the most recent presidential polls in Arizona. Be aware that a few of the polls at the top include matchups besides Biden and Trump. On average, Biden is trailing Trump by 5 to 6 points in Arizona, according to these polls. Not one shows a lead for Biden, or even a tied race.
One potential explanation for Biden’s poor polling matter is that Arizona has a large Hispanic population, roughly 31% of the state (though Hispanics make up only about 24% of the state’s eligible voting population). In 2020, no other group moved so dramatically against Biden compared to Clinton four years earlier.
Given that we’re almost certain to have the same pair of candidates from 2020, it’s very likely that the electoral trends seen among certain voting groups between 2016 to 2020 will continue, and perhaps that reality is already showing up in polling. In general, Biden’s polling among non-whites has been quite weak, so the snapshot provided by these polls of Arizona is less surprising when we realize that it’s part of a bigger picture.
So, both from recent polling and what we witnessed in 2020, it’s reasonable to assume that Hispanic voters may very well continue moving towards Republicans in 2024. And given how thin Biden’s winning margin was in Arizona in 2020, it’s not implausible that a modest shift among Hispanic voters could flip the state red.
It’s worth noting, though, that the Hispanic shift towards the GOP seen in 2020 was a bit less dramatic in Arizona than in other parts of the country. Specifically, in urban/suburban Arizona, Hispanic voters did not move towards Republicans quite as much as Hispanics did elsewhere. This might be a little too nerdy for some, but below, I have a map from Dave’s Redistricting, a program that I often use for electoral analysis.
In this map, I shaded in blue a section of Maricopa County that is almost entirely within Phoenix’s city lines. As seen in the box on the left (sorry if the numbers are hard to read, which I’m sure they are), this area in blue is home to a bit over 600,000 people, nearly 70% of whom identified as Hispanic according to the 2020 Census. What’s key here is that despite being overwhelmingly Hispanic, this group of precincts barely shifted Republican from 2016 to 2020; Clinton won these voters by 50 points in 2016, while Biden won them by about 49 points.
I’ve spent more time in this mapping program than I care to admit, and I’ve almost never come across an area that is majority Hispanic that didn’t shift at least 5 or so points towards Republicans from 2016 to 2020. The stability in Hispanic voting behavior, unique to Phoenix (in Tucson, the Hispanic shift towards the GOP was a bit more pronounced, though still modest compared to most other places), is incredibly noteworthy, and isn’t something I’ve really seen remarked upon elsewhere. This alone is reason to be skeptical of any analyst who argues that Arizona is bound to flip back to Trump based solely on the Hispanic shift to the right in 2020.
In contrast to Arizona’s urban Hispanic population, the state’s rural Hispanic voters did indeed vote more noticeably Republican from 2016 to 2020. The shift was quite significant, and in two majority-Hispanic counties along the Mexican border, Yuma and Santa Cruz County, Biden did about 5 and 12 points worse, respectively, than Clinton had in 2016. Most likely, this is yet another manifestation of college attainment being a major factor in voting behavior.
Ultimately, Yuma and Santa Cruz County, which combined have just slightly over 250,000 people, can only affect the statewide vote so much, even if they continue trending Republican. Of course, the fact that Biden generally held firm with Arizona’s urban Hispanic voters in 2020 does not mean that he automatically will again this fall, but he’d have much more reason to worry if the Hispanic population of Maricopa and Pima County had defected from Democrats, as seen in so many other places across the country.
In the end, Arizona is an interesting example of a state where most of the fundamentals (recent electoral trends, educational attainment, and the 2022 midterm election results) point to a clear positive trajectory for Democrats, despite what polling indicates. Even so, I would never tell someone to ignore poll results just because they contradict our preconceptions.
As such, I think Arizona should be thought of as a state that tilts Democratic this fall. The educated, suburban voters in its two major metro areas should provide Biden with real potential upside that could counteract his likely continued erosion with Hispanic voters. However, if the polls of Arizona still show Trump leading by a decent margin come June (the official end of primary season), I won’t hesitate to move the state to tilts Republican.
I plan to take the next few weeks off from writing, as I posted a bit more in February than I had planned, and I don’t want to flood people’s inboxes. My next post won’t be until mid-March, so until then, thanks so much for reading, and happy Super Tuesday, for those who celebrate!
Brad, An impressive trend analysis. As a part-time Arizonan, allow me to address two points that lead me to a different conclusion.
1, the 2022 midterm primary results were striking and key to understanding the outcome. Trump endorsed a whole slate of primary candidates for the statewide offices. Given all we know about low and unrepresentative turnouts in primaries, they all won their nominations, but they were terrible candidates. They were just objectively much weaker, less experienced, less known, less qualified than mainstream candidates like Beau Lane and Andrew Gould. But of course Trump's sole criterion for office was MAGA-loyalty. Democrats swept all these races in the general election, but this was not a demonstration of Democratic strength in Arizona, but of MAGA weakness. Trump runs weakly in AZ.
2, border insecurity. Arizona has been badly hurt by this huge influx, especially in the south. Biden gets most of the blame and certainly deserves some share of it. regardless of any broader trends in party preference, he will run well below par in AZ in 2024.
Trump weak, Biden weak. Another lesser of two evils choice? Maybe, and then I wouldn't hazard a guess who will be weaker. But the major parties' weakness will certainly mean a stronger showing for the Unity Ticket being offered by No Labels. Of course, we'll know much more after that ticket has faces, but you are talking trends. All else being equal, those trends favor neither Trump nor Biden, but a Unity Ticket. Win or lose, this won't be a two-horse match race this year, at least in AZ. Stay tuned.
Keep up your good work, and don't stay away too long. Owen