Electoral College Outlook: Biden Has the Edge

July 17, 2020 · 10:30 AM EDT

7/24/20, 12:30pm: This article has been updated with more recent polling averages.

A month ago, a public poll showed President Donald Trump leading Joe Biden by just 2 points in Arkansas. That would be easy to dismiss as an outlier, considering Trump won the state by 27 points four years ago, except for the growing mountain of data elsewhere that is evidence of the president’s increasingly steep climb to a second term. 

For three-and-a-half years, Trump’s job rating was arguably the most stable part of his presidency. With a committed and loyal base of Republicans for the president and a slightly larger committed and loyal group of Democrats against him, the country was on a trajectory to experience a close and competitive Electoral College contest with both parties fighting over a half-dozen or so key states. 

That outlook has changed.

While the precise cause can be argued, Trump’s job rating has been on a precipitous decline over the last two months, not only putting a second term increasingly out of reach but potentially wreaking havoc on GOP candidates down the ballot. 

On May 8, Trump’s job rating was 45 percent approve/51 percent disapprove, according to the RealClearPolitics national average, a difference of 6 points. On July 23, Trump’s job rating was 42 percent approve/56 percent disapprove, a difference of 14 points. It’s a similar story on the ballot test. Biden led Trump 47-42 percent in early May in the national average compared with 50-41 percent in mid-July. 

Of course we don’t have national elections in this country. That should have been one of the biggest lessons from 2016, not that we should dismiss or distrust all polling data. But the survey data in individual states and districts are no better, and arguably worse, for Trump. 

For example, the president won Montana by 20 points in 2016. Public and private survey data there show Trump and Biden are within a few points of each other. It’s a similar story in Alaska, which Trump won by 15 points and  looks like a competitive state today. That doesn’t mean that Montana, Alaska, or Arkansas are swing states, or will even be close in November. But it does cast considerable doubt on the president’s ability to win previously agreed upon swing states that have much less of a GOP cushion.

Four years ago, Trump narrowly won Michigan and narrowly lost Minnesota. He’s likely down double-digits in both today. States he won by wider margins, such as Arizona, North Carolina, Texas, Georgia, Iowa, and Ohio are all at considerable risk for the president. Biden probably just needs to win one of them to secure the White House.

Analyzing a combination of partisan and nonpartisan, public and private, national and state-level polling, we’re changing our presidential rating in 17 states- all in favor of Biden. With those changes, Biden leads Trump in our Electoral College projection 319 to 187, when 270 is needed to win. There are some key states, such as North Carolina, where Trump doesn’t appear to be hemorrhaging voters at the same rate, but there just isn’t significant evidence that there is a single state getting better for Trump right now.

Biden’s total includes winning all the states Hillary Clinton carried in 2016, along with Michigan, Pennsylvania, Arizona, Florida, Wisconsin, and Nebraska’s 2nd District — all of which Trump carried four years ago. Georgia, North Carolina, and Maine’s 2nd District make up the Toss-up category. Iowa, Ohio, and Texas join the Tilt Republican category, while Alaska and Montana are now rated Lean Republican. The Likely Republican category is evidence of the expanding softness in Trump’s support and now includes Kansas, Missouri, South Carolina and Utah. There could be more candidates for this category, but there just isn’t a lot of data in most of the other states assumed to be going Republican.

Critics will claim that Inside Elections has counted Trump out and that we’ve declared the president can’t win. That is not true. Just like the race wasn’t over when neither candidate was projected to win more than 270 votes, the race isn’t over when current projections put Biden over 270. But the fear of projecting the race incorrectly shouldn’t cause us to ignore the preponderance of data which show a Biden win is more likely than a Trump win at this stage of the race. 

Ah, but what if the data are wrong? The question about “Shy Trump” voters comes up almost as often as the president tweets. Qualitatively, it seems like the president has emboldened his supporters, and people who might have been sheepish about vocalizing their support for Trump four years ago are now happy with what he’s done in office and unafraid to tell the world. Quantitatively, multiple GOP pollsters admit that those voters could still exist, but make up potentially 1 to 3 percent of the electorate and would have to manifest themselves in just the right states to matter.

On a broader level, to dismiss the current presidential data would be to assume that virtually all pollsters (partisan and nonpartisan) are independently making the same methodological mistake in the same direction. That’s possible, but not likely. Up to this point, private GOP polling doesn’t paint the picture of a fundamentally different race, just closer margins in some areas. 

Rather than argue about the current state of play, Republicans will argue about the future. They believe the economy (and the public’s confidence in the president on the economy) will supersede other issues, Trump’s law and order approach will assure skittish independent voters, and that Biden will make a fatal mental mistake that will discredit him as a credible alternative for voters looking for a change. 

Not only is there considerable doubt the economy will be able to recover enough by the fall for that dynamic to benefit Trump, positive coronavirus cases continue to climb higher around the country, threatening whatever plans for re-opening are in the works and even causing governors to send their states back into various degrees of lockdown. 

And the president has shown an inability to focus on the economy, the one area where voters still may view him as more capable than Biden. He keeps coming back to cultural issues, such as statues, Confederate battle flags, even canned beans, and settling scores with his critics. Voters just don’t trust him (compared to Biden) to handle some of the biggest issues facing the country right now including coronavirus and race relations. More than anything, it appears the president has lost the benefit of the doubt with a majority of voters, and that is difficult to recover.

Republicans and political observers will also bring up time as the wild card. “Four months is an eternity,” is the common refrain. But it’s not clear Trump and the Republicans have enough time to significantly recover. Just like the president’s decline has been steady, his recovery would likely take time in the other direction. With the polarized state of the electorate, it’s very unclear what single news event would cause Trump’s standing to improve dramatically, considering 50 percent of registered voters said there was “no chance at all” they might support the president, according to the latest NBC News/Wall Street Journal poll.

And Trump doesn’t have until Nov. 3 to recover. Four years ago, approximately 40 percent of voters cast their ballots before Election Day. That percentage is likely to grow significantly with increased access to mail-in balloting in response to the coronavirus, and some voters will start receiving their ballots as early as September, as explained by Grace Panetta of Business Insider. 

If the data does change dramatically between now and November, then we’ll adjust our ratings and analysis accordingly. These are projections rooted in data, not predictions of the future. 

The biggest mistake of 2016 was not underestimating Trump’s support but a failure of imagination. Too many people couldn’t comprehend how he could win. Four years later, too many people can’t comprehend how he could lose and underestimate how far he could possibly fall. 

Rating Changes

Alaska from Solid Republican to Lean Republican
Arizona from Toss-up to Tilt Democratic
Florida from Toss-up to Tilt Democratic
Georgia from Lean Republican to Toss-Up
Iowa from Lean Republican to Tilt Republican
Kansas from Solid Republican to Likely Republican
Maine (statewide) from Likely Democrat to Solid Democratic
Maine’s 2nd District from Tilt Republican to Toss-Up
Minnesota from Lean Democratic to Likely Democratic
Michigan from Tilt Democratic to Lean Democratic
Missouri from Solid Republican to Likely Republican
Montana from Solid Republican to Lean Republican
Nebraska’s 2nd District from Toss-up to Tilt Democratic
Ohio from Likely Republican to Tilt Republican
Pennsylvania from Tilt Democratic to Lean Democratic
South Carolina from Solid Republican to Likely Republican
Texas from Likely Republican to Tilt Republican
Utah from Solid Republican to Likely Republican
Wisconsin from Toss-up to Tilt Democratic