Welcome to Pollapalooza, our weekly polling roundup.
Poll(s) of the week
Joe Biden, the presumptive Democratic nominee for president, leads President Trump in most national polls, and surveys conducted even this far out have tended to roughly resemble the eventual general election results, as FiveThirtyEight’s Geoffrey Skelley explained in an article this week. Of course, national polls measure the national popular vote, which is really only indirectly related to who will win the White House — Democrats have won the popular vote but lost the Electoral College in two of the last seven elections and could do so again in 2020. U.S. presidential elections are really a contest of states.
So what’s going on in the states? We thought it was worth taking a look at that question now, as April featured some high-quality, state-level polling (though not a ton of it1). It also happened to be the month when the general election came into sharper focus: It’s now clear who the Democratic nominee will be and that the novel coronavirus pandemic will be a major part of the campaign.2
Here’s what we learned looking at every state-level poll conducted in April:3
Biden’s leads narrowly in several key states
Several polling firms released surveys of Florida, Michigan, Pennsylvania and Wisconsin in April. Former President Barack Obama carried all four states in 2012. Trump flipped all four in 2016 (as well as Ohio and Iowa, neither of which has much recent polling.) And Biden appears to lead in all four now. (North Carolina, which has gone Republican in both of the last two cycles, was also polled pretty often in April, with Trump and Biden looking basically tied there.)
|State||Number of polls||Biden||Trump||Average Margin|
This data highlights a few things. First, at least at the moment, Florida, Michigan, Pennsylvania and Wisconsin are very close to the national tipping point — so they’re likely to be among the more determinative states this November. Second, the former vice president’s lead nationally is big enough to carry these states. This is important — if Biden wins all of the states Hillary Clinton won in 2016 plus any combination of three of these four, he would be elected president.
But crucially, Biden’s margins in these states are slightly smaller than his advantage in national polls. It’s worth thinking about the race at the state level in these relative terms because there’s still so much time for things to shift. If Biden’s lead nationally narrowed to 2 to 3 percentage points, these states would likely be much closer, if not lean toward Trump. Also, as The New York Times’ Nate Cohn wrote recently, Trump is likely to look stronger when pollsters start limiting their results to “likely voters.” Most of the April surveys in these four states were conducted among registered voters or all adults, two groups that include some people who may not vote in November.4
In other words, this data suggests Trump may have an Electoral College advantage again — he could lose the popular vote and win the election. Of course, this data also suggests that if Biden is winning overall by a margin similar to his advantage now, Trump’s potential Electoral College edge really won’t matter.
Biden polling particularly strong in Michigan and Pennsylvania
It’s not worth making too much of small differences among states — we don’t have a lot of polling and things will likely shift between now and November anyway. But it seems pretty clear Biden is doing better in Michigan and Pennsylvania than Florida and Wisconsin. Fox News, Ipsos and Public Policy Polling all recently polled several swing states. All three found Biden had a larger lead in Michigan and Pennsylvania than in other swing states they surveyed (Florida for Fox, North Carolina and Wisconsin for PPP, Wisconsin for Ipsos.)
The race is unclear in other Trump 2016 states
Outside of the five states I discussed above, a recent poll of Arizona was great for the former vice president, while he wasn’t looking as strong in a survey of Georgia. But I would be very cautious about states in which we only have one or two recent polls. So please keep that caveat in mind, but here are all the states for which we got polls in April — even the ones with just one poll:
|State||Number of polls||Biden||Trump||Average Margin|
I do think Biden could win Arizona, but that’s based more on Democrats’ winning a U.S. Senate race there in 2018 than the recent OH Predictive Insights survey that showed Biden with a 9-point lead in the state.
It’s also unclear in Clinton 2016 states
It’s likely that Trump and Republicans will target Maine, Minnesota, Nevada and New Hampshire in particular — looking for an upset in states Democrats could overlook because they might assume Biden won’t lose where Clinton won four years ago. Trump narrowly lost in those four states in 2016.5 Maine, Minnesota and New Hampshire are states in which a disproportionately high percentage of the electorate is non-Hispanic white voters without college degrees, an electoral bloc in which the president was especially strong in 2016 and could make even more gains in 2020.6
But we just need more polling to get a sense for how these states are shaping up relative to the country overall. All we have is one recent poll from any of these four states — a survey showing the former vice president ahead by 8 points in New Hampshire. Biden’s prospects would be brighter if he ends up outperforming Clinton among non-college whites, making these states less Republican-leaning relative to the country than they were in 2016.
Other polling bites
- Among registered voters in New York, 32 percent said they personally know someone who has died as a result of the coronavirus, according to a new Siena College Research Institute poll, compared to 67 percent who said they did not. That first number is higher among the state’s black (48 percent) and Latino registered voters (52 percent), as well as those who live in New York City (46 percent.)
- About 37 percent of American adults said they had a positive view of capitalism, compared to 41 percent with a neutral view and 21 percent who viewed capitalism negatively, according to a new poll from the left-leaning firm Data for Progress. Socialism is less popular, at 18 percent positive, 44 percent neutral, 38 percent negative.
- More than half of American adults said their employment status had been negatively affected by the COVID-19 outbreak, according to a recent Gallup poll. The effects included a loss of income (26 percent), hours being reduced (15 percent), being temporarily laid off (10 percent) and losing their job permanently (2 percent). Of the adults who were surveyed who live in households with incomes below $35,000, 68 percent said they had experienced at least one of these four negative effects. Among those in households with incomes over $90,000 per year, 49 percent said they had experienced at least one.
- 30 percent of American adults said they thought Trump was kidding when he suggested that scientists study the idea of people injecting disinfectants into their bodies as a way to fight the coronavirus, according to a recent YouGov survey. The majority of Americans (54 percent) did not think Trump was joking. There was a clear partisan split: 68 percent of Republicans said the president was kidding, compared to 16 percent who said he was not and another 16 percent who said they were not sure.
- Black adults (28 percent) were more likely than Hispanic (13 percent) or non-Hispanic white (9 percent) adults to say they know someone who has died from COVID-19, according to a recent Axios-Ipsos survey. People in the Northeast (25 percent) were more likely to know someone who has died of the virus than those in the South (12 percent), Midwest (8 percent) or West (7 percent.) Democrats (15 percent) were more likely to know someone who has died than Republicans (8 percent.)
- A recent Washington Post/University of Maryland survey found 65 percent of American adults supported temporarily blocking immigration to the U.S. during the COVID-19 outbreak. (Trump recently enacted a temporary halt to the issuing of green cards, although the pollsters who conducted this survey did not refer to the president’s announcement in their question.) Republicans (83 percent) and independents (67 percent) overwhelmingly favored this idea, while Democrats were split (49 percent supported it, 49 percent opposed it.)
- Relatively few registered voters (31 percent) said they thought Trump is “honest and trustworthy,” according to a new USA Today/Suffolk University poll. But Trump gets higher marks on standing up for American interests (53 percent of voters agreed with this notion), knowing how to get things done (51 percent), being a strong leader (45 percent), working with foreign leaders (45 percent) and caring about “people like me” (39 percent.) Biden was rated the highest by voters on being able to work with foreign leaders (64 percent said they agree with this notion) and was ahead of Trump in three other measures as well. But voters weren’t that confident that Biden is a strong leader (43 percent) or can get things done (48 percent.)
- A total of 67 percent of adults said they thought COVID-19 would “significantly disrupt” Americans’s ability to vote in November’s elections, according to a new Pew Research survey.
According to FiveThirtyEight’s presidential approval tracker, 42.9 percent of Americans approve of the job Trump is doing as president, while 52.4 percent disapprove (a net approval rating of -9.5 points). At this time last week, 43.4 percent approved and 52.6 percent disapproved (for a net approval rating of -9.2 points). One month ago, Trump had an approval rating of 45.8 percent and a disapproval rating of 49.7 percent, for a net approval rating of -3.9 points.
In our average of polls of the generic congressional ballot, Democrats currently lead by 7.7 percentage points (48.0 percent to 40.3 percent). A week ago, Democrats led Republicans by 7.6 points (48.1 percent to 40.5 percent). At this time last month, voters preferred Democrats by 7.8 points (48.9 percent to 41.2 percent).