Contrary to popular belief, “May you live in interesting times” isn’t actually a Chinese curse. And even if it were, I can easily think of a lot of interesting times in which the good outweighed the bad.
So it doesn’t bother me that 2020 is turning into an interesting year. Today I happened to be in my old neighborhood and saw “Black Lives Matter” signs in front of a majority of houses on the street where I used to live, and that street is overwhelmingly white.
Polling by UCLA/Nationscape has shown a substantial shift in support for Black Lives Matter over the past three years. In June 2017 Americans who said they supported the Black Lives Matter were outnumbered by those who didn’t by a margin of -4 percentage points. Net support became positive a little over two years ago and has now reached +28 points, with a massive shift over just the past two weeks. Of course, over this time the general public has learned of more and more cases of African Americans being harassed, threatened, and even killed by white vigilantes and the police, increasingly often backed up by graphic video, making the problem more and more obvious. We've also seen law enforcement personnel violently assault peaceful demonstrators and journalists, very often including white people, for no justifiable reason.
On most other subjects polled — gun control, immigration, the environment, etc. — changes in opinion have been less dramatic, either because they have been modest or because they have reflected a strengthening of what was already the majority view.
For example, since June 2017 favorable opinion toward Russia as an ally have fallen from a net negative of -39 percentage points to a net negative of -59, which is a shift of 20 points but in the same direction. Support for legalizing marijuana has similarly risen from from +32 to +43 over that time frame, and support for same-sex marriage from 28 to 35. Both were pretty strongly opposed by majorities of the public in the past, but the big shift in opinion from opposition to support was some years ago.
The case of same-sex marriage is interesting in that much of the opposition was based on the notion that its legalization would somehow harm traditional marriage and the family. When the Supreme Court legalized same-sex marriage and nothing disastrous happened it was hard to maintain that belief. The 6-3 Supreme Court's ruling that the 1964 Civil Rights Act prohibits discrimination in employment on the basis of sexual orientation and gender identity may go the same way.
Another recent shift that can be said to reinforce a majority attitude is the drop in support for Donald Trump. A week ago Nate Cohn in The New York Times (echoed the same day by Greg Sargent in The Washington Post) pointed out Trump’s decline in approval ratings over the past couple of months. His net approval across fell from -6.7 percentage points on April 15 to an even worse -13.2 on June 7, based on an average of recent polls compiled by fivethirtyeight.com. As I write this it has deteriorated further to -14.3.
In keeping with this, Joe Biden’s lead over Trump in national polls has reached about 10 percentage points, up from 6 in March and April.
The Washington Post has a summary of recent polling as of earlier today, June 16, here. Among other things it notes that Joe Biden now leads in Michigan (which Trump narrowly won in 2016) by 12 points in one poll and by 16 in another. Ann Selzer, considered the gold standard for pollsters in Iowa, has Trump ahead in that state by just 1 point. (He won by 9 in 2016.) A May Fox News poll had Biden up by 8 nationally after being tied in April, pretty much in keeping with polls for The Washington Post/ABC and CNBC. No major poll shows Biden's lead declining.
Obviously the election is still a good 140 days away, and those polls could change in either direction in the next 20 weeks. For most of 2016 the polls showed Hillary Clinton in the lead and she ended up losing the Electoral College. On the other hand, she won the popular vote by an amount quite close to what the polls were predicting just before the election, and Biden is generally doing better than she was at the same point in 2016.
It’s particularly interesting to look at subsets of the population. In the fall of 2016 polls showed people 65 and older favoring Trump over Clinton by 5 points, but this time the 65+ cohort is pro-Biden by an average of 7 points, a net shift of 12 percentage points toward the Democrats. In 2016 women supported Clinton, the first female major party presidential candidate, by 14 points over Trump, but they now favor Biden over Trump by 25.
On the other hand, Trump is currently doing better among nonwhite voters than he did in 2016. Back then they preferred Clinton by 50 percentage points but Biden’s lead is a narrower (but still impressive) 45. Trump leads among white voters as he did in 2016, though by only 5 points now versus 13 back then. And as in 2016, it’s white voters without a college diploma who give him that lead. White college graduates prefer Biden by 20 points (up from Clinton’s 12 in 2016).
What about the electoral vote? Polls of the larger swing states conducted May 1 through June 12 showed Trump ahead by an average of 1.5 points in Texas and 1 point in Georgia, which is far worse than he did in 2016. But Biden leads in Pennsylvania, North Carolina, Florida, Arizona, Wisconsin, and Michigan, states that went for Trump last time.
The website electoral-vote.com, which aggregates an immense amount of state-by-state polling data (and also has an excellent if snarky daily compilation of political news), today shows Biden ahead of Trump in the electoral vote by 352 to 148 based on the the most recent state-level polls (with Texas — Texas! — shown as tied, something the people running the site seem not to believe even though that’s what the polls indicate).
What’s more interesting is the breakdown by the size of the candidates' lead in each state. If it’s at least 10 points the site rates the state as “Strongly" Democratic or Republican. If it’s at least 5 but less than 10 it’s considered “Likely.” if it’s less than five it’s “Barely.”
As I write this, Biden leads by at least 10 points in states totaling 229 electoral votes, and by 5-10 in states worth another 44. This means that if he wins only those states in which he has at least a 5-point lead and loses all the rest, he ends up with 273 electoral votes, more than the 270 needed to win. In contrast, Trump leads by at least 5 points in states with a total of just 104 electoral votes. To win Trump would have to take every state in which he currently has a lead, plus every state in which Biden currently leads by less than 5 points, and at least one additional state.
Again, the 10 fortnights from now until the election is a long time, and none of this means Trump won’t win a second term. In fact, there are a lot of pro-Trump Republicans who are predicting a Trump landslide. We'll see.