But we can and should do everything we can to take the pulse of the nation. We went to Jennifer Agiesta, CNN's director of Polling and Election Analytics, who runs CNN's decision desk on Election Night and knows as much or more about how to actually read polls than anyone else.
Her tips on what to look for when you read a poll are worth reading (and bookmarking). Our email conversation is below.
How real is Biden's lead?
What Matters: Polling consistently shows Joe Biden with a clear lead over Donald Trump. Is there any precedent for a presidential candidate coming back from that kind of deficit four months from Election Day?
JA: It is a strikingly large deficit this close to the election, but as close as we are to the election, there is still a lot of time for opinions to shift, and if 2020 has proven anything it's that the news can give you whiplash at just about any moment. And large shifts in the presidential horse race can happen late in the game if the news is big enough.
In 2008, for example, CNN polling found a near neck-and-neck race for much of the year, until the financial crisis really hit in mid-September. Just after Labor Day and the Republican convention, the race stood at 48% for Barack Obama to 48% for John McCain among registered voters. After the stock market crash, Obama quickly gained a double-digit advantage over McCain among that same group, maxing out in CNN's polling at a 14-point Obama lead in an October 3 through 5 poll. That poll came less than a month after the tied result. There wasn't a single reputable national poll which found Obama behind McCain for the remainder of the campaign.
Beyond the ability of news to change voters' perspectives, it is important to remember that much of the polling which shows Trump lagging so far behind Biden is national polling, while US presidential elections are decided state-by-state via the electoral college. Running up the score in population centers like California and New York wouldn't do Biden much good in his bid to take the White House. Recent polling in critical battlegrounds seems to confirm what the national polling is showing at the moment, however, with Biden holding edges in several of the states he must win in order to gain an electoral college advantage and competitive in states where the President won handily last time around. That all suggests that Biden's current edge is a meaningful one. But polling in some of those same states just a few months ago found much tighter contests, so again, there is room for things to change.
It is also worth noting that at this point in the race, pollsters are largely measuring the opinions of registered voters. Typically, looking only at those who are likely to vote rather than the full pool of registered voters yields results which are slightly more favorable to the Republicans. Polling in 2018 found a different pattern, though, and it's possible that 2020 would continue that shift. At this point in time, it is really too early to say with any certainty who is a likely voter and who is not. That's true in a normal year, but in the midst of a pandemic, many Americans don't even know yet how elections will be administered where they live, let alone whether they are comfortable participating in them.
The pandemic election
What Matters: How is coronavirus, which has touched every part of American life, affecting voters?
JA: For one thing, it's changing the way Americans vote and the way that election results are reported. Recent primary elections in New York, New Jersey, Kentucky and elsewhere have featured a sharply increased share of the public choosing to vote by mail rather than in-person at a polling place. Given the complexities of counting and verifying by-mail votes, election officials are taking longer periods of time to release results from those elections. What we normally think of as Election Day might stretch out in to Election Week or Election Month if similar changes in how people cast votes happen in November.
I think the impact it could have on who they choose to vote for is still unknown depending on what happens with the outbreak between now and November, though our colleague Harry Enten has some theories.
How solid is Trump's base?
What Matters: One theme a lot of people took out of the 2016 election was that Trump won because he consolidated the support of White, working-class voters. That helped him win the Rust Belt and the White House. Has that group, broadly speaking, stayed with him?
JA: White voters without college degrees remain one of the groups most apt to back Trump in most polling. There are several question marks around this group, however.
The first is whether Trump can maintain his support among White women who do not have college degrees. The President's ratings among women generally are abysmal, and if that carries over to White women without degrees, it could be enough to sink the President in some places.
The second is how large a share of the electorate they will be in critical states this fall. States across the Sun Belt which previously were reliably Republican are seeing their share of White voters without college degrees decline, a shift which may make some of those states more competitive.
In the Rust Belt, Whites without college degrees generally remain a large share of the voter base, so can have more of an impact on the outcome of elections there.
How to read polls
What Matters: I'm always very focused on the horse race numbers in polling, but I've heard you say that's not the best way to think about polls. What broad themes should people be pulling out of recent polls?
JA: The horse race isn't the worst thing one could look at in a poll, so if you're going to pay attention to it, do yourself a favor and look for a few things beyond the margin between the two candidates:
Does the leading candidate have majority support? If neither candidate is over 50%, that suggests their support is a little softer than it may need to be to win.
How many voters are undecided? It could be a sign of a volatile electorate (and I think was underplayed in 2016) but it can also vary based on methods (many online polls have higher undecided shares than phone polls).
What's the general trend for an individual candidate's support across quality polls? One notable thing in recent polling has been how consistent support for the two candidates has been. Across the nine national phone polls which meet CNN standards that have been released since June 1, seven pegged his support between 41% and 44%, Biden's support has a slightly wider range, but generally hovers between 50% and 55% in those same polls, with two placing his support at 49%.
Looking outside the horserace. Watch to see which of the most pressing issues facing the country right now hold on as meaningful for voters this fall, whether Trump's approval rating for handling the economy remains in positive territory, and whether any of the attacks on Biden dent his strong ratings for leadership. On the issues that matter most to voters, notably, CNN's June survey found two-thirds of Americans considered racism in our society a big problem. That was up sharply compared with the 2016 presidential election cycle, while 42% of registered voters called race relations extremely important to their vote for president, the highest number for any of the four issues tested in the poll and also well ahead of the share who said the same in the previous presidential cycle. Coronavirus rated below race relations (31% called that extremely important).
Are there actually any undecided voters?
What Matters: I don't know anyone who doesn't have an opinion about Trump. Is there a universe of undecided people who will decide the election?
JA: Not very many. And the downside for Trump is that as an incumbent, views on him are likely to matter far more than views on Biden. One of the most fascinating things about the 2016 exit poll results was that voters who view both Trump and Hillary Clinton unfavorably broke heavily toward Trump, and there were even a decently sized set of voters who felt Trump was not qualified to be president, but still voted for him.
As a sitting President, it's unlikely that Trump would get the benefit of the doubt from those voters again has he not convinced them he deserves to remain in office.
Everything is about Trump
What Matters: My former colleague Amy Walter has said pollsters should not only ask who voters support, but also if they're enthusiastically voting against someone (Trump). What's your thought on that?
JA: The last CNN poll conducted by SSRS included a very similar question. Biden supporters were asked whether their preference was more a vote for Biden or against Trump, and Trump supporters were asked the same about their support. Among Biden's backers, 60% said their vote was against Trump, while among Trump supporters, 70% said their backing was to vote for the president.
Looking at Biden supporters by their enthusiasm to vote, those who are most fired up about going to the polls in November are a bit less likely to say that their vote is in opposition to the President (56%) than are those who say they are somewhat enthusiastic about voting or less (69%). Whether that anti-Trump sentiment is enough of a motivator for that group to show up at the polls (or mail in their ballots) will be a key question to watch through the fall.
One last question
What Matters: I'm asking for a wild guess here, but what are the odds we know the identity of the next President on Election Night?
JA: Oh to know the answer to that question!
At this point, it seems unlikely that we'll know the outcome on Election Night itself. While a number of key battlegrounds are used to processing a lot of absentee or early votes (Florida and Arizona most notably, where majorities typically vote absentee or early), several other critical states for the electoral college count are relatively new to the absentee ballot process. In Pennsylvania, for example, just 8% of voters cast ballots absentee or early in 2016, in New Hampshire, it was 10%, in Virginia, 13%, and in Michigan, Minnesota and Wisconsin, it was around a quarter of voters -- which suggests their counts could be slower than usual if those numbers spike dramatically.
One growing trend in recent elections is that vote preferences among those who vote in-person are often different from those who vote absentee or early, and the politicization of behavior in the coronavirus era seems as if it could exacerbate those differences, making it difficult to project races until there is a solid number of reports from both groups of voters.
"right" - Google News
July 11, 2020 at 06:46PM
https://ift.tt/2ANwry4
What we know right now about the November election - CNN
"right" - Google News
https://ift.tt/32Okh02
Bagikan Berita Ini
0 Response to "What we know right now about the November election - CNN"
Post a Comment