Politics with The Fix: What do you want to know about the latest polls?

Sep 17, 2020

Got a burning politics question, or just something you’re curious about? Each week, starting Thursday at noon, The Fix team chats with readers about the big stories in politics.

This week, the team was joined by Scott Clement and Emily Guskin, who run The Post's polling team. Their latest Post-ABC poll found Biden narrowly leading over Trump in Wisconsin, but ahead by a wide margin in Minnesota. Another recent poll found that most Americans want to vote before Election Day.

Emily and Scott work on deciding how samples for surveys are drawn, what questions are asked and analyze what the results show, in collaboration with polling partners, including ABC News. They also help others in the newsroom write about polls.

You're reading a transcript of our chat. Check out our previous discussions.

Sign up for our newsletter, The 5-Minute Fix, a must-read politics cheat sheet sent every weekday afternoon.

Understanding polls takes more than just "yay/nay" -- what other kind of questions help out in interpreting the voter's perspectives?

Great question! You're right, it's not just horserace. We like to ask all sorts of questions to get voters' opinions. For one, we ask standard demographic questions like region, gender, race/ethnicity and age. That allows us to run crosstabs on substantive questions -- like vote choice -- and see how different groups of people think and compare them to previous polls. We also love to ask issue questions -- what do Americans think about various issues? Do they support or oppose particular policies? These questions together help us get a good view of Americans' opinions.

A good example of this was a poll out yesterday that Emily and Scott worked on, asking voters in Wisconsin whether they're worried they or their family members might catch coronavirus. They found that, yes, a substantial amount are. And of those worried, they largely back Biden over Trump. That's helpful information if, say, closer to November there's a coronavirus outbreak (or the opposite) in that important swing state.  

Lots of us lost confidence in polls after 2016. Did they really get it wrong in 2016 or was it a "margin of error" thing? Why are the 2020 polls more reliable? Bob Feldman Baltimore, MD

I certainly understand this perspective, especially with some election forecasts that gave Clinton a greater than 90% chance of victory. On poll accuracy, national polls were pretty accurate but state polls did underestimate Trump's support across a number of important states. The reasons? A post-election report by the American Association for Public Opinion Research (which I took part in) looked at several factors, but noted that many state polls did not weight samples by education as they did with other demographics. This led to an under-representation of voters without college degrees, including White voters without college degrees who supported Trump by a wide margin. Exit polls also suggested late-deciding voters helped Trump, a group that polls always struggle to reach.

Are 2020 polls better? Many polling firms have made improvements, including weighting samples by educational attainment. But one big takeaway from 2016 and past elections is that state polls are historically less accurate than national polls generally, so we should not expect them to be similarly precise.

Scott is definitely the authority on this, and I second many of his points. I've also written about how I think the criticism is overzealous.

Another part of the issue in 2016 was the sheer dearth of polls in the states that wound up being decisive. We'll have a lot more data in those states this time around, which will allow us to parse the polls better and provide more relevant polling averages in those states.

But caution is always the watchword.

What are the chances a big name (Bush, Mattis, Kelly) comes out and publicly endorses Biden between now and election day?

Low. George W. Bush refused to endorse a candidate in 2016 and there is little evidence that his stance has changed in 2020. Mattis has been more willing to publicly criticize Trump, but that doesn't mean he will endorse Biden. Kelly echoed Mattis's criticism earlier this year, but has been noticeably silent following recent reporting about Trump disparaging veterans.

I'd say if anyone did it, it would probably be Mattis. Bolton hasn't been willing to go that far either, but he's more ideologically entrenched in the GOP.

The holdup with Mattis would be that he wouldn't necessarily think it's a military man's job to endorse -- so much as say what he knows about the candidates. Even before doing that, he made clear he was reluctant to get involved.

Any chance for Amy?

The most recent poll had her down by 12 points -- which, alongisde a bunch of other much-rosier polls for Dems in more competitive states would seem to argue for Democrats focusing their priorities elsewhere. I'm just not sure how the presidential race will be close enough for Dems to pull a Senate seat in Kentucky. In a non-presidential year, it'd be a more logical proposition.

I care about Senate races. We know the big four -- Arizona, Colorado, North Carolina, and Maine (with Alabama all but certain to flip to GOP). Dems seem favored in Arizona and Colorado. Tillis seems to be struggling in North Carolina, but the race is still considered a tossup. And both sides are projecting confidence about Maine, so who knows what to believe. But how are Democrats looking in the next tier of races? Iowa has moved right in recent years, but Democrats had a strong 2018 and got a decent candidate in Theresa Greenfield. They got their A-list candidate in Steve Bullock in Montana, but Montana is still Montana. Jon Ossoff is probably a C-level candidate, at best, in Georgia, but the state is moving toward the Democrats, though, the GOP can count on the 50%+1 rule to help. And then there's Alaska, which is somewhat quixotic politically. So where do Democrats stand in those four states?

I just looked at the top 15 or so races. I'll put it this way: One Democratic strategist said she predicts the battle for the Senate majority will come down to Democrats toppling Susan Collins (R) in Maine and Joni Ernst (R) in Iowa. It's possible -- Collins is starting to look more vulnerable in recent polls this week -- but Ernst is polling in a new AARP poll at 50%, which makes Republicans feel good.

I go into more detail on the other races Democrats have opportunities in to take back the Senate majority. Right now I have Montana and the main Georgia Senate races (there's also a special election) as toss ups. The rest are stretches for Democrats -- Georgia's special, South Carolina (where two back-to-back polls have showed the race tied). 

Are shy Trump voters a real thing to any significant degree, or just wishful thinking on the part of Republicans?

For the most part, wishful thinking. Pew looked at this after the 2016 election and found "little to no evidence" to support this theory. My colleague Philip Bump further debunked this theory in a piece yesterday.

The most recent polling I've seen had Gideon up by 12. The Cook Political Report still has it as a toss-up. How do you see it?

I'm definitely skeptical of the idea that she's down 12. That same pollster a month and a half ago had her down just four.

It's possible the race has shifted that much, given we've seen some other similarly good swings for Democrats in some other key states, but it's too early to tell. 

I would say that, given Gideon has led virtually every recent poll, I'd be inclined to maybe put it at "lean Democrat." As someone who has rated races before, the inclination is to put anything close in the "toss-up" category, but when one candidate consistently leads ...

I keep reading that mail-in ballots cannot be counted until after Election Day? Why not? I also read once, several years ago, that they are not counted at all if they won't make a difference in the outcome. Who decides that they won't?

It depends on the state. Different states have different rules on when mail-in ballots can be counted.

The National Conference of State Legislatures has a helpful guide on when each state can begin counting absentee and mail ballots.

There’s been speculation that polls can’t capture “shy voters” — that enough Trump supporters would refuse to take a poll ran by the “fake news” that it would affect the result, even if the poll is not with a live caller. Is there any evidence to back this up, and are there any precautions taken to account for the types of people who might refuse to answer polls based on perceived political bias?

While polls underestimated Trump's support in key states in 2016, there is little evidence to back the "Shy Trump" or "Shy Conservative" voter theory. The American Association for Public Opinion Research evaluated 2016 pre-election polls and conducted several tests of this question, including comparing surveys administered by a live phone interviewer to web-administered polls where reluctant Trump supporters might have felt more comfortable. They found live interviewer polls did not consistently underestimate Trump's support, just one test of this theory. Our colleague, Philip Bump published a piece about this issue yesterday.


Read it - didn't agree at all - but curious as to the timing, when it had been announced Trump wanted the bigger package now. With Pelosi on TV every single day touting the Republicans' intransigance I thought the pressure was solely on the Republicans. I also didn't get your point about swing districts, since no one I've read in the mainstream thinks the House is going back to Republican Majority this cycle.

You're talking about my analysis of why House Democrats have good reason to be anxious about no coronavirus relief bill.

In the piece, I say that Republicans are a longshot to take back the House majority that Democrats took in 2018, but Democrats don't want to lose hard-won seats that helped them win that majority. The longer they can keep holding districts like Oklahoma City or rural New Mexico or in South Carolina, well, the longer they can keep them.

As for the president saying he wants a bigger package, at first it was totally unclear what he meant. How big? Would he meet Pelosi at $2 trillion? He clarified last night, and I've since updated the story, that he seems to be supportive of a bipartisan proposal of about $1.5 trillion.

If Trump can convince skeptical Republicans to spend that much, I think it helps Republicans message that Democrats are the one being intransigent, kinda taking away a message that Democrats had relied on since they passed their $3.5 trillion package in the House in May. 

Why would they vote for Trump? What has he done for them? Or what do they think he has done for them?

Polls show that many Latinos in Florida that back Trump identify as white, the one racial/ethnic group that has consistently supported the president since his earliest days in office. 

Many Florida Latinos are from families that came to the U.S. from communist countries. When Trump claims that the left embraces socialism and wants to move the U.S. towards a direction that resembles the Latin American countries that the families of these voters fled, many of these voters decide that the GOP is a more attractive option for them.

Lastly, many Latino voters are Catholic, evangelical or practice another form of conservative Christianity that takes positions on abortion, LGBT issues and marriage that are more consistent with the GOP. 

Ultimately, these reasons lead these voters to back Trump.

Is Harrison going to be a threat to Graham?

I'd say he is a threat to Sen. Lindsey Graham (R-S.C.). 

Democrat Jaime Harrison is raising tons of money, and Democratic strategists say he's not just making inroads with the Black, more Democratic-voting community in South Carolina, but also White moderate voters. And there's evidence that some of these voters think Graham is too much of a Trump apologist. Two back-to-back Quinnipiac University polls (read: high quality) have had the race tied. 

But! I agree with Republican strategists when they question whether there are enough Democratic-leaning voters to get Harrison over the top and to win. South Carolina is not Georgia yet.

For more, check out this story exploring the race in detail, with interviews from Harrison. 

Are polling results adjusted for state districting? I keep hearing that Biden's numbers must be reduced by 5-6 points because of the gerrymandering of voting districts. If polls are not adjusted for this, Why not? By not adjusting, you may be contributing to a false sense of security that may keep some voters from voting, a la 2016.

The presidential election in each state is decided based on who wins the popular vote in the state and in all but two states, electoral votes are allotted to whichever candidate wins the majority of votes in that state. In Maine and Nebraska, electoral votes are allocated based on Congressional district, but those are the only two states where electoral votes are distributed this way -- and I'm not familiar enough with the Congressional districts in those two states to judge their level of gerrymandering.

What's the best way for Biden to respond during the debates when Trump tells 30 lies in one answer ("Biden will shut us down, kill the suburbs, create anarchy in the streets, and more people will die" etc.). How does he respond without enumerating each lie, and sounding defensive? We all know the regular lines Trump always gives and will likely repeat in the debates. How does Biden counteract that effectively?

I think Biden is struggling with how to do this! I was reading an Associated Press article on this the other day. 

Biden has said he wants to be a fact checker for the president, but he more recently said "people know what a liar" the president is and there's no need for Biden to always point that out.

Biden also said he didn’t want to get pulled into a personal fight on issues like family and appear angry onstage. “I hope I don’t get baited into a brawl with this guy, because that’s the only place he’s comfortable.

Hello, is there evidence that Gen Milley was aware of the government's interest in using the heat ray weapon on Lafayette Square protesters in June?

At this point, that's unclear. Milley has previously said he did not know who gave the order to clear protesters from the square. He has also said he did not know about Trump's planned church photo but did know that they would be "observing damage" at St. John's Church. I previously walked through what various officials have said they knew about the operation. The Fix will continue to update that post if new information is revealed.

Given that the electoral college chooses the president based on state-by-state voting, is there any significance to the national polling numbers and is there generally any correlation between those numbers and those of the swing states?

It's good to see what the nation as a whole thinks! And the president is the president of all of America so our opinions as a country matter. And not just about vote choice but the myriad issues that affect us every day, from the coronavirus to racism to the economy. So yes, it's important to poll Americans nationally even if that's not how the U.S. elects a president.

The Attorney General yesterday stated that he has the right to overrule any individual prosecutor on a case. This is quite true but normal procedure would be to make the determination as to whether to prosecute before the case is tried. It is improbable that a lowly prosecutor determined to try,, say, Roger Stone without the hierarchy in DOJ being aware. This is why I wonder why Barr intervened so late in the process, after the trial and conviction of Stone.

Well, in the Stone case, the intervention dealt only with the sentencing recommendation, which came after he was convicted. Barr said even afterward that Stone's prosecution was "righteous"; he just disagreed with how much jail time it warranted.

And in the case of Michael Flynn, the prosecution was brought before Barr was attorney general. So maybe he could have intervened earlier, but he couldn't have overruled it from the start.

this year, that they have made some tweaks after 2016 that will improve their accuracy? Or is it more valuable to educate people about what the MOE really means in a close race?

Knowing what margin of error means is super important to understanding what the difference is -- and whether there even is a difference -- between two candidates' level of support in a poll. The Pew Research Center wrote a helpful guide for understanding margin of error.

But to get to your larger question that Scott also answered above, many polling firms have improved by weighting samples by educational attainment.

Great news that Louisville will pay up for Breonna Taylor's unjustified killing hopefully, there will also be changes in law enforcement. I was horrified this morning when reading the account of officials wanting to use a microwave- type weapon against the protestors in Washington. Some officials seem unfazed that it had not been used on humans before because of "humanitarian" reasons, but they thought it would be okay to use it on American protesters. Who are these people?

One thing this summer has made more clear is how lowly some people view demonstrators and how few protections some people believe that Americans exercising their right to  protest their government should have. I expect this to increasingly be an issue moving forward as police violence against black people continues to be an issue that moves people to gather in the streets seeking justice.

As a Fix Chat rookie, I accidentally deleted an interesting question about whether polls have become less accurate over time. There were a couple big studies on this after the 2016 election, and their basic conclusion was "no." In a study of 351 elections in 45 countries, political scientists Will Jennings and Chritopher Wlezian found “there is no evidence that poll errors have increased over time, and the performance in very recent elections is no exception.” State polls in 2016 did have larger and more systematic errors than in previous presidential elections, although the 2018 midterm elections were fairly accurate. More about that study and an analysis by Nate Silver here.

Since independents make up a substantial part of the electorate, why do we see polls only report how democrats and republicans view various subjects.

While the analysis of the polls may focus on partisans, all polls include independents too. We could all probably do a better job of focusing on them, but oftentimes the partisan differences are the most interesting.

Your response to my question about Latinos makes sense. I had not considered their country of origin or religious beliefs

Thanks for coming to our chat! :)

What chance does the GOP have of locking Democrats out of the runoff for the special election in the Georgia Senate race? Warnock doesn't seem to be pulling away from Lieberman and Tarver enough.

Actually, a recent AARP poll of likely voters has one Republican (Sen. Kelly Loeffler) leading at 24 percent, another Republican (Rep. Doug Collins) behind at 20 percent, and Democrat Raphael Warnock (the party's preferred Democratic candidate) just behind at 19 percent. 

It's just one poll, but Republicans and Democrats watching this race say they expect Warnock to make it to the runoff -- especially with all the presidential turnout in a state Biden could win. 

Per Georgia election rules, which throw all the candidates together on one ballot (rather than separate party primaries), if no candidate gets above 50%, it will go to a runoff in January among the top-two vote getters. That's where things get difficult for Democrats. They haven't had much luck in runoffs in Georgia, and Joe Biden won't be on the ballot (or rather, voting against Trump won't be on the ballot) to motivate them to come out. 

Why is there no longer a link on the WaPo home page to the The Fix chat? Or any other chat for that matter? The redesign of the home page is very crowded, but surely they could have found room for a way to guide the readers to the much-loved chats? Please advise.

There have been some changes, though we try to get the word somewhere on the homepage each week. But, for you regulars (love you!), maybe it's easiest to bookmark this page, where you can find a link to this week's chat and all the preceding ones.

How long does Robert Redfield have left as Director of CDC? it seems that he doesn't understand the rules in the current administration: Never contradict Donald Trump, regardless of the facts.

Well, this isn't the first time he's said something major that Trump didn't like. Redfield told The Washington Post's Lena Sun back in April that he's worried tamping down on coroanvirus could get more difficult this fall with cold and flu season: "There’s a possibility that the assault of the virus on our nation next winter will actually be even more difficult than the one we just went through."

Trump didn't like that. Redfield didn't back down. And Redfield is still around. 

There are similar dynamics this time on the coronavirus vaccine timeline. It's notable that Redfield tweeted from his CDC account after Trump tried to say the director misspoke about the virus timeline, and he didn't at all take the opportunity to say "oh yeah, I misspoke."

It is also worth noting that what Redfield said yesterday about when a vaccine might be widely available to Americans was nearly identical to what Trump's vaccine czar Moncef Slaoui and Anthony Fauci have said in recent weeks: Essentially, a vaccine won't be available to most Americans until the first or second quarter of 2021, at the earliest.

I am not talking about voting intentions, but about a rise in corona virus cases among the people who attended maskless and without physical distancing. I also hope that someone is keeping an eye on the aftermath of other events such as the recent indoor rally in Arizona.

This is an area of interest of mine particularly following the death of Herman Cain after he attended the Trump rally in Tulsa, Oklahoma. The president, I imagine, is hoping that these events result in no or few cases thus proving to his critics that gatherings are not as dangerous as some public health experts have claimed. But time will tell soon. And it will be worth paying attention to whether or not the outcome leads to any changes in the behavior of the campaign or the people who support it.

Do any of the QAnon-leaning GOP candidates still have a chance of winning House or Senate seats in November? Or will they all go the way of Christine "I am not a witch" O'Donnell? In particular, I'm interested in whether the Georgia Democratic party can get a new candidate to replace drop-out Kevin Van Ausdal -- one who will be able to "bring it" to Republican Marjorie Taylor Greene in an otherwise bright red Congressional district.

Greene seems like the most likely to make it to Congress. (And I certainly haven't heard of any whispers to try to get a new Democratic candidate in that race. It's a really tough one for Democrats to begin with.)

Another one with ties to QAnon to watch is Lauren Boebert in Colorado. She ousted Rep. Scott Tipton (R) in a primary in a Republican-leaning district. 

There is another candidate who won her primary in a Senate race in Oregon, but she's not going to beat Sen. Jeff Merkley (D).

Here's more exploring why QAnon candidates are gaining some traction politically right now. But experts said one or two winning is not necessarily a trend.

Why would Catholics prefer the irreligious Trump over Biden?

The most recent polling on this actually shows the Catholic vote very close -- 50-49 for Trump.

A lot of it boils down to cultural conservatism on issues like abortion. This line from the poll's write-up is also relevant here: "Just 14% of Catholics say it is very important to them to have a president who shares their own religious beliefs."

Right-wing voices have suggested that they would use violence to prevent a fair counting of ballots or to keep Trump in the White House at all costs. What is the likelihood of this happening?

That's not clear right now. But there is reason to believe that there are Trump supporters who will respond violently if the president doesn't win re-election because they've indicated that they would. And because we have seen a response to anti-racist demonstrators from people on the right this summer that have ended violently. Much attention will be paid to how law enforcement and political leaders respond to whatever actions surface this November as a result of some people's dissatisfaction with the election.

What is the makeup of the typical undecided voter at this point? White? Minority? College-educated or not? Man? Woman? Religious or not? Opinions, if any, about Trump and Biden? How do they compare to undecideds from 2016?

A very timely question, and there's some recent polling on this very question. A survey released this morning by the Kaiser Family Foundation (KFF) and the Cook Political report in Arizona, Florida and North Carolina classified nearly 1 in 4 voters as "swing voters" who had not yet made up their mind on who to support. The report wrote:

On most demographics, swing voters look very similar to their counterparts (voters who say they have already decided who they are going to vote for in the 2020 election), but they differ on three key variables: age, party identification, and ideology. Swing voters generally are more likely to say they are moderate in terms of their ideology (61%) and a larger share identify as political independents (43%) than their decided counterparts (10%). In addition, swing voters are slightly younger as a whole with about six in ten under the age of 50. In addition, nearly one in four swing voters are Hispanic voters (22%) compared to 13% of decided voters.

Those findings align with common sense -- younger voters have had less time to develop their political opinions, and independents are by-definition reluctant to affiliate with a party. There are other factors in decision making which are important as well, such as whether people choose to turnout.

What strategies do you use to reach representative population samples, especially by phone? Certainly, different segments on the population are more likely than others to answer phonecalls from unknown numbers.

This question is central to the ability of polls to produce accurate estimates of the population. Our national and state Post-ABC polls start with a systematic sample of all cellphone and landline numbers, and about 75 percent of our interviews are conducted by cellphone, so that we can ensure everyone has a chance of being selected. While many people do not answer their phones, completing a large percentage of interviews on cellphones helps improve the representativeness of the sample. Still, some groups are more likely to pick up, particularly older adults, people with more education and those who are more politically engaged. Once interviews are completed, the sample is weighted to match the demographics of the population according to the best-available benchmarks from the Census Bureau for the population of interest.

How do Americans feel about POTUS' stand on these two issues?

Our August Post-ABC poll found 59 percent disapproving and 40 percent approving of how Trump was handling the coronavirus outbreak. That included 50 percent who said they strongly disapproved.


We haven't recently asked how he's handling the environment, but a 2019 Post-ABC poll found that 62 percent disapproved of how Trump was handling climate change.



How are the polls in those 3 crucial states that Clinton lost in 2016? Is Biden expected to carry even one of them?

Biden leads in all three by between 4-7 points, according to the polling averages. The most recent ones suggest Wisconsin is the most likely to flip Biden. He most likely needs to win two of them, or one of them if he also flips Florida.

Thanks for taking my question, which concerns polls. Please help me. I've been avidly reading poll results this year, but after 2016 I have my doubts. I know, from sad experience, that national polls have little significance in a system like ours that gives greater weight to smaller population states than to larger ones. But I'm reading that some polls are better than others, that some polls measure voting likelihood differently, and some polls account for cell vs. landline use differently (but not texting). For example, should I be concerned that Trump voters are more motivated to vote for their guy than are Biden voters, but it equals out if you compare Trump voters to anti-Trump voters. So how does one know how much confidence to have in polls, which polls to believe and how much faith to put in polls at this point in time, less than two months to go to the election, and after the convention "bounce" of both candidates?

It's a good question, and one we get alot.

I would say three things:
1. Remember that polls are a snapshot in time, not predictive of the future. So that helps you weigh how much to put stock in them.
2. The presidential election is won and lost at the state level. So those polls can tell us more about where the race is -- but state polls can vary in quality (which we saw in 2016 when they failed to capture Trump's support in Michigan and Wisconsin and Pennsylvania, for example).
3. Our polling team, Scott and Emily, have very high standards for what polls The Washington Post can use in its reporting and analysis. So trust what you read here!

I'm sure Emily and Scott have more thoughts, if they see this question.

What was your normal workday six years ago? Complained that you missed the green traffic light, your kid missed the school bus, you forgot your wallet at home? Now we scour stores for Clorox wipes while the West Coast burns and funeral home employees try to convince others that Corona virus is real and it kills.

I'm not used to this. Anyone else? The day-to-day stuff is so much harder, and more draining. And that's if you and your family are healthy with a roof over your head.

Numbers don't add up for me, whats the reason Trump is getting credit when all he's done is ride on what was already growing when he took office? The tax cut helped corporations, and truly wealthy ppl, but what was in it for the middle class (according to two CPA friends of mine) was front loaded and will end. Manipulation of the Fed. is also not for the greater population...

This is the reason that many voters on the left do not actually give Trump credit for most of the successes he claims and instead attribute some of the upticks in the economy from Trump's earliest days to the Obama administration. But in fact, these voters credit Trump for many of the failures of the economy that resulted from the coronavirus pandemic -- and I expect we will continue to hear more of this during the final weeks of the campaign.

Visiting with my neurologist this morning, he said he does not expect a Covid vaccine to be widely available before Q3 of 2021, and more likely Q4. So, not before the election. I assume Trump would continue his current way of dealing with it, but do we know if Joe Biden has plans for how his new administration would potentially deal with the ongoing situation?

Three points here are key to stress:

1) CDC Director Robert Redfield, Trump's vaccine czar Moncef Slaoui and Anthony Fauci have all said they expect to have a vaccine before the end of 2020. But each have also said there will be a limited supply of the vaccine that would likely be distributed to first responders, front-line workers and the most vulnerable Americans.

2) Redfield, Slaoui and Fauci have also all said that widespread availability of a vaccine, when hundreds of millions of doses would be available, will not happen until 2021. Each have generally said they expect widespread availability to begin late in the first quarter of 2021 or early in the second quarter of 2021.

3) State officials told The Washington Post in August that they were in the dark on the Trump administration's vaccine distribution plan. And The Post reported yesterday that states still lack the billions of dollars that Redfield has said will be needed to distribute the vaccine.

In terms of how Biden would address the pandemic and work to quickly distribute a vaccine, The Post had a good rundown of this on Friday.

Just curious if you have any funny moments on live TV when you are doing it remotely from home?

There was one FUN time when I locked myself in a room and the 5-year old proceeded to knock on the door for like 3 solid minutes and whisper loudly at me, "DADDY! DAAAADDDYYYYY!"

Do you know of any polls that track the general public's intention to vote? By that I mean predicting the overall turnout. Is the public's intention to vote strengthening or weakening?

An August. Post-U. Md. poll conducted by Ipsos found 78 percent of citizens overall, including 90 percent of registered voters, saying they were "absolutely certain to vote." That being said, generally more people predict that they are certain to vote than actually do vote.

Is this thing on?

It is!

The chats are the highlight of my week. I appreciate the opportunity to ask questions and see those of other followers. The Fix, Weingarten and Robinson make my annual subscription worthwhile. I'm not sure I'd continue to pay if the chats stopped. I've always had a hard time finding the link on the home page, so just type The Fix into the search option, ditto for Robinson and Weingarten.

THANK YOU for the kind words. We hope to add some value to your Post experience here, and I'm glad to see that we're doing it.

The easiest way to always find our live chats is to bookmark this link. Meantime, I'll pass along the message about them being difficult to find.

How many seats Democrats are likely to gain as per the last few polls. Please consider democrats gaining/losing seats if it is outside the margin of error.?

Well, they're likely to lose one, Alabama. And they have a good shot at picking up two, Colorado and Arizona. They're likely to hold onto Michigan.

But they need three more and to win the White House to take the majority. They have a shot in North Carolina, Maine, Iowa, Georgia, Montana and maybe Georgia's special election and South Carolina. Kentucky, Texas and Alaska are the other potentially competitive races and they're bigger stretches.

The races where the majority will truly be decided are also true toss ups. So I can't predict how many they're likely to pick up!

Trump did not perform particularly well in this week's town hall meeting on ABC. One of the three presidential debates follows this same format. Can he do something different to improve his performance?

I think debate prep -- assuming Trump does it -- will better prepare the president for these upcoming events. Town halls are a bit different in the sense that the questions are coming from voters and show how Trump responds to the people he is expected to serve and represent. Debates put Trump against the media, a format that works in his favor at least with his base. And we should expect to see him attempt to take advantage of that. 

How is the indigenous vote being courted?

I don't know the answer to that. It's a difficult community to get in touch with, even moreso in a pandemic, I imagine. 

I didn't think I'd have to say this, but apparently given the questions so far, I feel I do. Thanks to all pollsters -- and to explainers like Emily and Scott -- for being so accurate! More accurate than the rest of us!

Thank YOU!

We hear that Presidential election results will take days or weeks to tally. Which states will report on election night? (Asked another way: Which states will count mail-in ballots ahead of Election Day?)

This page, from the National Conference of State Legislatures, fills you in on all 50 states.

Business Insider has floated this possibility. That'd be a smart move on the president's part, but we all know intelligence is not his strong suit. My question is, can the president be pardoned for crimes for which he's been charged but not yet convicted?

Can he? I think so. President Ford pardoned Richard Nixon for any crimes he may have committed during Watergate.

Would he? I'd be very surprised if Trump resigns, and if he did, Pence doing such a controversial thing (it wasn't popular when Ford did it) could be detrimental to any future political career Pence would want.

I love the PBS NewsHour's cats in the background at reporters' homes -- especially William Brangham's, although I also spied one at Lisa Desjardin's this week, bathing itself on the couch, then taking a nap.

I have parakeets at home and they get to go on "vacation" in their cage on the balcony when I do a radio or a TV hit. They enjoy taking part in every call and have a lot to add about survey weighting methods!

All poll projections have a certain amount of bias. Which seem biased the least?

Trick question! Though I doubt you meant it as such.

First, polls are different from projections or forecasts estimating candidates' probability of victory. Their purpose is to measure what people think at the time the survey was conducted -- a snapshot in time -- but people can change their minds about an issue or candidate. Election polls have the added complication that we don't know who will actually vote, so the population of "likely voters" is a survey's best estimate.

On the question of bias, this usually refers to the direction in which polls miss. While some polls do tend consistently overestimate/underestimate support for one party across elections, it's extremely hard to predict whether polls overall will be biased in a given election. While state polls underestimated Trump in 2016, polls also underestimated Obama's support just four years earlier. Many people predicted polls were overestimating Obama's support that year, and the result was just the opposite!

That's a wrap. Thanks to all of you and huge thanks to Emily and Scott, who are always honorary Fix team members, for sharing their time with us today. 

In This Chat
Amber Phillips
Amber Phillips writes about politics for The Fix. She was previously the one-woman D.C. bureau for the Las Vegas Sun and has reported from Boston and Taiwan.
Aaron Blake
Aaron Blake is senior political reporter for The Fix. A Minnesota native, he has also written about politics for the Minneapolis Star Tribune and The Hill newspaper.
Eugene Scott
Eugene Scott writes about identity politics for The Fix. He was recently a fellow at the Georgetown University Institute of Politics. And prior to joining the Post, he was a breaking news reporter at CNN Politics.
JM Rieger
JM Rieger is the video editor for The Fix, covering national politics. He joined The Washington Post in 2018. Previously, Rieger worked as a video producer covering national politics for HuffPost. He began his career as a video editor covering Congress for Roll Call.
Natalie Jennings
Natalie Jennings is editor of The Fix. She has been at The Washington Post since 2010 and was previously a senior producer for Post Video.
Emily Guskin
Emily Guskin is the polling analyst at The Washington Post, specializing in public opinion about politics, election campaigns and public policy. Before joining The Post in 2016, she was a research manager at APCO Worldwide and prior to that, she was a research analyst at the Pew Research Center's Journalism Project.
Scott Clement
Scott Clement is the polling director for The Washington Post, conducting national and local polls about politics, elections and social issues. He began his career with the ABC News Polling Unit and came to The Post in 2011 after conducting surveys with the Pew Research Center's Religion and Public Life Project.
Recent Chats
  • Next: