Polling Is Ubiquitous, But Is It Bad For Democracy? New Yorker writer Jill Lepore examines the history of polling in America. She tells Fresh Air's Terry Gross that today's polls may be less reliable — and more influential — than ever before.

Polling Is Ubiquitous, But Is It Bad For Democracy?

Polling Is Ubiquitous, But Is It Bad For Democracy?

  • Download
  • <iframe src="https://www.npr.org/player/embed/466405233/466437078" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

New Yorker writer Jill Lepore examines the history of polling in America. She tells Fresh Air's Terry Gross that today's polls may be less reliable — and more influential — than ever before.


This is FRESH AIR. I'm Terry Gross. Now that the presidential primaries are underway, there are new political polling results just about every day. Is polling good or bad for democracy? That question frames the article Jill Lepore recently wrote in The New Yorker, called "Politics And The New Machine."

She reports that polling may never have been less reliable or more influential than it is now.

She writes (reading) from the late 1990s to 2012, 1,200 polling organizations conducted nearly 37,000 polls by making more than 3 billion phone calls. Most Americans refused to speak with them. Lepore is a staff writer for The New Yorker and a professor of American history at Harvard University. Later, we'll hear from Nate Silver, founder of the popular website FiveThirtyEight, which analyzes political polls and predicts outcomes.

Jill Lepore, welcome to FRESH AIR. So, you know, you write in your article that polls may be more influential than ever. You also write that you think polls may be less reliable than ever. A typical response rate to polls is in the single digits. Less than 10 percent of people who are asked to respond to a poll actually follow through and respond?

JILL LEPORE: Yeah. Isn't that a little staggering to think about?

GROSS: Is it the low single digits or the high single digits?

LEPORE: (Laughter) There's a range. I think nine is not uncommon. I mean, it can be very much at the high end of the range. I think that - to know whether that's a crisis or not requires understanding what pollsters do or what public opinion researchers do more broadly. They use a sample survey method, so they conduct interviews with what is supposed to be a statistically representative sample of the electorate. That sample is going to be tiny, and a sample works to represent all of our views as an electorate, so long as it is well-chosen. Choosing that (laughter) sample is harder when the response rate is lower. When public opinion survey research started in the 1930s, the response rate was well above 90. People considered it a civic duty. It was like a thrill. Someone would come knock on your door and want to talk to you for 45 minutes about politics. Now I think if someone calls you at 6:01 p.m. (laughter) and wants to talk to you about politics while you're trying to get the, you know, the chicken fried, you hang up the phone. There are a whole bunch of forces that are behind that. There's just a fatigue. There are so many pollsters. And then there's - just people don't have landlines anymore. I think it under - only 40 percent of Americans have land lines, and you can't do random dialing to cell phones. The FCC made that rule, and they recently reinforced it.

And then you have the problem of the people who answer the phone are really different from a random sampling of the electorate, right. The people who answer the phone tend to be older. They tend to be more conservative. Those are the people who have landlines. And then the people who answer the phone and are going to really participate (laughter) in your survey, they tend to be people who have a very strong sense of civic obligation, and they're involved in their communities. They tend to be voters. You know, they're very goodhearted (laughter), civic-minded people, and they're overrepresented, unless you work harder to represent the smaller number of people that you can get who are not those people. So pre-election polling, in particular, is just in crisis.

So you know what pollsters will say - because pollsters want to get the right answer, right? It's not actually a fraudulent industry. Their bread-and-butter is making an accurate prediction. They will say they can accommodate and moderate the ill effect of a very low response rate by weighting their responses and by doing more calling. And so that means polls get more expensive and the results don't really bear out those promises. So there's this whole other kind of cottage industry in trying to figure out which polls might be reliable.

GROSS: So if the people who actually respond to polls are older, they're more civic-minded. They're more likely to be a churchgoer. What's the population of people who are less likely to respond to polls? In other words, who's being left out of the polls?

LEPORE: The representative sample in a poll should represent the possible eligible voters. What happens with a really low response rate is that the representative sample in a poll represents the likeliest voters. That is to say the people who (laughter) are mostly be to answer a poll are also most likely to vote. So this, for pre-election polls, where really what you're trying to do is figure out what's going to happen in an election, in fact, in some ways, the low response rate makes those polls more likely to be accurate, right, because this a big problem of predicting an election. You can get people to tell you who they might vote for, but you can't tell whether they're going to vote or not. But actually, the bigger question then is - isn't that actually completely inimical to representative democracy?

Like, should we really be - then we're disenfranchising everybody who doesn't answer the poll. Do you know what I mean? Like, there's so many forms of disenfranchisement in our political culture, and polls have, in a weird and completely unintentional way, become yet another one of them. So in a world of widening political inequality and economic inequality, polls have become - you know, they're just riding that same wave.

GROSS: Let's talk a little bit about polling history. You write that pollsters rose to prominence by claiming that measuring public opinion is good for democracy. What is that argument and who first made it?

LEPORE: That argument was made in the 1930s by George Gallup who founded the first public opinion polling research institution in Princeton, N.J. in 1935. And before Gallup began conducting his public opinion surveys, there really weren't polls in anything like we know them today. Polls - they're so ubiquitous it's easy to forget that there was a time before there were polls. Like, people are always calling elections, like, you know, it'll be the 1828 election between Jackson and John Quincy Adams and people - I wonder who's going to win. I think I know who's going to win. And you might even go around to your precinct and ask people, knock on their doors - they're like, who you voting for? Go down to the bar now - who you voting for? Trying to make a guess and deciding maybe where to put more resources if you were running for office.

But doing what Gallup did, which is using the sample survey method of social science which was new at the time to predict elections, was only really done by him in the 1930s. Gallup actually wanted to be a journalist. He wanted to be a newspaper editor. But when he was in college, there was no journalism program, so he studied psychology instead. And then he went and got a PhD in psychology. But think about the 1930s then when he starts doing this and saying, you know what? Public opinion - if we could - elections come only every two years. But we in this fast-moving world, he said, we need to be able to measure public opinion every minute instantly. We need to know the will of the people. And part of the reason for that was there was tremendous anxiety in the 1930s about basically the kind of demagoguery that Americans feared that the radio - here I am talking to you on the radio - might be introducing. Radio was greeted with great celebration, but it was also very gravely feared.

I mean, the war of the worlds is just a few years after George Gallup founds the American Institute for Public Opinion. Public opinion, it can be contorted by a voice coming into your living room and telling you what to think. And that's how fascism grew in Europe. So a countervailing method - they would go the opposite direction from the broadcast of a radio - would be to actually call somebody or go to their door and ask them what they think and have them tell you. And then you would report that. That is, you could turn the radio upside down, and this would be the way democracy could fight fascism almost in this kind of very mechanical way. But Gallup, when he started out, he wanted to measure what he called public opinion - that is, the views that ordinary Americans held. He actually didn't want to forecast elections or conduct polls - pre-election polls. He was doing opinion surveys. But he found that no one really believed his opinion survey results because how are you going to check that? Like, 47 percent of Americans believe in God (laughter). Well, I don't know - says George Gallup. Why should I believe George Gallup? I mean, he made a bunch of phone calls? I don't know.

So he - to prove that his method was accurate, that he was actually measuring something real, he started asking people how they were going to vote so that he could then predict an election just basically as a stunt in order to convince people that his public opinion research had teeth. And then he found -kind of, I think, somewhat he was disheartened to find people loved these predictions. They, you know, people eat that stuff up. (Laughter) Obviously we do. People love to read election predictions. And so that, more or less, took on a life of its own. But he really didn't defend that as good for democracy. He defended public opinion measurement as good for democracy.

GROSS: You know, there are so many newspapers and TV stations and radio stations and other media outlets who sponsor their own polls now. When did that proliferation in polling start to catch on?

LEPORE: It really started in the 1970s. So for a long time, Gallup was the poller. There was also Roper and then the Harris Poll followed soon after. So there was significance - in almost a way there were kind of, you know, the network television stations. There's kind of the big three. There were the big three for a long time in the '40s, '50s and '60s. And then there grew to be a lot of concern about what seemed to be the proliferation of smaller polling organizations. So much so that in 1972, Congress debated a Truth-In-Polling Act because they wanted to help voters discriminate between good polls and bad polls. And the act didn't pass. And then in 1975, The New York Times and CBS, I believe, collaborated on the first basically, you know, major newspaper poll. And it's been off to the races ever since. I've, you know, talked about this piece on, like, a local radio station here in Boston. And that local radio station had just launched its own polling collaboration with a local college. And I said, like, how do you justify that?

Like, in the 1970s, people said you're actually creating news and your job is to report news. But it was tremendously profitable for those newspapers and those television stations and radio stations to pay for polls to be done in order to report those polls. I think the proliferation most recently, like in the last 10 years, the number of those kind of smaller organizations - small colleges, for instance, that get into this industry, smaller radio stations, smaller town newspapers - like, I think the Union Leader in New Hampshire has its own pollster. That is actually about the incredible shrink in sources of deep journalism. So it's about the crisis of newspapers and the crisis of news more generally with the rise of the Internet and kind of niche reporting. It's very cheap to conduct a poll relative to doing the kind of deep reporting that would be required to actually get some sense of an electorate and report in a - using something other than pure quantification.

GROSS: If you're just joining us, my guest is Jill Lepore, and we're talking about her recent article in the New Yorker called "Politics And The New Machine: What The Turn From Polls To Data Science Means For Democracy." And Jill Lepore is a staff writer for The New Yorker and a professor of history at Harvard. Let's take a short break here, then we'll talk some more. This is FRESH AIR.


GROSS: This is FRESH AIR, and if you're just joining us, my guest is Jill Lepore. We're talking about her recent article in The New Yorker about polls, called "Politics And The New Machine." She's a professor of history at Harvard and a staff writer for The New Yorker.

So everybody paid so much attention for so long to Iowa - to predicting what the outcome of the Iowa caucuses will be and then reporting on the outcome and then trying to read the tea leaves to predict what the outcome of the election is going to be. And I realize that in Iowa its caucuses. It's not primaries. But, you know, in America we pride ourselves on having a secret ballot. You go to the voting booth. People aren't allowed to give you campaign literature within a certain number of feet of the polling booth. You go in. You close the curtain. You vote. No one knows what you vote unless you tell them. It's your decision whether you want to tell anybody or not.

You're under no obligation to. In the Iowa caucuses, you're in a gym or you're in a room. Everybody who's there knows what you're doing because you're telling them. Or you're walking to a side of the room to indicate what you're doing. And there's something about it that just seems so counter to the idea of it's an American democracy with a secret ballot. There's so - I mean, there's such possibility for, like, peer pressure, for intimidation, for, like, wanting to go with the crowd, you know, wanting to go along with your spouse or your best friend because it's...

LEPORE: Yeah, and it's so funny that you...

GROSS: ...Visible and public.

LEPORE: Yeah, but it's so funny that you would have a kind of suspicion that that's somehow anti-American but that polling, which is, of course, a version of that, like, that bandwagon effect, wouldn't necessarily leap out at you in that same way. But in fact, the secret ballot is a latecomer in American political culture. It isn't really adopted - the first presidential election where the secret ballot is the preponderant mode of voting is 1896, which is also the first presidential...

GROSS: Wow, really?

LEPORE: ...First presidential election where some is not killed on Election Day. No, voting in public is far more ancient and more American in that sense. So the whole idea of being a good citizen requires publicly exercising your franchise. So people - you go to the polls and you would - the government didn't supply a ballot. You would - you know, first you'd have to - first, they were all viva voce. You'd basically be like a caucus. You'd go to the polling place and be like, OK, if you are voting for Smith, stand over here against the butcher's. And if you're in favor of Jones, stand over there down by the library. And that's how the votes were made. And the polls would be counted. That - poll means the top of your head, so people would count the tops of people's heads, and that's why it was called the polls (laughter).

So the call to reform public voting, or what was known as open voting, was super controversial because - Massachusetts was the first to do it. And they had this idea that the government would supply these envelopes, and you could bring - so people would start - when the party system got really strong, and newspapers were partisan, like, the Republicans would print a ballot - like, a whole party ticket. It would be, like, we'll say it's red. And the democratic newspaper would print a blue party ticket. And so you'd go to the town hall - this is when oral voting had kind of been replaced by paper voting because people were literate. But still, you'd have this giant, long sheet, like a railway ticket, like two-foot long. It would be brightly colored. And so people would try to prevent you from getting to the ballot box and casting your vote. The parties would hire these thugs to go down there. Democratic thugs, you know, would beat up all the Republicans with their blue tickets and prevent them - but people would die.

People were killed every election (laughter) in these incredible battles over trying to get to the - then there was this thing called vest-pocket voting. So this was a little bit sneakier. If you wanted to keep your vote private, you'd fold up your long, you know, blue ticket, and you'd stick it in your vest pocket and try to get to the polls without anybody, you know, knocking you on it. But this was considered unmanly. So when Massachusetts in the 1850s tried to say, well, we'll just supply envelopes and people can put their tickets in the envelopes before they come to the ballot box, it was repealed because people said that only cowards would use an envelope to vote, you know. So it was a really controversial thing. It takes years for the secret ballot to be adopted, and it's part of, actually, a lot of forces that are not - people wanted to - didn't realize there was a lot of corruption as a result. All kinds of party machine nonsense - people are being beaten up, you know, that's obviously not good. Also, women wanted the right to vote, and they were like, we wouldn't need to vote in secret. We don't want to get hit in the head. So suffragists were sort of supporting the secret ballot. And it was first passed in Massachusetts and New York in the 1880s.

But then for years, the only other places that adopted the secret ballot - which is a written ballot supplied by the government to each voter - was the South after Reconstruction because it was a way to disenfranchise newly-enfranchised black men who - none of them knew how to read. I mean, they'd been, you know, raised in slavery, lived their entire lives as slaves on plantations. And so it was - the real success of the secret ballot as a national political institution had to do with the disenfranchisement of black men.

GROSS: So the secret ballot was a way of helping them get the vote.

LEPORE: No, it's helping - it was preventing them from voting. If you could cut your ballot out of the newspaper, and you're going to vote a party ticket, and knew you wanted to vote Republican, and that ticket was going to be read, you didn't have to know how to read to vote. Immigrants could vote. Newly-enfranchised black men in the South could vote. It actually was a big part of expanding the electorate. But people in the North were like, hey, we don't really like when all those immigrants vote. And people in the South were like, we really don't want these black guys to vote, so they, in a sense, kind of colluded over - and there were good reasons for the secret ballot too. But they - very much motivated by making it harder for people who were illiterate to vote. It's essentially a de facto literacy test. And so...

GROSS: Because in the polling place - like, in the election booth - it wouldn't be, like, code-colored like that, or you couldn't ask anybody to read it to you. Is that what you're saying?

LEPORE: You couldn't - right. And so there's some counties in Virginia, I think it is, that in the 1890s they print some regular ballots. But then they print ballots in Gothic type - like, deep medieval Gothic type. And they give all those ballots to the black men. It's a completely illegible ballot. So there's a very - so anyway, people did - people debated it a lot of the time for those reasons, and also - I mean, if you think about it, why does Congress vote in the open? Well, it could be if you're a congressperson, your vote should be known to the public. That's part of transparency that we believe in, and it seems obvious to us. People believed that way about ordinary citizens voting as well for a long, long time. So the caucus in Iowa, that method which is so kooky and kind of fascinating - that's deliberative democracy. That is a certain kind of exercise of civic virtue that needs to be conducted in public.

GROSS: So I am stunned by everything you have just told me because I thought the secret ballot was one of the principles American democracy was built on. Did you always know this did you find that out later in life?

LEPORE: No, I totally didn't know. I always have to find things out because I just get curious about them. Remember the election with the Florida ballot - the hanging chad and everything? So I was asked to write a piece about voting machines or technologies in voting because we were looking like we were maybe moving to Internet voting. And I thought, I don't know if that's - going from paper voting to Internet voting doesn't seem that big of a deal, but what's a really big deal would be going from oral voting to paper voting. I wonder what that was like. And so then I did all this research and realized that, oh, the secret ballot is just such an incredible latecomer. And it's - the whole story of its origins utterly shocked me, and was really illuminating because it made me think about how Victorian our voting is. Do you know what I mean? It's like, go in this little booth with a little curtain, and you've got to be alone, and it's going to be dainty and private. It is actually a kind of Victorian domestic, quiet, sacred, middle-class space. And that was carved out by middle-class reformers, you know, against the kind of hurly-burly of, you know, the rowdy, exciting, drunken voting day.

GROSS: My guest is Jill Lepore. Her article "Politics And The New Machine," about whether polling is good or bad for democracy, was published in The New Yorker. After a break, we'll talk about how campaign season got so long. Also, we'll talk with Nate Silver, the founder and editor of the website FiveThirtyEight, which for many politics-watchers is the go-to place for analysis of the polls and predictions of election outcomes. I'm Terry Gross, and this is FRESH AIR.


GROSS: This is FRESH AIR. I'm Terry Gross back with Jill Lepore. She wrote a recent article in The New Yorker about whether political polls are good or bad for democracy. We've been talking about that and about some fascinating election history. She's a staff writer for The New Yorker and a professor of American History at Harvard.

So here's another American history question for you. It always seems that the presidential campaign gets longer and longer and longer. And I don't even mean the permanent campaign where after you win the election, you are campaigning immediately for reelection. I just mean like the relentless press coverage of the horse race. It seems longer and longer each presidential election. So if we look back historically, how much time was devoted to the actual campaigning?

LEPORE: So campaigning used to be considered completely gauche. So there was no presidential campaigning at all until Andrew Jackson. So people thought if you actually tried to convince the people to vote for you that you were a demagogue, that you should be such an esteemed man in society that everyone would know who you are and would be familiar with your accomplishments and let it to them to decide. This is very kind of courtly notion. This is a very small electorate of, you know, propertied white men. They're the only people who can vote anyway. Those people do know who you are. How you get to run is to do with, you know, your party nominates you in this completely closed proceeding. Andrew Jackson comes along. Andrew Jackson is not a natural candidate.

GROSS: What year are we talking here?

LEPORE: 1824. He's not a son of - he's not a founding father and he's not a son of a founding father. And everybody who had been president so far had been either a founding father or a founding son. And he's - you know, he's kind of like an orphan, brutal soldier guy. But he's famous for this battle. And he's - I'm going to start campaigning. And he hires a biographer to write the story of his life. It's the first campaign biography. And then he starts, like, talking to people. You know, he invites people. When he is inaugurated, he invites, like, ordinary people to come see him be inaugurated. I mean, this is the beginning of - so much of our political culture is - can be traced to Jackson. So then there's this great age of campaigning, but it's very - it is absolutely last-minute. In most (unintelligible), you've got to have someone write your biography about it maybe a year in advance. You get those books out. And then you got to buy a lot of whiskey. And you get to go - there is, like - when people become the log cabin candidates - the Whig Party - they hired these, like, rolling log cabins and like, yeah.

William Henry Harrison I think campaigns in these rolling log cabins like John McCain's bus tour or something. So there's a bunch of campaigning in the 19th century. In fact, it retracts in the 20th century because - (unintelligible) Wilson is the first guy to, like, have a record - he is - or Harding maybe - has a record. He cuts an album on his campaign speech. So he says, I'm not going to travel. Like I could travel by railroaded. I'll just do these, you know, whistle stops. But there's this sort of weird retraction. FDR, you know, he does newsreels. He is the first newsreel president. From 1932 he uses newsreels. So there's a kind of remoteness almost. The campaign stretches in time. But the candidate becomes more remote from the people, which is in some ways what the caucuses and primary movement is about, right? Getting these candidates to actually come to your states.

GROSS: Yeah, maybe FDR did the newsreels because of his polio.

LEPORE: Oh, yeah. Absolutely, absolutely. But then, by then, by 1932, that's the first newsreel. 1933 is the first political consulting firm, this Campaigns Inc., they start using the radio. And they - you know, they will come into your living room. And they will tell you you should never vote for FDR. The man will doom, doom, doom the country, you know what I mean? Like, they have this whole other - they have actually an amplifier.

GROSS: Right, and that gets to where we started (laughter)...


GROSS: ...Which is that public polling was seen as a way to combat radio 'cause through - strangers coming into your home and telling you who to vote for, it could lead to fascism.


GROSS: But if we had public polling, the public opinion would be represented, and you wouldn't be relying on that voice intruding.

LEPORE: Yeah, the voice of the people is the answer to propaganda.

GROSS: Right, OK. Jill Lepore, thank you so much.

LEPORE: Thanks for having me, Terry. It's been a blast.

GROSS: Jill Lepore's article about whether polling is good or bad for democracy was published in The New Yorker.

Copyright © 2016 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.