ForReal - Exit Polls, and Why the Primary Was Not Stolen From Bernie Sanders
Waiting to hear Bernie Sanders speak in May at Rancho Buena Vista High School in Vista, Calif. Sam Hodgson for The New York Times
By Nate Cohn
June 27, 2016
The background music of my life is the steady drumbeat of tweets about how Hillary Clinton stole the presidential primary from Bernie Sanders.
It’s there when I wake up in the morning. It’s there when I go to sleep at night. The constant thrumming of election fraud conspiracists is like the noise ..
.. made by that board game in the movie “Jumanji.”
I didn’t write about this during the primary season, since I didn’t want to dignify the views of conspiracy theorists. But they’re still going.
The allegations are remarkably consistent. They go like this: Mr. Sanders did better in the early exit polls than he did in the final result. Therefore, Mrs. Clinton probably stole the election. The exit polls are a sufficient basis to make this determination, in the eyes of the conspiracists, because exit polls are used internationally to detect fraud. They’re supposedly very accurate and “well controlled” (where this phrase comes from, I don’t know). Furthermore, they say, the exit polls were right on the G.O.P. side — confirming the underlying validity of the methodology and raising suspicions about the Democratic vote count.
All of this starts with a basic misconception: that the exit polls are usually pretty good.
I have no idea where this idea comes from, because everyone who knows anything about early exit polls knows that they’re not great.
We can start in 2008, when the exit polls showed a pretty similar bias toward Barack Obama. Or in 2004 .. http://abcnews.go.com/images/Politics/EvaluationofEdisonMitofskyElectionSystem.pdf , when the exit polls showed John Kerry easily winning an election he clearly lost — with both a huge error and systematic bias outside of the “margin of error.” The national exits showed Kerry ahead by three points (and keep in mind the sample size on the national exit is vastly larger than for a state primary exit poll) and leading in states like Virginia, Ohio and Florida — which all went to George W. Bush.
The story was similar in 2000. The early exit polls showed Al Gore winning Alabama, Arizona, Colorado and North Carolina. Mr. Bush won these states by between six and 15 points. The exit polls showed Mr. Gore winning Florida by six points — leading the networks to call the race before 8 p.m. in the East.
IMAGE
The same thing happened in 1996. It was actually even worse in 1992. The exit polls had Bill Clinton winning Texas, which went to George H.W. Bush, and basically everywhere.
Now, how can this happen? There are a lot of sources for exit poll error — even more than in an ordinary poll. Here are a few:
* Differential nonresponse, in which the supporters of one candidate are likelier to participate than those of another candidate. Exit polls have limited means to correct for nonresponse, since they can weight only by visually identifiable characteristics. Hispanic origin, income and education, for instance, are left out.
* Cluster effects, which happen when the precincts selected aren’t representative of the overall population. This is a very big danger in state exit polls, which include only a small number of precincts. As a result, exit polls have a larger margin of error than an ordinary poll of similar size. These precincts are selected to have the right balance of Democratic and Republican precincts, which isn’t so helpful in a primary.
* Absentee voters aren’t included at all in states where they represent less than 20 percent or so of the vote.
For all these reasons, exit polls can be very inaccurate and systematically biased. With this kind of history, you can see why no one who studies the exit polls believes that they can be used as an indicator of fraud in the way the conspiracy theorists do.
But why were exit polls so tilted toward Mr. Sanders? It’s impossible to be 100 percent sure, but the best-known bias in the exit poll offers a very good explanation: young voters.
Young voters are far likelier to complete the exit polls than older voters, according to data from Edison Research, the organization that conducts the exit polls. The gap is particularly pronounced when the interviewers are also younger, but the gap persists even when older interviewers are conducting the exit interviews.
The exit polls try to correct for this bias by giving more weight to older respondents. The way it works is pretty novel: Interviewers guess the age of voters as they leave the polling place. Then the responses are weighted to match the age of all the voters who showed up, based on the guesses of the interviewer.
IMAGE
This could work, in theory. In practice, it falls short. There’s a persistent, decades-long bias toward young voters in the exit polls — even in the final, “adjusted” data — when compared with census or voter file data.
You can see that in data from the 2012 presidential election, which I wrote about .. http://www.nytimes.com/2016/06/10/upshot/there-are-more-white-voters-than-people-think-thats-good-news-for-trump.html .. a few weeks ago. Over all, the exit polls showed that 19 percent of voters were ages 18 to 29, compared with around 15 percent in census and voter file data. Notably, the census is an extremely high-quality survey — so you can just pack away any theory that election administrators are tossing the votes of young voters in basically every jurisdiction across the country.
Why does this bias exist, despite the exit poll effort to adjust for nonresponse by age? It’s hard to say; if the exit polls had the data to identify the cause, they presumably could fix it. There are at least a few possibilities — like biased guesses, trimming weights (where they’re not weighting older voters enough), or the absence of old absentee voters in some states — but there’s no way to be really sure.
What’s clear is that this bias persisted in the 2016 Democratic primary. Voter registration files are just starting to be updated, and they all show that Democratic primary voters were far older than the exit polls suggested.
Mr. Sanders, of course, is a candidate with historic strength among young voters — so it should be no surprise that the exit polls were particularly biased in his direction. Nor should it be a surprise that the exit polls were also biased toward Mr. Obama in 2008, or Democrats in many recent elections.
I’d also note that the age bias of the exit polls wouldn’t have much of an effect on the Republican results: There are far fewer young voters in the Republican primary, and there wasn’t much of a split between older and younger Republicans.
There are other challenges with exit polls in the primaries. Usually, the exit polls select precincts by partisanship — ensuring a good balance of Democratic and Republican precincts. This helps in a general election. It doesn’t do as much good in a primary.
If you’re looking for the exit polls to identify fraud in the United States, you’re out of luck. They would need to be redesigned: sampling many more precincts, calling more early voters and taking steps to boost response rates — like short questionnaires, rather than the long ones used in the United States. That’s not going to happen. The news media uses exit polls to get a sense of “why” voters did what they did, not to validate the election results. It’s not going to spend millions more dollars to increase the accuracy of early exit polls.
What should be telling, though, is that the news media organizations that commission the exit polls, and Edison Research, which conducts them, do not believe that their own data is good enough to call the Democratic primary results into question.
Your conservative line that the 2016 election was stolen from Bernie is a well debunked myth.
.. just here now on Tornado Alley because i put it in another place, and no doubt we'll be hearing more of rigging from Trump supporters .. as you know The Nation supported Bernie Sanders .. it's a decent history of the 2016 Democratic primary, too ..
But now Clinton and Sanders supporters should unite on making it easier to vote.
By Ari Berman June 16, 2016
A maze of political signs for Bernie Sanders and Hillary Clinton in Charleston, South Carolina. (Reuters / Randall Hill)
First off, the party’s rules were not the deciding factor. Sanders has rejected the idea that the nomination was “rigged” but has repeatedly criticized things like superdelegates and closed primaries, in which Independent and unaffiliated voters can’t participate. Here’s what he told Face the Nation in late May:
-- What has upset me, and what I think is—I wouldn’t use the word rigged, because we knew what the rules were—but what is really dumb is that you have closed primaries, like in New York state, where three million people who are Democrats or Republicans could not participate, where you have situation where over 400 superdelegates came on board Clinton’s campaign before anybody else was in the race, eight months before the first vote was cast.
That’s not rigged. I think it’s just a dumb process, which has certainly disadvantaged our campaign.
-- Sanders was generally hurt by closed primaries. By our estimates, he did about 3.5 points worse in such contests. But most states aren’t closed, so Mr. Sanders wasn’t hurt that badly over all. And Mrs. Clinton was hurt more by caucuses, where she did about 10 points worse, according to the same model.
If every contest in the country had been an open primary, Mrs. Clinton’s delegate lead would have grown. She would have lost ground in some of the contests, gained ground in the states with large numbers of anti-Obama registered Democrats (Oklahoma, West Virginia and Kentucky), and gained lots of ground in Western caucuses—where Mr. Sanders earned most of his big delegate hauls.
Over all, Mrs. Clinton would have about a 12-point lead in pledged delegates if every state had an open primary, according to our estimates. --
Nor did superdelegates decide the nomination for Clinton. They gave her a symbolic early lead and momentum, but Clinton’s pledged delegate lead over Sanders was three times larger than Obama’s margin over Clinton in 2008, under the same rules. I’m in favor of abolishing superdelegates or curtailing their influence, but it’s worth remembering that they’ve followed the pledged-delegate winner in every presidential contest since their creation in 1984.
I could go on and on, but the point is that the outlandish claims fall apart when subjected to closer scrutiny.
“To the extent that the nomination was rigged in the sense that there was illegal activity going on that was directed by the Democratic Party or the Clinton campaign to sabotage Bernie Sanders’s chances, I’ve seen no credible evidence of that,” says law professor Rick Hasen of the University of California–Irvine.
There were problems in the primaries, without a doubt, and Sanders is right that the nominating process should be reformed. He’s called for open primaries, same-day registration, and abolishing superdelegates—worthwhile ideas that merit more discussion.
Yet not all of these reforms would benefit a future Bernie Sanders. Open primaries, for example, tend to produce more moderate candidates for local offices and can be an impediment to progressive insurgents like Sanders.
-- By their very nature, caucuses disenfranchise voters: they’re confusing, stressful and tiring to participate in, they have no voter secrecy or privacy, they require at least an hour of the voter’s time and often more at a very specific time of day, there is no ability to caucus by mail, etc. Caucuses also require chairs, secretaries, check-in volunteers and other officials who count votes, make tallies and determine eligibility–all of whom are are chosen by party leaders who can and do stack the deck for chosen establishment candidates in ways that range from unseemly to blatantly illegal. --
The eight states with the lowest voter turnout .. http://www.electproject.org/2016P .. in 2016 were all caucuses, according to political scientist Michael McDonald of the University of Florida, with an average turnout of 8.4 percent. Turnout was three times lower in caucuses than primaries in 2016. Yet Sanders has refrained from criticizing caucuses because he won 12 out of 18, compared to 28 of 38 primaries for Clinton.
“The caucus is the most voting suppressive system we have,” says Hasen. “The fact that this is off his list makes me question his integrity in making these reforms.”
The bigger danger facing Democrats are not their own rules, but new GOP voting restrictions .. http://www.brennancenter.org/new-voting-restrictions-2010-election .. aimed at depressing the vote among Democratic-constituencies. “What is happening is a sweeping effort to disempower and disenfranchise people of color, poor people and young people from one end of our country to the other,” Clinton said in a major speech on voting rights .. http://www.thenation.com/article/hillary-clintons-bold-plan-voting-rights/ .. in June 2015, at the beginning of her campaign. Clinton emphasized this issue before Sanders, and lawyers affiliated with her campaign have filed lawsuits against GOP voting restrictions in key battleground states .. https://electionlawblog.org/?p=72945 .
Both campaigns can unite on this issue. Clinton supporters should join Sanders in reforming the Democratic Party, while Sanders supporters should help Clinton combat obstacles to the ballot box like strict voter ID laws, cutbacks to early voting and restrictions on voter registration.
If you hadn't noticed, there is a bit of a split on the con side between trump cultists and the remnants of the republican party that are believers in the constitution, science, and civility. Those folks will likely prefer biden over your Douchenfuhrer.
So you ought to do some reflection instead of sowing every bit of divisiveness you can.
Market Data 
Markets 
AI
