In the Dark

How Do We Know Our Elections Are Fair?

In any other country, the U.S. State Department would declare the presidential election results a hoax: Clinton won initial exit polls—usually dead accurate—in four swing states.

1. Paper ballots and audits should be routine

Jill Stein, the Green Party candidate for president, has done a public service in demanding that the votes be audited in Pennsylvania, Wisconsin, and Michigan. Even where paper trails exist, most states do not audit elections. It is important to know that the vote each of us casts is counted fairly.

In Wisconsin, Donald Trump’s narrow margin over Hillary Clinton was 0.7 percent, in Pennsylvania 1.2 percent, and in Michigan 0.2 percent. In Michigan, for example, Trump won by 10,708 votes out of 4,799,284. Further, 80,000 submitted ballots with no mark for president, which, at minimum, should be checked by the human eye.

Computer experts like Professor J. Alex Halderman at Michigan urged Clinton to join Stein’s challenge. Computerized voting machines—“touch screens” or DREs (Direct Recording Electronics)—which leave no paper trail, are notoriously unsafe. Both Halderman and Professor Andrew Appel of Princeton have altered their programs quickly—Appel in seven minutes with a screwdriver—and inserted malware. Such machines are used widely in Pennsylvania.

Further, for 2016, Halderman underlines that Russian or other interference has already broken into the Democratic National Committee’s and Clinton campaign chief John Podesta’s emails, as well as into voter registration systems in Illinois and Arizona. Others point to a 7 percent statistical deviation toward Trump in Wisconsin counties that use DREs with no paper trail. Clinton has now joined Stein’s effort.

Note that in swing states with narrow margins, one would not have to falsify a wide part of the election. Altering the vote in one or two main counties would, very likely, do. The reassurance offered by President Obama that America has a decentralized voting system, hard to tamper with as a whole, is, unfortunately, undercut by these facts.

Moreover, Trump was defeated in the actual voting by a startling—and still growing—2,676,670 votes. As of Dec. 6, Clinton had won 65,534,951 votes nationwide, Trump 62,858,281 (see the nonpartisan Cook Political Report). Clinton’s popular vote victory margin is now 2 percent, thus handing Trump the largest defeat suffered by a candidate elevated to the presidency by the Electoral College in modern history. Clinton will soon catch up to Barack Obama's 2012 total, the second highest popular vote ever in favor of a president. Further, given 7.9 million votes cast for third party candidates, Trump’s current percentage—46.2 percent and shrinking—is the lowest for any popular vote-defeated president. Even Rutherford Hayes, who gained Electoral College victory in 1876 by promising a Jim Crow South, had 47.9 percent.

Moreover, Clinton’s margin dwarfs President John F. Kennedy’s popular vote victory in 1960 (more than 15 times) and Richard Nixon’s in 1968 (more than four times).

Now, at its founding, the United States had a Bill of Rights for white men, but not for non-whites, the poor, and women. The regime initially protected the power of slave-masters. The Constitution counted those in bondage as three-fifths of a human to multiply the votes of Southern “man-owners.” This beginning has long skewed American politics away from the basic democratic principle: one person, one vote.

From the standpoint of modern democracies (regimes that respect individual rights, including the right to vote of each person), Donald Trump did not win. Even in this regime, he limped, barely, across the finish line to august power, by narrow margins. Were the audits to reverse the outcomes in three states, Trump would lose the Electoral College.

2. Initial exit polls are used by the U.S. State Department to test electoral fairness in other countries

There is, however, a far more damning point about Trump’s victory by U.S. State Department and international standards, which has been suppressed by the corporate media in the National Election Pool (CBS, NBC, ABC, Fox, AP, and CNN). The initial exit polls show Clinton to have won the swing states of Florida, Pennsylvania, Wisconsin, and North Carolina, and to be close—within one-tenth of a point—in Ohio. Initial exit polls are taken with random voters—usually 1 in 10 is asked—at randomly selected polling stations. Only responses from actual voters are used. When the last vote is cast, polling ends.

Get The Beast In Your Inbox!

Daily Digest

Start and finish your day with the top stories from The Daily Beast.

Cheat Sheet

A speedy, smart summary of all the news you need to know (and nothing you don't).

By clicking “Subscribe,” you agree to have read the Terms of Use and Privacy Policy
Thank You!
You are now subscribed to the Daily Digest and Cheat Sheet. We will not share your email with anyone for any reason.

In Germany, Canada, and many other countries, an initial exit poll is released. And then paper ballots are counted. Where this procedure is used, there is no controversy. If the election is very close, the ballots can easily be recounted.

Further, since 2000, the U.S. State Department has used initial exit polling to test the fairness of elections in 14 “transitional” democracies, and challenged what happened in five of them. That is more than 33 percent.

As Eric Bjornlund and Glenn Cowan’s Vote Count Verification: A User’s Guide for Funders, Implementers, and Stakeholders, prepared for U.S. AID in 2011, underlines:



In recent years, domestic and international organizations have increasingly turned to exit polls to verify the officially reported results in the transitional elections of emerging democracies. Outside observers have credited exit polls with playing a key role, for example, in exposing fraud in Serbia and Mexico in 2000, Georgia in 2003, and the Dominican Republic and Ukraine in 2004. U.S.-funded organizations have sponsored exit polls as part of democracy assistance programs in Macedonia (2002), Afghanistan (2004), Ukraine (2004), Azerbaijan (2005), the West Bank and Gaza Strip (2005), Lebanon (2005), Kazakhstan (2005), Kenya (2005, 2007), and Bangladesh (2009), among others.

For example, according to Volodymyr Paniotto, director of the Kiev Institute of Sociology, in Ukrainian elections Nov. 23, 2004, exit polls by the Social Monitoring Center showed that Viktor Yushchenko, the opposition candidate, won by 3.5 percent: Yushchenko 49.4 percent, Viktor Yanukovych 45.9 percent. When the government declared Yanukovych the winner, 300,000 people—the Orange Revolution—protested in Kiev’s main square. Now exit polls for the first round of voting Oct. 31 found discrepancies of only 1.5-2 percent with the final results, within the margin of error. Exit polling was funded by the U.S embassy.

In Ohio in 2004, the initial exit poll showed a Kerry victory by 4.4 percent, with a large swing—6.7 percent—in the machine-recorded vote to Bush. The similarity in these cases—though not in the State Department’s verdicts abroad and at home—is striking.

In fact, the contradiction between U.S. AID’s use of initial election polls in other countries—often advised by the Edison Corporation—and the silence about initial exit polls by the NEP relying on the Edison Corporation inside the United States, is startling.

3. Clinton’s initial exit poll margins

In contrast, this Nov. 9, The New York Times and the NEP published only tedious demographies of characteristics of voters, e.g., how Latinos with incomes less than $60,000 voted. Aside from the threat of Russian hacking, they pretended that American elections “must” be fair.

Way down its list of characteristics, however, CNN did reveal the central information—who won and who lost—for 28 states. Clinton’s initial exit poll leads were as follows:

—North Carolina: 48.6 to Trump’s 46.5, 2.1 percent

—Pennsylvania, 50.5 to 46.1, 4.4 percent

—Wisconsin, 48.2 to 44.3, 3.9 percent

—Florida, 47.7 to 46.4, 1.3 percent

On machine-recorded results, however, Trump’s victory margin in North Carolina was 3.8 percent, 50.54 to 46.7, a shift of a very large 5.9 percent. An exit poll usually has a 2.5 percent margin of error in either direction. This result is a suspicious 2.3 percent beyond the exit poll’s margin of error.

In Pennsylvania, Trump’s margin was 1.1 percent, 48.79 to 47.65, a shift of 5.5 percent. That is 1.6 percent beyond the exit poll’s margin of error.

In Wisconsin, Trump’s margin was 0.9 percent, 47.87 to 46.94, a shift of 4.8 percent. This is 1.4 percent beyond the exit poll’s margin of error.

In Florida, Trump’s margin was 1.27 percent, 49.06 to 49.79, a shift of 2.6 percent. This result, though particularly doubtful in the context of the other deviations, is within the exit poll’s margin of error.

Note, however, that Clinton’s reported victory margin, particularly in Florida, may have been diminished on initial exit polls by the Edison Company immediately mixing in uncheckable precinct results from computerized machines (polls close at different times in two parts of that state).

The U.S. State Department would challenge these suspicious results in any other country.

Additionally, on the Edison exit poll for Ohio, Clinton was behind by one-tenth of a point, 47.0 to 47.1. To protect against a repeat of the 2004 fiasco, in 2008, Ohio Secretary of State Jennifer Brunner had purchased optical scan machines. But for the 2016 election, Republican Secretary of State Jon Husted turned off their “security”—their independent photographic record of how each person voted, the point of having an optical scan. Unsurprisingly, the machine-recorded results are for Trump 52.05 to 43.51, a very large swing of 8.4 percent. This result is outside the exit poll’s margin of error by 5 percent, one-tenth of a point higher even than North Carolina.

For Michigan, which is being audited, the exit poll showed a tie at 46.8 percent, the recorded result—47.59 to 47.25—a three-tenths of a point Trump victory. That is within the exit poll’s margin of error.

Despite the reporting of Clinton’s popular vote margin of 2.7 million as well as an audit of three close states, this startling set of exit poll deviations is, until this article in The Daily Beast, highlighted nowhere in the corporate media.

In a 2015 Harvard/University of Sydney study of fairness in elections conducted under the leadership of Pippa Norris, the U.S. ranks 60th, below all of Europe, Canada, and even Rwanda, Argentina, Tunisia, and Brazil in terms of transparent democratic practices.

Professor Halderman notes that counting optical scan machines, nearly 70 percent of American voters leave some paper trail. That is still 30 percent—an enormous one-third—who leave none. Underlining the public significance of Jill Stein’s call for an audit, Halderman stresses the possibility of computer hacks: “There’s just one problem, and it might come as a surprise even to many security experts: No state is planning to check the paper in a way that would reliably detect that the computer-based outcome was wrong. About half the states have no laws that require manual examination of paper ballots, and most other states perform only superficial spot checks. If nobody looks at the paper, it might as well not be there. A clever attacker would exploit this.”

4. The Edison Company’s reliability on initial exit polling and doctoring of American “results”

One company, initially called Edison Mitosky, now Edison, has run 11,000 exit polls in 38 countries since Warren Mitofsky developed them in 1967. There are some complexities in choosing polling sites as well as getting some voters to assess elections. For instance, 40-50 percent of randomly selected voters decline to be interviewed. But Edison representatives have perfected weighting and reweighting types of voters so they often get the recorded vote right to within 1 percent.

Edison is today led by a single expert, Joseph Lenski, its executive vice president and sole public spokesman. He worked closely with Mitofsky. On election days, Lenski employs 3,000 people to compile poll results. “The exit poll,” Lenski says, “is the one survey out there where you get evaluated immediately. We’ll process over 100,000 interviews, and within two hours we know how well we did.”

As early and absentee voting have become more important, Lenski reports, Edison has learned how those alternatives differ from each other and to correct for them. “Historically, by which I mean before 2000, early voting was almost entirely absentee by mail, and it tended to be older voters,” he says. “Since the 2000 election, the Democratic Party has put emphasis on early in-person voting. The in-person voting tends to be encouraged by voter outreach and to be more Democratic, and the absentee tends to be older voters or people traveling out of state, and to be more Republican.”

Nonetheless, Edison also makes profits under contract from the National Election Pool. It thus does what six huge media corporations—its employers—request. Further, as a corporate executive, Lenski is cautious in what he says. To ensure Edison’s business, he cannot challenge the loud and unjustified assertions of fairness of American elections, despite what exit polling may show, by NEP reporters.

In other countries, Edison conducts only initial exit polls. Solely in the United States, however, Edison immediately doctors these results to match machine-recorded results. But touchscreen machines leave no paper trail at all, and though better, optical scan machines, which take separate photos of how individuals voted, can have “the security” turned off—as in Ohio in 2016—or be hacked.

Edison’s procedure in the U.S. thus inverts the purpose of initial exit polling from testing fairness within a range of reasonable results—something initial exit polls are adept at—to adjusting in the blink of an eye to whatever numbers appear on an often unverifiable computer screen. Edison’s Mitofsky adopted this odd copy of “exit” polling only after the 2004 election in Ohio. He recrafted Edison’s initial poll for Kerry at 4 p.m. Tuesday by 6.7 percent to undergird Wednesday morning’s machine-recorded results.

But such doctored results are not exit polls. They are no check for fairness. Instead, they smoothly ornament already existing error or fraud. In contrast, initial exit polls are what the State Department uses to challenge such results abroad.

Today, according to Lenski, the NEP is interested in these adjusted results solely to get a demography of how different categories of people vote. But what Edison misnames “exit” polls for the corporate media inside the United States is the opposite of initial exit polls here and abroad.

Now NEP reporters maintain a fiction that American elections are “always fair.” With widespread Republican suppression of nonwhite, poor, and elderly voters in many states, this fiction is, on the face of it, bizarre. On behalf of the American people, the news media could have demanded fairness. Instead, in the 2016 presidential election, NEP corporations—excepting CNN—ceased to report Edison’s initial exit polls at all, even though Edison conducted such polls in 28 states (Theodore de Macedo Soares, co-author of the Election Justice report on the Democratic primaries, recorded these numbers).

In spring 2016, a furor arose concerning conflicts of initial exit polls with machine-recorded results in 10 out of 22 Democratic primaries. But the same exit pollers in the same places talking to the same voters got the Republican race right in 20 out of those same 22 primaries, and the order of finish of the top three candidates in all of them.

As a result of the highly controversial New York Democratic primary, Lenski gave interviews to The Washington Post and Raw Story, saying that exit polls in the United States differ from those in “transitional democracies.” In new democracies, exit pollers ask a higher percentage of voters—Lenski did not specify a number—a simple question: Who did you vote for?

In contrast, Lenski suggests, the NEP is, once again, only interested in a demography of voter characteristics. And Edison uses longer questionnaires. Voters, Lenski claimed, are just concerned about, say, how college educated women over 60 vote. No one cares who won or lost.

5. Talk about a tall story…

At Raw Story, Joshua Holland writes, “As for using [Lenski’s] results to suss out fraud [as in other countries], he says that American exit polls are ‘just not designed for that type of precision. They’re surveys, and like any other survey, they have a margin of error [as do exit polls used abroad—AG]. The precision that a lot of these people are talking about just doesn’t exist with our polls.’”

But then, perhaps as a matter of pride—after all, the NEP invests a lot of money in exit polling and Lenski is a genuine expert—Lenski told The Washington Post that against the pre-primary polls, which predicted as much as a 15 point Clinton victory, Edison’s initial exit polls got Sanders’s unexpected triumph in Michigan right.

“In Michigan,” Lenski said, “we actually had exit polling all day showing Bernie Sanders up by two points, though every pre-election poll had Clinton up by 10 points or more… In that case, the exit poll was right.”

Lenski also invokes evidence about Trump: “While everyone is talking about the Democratic side, we went out at 9 o’clock saying that Trump was going to get 58 percent of the vote. He got just about 60 percent. Everything we did on the Republican side hit the mark.”

But these points rule out Lenski’s implausible claim that American exit polls only seek to measure voter characteristics.

If exit polls got Trump’s margins of victory right in the Republican primaries, how come these same polls conducted by the same methods by the same Edison Corporation got Trump’s margin decidedly wrong in five states that decided the presidential election?

6. How Trump out-foxed the Democrats

I should also underline two telling political points about the significance of initial exit polls. First, Trump tweets late at night “bigly” that the elections were “rigged against me” because of alleged voter fraud: people voting more than once. But this is almost non-existent in American elections: 31 possible cases out of over a billion votes cast since 2000, according to a study by Justin Levitt, a Loyola Law School professor.

Yet Republicans have long used this false claim to disenfranchise large numbers of poor and nonwhite voters, particularly in swing states. Further, in contrast to Europe, which has elections on weekends and where nearly everyone votes, only 50 percent of those eligible vote on Tuesday, a working day, in American presidential elections.

Because of the anti-democratic electoral college, presidential campaigns are waged in a dozen or so “battleground states.” Now in those states for 2016, the turnout rises, despite obstacles to voting, to 58.7 percent, according to NPR.

Still, the years-long Republican campaign of legal and quasi-legal disenfranchisment of ordinary voters clearly benefited Trump.

In addition, if the results went the wrong way, Trump riled up his supporters to attack people at the polls and afterward.

In response, President Obama suggested that the election must be fair because American elections are so decentralized. But once again, elections are decided in a few states, and often a few counties. For the 2016 election, initial exit polls showed that these “results” are very doubtful. The brilliance of Trump’s political maneuver deserves to be taken in by those who mock him as well as by those of us who value fair elections. For the Democrats and the corporate media denounce any claim of fraud or error. They “hear no evil, see no evil…”

Second, Hillary Clinton benefited from corporate media suppression of initial exit polls in the Democratic primaries. Pro-Clinton reporters for the NEP routinely mocked those who rightly took initial exit polls seriously and protested. So Clinton is in a very weak position to raise this issue about the presidential election in which she has, so far, been out-Trumped.

Now, Clinton’s campaign and even Halderman’s challenge was based on an unproven, probably false claim of interference by the big, bad Russians. At a minimum, however, the trouble instead arises from several sources: the dishonesty of the Supreme Court decision to disable section 4(b) of the Voting Rights Act in Shelby County v. Holder, the anti-democratic legal maneuvers of the Republicans, and given initial exit polling, intervention by computer experts very likely nearer to home.

As Halderman points out more aptly, “It doesn’t matter whether the voting machines are connected to the internet. Shortly before each election, poll workers copy the ballot design from a regular desktop computer in a government office, and use removable media (like the memory card from a digital camera) to load the ballot onto each machine. That initial computer is almost certainly not well secured, and if an attacker infects it, vote-stealing malware can hitch a ride to every voting machine in the area. There’s no question that this is possible for technically sophisticated attackers. If anyone reasonably skilled is sufficiently motivated and willing to face the risk of getting caught, it’s happened already.”

The fundamental issue is not only whether American election results have been reversed to declare the wrong winner. The issue is whether, aside from initial exit polls, citizens have any way of judging whether our elections are fair. For unlike Europe or Canada, there is no adequate paper trail. The answer is, sadly, no, not a chance.