NATIONAL | 2012 | Republican Presidential Primary

Statistical Evidence Suggests 1.2 Million Votes Were Flipped to Romney Electronically

Such an alleged election fraud could be accomplished by only a single, highly clever computer programmer with access to voting machine software updates.
— François Choquette and James Johnson


  • Suspicious Results: Mitt Romney had suspicious vote gains in precincts where more people voted.
  • Statistical Evidence: Two statisticians detected a linear pattern consistent with electronic vote flipping.
  • Computer Manipulation: Algorithms are a likely explanation for the patterns observed in the data.
  • No Accountability: Many of the voting machines in use were riggable touchscreens that are impossible to audit.
  • Strategic Timing: The trend favoring Romney was stronger in the important early primaries that influence national opinion.
  • Partisan Technology: Romney's family and campaign had business and personal connections to the voting machine vendor Hart InterCivic.
  • Election Defense: A watchful citizen noticed that the Iowa GOP overstated Romney's vote total by a margin large enough to flip the Iowa caucus.

2012 Republican Primary

Estimate of Total Votes Flipped

Researchers used statistical analysis to detect an upward trend in Mitt Romney's vote totals across the 2012 Republican presidential primaries. By calculating Romney's divergence from expected results, they estimate that more than 1.2 million votes were flipped electronically, mostly from Santorum and Gingrich.


What caused the anomalous wins for Mitt Romney during the 2012 Republican primary?

Researchers François Choquette and James Johnson became interested in this question when South Carolina election workers noticed that Mitt Romney had unusual gains in larger precincts.

Choquette and Johnson looked at data from the 2012 Republican primaries and found a surprising trend: Mitt Romney’s share of the Republican vote increased in proportion to precinct size (as measured by number of votes).

New Hampshire 2012 Republican Primary

Iowa 2012 Republican Primary Caucus

Wisconsin 2012 Republican Primary

Ohio 2012 Republican Primary


This violated their statistical expectations. Typically what happens in large precincts is that a candidate’s vote share levels off to an average. But what they were seeing was a linear upward trend.

Choquette and Johnson investigated demographic explanations, yet the pattern persisted. Wealth, gender, and race didn’t correlate; the trend was present in urban as well as rural districts; and in Ohio they tested “republicanness”—the ratio of registered Republicans to Democrats—and found it didn’t correspond to precinct size either.

What the Data Should Look Like

Palm Beach County vs. Duval County

That the vote gain is linear as a function of cumulative precinct size indicates a computer algorithm could be at play.
— François Choquette and James Johnson


Choquette and Johnson also looked at previous Republican races. Romney benefited from a similar lift in the 2008 Republican primary—until he dropped out of race, when the large-precinct effect shifted to John McCain. Looking back at other elections they found no significant pattern for other GOP candidates.

They also looked at a wide range of state and local races—all of them leveled off as expected. The 2012 Democratic Party elections leveled off, too. In fact, Choquette and Johnson didn’t find the trend in any elections that didn’t include Republicans.

The one state primary that did not show a pro-Romney trend was Utah—the Mormon stronghold where he was expected to dominate.

So what was behind Romney’s surge against Rick Santorum, Newt Gingrich, and Ron Paul?



Choquette and Johnson noticed two other things: Counties without central tabulators didn’t exhibit this trend, and neither did hand-counted paper ballots. They theorize that electronic voting machines were the problem, and that someone flipped more than 1.2 million votes across the 2012 Republican primaries.

Anti-fraud measures typically look for ballot box stuffing. But with electronic vote flipping the total number of votes is kept intact by reallocating a proportion of votes between candidates, making it hard to detect.

Vote flipping is more efficient in large precincts because more votes can be shifted and fewer precincts have to be targeted. Small precincts and low-polling candidates aren’t targeted so as to avoid suspicion.


Estimating Flipped Votes

Choquette and Johnson employed a conservative method for estimating the number of votes gained or lost by each candidate. The first 5 percent of each cumulative vote chart is rejected because small precincts are highly variable. Next a median value is calculated between 5 percent and 20 percent of the precincts to establish a baseline. The baseline value is subtracted from the final result (at the rightmost point of the chart).

Vote Flipping Strategy

  • Vote flipping is best accomplished electronically.

  • Algorithms are used to reallocate a fraction of votes between candidates.

  • The total vote count is kept intact to make the result look legitimate.

  • Crooks target large precincts because: There are more votes to gain, and fewer precincts are needed to flip the result.

  • Flipping starts after a minimum threshold of votes to avoid detection in small precincts where anomalies might stand out.

  • Minor candidates are left alone to prevent negative vote totals.

  • Computer code can be written to self-erase.


Choquette and Johnson also noticed that the pro-Romney trend was stronger in the important early elections that influence public opinion. Based on their analysis, Romney should have come in third in Iowa (as opposed to second), and second in New Hampshire (as opposed to first).

The probability of this trend being a chance outcome is “beyond typical or even extreme,” write Choquette and Johnson.



Their research deserves further investigation since so many of America’s votes are cast on notoriously insecure electronic voting machines without an auditable paper trail. These deficient machines expose American elections to manipulation by malicious actors.  

“This is not a large conspiracy involving a complex network of perpetrators,” write Choquette and Johnson. “Such an alleged election fraud could be accomplished by only a single, highly clever computer programmer with access to voting machine software updates.”

American voters deserve to know that their votes count.

It is relatively simple to see that a large number of votes are being exchanged for the benefit of Republican candidates and never the reverse.
— François Choquette and James Johnson

Partisan Technology

Romney CONNECTED to Voting Machine Vendor HART INTERCIVIC

The Romney family and campaign had investment and personal connections to the voting machine vendor Hart InterCivic during the 2012 presidential campaign.

HIG Capital was the nexus of this blatant conflict of interest. According to Craig Unger, seven directors of HIG were former employees of Bain & Co., the consulting firm where Mitt Romney had once been CEO.

HIG Capital bought an ownership stake in Hart InterCivic, seating three out of five members of the board of directors.

HIG and Hart board members were also donors to the Romney campaign and actively bundling major donations for him. The private equity firm Solamere, run by Romney’s son, had investments with HIG.

Hart InterCivic voting machines were labeled highly insecure by a major research project in 2007.

Voting technology companies should be held to the highest standard of business ethics and technological performance given their important role in determining American governance.

Election Defense

Whistleblower Reports Inflated Romney Vote count in Iowa, Reversing the outcome

Mitt Romney was initially reported as the winner of the 2012 Republican primary in Iowa, but watchful citizen Edward True noticed that Rick Santorum was in fact the true winner—the Iowa GOP had misreported the vote count by enough votes to flip the caucus in Romney's favor

Whether accidental or deliberate, this incident highlights the fact that elections are wide open manipulation when results are tabulated and reported electronically. The Wall Street Journal and other news outlets covered True's revelation, but only after Romney had been portrayed as having electoral momentum.

Losing Candidate

Biggest vote flips from Gingrich Were in States with a History of Election Fraud

Choquette and Johnson estimate that Newt Gingrich lost 500,000 votes that were flipped to Romney. The biggest flips were in states where Gingrich got the most votes: Alabama, Florida, Georgia, and South Carolina.

Of those states, only Alabama uses optical scanners with auditable paper ballots. The other three all have some electronic voting machines without a paper trail—South Carolina and Georgia deploy these deficient systems across the entire state.

All four states are known for election fraud and voting machine manipulation: the 2000 presidential race in Florida, the Alabama governorship stolen from Don Siegelman in 2002, the 2002 Senate race in Georgia, and the 2010 Democratic U.S. Senate primary in South Carolina.